diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,151991 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 21707,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.6068088635002536e-05,
+      "grad_norm": 2.8885994615284454,
+      "learning_rate": 3.0674846625766876e-07,
+      "loss": 2.0386,
+      "step": 1
+    },
+    {
+      "epoch": 9.213617727000507e-05,
+      "grad_norm": 2.4853857454833372,
+      "learning_rate": 6.134969325153375e-07,
+      "loss": 2.0464,
+      "step": 2
+    },
+    {
+      "epoch": 0.0001382042659050076,
+      "grad_norm": 5.947029380866053,
+      "learning_rate": 9.202453987730062e-07,
+      "loss": 3.0483,
+      "step": 3
+    },
+    {
+      "epoch": 0.00018427235454001014,
+      "grad_norm": 1.9604672986800018,
+      "learning_rate": 1.226993865030675e-06,
+      "loss": 1.6708,
+      "step": 4
+    },
+    {
+      "epoch": 0.00023034044317501266,
+      "grad_norm": 3.7525376073921013,
+      "learning_rate": 1.5337423312883435e-06,
+      "loss": 2.4233,
+      "step": 5
+    },
+    {
+      "epoch": 0.0002764085318100152,
+      "grad_norm": 2.1055755212213287,
+      "learning_rate": 1.8404907975460124e-06,
+      "loss": 1.9335,
+      "step": 6
+    },
+    {
+      "epoch": 0.0003224766204450177,
+      "grad_norm": 1.909676022212204,
+      "learning_rate": 2.147239263803681e-06,
+      "loss": 2.1367,
+      "step": 7
+    },
+    {
+      "epoch": 0.0003685447090800203,
+      "grad_norm": 2.4600756165701916,
+      "learning_rate": 2.45398773006135e-06,
+      "loss": 2.2419,
+      "step": 8
+    },
+    {
+      "epoch": 0.0004146127977150228,
+      "grad_norm": 2.9572675617069697,
+      "learning_rate": 2.7607361963190186e-06,
+      "loss": 2.2226,
+      "step": 9
+    },
+    {
+      "epoch": 0.0004606808863500253,
+      "grad_norm": 2.35991144332441,
+      "learning_rate": 3.067484662576687e-06,
+      "loss": 1.848,
+      "step": 10
+    },
+    {
+      "epoch": 0.0005067489749850278,
+      "grad_norm": 3.3092518791016183,
+      "learning_rate": 3.374233128834356e-06,
+      "loss": 2.4174,
+      "step": 11
+    },
+    {
+      "epoch": 0.0005528170636200304,
+      "grad_norm": 2.8722596282010584,
+      "learning_rate": 3.680981595092025e-06,
+      "loss": 2.4017,
+      "step": 12
+    },
+    {
+      "epoch": 0.000598885152255033,
+      "grad_norm": 2.3586147168480265,
+      "learning_rate": 3.987730061349693e-06,
+      "loss": 2.1985,
+      "step": 13
+    },
+    {
+      "epoch": 0.0006449532408900354,
+      "grad_norm": 2.45748551895916,
+      "learning_rate": 4.294478527607362e-06,
+      "loss": 2.4288,
+      "step": 14
+    },
+    {
+      "epoch": 0.000691021329525038,
+      "grad_norm": 2.8873996547834166,
+      "learning_rate": 4.601226993865031e-06,
+      "loss": 2.2591,
+      "step": 15
+    },
+    {
+      "epoch": 0.0007370894181600406,
+      "grad_norm": 1.3361297459075852,
+      "learning_rate": 4.9079754601227e-06,
+      "loss": 2.1127,
+      "step": 16
+    },
+    {
+      "epoch": 0.000783157506795043,
+      "grad_norm": 2.0845066214353363,
+      "learning_rate": 5.214723926380368e-06,
+      "loss": 1.8089,
+      "step": 17
+    },
+    {
+      "epoch": 0.0008292255954300456,
+      "grad_norm": 1.0891753417218708,
+      "learning_rate": 5.521472392638037e-06,
+      "loss": 1.751,
+      "step": 18
+    },
+    {
+      "epoch": 0.0008752936840650482,
+      "grad_norm": 2.236673170554481,
+      "learning_rate": 5.828220858895706e-06,
+      "loss": 2.2301,
+      "step": 19
+    },
+    {
+      "epoch": 0.0009213617727000506,
+      "grad_norm": 2.2744396349837324,
+      "learning_rate": 6.134969325153374e-06,
+      "loss": 1.9663,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009674298613350532,
+      "grad_norm": 2.8223080207746913,
+      "learning_rate": 6.4417177914110434e-06,
+      "loss": 1.8689,
+      "step": 21
+    },
+    {
+      "epoch": 0.0010134979499700557,
+      "grad_norm": 2.3147856521490193,
+      "learning_rate": 6.748466257668712e-06,
+      "loss": 2.1413,
+      "step": 22
+    },
+    {
+      "epoch": 0.0010595660386050582,
+      "grad_norm": 2.1967873852880015,
+      "learning_rate": 7.05521472392638e-06,
+      "loss": 2.22,
+      "step": 23
+    },
+    {
+      "epoch": 0.0011056341272400608,
+      "grad_norm": 0.6177216024479529,
+      "learning_rate": 7.36196319018405e-06,
+      "loss": 1.3658,
+      "step": 24
+    },
+    {
+      "epoch": 0.0011517022158750634,
+      "grad_norm": 1.4023335936504202,
+      "learning_rate": 7.668711656441718e-06,
+      "loss": 1.8072,
+      "step": 25
+    },
+    {
+      "epoch": 0.001197770304510066,
+      "grad_norm": 0.8788482392639293,
+      "learning_rate": 7.975460122699386e-06,
+      "loss": 1.5616,
+      "step": 26
+    },
+    {
+      "epoch": 0.0012438383931450685,
+      "grad_norm": 0.5875247177968446,
+      "learning_rate": 8.282208588957055e-06,
+      "loss": 1.6255,
+      "step": 27
+    },
+    {
+      "epoch": 0.0012899064817800709,
+      "grad_norm": 1.2835039713615897,
+      "learning_rate": 8.588957055214725e-06,
+      "loss": 1.8658,
+      "step": 28
+    },
+    {
+      "epoch": 0.0013359745704150734,
+      "grad_norm": 1.2747503775435078,
+      "learning_rate": 8.895705521472392e-06,
+      "loss": 1.9026,
+      "step": 29
+    },
+    {
+      "epoch": 0.001382042659050076,
+      "grad_norm": 1.023457664442394,
+      "learning_rate": 9.202453987730062e-06,
+      "loss": 1.9068,
+      "step": 30
+    },
+    {
+      "epoch": 0.0014281107476850786,
+      "grad_norm": 1.4381142777661258,
+      "learning_rate": 9.509202453987731e-06,
+      "loss": 2.1089,
+      "step": 31
+    },
+    {
+      "epoch": 0.0014741788363200811,
+      "grad_norm": 1.4272368979226142,
+      "learning_rate": 9.8159509202454e-06,
+      "loss": 1.5095,
+      "step": 32
+    },
+    {
+      "epoch": 0.0015202469249550837,
+      "grad_norm": 1.3834343843564674,
+      "learning_rate": 1.0122699386503068e-05,
+      "loss": 1.5148,
+      "step": 33
+    },
+    {
+      "epoch": 0.001566315013590086,
+      "grad_norm": 0.7798213619632742,
+      "learning_rate": 1.0429447852760736e-05,
+      "loss": 1.7087,
+      "step": 34
+    },
+    {
+      "epoch": 0.0016123831022250886,
+      "grad_norm": 1.1658290335447212,
+      "learning_rate": 1.0736196319018407e-05,
+      "loss": 1.7121,
+      "step": 35
+    },
+    {
+      "epoch": 0.0016584511908600912,
+      "grad_norm": 1.0724729665462527,
+      "learning_rate": 1.1042944785276074e-05,
+      "loss": 1.9625,
+      "step": 36
+    },
+    {
+      "epoch": 0.0017045192794950938,
+      "grad_norm": 1.77426786426319,
+      "learning_rate": 1.1349693251533742e-05,
+      "loss": 1.3629,
+      "step": 37
+    },
+    {
+      "epoch": 0.0017505873681300963,
+      "grad_norm": 0.8387142524311813,
+      "learning_rate": 1.1656441717791411e-05,
+      "loss": 1.6369,
+      "step": 38
+    },
+    {
+      "epoch": 0.001796655456765099,
+      "grad_norm": 0.8738767732090972,
+      "learning_rate": 1.196319018404908e-05,
+      "loss": 1.9258,
+      "step": 39
+    },
+    {
+      "epoch": 0.0018427235454001013,
+      "grad_norm": 0.7846628224248722,
+      "learning_rate": 1.2269938650306748e-05,
+      "loss": 1.569,
+      "step": 40
+    },
+    {
+      "epoch": 0.0018887916340351038,
+      "grad_norm": 1.5987980504172683,
+      "learning_rate": 1.2576687116564418e-05,
+      "loss": 1.7139,
+      "step": 41
+    },
+    {
+      "epoch": 0.0019348597226701064,
+      "grad_norm": 0.9737408878604157,
+      "learning_rate": 1.2883435582822087e-05,
+      "loss": 1.8204,
+      "step": 42
+    },
+    {
+      "epoch": 0.001980927811305109,
+      "grad_norm": 0.7376015462269349,
+      "learning_rate": 1.3190184049079754e-05,
+      "loss": 1.478,
+      "step": 43
+    },
+    {
+      "epoch": 0.0020269958999401113,
+      "grad_norm": 0.7934221929590307,
+      "learning_rate": 1.3496932515337424e-05,
+      "loss": 1.4678,
+      "step": 44
+    },
+    {
+      "epoch": 0.002073063988575114,
+      "grad_norm": 0.9971279324447689,
+      "learning_rate": 1.3803680981595093e-05,
+      "loss": 1.3803,
+      "step": 45
+    },
+    {
+      "epoch": 0.0021191320772101165,
+      "grad_norm": 0.6821071761609323,
+      "learning_rate": 1.411042944785276e-05,
+      "loss": 1.5055,
+      "step": 46
+    },
+    {
+      "epoch": 0.0021652001658451193,
+      "grad_norm": 0.8908945329724476,
+      "learning_rate": 1.441717791411043e-05,
+      "loss": 1.8296,
+      "step": 47
+    },
+    {
+      "epoch": 0.0022112682544801216,
+      "grad_norm": 0.7473167209551226,
+      "learning_rate": 1.47239263803681e-05,
+      "loss": 1.2616,
+      "step": 48
+    },
+    {
+      "epoch": 0.002257336343115124,
+      "grad_norm": 0.6602531119536266,
+      "learning_rate": 1.5030674846625767e-05,
+      "loss": 1.4229,
+      "step": 49
+    },
+    {
+      "epoch": 0.0023034044317501268,
+      "grad_norm": 0.7487819515993295,
+      "learning_rate": 1.5337423312883436e-05,
+      "loss": 1.768,
+      "step": 50
+    },
+    {
+      "epoch": 0.002349472520385129,
+      "grad_norm": 0.8349944902040044,
+      "learning_rate": 1.5644171779141108e-05,
+      "loss": 1.3532,
+      "step": 51
+    },
+    {
+      "epoch": 0.002395540609020132,
+      "grad_norm": 1.0314123260394925,
+      "learning_rate": 1.5950920245398772e-05,
+      "loss": 1.4643,
+      "step": 52
+    },
+    {
+      "epoch": 0.0024416086976551342,
+      "grad_norm": 0.9284303181119671,
+      "learning_rate": 1.6257668711656443e-05,
+      "loss": 1.3444,
+      "step": 53
+    },
+    {
+      "epoch": 0.002487676786290137,
+      "grad_norm": 0.7226389963220002,
+      "learning_rate": 1.656441717791411e-05,
+      "loss": 1.2686,
+      "step": 54
+    },
+    {
+      "epoch": 0.0025337448749251394,
+      "grad_norm": 0.7386803623253643,
+      "learning_rate": 1.6871165644171778e-05,
+      "loss": 1.217,
+      "step": 55
+    },
+    {
+      "epoch": 0.0025798129635601417,
+      "grad_norm": 0.7715436531432492,
+      "learning_rate": 1.717791411042945e-05,
+      "loss": 1.8426,
+      "step": 56
+    },
+    {
+      "epoch": 0.0026258810521951445,
+      "grad_norm": 0.6712272114948813,
+      "learning_rate": 1.7484662576687117e-05,
+      "loss": 1.5499,
+      "step": 57
+    },
+    {
+      "epoch": 0.002671949140830147,
+      "grad_norm": 0.8639681702420693,
+      "learning_rate": 1.7791411042944784e-05,
+      "loss": 1.2879,
+      "step": 58
+    },
+    {
+      "epoch": 0.0027180172294651497,
+      "grad_norm": 0.6904441269764897,
+      "learning_rate": 1.8098159509202455e-05,
+      "loss": 1.5952,
+      "step": 59
+    },
+    {
+      "epoch": 0.002764085318100152,
+      "grad_norm": 0.7838802034334353,
+      "learning_rate": 1.8404907975460123e-05,
+      "loss": 1.4813,
+      "step": 60
+    },
+    {
+      "epoch": 0.0028101534067351544,
+      "grad_norm": 0.8301478444670534,
+      "learning_rate": 1.8711656441717794e-05,
+      "loss": 1.2703,
+      "step": 61
+    },
+    {
+      "epoch": 0.002856221495370157,
+      "grad_norm": 0.6072030856856849,
+      "learning_rate": 1.9018404907975462e-05,
+      "loss": 1.075,
+      "step": 62
+    },
+    {
+      "epoch": 0.0029022895840051595,
+      "grad_norm": 0.9399462674501765,
+      "learning_rate": 1.932515337423313e-05,
+      "loss": 1.5042,
+      "step": 63
+    },
+    {
+      "epoch": 0.0029483576726401623,
+      "grad_norm": 1.0748386253640623,
+      "learning_rate": 1.96319018404908e-05,
+      "loss": 1.041,
+      "step": 64
+    },
+    {
+      "epoch": 0.0029944257612751646,
+      "grad_norm": 0.8759067264921214,
+      "learning_rate": 1.9938650306748465e-05,
+      "loss": 1.501,
+      "step": 65
+    },
+    {
+      "epoch": 0.0030404938499101674,
+      "grad_norm": 0.5677778194060107,
+      "learning_rate": 2.0245398773006136e-05,
+      "loss": 1.6372,
+      "step": 66
+    },
+    {
+      "epoch": 0.00308656193854517,
+      "grad_norm": 0.6306116113124961,
+      "learning_rate": 2.0552147239263807e-05,
+      "loss": 1.29,
+      "step": 67
+    },
+    {
+      "epoch": 0.003132630027180172,
+      "grad_norm": 0.7061779485924772,
+      "learning_rate": 2.085889570552147e-05,
+      "loss": 1.2191,
+      "step": 68
+    },
+    {
+      "epoch": 0.003178698115815175,
+      "grad_norm": 0.9238831209661453,
+      "learning_rate": 2.1165644171779142e-05,
+      "loss": 0.8881,
+      "step": 69
+    },
+    {
+      "epoch": 0.0032247662044501773,
+      "grad_norm": 0.6731632831571692,
+      "learning_rate": 2.1472392638036813e-05,
+      "loss": 1.6979,
+      "step": 70
+    },
+    {
+      "epoch": 0.00327083429308518,
+      "grad_norm": 0.5785625144198001,
+      "learning_rate": 2.1779141104294477e-05,
+      "loss": 1.3842,
+      "step": 71
+    },
+    {
+      "epoch": 0.0033169023817201824,
+      "grad_norm": 0.5595957224497377,
+      "learning_rate": 2.208588957055215e-05,
+      "loss": 1.3164,
+      "step": 72
+    },
+    {
+      "epoch": 0.0033629704703551848,
+      "grad_norm": 0.6637814500101477,
+      "learning_rate": 2.239263803680982e-05,
+      "loss": 1.4102,
+      "step": 73
+    },
+    {
+      "epoch": 0.0034090385589901876,
+      "grad_norm": 0.7078963685671174,
+      "learning_rate": 2.2699386503067484e-05,
+      "loss": 1.3789,
+      "step": 74
+    },
+    {
+      "epoch": 0.00345510664762519,
+      "grad_norm": 0.6188898506780091,
+      "learning_rate": 2.3006134969325155e-05,
+      "loss": 1.1374,
+      "step": 75
+    },
+    {
+      "epoch": 0.0035011747362601927,
+      "grad_norm": 0.7123580209322591,
+      "learning_rate": 2.3312883435582822e-05,
+      "loss": 1.6773,
+      "step": 76
+    },
+    {
+      "epoch": 0.003547242824895195,
+      "grad_norm": 0.6416400845419364,
+      "learning_rate": 2.361963190184049e-05,
+      "loss": 1.5887,
+      "step": 77
+    },
+    {
+      "epoch": 0.003593310913530198,
+      "grad_norm": 0.8472064113816509,
+      "learning_rate": 2.392638036809816e-05,
+      "loss": 1.4512,
+      "step": 78
+    },
+    {
+      "epoch": 0.0036393790021652,
+      "grad_norm": 0.5690622716386861,
+      "learning_rate": 2.423312883435583e-05,
+      "loss": 1.3825,
+      "step": 79
+    },
+    {
+      "epoch": 0.0036854470908002025,
+      "grad_norm": 0.6612031488359771,
+      "learning_rate": 2.4539877300613496e-05,
+      "loss": 1.2815,
+      "step": 80
+    },
+    {
+      "epoch": 0.0037315151794352053,
+      "grad_norm": 0.5838617309456415,
+      "learning_rate": 2.4846625766871167e-05,
+      "loss": 1.5907,
+      "step": 81
+    },
+    {
+      "epoch": 0.0037775832680702077,
+      "grad_norm": 0.6831568868552844,
+      "learning_rate": 2.5153374233128835e-05,
+      "loss": 1.1894,
+      "step": 82
+    },
+    {
+      "epoch": 0.0038236513567052105,
+      "grad_norm": 0.8666771269835946,
+      "learning_rate": 2.5460122699386503e-05,
+      "loss": 1.319,
+      "step": 83
+    },
+    {
+      "epoch": 0.003869719445340213,
+      "grad_norm": 0.6210786965278764,
+      "learning_rate": 2.5766871165644174e-05,
+      "loss": 1.1669,
+      "step": 84
+    },
+    {
+      "epoch": 0.003915787533975215,
+      "grad_norm": 0.7018582293018187,
+      "learning_rate": 2.607361963190184e-05,
+      "loss": 1.2995,
+      "step": 85
+    },
+    {
+      "epoch": 0.003961855622610218,
+      "grad_norm": 1.6080189872276471,
+      "learning_rate": 2.638036809815951e-05,
+      "loss": 1.5293,
+      "step": 86
+    },
+    {
+      "epoch": 0.004007923711245221,
+      "grad_norm": 0.6192138102568777,
+      "learning_rate": 2.668711656441718e-05,
+      "loss": 1.1555,
+      "step": 87
+    },
+    {
+      "epoch": 0.004053991799880223,
+      "grad_norm": 0.6977644881663772,
+      "learning_rate": 2.6993865030674848e-05,
+      "loss": 1.5977,
+      "step": 88
+    },
+    {
+      "epoch": 0.0041000598885152255,
+      "grad_norm": 0.833586077535666,
+      "learning_rate": 2.7300613496932515e-05,
+      "loss": 1.531,
+      "step": 89
+    },
+    {
+      "epoch": 0.004146127977150228,
+      "grad_norm": 0.684665427155929,
+      "learning_rate": 2.7607361963190186e-05,
+      "loss": 1.492,
+      "step": 90
+    },
+    {
+      "epoch": 0.00419219606578523,
+      "grad_norm": 0.4817174083175216,
+      "learning_rate": 2.7914110429447854e-05,
+      "loss": 0.9086,
+      "step": 91
+    },
+    {
+      "epoch": 0.004238264154420233,
+      "grad_norm": 0.6164597285614591,
+      "learning_rate": 2.822085889570552e-05,
+      "loss": 1.2594,
+      "step": 92
+    },
+    {
+      "epoch": 0.004284332243055236,
+      "grad_norm": 0.7856198371013854,
+      "learning_rate": 2.8527607361963193e-05,
+      "loss": 1.4152,
+      "step": 93
+    },
+    {
+      "epoch": 0.0043304003316902385,
+      "grad_norm": 0.7100667625007688,
+      "learning_rate": 2.883435582822086e-05,
+      "loss": 1.4944,
+      "step": 94
+    },
+    {
+      "epoch": 0.0043764684203252404,
+      "grad_norm": 0.6149503874285197,
+      "learning_rate": 2.9141104294478528e-05,
+      "loss": 1.1667,
+      "step": 95
+    },
+    {
+      "epoch": 0.004422536508960243,
+      "grad_norm": 1.8226457809928336,
+      "learning_rate": 2.94478527607362e-05,
+      "loss": 1.7735,
+      "step": 96
+    },
+    {
+      "epoch": 0.004468604597595246,
+      "grad_norm": 0.6186089181012165,
+      "learning_rate": 2.9754601226993867e-05,
+      "loss": 1.5332,
+      "step": 97
+    },
+    {
+      "epoch": 0.004514672686230248,
+      "grad_norm": 0.5807301295288216,
+      "learning_rate": 3.0061349693251534e-05,
+      "loss": 1.3597,
+      "step": 98
+    },
+    {
+      "epoch": 0.004560740774865251,
+      "grad_norm": 0.8064744886325684,
+      "learning_rate": 3.0368098159509205e-05,
+      "loss": 1.6612,
+      "step": 99
+    },
+    {
+      "epoch": 0.0046068088635002535,
+      "grad_norm": 0.7416955797006362,
+      "learning_rate": 3.067484662576687e-05,
+      "loss": 1.5032,
+      "step": 100
+    },
+    {
+      "epoch": 0.004652876952135256,
+      "grad_norm": 0.6646870457942197,
+      "learning_rate": 3.098159509202454e-05,
+      "loss": 1.0803,
+      "step": 101
+    },
+    {
+      "epoch": 0.004698945040770258,
+      "grad_norm": 0.7532145052689211,
+      "learning_rate": 3.1288343558282215e-05,
+      "loss": 1.3275,
+      "step": 102
+    },
+    {
+      "epoch": 0.004745013129405261,
+      "grad_norm": 0.6111110939858769,
+      "learning_rate": 3.159509202453988e-05,
+      "loss": 1.6855,
+      "step": 103
+    },
+    {
+      "epoch": 0.004791081218040264,
+      "grad_norm": 0.7062427932249076,
+      "learning_rate": 3.1901840490797544e-05,
+      "loss": 0.9547,
+      "step": 104
+    },
+    {
+      "epoch": 0.004837149306675266,
+      "grad_norm": 0.6767567349988785,
+      "learning_rate": 3.220858895705521e-05,
+      "loss": 1.3032,
+      "step": 105
+    },
+    {
+      "epoch": 0.0048832173953102685,
+      "grad_norm": 0.6645286415811804,
+      "learning_rate": 3.2515337423312886e-05,
+      "loss": 1.046,
+      "step": 106
+    },
+    {
+      "epoch": 0.004929285483945271,
+      "grad_norm": 0.7087126373716247,
+      "learning_rate": 3.282208588957055e-05,
+      "loss": 1.1283,
+      "step": 107
+    },
+    {
+      "epoch": 0.004975353572580274,
+      "grad_norm": 0.7301061932197864,
+      "learning_rate": 3.312883435582822e-05,
+      "loss": 1.3343,
+      "step": 108
+    },
+    {
+      "epoch": 0.005021421661215276,
+      "grad_norm": 0.6424112948228664,
+      "learning_rate": 3.3435582822085895e-05,
+      "loss": 1.2865,
+      "step": 109
+    },
+    {
+      "epoch": 0.005067489749850279,
+      "grad_norm": 0.5953421644665682,
+      "learning_rate": 3.3742331288343556e-05,
+      "loss": 1.2971,
+      "step": 110
+    },
+    {
+      "epoch": 0.0051135578384852816,
+      "grad_norm": 0.720132742910124,
+      "learning_rate": 3.4049079754601224e-05,
+      "loss": 1.2389,
+      "step": 111
+    },
+    {
+      "epoch": 0.0051596259271202835,
+      "grad_norm": 0.7363404655824318,
+      "learning_rate": 3.43558282208589e-05,
+      "loss": 1.407,
+      "step": 112
+    },
+    {
+      "epoch": 0.005205694015755286,
+      "grad_norm": 0.7284798145674991,
+      "learning_rate": 3.4662576687116566e-05,
+      "loss": 0.9115,
+      "step": 113
+    },
+    {
+      "epoch": 0.005251762104390289,
+      "grad_norm": 0.5160058636188735,
+      "learning_rate": 3.4969325153374234e-05,
+      "loss": 1.3814,
+      "step": 114
+    },
+    {
+      "epoch": 0.005297830193025291,
+      "grad_norm": 0.7810931947007471,
+      "learning_rate": 3.527607361963191e-05,
+      "loss": 1.3892,
+      "step": 115
+    },
+    {
+      "epoch": 0.005343898281660294,
+      "grad_norm": 1.1390115963276881,
+      "learning_rate": 3.558282208588957e-05,
+      "loss": 1.3394,
+      "step": 116
+    },
+    {
+      "epoch": 0.0053899663702952965,
+      "grad_norm": 0.6511536411964898,
+      "learning_rate": 3.5889570552147236e-05,
+      "loss": 1.2912,
+      "step": 117
+    },
+    {
+      "epoch": 0.005436034458930299,
+      "grad_norm": 0.6156704331014677,
+      "learning_rate": 3.619631901840491e-05,
+      "loss": 1.6418,
+      "step": 118
+    },
+    {
+      "epoch": 0.005482102547565301,
+      "grad_norm": 0.6925596507860653,
+      "learning_rate": 3.650306748466258e-05,
+      "loss": 1.6097,
+      "step": 119
+    },
+    {
+      "epoch": 0.005528170636200304,
+      "grad_norm": 0.8030180549468483,
+      "learning_rate": 3.6809815950920246e-05,
+      "loss": 1.4437,
+      "step": 120
+    },
+    {
+      "epoch": 0.005574238724835307,
+      "grad_norm": 0.6158658459086824,
+      "learning_rate": 3.711656441717792e-05,
+      "loss": 1.3353,
+      "step": 121
+    },
+    {
+      "epoch": 0.005620306813470309,
+      "grad_norm": 0.6335028476264791,
+      "learning_rate": 3.742331288343559e-05,
+      "loss": 1.3575,
+      "step": 122
+    },
+    {
+      "epoch": 0.0056663749021053115,
+      "grad_norm": 0.7141078537461296,
+      "learning_rate": 3.773006134969325e-05,
+      "loss": 1.4996,
+      "step": 123
+    },
+    {
+      "epoch": 0.005712442990740314,
+      "grad_norm": 0.6376448303538451,
+      "learning_rate": 3.8036809815950924e-05,
+      "loss": 1.2437,
+      "step": 124
+    },
+    {
+      "epoch": 0.005758511079375317,
+      "grad_norm": 0.7810764633480157,
+      "learning_rate": 3.834355828220859e-05,
+      "loss": 1.4916,
+      "step": 125
+    },
+    {
+      "epoch": 0.005804579168010319,
+      "grad_norm": 0.6247377260042919,
+      "learning_rate": 3.865030674846626e-05,
+      "loss": 1.3042,
+      "step": 126
+    },
+    {
+      "epoch": 0.005850647256645322,
+      "grad_norm": 0.5011246604985055,
+      "learning_rate": 3.895705521472393e-05,
+      "loss": 1.1107,
+      "step": 127
+    },
+    {
+      "epoch": 0.005896715345280325,
+      "grad_norm": 0.6021163174414093,
+      "learning_rate": 3.92638036809816e-05,
+      "loss": 1.0084,
+      "step": 128
+    },
+    {
+      "epoch": 0.0059427834339153265,
+      "grad_norm": 0.919164352959516,
+      "learning_rate": 3.957055214723926e-05,
+      "loss": 1.2499,
+      "step": 129
+    },
+    {
+      "epoch": 0.005988851522550329,
+      "grad_norm": 0.7269027536460928,
+      "learning_rate": 3.987730061349693e-05,
+      "loss": 1.5578,
+      "step": 130
+    },
+    {
+      "epoch": 0.006034919611185332,
+      "grad_norm": 0.7439218527853402,
+      "learning_rate": 4.0184049079754604e-05,
+      "loss": 1.2615,
+      "step": 131
+    },
+    {
+      "epoch": 0.006080987699820335,
+      "grad_norm": 0.717582575080114,
+      "learning_rate": 4.049079754601227e-05,
+      "loss": 1.521,
+      "step": 132
+    },
+    {
+      "epoch": 0.006127055788455337,
+      "grad_norm": 0.7655585604986982,
+      "learning_rate": 4.079754601226994e-05,
+      "loss": 1.2564,
+      "step": 133
+    },
+    {
+      "epoch": 0.00617312387709034,
+      "grad_norm": 0.999660999046999,
+      "learning_rate": 4.1104294478527614e-05,
+      "loss": 1.1935,
+      "step": 134
+    },
+    {
+      "epoch": 0.006219191965725342,
+      "grad_norm": 0.5823313516952822,
+      "learning_rate": 4.1411042944785274e-05,
+      "loss": 1.3401,
+      "step": 135
+    },
+    {
+      "epoch": 0.006265260054360344,
+      "grad_norm": 0.7496251306467454,
+      "learning_rate": 4.171779141104294e-05,
+      "loss": 1.6325,
+      "step": 136
+    },
+    {
+      "epoch": 0.006311328142995347,
+      "grad_norm": 0.8040657656598834,
+      "learning_rate": 4.2024539877300617e-05,
+      "loss": 1.546,
+      "step": 137
+    },
+    {
+      "epoch": 0.00635739623163035,
+      "grad_norm": 0.6188497962280631,
+      "learning_rate": 4.2331288343558284e-05,
+      "loss": 0.9043,
+      "step": 138
+    },
+    {
+      "epoch": 0.006403464320265352,
+      "grad_norm": 0.6911924162859143,
+      "learning_rate": 4.263803680981595e-05,
+      "loss": 1.1394,
+      "step": 139
+    },
+    {
+      "epoch": 0.0064495324089003546,
+      "grad_norm": 0.706981188313127,
+      "learning_rate": 4.2944785276073626e-05,
+      "loss": 1.3262,
+      "step": 140
+    },
+    {
+      "epoch": 0.006495600497535357,
+      "grad_norm": 0.6236099476645077,
+      "learning_rate": 4.3251533742331294e-05,
+      "loss": 1.3226,
+      "step": 141
+    },
+    {
+      "epoch": 0.00654166858617036,
+      "grad_norm": 0.6189147384850641,
+      "learning_rate": 4.3558282208588955e-05,
+      "loss": 1.3624,
+      "step": 142
+    },
+    {
+      "epoch": 0.006587736674805362,
+      "grad_norm": 0.7441504021914012,
+      "learning_rate": 4.386503067484663e-05,
+      "loss": 1.4312,
+      "step": 143
+    },
+    {
+      "epoch": 0.006633804763440365,
+      "grad_norm": 0.656012673762119,
+      "learning_rate": 4.41717791411043e-05,
+      "loss": 1.6262,
+      "step": 144
+    },
+    {
+      "epoch": 0.006679872852075368,
+      "grad_norm": 0.7668798837523663,
+      "learning_rate": 4.4478527607361964e-05,
+      "loss": 1.1004,
+      "step": 145
+    },
+    {
+      "epoch": 0.0067259409407103695,
+      "grad_norm": 0.524625452887265,
+      "learning_rate": 4.478527607361964e-05,
+      "loss": 0.7794,
+      "step": 146
+    },
+    {
+      "epoch": 0.006772009029345372,
+      "grad_norm": 0.5760201751085735,
+      "learning_rate": 4.5092024539877307e-05,
+      "loss": 1.4731,
+      "step": 147
+    },
+    {
+      "epoch": 0.006818077117980375,
+      "grad_norm": 0.7355877159114541,
+      "learning_rate": 4.539877300613497e-05,
+      "loss": 1.3468,
+      "step": 148
+    },
+    {
+      "epoch": 0.006864145206615378,
+      "grad_norm": 0.6666419833453587,
+      "learning_rate": 4.570552147239264e-05,
+      "loss": 0.9562,
+      "step": 149
+    },
+    {
+      "epoch": 0.00691021329525038,
+      "grad_norm": 0.7290806760830116,
+      "learning_rate": 4.601226993865031e-05,
+      "loss": 1.2349,
+      "step": 150
+    },
+    {
+      "epoch": 0.006956281383885383,
+      "grad_norm": 0.9227103349254063,
+      "learning_rate": 4.631901840490798e-05,
+      "loss": 1.4508,
+      "step": 151
+    },
+    {
+      "epoch": 0.007002349472520385,
+      "grad_norm": 0.55083708176283,
+      "learning_rate": 4.6625766871165645e-05,
+      "loss": 1.2044,
+      "step": 152
+    },
+    {
+      "epoch": 0.007048417561155387,
+      "grad_norm": 0.7912929727375952,
+      "learning_rate": 4.693251533742332e-05,
+      "loss": 1.1106,
+      "step": 153
+    },
+    {
+      "epoch": 0.00709448564979039,
+      "grad_norm": 0.6432400526838047,
+      "learning_rate": 4.723926380368098e-05,
+      "loss": 1.3203,
+      "step": 154
+    },
+    {
+      "epoch": 0.007140553738425393,
+      "grad_norm": 0.6651010161265403,
+      "learning_rate": 4.754601226993865e-05,
+      "loss": 1.2899,
+      "step": 155
+    },
+    {
+      "epoch": 0.007186621827060396,
+      "grad_norm": 0.7910931208807749,
+      "learning_rate": 4.785276073619632e-05,
+      "loss": 1.1118,
+      "step": 156
+    },
+    {
+      "epoch": 0.007232689915695398,
+      "grad_norm": 0.7401880428005956,
+      "learning_rate": 4.815950920245399e-05,
+      "loss": 1.6995,
+      "step": 157
+    },
+    {
+      "epoch": 0.0072787580043304,
+      "grad_norm": 0.6571139851425342,
+      "learning_rate": 4.846625766871166e-05,
+      "loss": 0.8387,
+      "step": 158
+    },
+    {
+      "epoch": 0.007324826092965403,
+      "grad_norm": 0.9692007769728087,
+      "learning_rate": 4.877300613496933e-05,
+      "loss": 1.1451,
+      "step": 159
+    },
+    {
+      "epoch": 0.007370894181600405,
+      "grad_norm": 0.9466986037847843,
+      "learning_rate": 4.907975460122699e-05,
+      "loss": 1.1712,
+      "step": 160
+    },
+    {
+      "epoch": 0.007416962270235408,
+      "grad_norm": 0.5715116600558664,
+      "learning_rate": 4.938650306748466e-05,
+      "loss": 1.0672,
+      "step": 161
+    },
+    {
+      "epoch": 0.007463030358870411,
+      "grad_norm": 0.7323648990183943,
+      "learning_rate": 4.9693251533742335e-05,
+      "loss": 1.2996,
+      "step": 162
+    },
+    {
+      "epoch": 0.007509098447505413,
+      "grad_norm": 0.5163347433102878,
+      "learning_rate": 5e-05,
+      "loss": 1.5478,
+      "step": 163
+    },
+    {
+      "epoch": 0.007555166536140415,
+      "grad_norm": 0.6865937355385883,
+      "learning_rate": 5.030674846625767e-05,
+      "loss": 1.2673,
+      "step": 164
+    },
+    {
+      "epoch": 0.007601234624775418,
+      "grad_norm": 0.5306487699533297,
+      "learning_rate": 5.061349693251534e-05,
+      "loss": 0.9874,
+      "step": 165
+    },
+    {
+      "epoch": 0.007647302713410421,
+      "grad_norm": 0.742610611177671,
+      "learning_rate": 5.0920245398773005e-05,
+      "loss": 1.2345,
+      "step": 166
+    },
+    {
+      "epoch": 0.007693370802045423,
+      "grad_norm": 1.4909522833434454,
+      "learning_rate": 5.122699386503068e-05,
+      "loss": 1.2877,
+      "step": 167
+    },
+    {
+      "epoch": 0.007739438890680426,
+      "grad_norm": 0.7625995233677858,
+      "learning_rate": 5.153374233128835e-05,
+      "loss": 1.0415,
+      "step": 168
+    },
+    {
+      "epoch": 0.007785506979315428,
+      "grad_norm": 0.6072134602644816,
+      "learning_rate": 5.1840490797546015e-05,
+      "loss": 1.753,
+      "step": 169
+    },
+    {
+      "epoch": 0.00783157506795043,
+      "grad_norm": 1.0113010389862784,
+      "learning_rate": 5.214723926380368e-05,
+      "loss": 1.5317,
+      "step": 170
+    },
+    {
+      "epoch": 0.007877643156585434,
+      "grad_norm": 0.8316756987923936,
+      "learning_rate": 5.245398773006135e-05,
+      "loss": 1.5424,
+      "step": 171
+    },
+    {
+      "epoch": 0.007923711245220436,
+      "grad_norm": 0.6122711249695125,
+      "learning_rate": 5.276073619631902e-05,
+      "loss": 1.3191,
+      "step": 172
+    },
+    {
+      "epoch": 0.007969779333855438,
+      "grad_norm": 0.575186231790355,
+      "learning_rate": 5.306748466257669e-05,
+      "loss": 1.4355,
+      "step": 173
+    },
+    {
+      "epoch": 0.008015847422490441,
+      "grad_norm": 0.5751874557217245,
+      "learning_rate": 5.337423312883436e-05,
+      "loss": 1.3828,
+      "step": 174
+    },
+    {
+      "epoch": 0.008061915511125443,
+      "grad_norm": 0.6391903506497365,
+      "learning_rate": 5.368098159509203e-05,
+      "loss": 1.3707,
+      "step": 175
+    },
+    {
+      "epoch": 0.008107983599760445,
+      "grad_norm": 0.4935099545429967,
+      "learning_rate": 5.3987730061349695e-05,
+      "loss": 0.9917,
+      "step": 176
+    },
+    {
+      "epoch": 0.008154051688395449,
+      "grad_norm": 0.6133616944494008,
+      "learning_rate": 5.429447852760736e-05,
+      "loss": 1.1758,
+      "step": 177
+    },
+    {
+      "epoch": 0.008200119777030451,
+      "grad_norm": 0.6447769398369667,
+      "learning_rate": 5.460122699386503e-05,
+      "loss": 1.147,
+      "step": 178
+    },
+    {
+      "epoch": 0.008246187865665453,
+      "grad_norm": 0.6055437570543893,
+      "learning_rate": 5.4907975460122705e-05,
+      "loss": 1.2724,
+      "step": 179
+    },
+    {
+      "epoch": 0.008292255954300456,
+      "grad_norm": 0.4997247955692959,
+      "learning_rate": 5.521472392638037e-05,
+      "loss": 1.0677,
+      "step": 180
+    },
+    {
+      "epoch": 0.008338324042935458,
+      "grad_norm": 0.6963439101408938,
+      "learning_rate": 5.552147239263804e-05,
+      "loss": 1.6119,
+      "step": 181
+    },
+    {
+      "epoch": 0.00838439213157046,
+      "grad_norm": 0.6736283232839384,
+      "learning_rate": 5.582822085889571e-05,
+      "loss": 1.2911,
+      "step": 182
+    },
+    {
+      "epoch": 0.008430460220205464,
+      "grad_norm": 0.6626058588056297,
+      "learning_rate": 5.6134969325153376e-05,
+      "loss": 1.3027,
+      "step": 183
+    },
+    {
+      "epoch": 0.008476528308840466,
+      "grad_norm": 0.6084146403327817,
+      "learning_rate": 5.644171779141104e-05,
+      "loss": 1.145,
+      "step": 184
+    },
+    {
+      "epoch": 0.00852259639747547,
+      "grad_norm": 0.6616488962349748,
+      "learning_rate": 5.674846625766872e-05,
+      "loss": 1.1377,
+      "step": 185
+    },
+    {
+      "epoch": 0.008568664486110471,
+      "grad_norm": 0.5784116149188117,
+      "learning_rate": 5.7055214723926385e-05,
+      "loss": 1.4958,
+      "step": 186
+    },
+    {
+      "epoch": 0.008614732574745473,
+      "grad_norm": 0.662803931158815,
+      "learning_rate": 5.736196319018405e-05,
+      "loss": 1.5426,
+      "step": 187
+    },
+    {
+      "epoch": 0.008660800663380477,
+      "grad_norm": 0.5539879147736603,
+      "learning_rate": 5.766871165644172e-05,
+      "loss": 1.0627,
+      "step": 188
+    },
+    {
+      "epoch": 0.008706868752015479,
+      "grad_norm": 0.6837886155548699,
+      "learning_rate": 5.797546012269939e-05,
+      "loss": 1.3584,
+      "step": 189
+    },
+    {
+      "epoch": 0.008752936840650481,
+      "grad_norm": 0.5644017805759372,
+      "learning_rate": 5.8282208588957056e-05,
+      "loss": 1.5029,
+      "step": 190
+    },
+    {
+      "epoch": 0.008799004929285485,
+      "grad_norm": 0.6491429573313913,
+      "learning_rate": 5.8588957055214724e-05,
+      "loss": 1.5192,
+      "step": 191
+    },
+    {
+      "epoch": 0.008845073017920486,
+      "grad_norm": 0.5392169549461656,
+      "learning_rate": 5.88957055214724e-05,
+      "loss": 1.2997,
+      "step": 192
+    },
+    {
+      "epoch": 0.008891141106555488,
+      "grad_norm": 0.7494238742353582,
+      "learning_rate": 5.9202453987730066e-05,
+      "loss": 1.317,
+      "step": 193
+    },
+    {
+      "epoch": 0.008937209195190492,
+      "grad_norm": 0.6184364969104925,
+      "learning_rate": 5.950920245398773e-05,
+      "loss": 1.281,
+      "step": 194
+    },
+    {
+      "epoch": 0.008983277283825494,
+      "grad_norm": 0.8851054066161518,
+      "learning_rate": 5.98159509202454e-05,
+      "loss": 1.223,
+      "step": 195
+    },
+    {
+      "epoch": 0.009029345372460496,
+      "grad_norm": 0.5828483996318384,
+      "learning_rate": 6.012269938650307e-05,
+      "loss": 1.1829,
+      "step": 196
+    },
+    {
+      "epoch": 0.0090754134610955,
+      "grad_norm": 0.515064401434602,
+      "learning_rate": 6.0429447852760736e-05,
+      "loss": 1.4175,
+      "step": 197
+    },
+    {
+      "epoch": 0.009121481549730501,
+      "grad_norm": 0.6396071912882595,
+      "learning_rate": 6.073619631901841e-05,
+      "loss": 1.1819,
+      "step": 198
+    },
+    {
+      "epoch": 0.009167549638365503,
+      "grad_norm": 0.7183317170306112,
+      "learning_rate": 6.104294478527609e-05,
+      "loss": 1.2914,
+      "step": 199
+    },
+    {
+      "epoch": 0.009213617727000507,
+      "grad_norm": 0.6423564725505114,
+      "learning_rate": 6.134969325153375e-05,
+      "loss": 1.3627,
+      "step": 200
+    },
+    {
+      "epoch": 0.009259685815635509,
+      "grad_norm": 0.5162597328404863,
+      "learning_rate": 6.165644171779141e-05,
+      "loss": 1.2203,
+      "step": 201
+    },
+    {
+      "epoch": 0.009305753904270513,
+      "grad_norm": 0.8676238754697555,
+      "learning_rate": 6.196319018404908e-05,
+      "loss": 1.0711,
+      "step": 202
+    },
+    {
+      "epoch": 0.009351821992905514,
+      "grad_norm": 0.5210090820619856,
+      "learning_rate": 6.226993865030674e-05,
+      "loss": 1.2501,
+      "step": 203
+    },
+    {
+      "epoch": 0.009397890081540516,
+      "grad_norm": 0.7963467904772505,
+      "learning_rate": 6.257668711656443e-05,
+      "loss": 1.4058,
+      "step": 204
+    },
+    {
+      "epoch": 0.00944395817017552,
+      "grad_norm": 0.5090022737582941,
+      "learning_rate": 6.288343558282209e-05,
+      "loss": 1.5013,
+      "step": 205
+    },
+    {
+      "epoch": 0.009490026258810522,
+      "grad_norm": 0.5168324015317822,
+      "learning_rate": 6.319018404907977e-05,
+      "loss": 1.168,
+      "step": 206
+    },
+    {
+      "epoch": 0.009536094347445524,
+      "grad_norm": 0.6188000061937989,
+      "learning_rate": 6.349693251533743e-05,
+      "loss": 1.4175,
+      "step": 207
+    },
+    {
+      "epoch": 0.009582162436080528,
+      "grad_norm": 0.7586415040107253,
+      "learning_rate": 6.380368098159509e-05,
+      "loss": 1.3299,
+      "step": 208
+    },
+    {
+      "epoch": 0.00962823052471553,
+      "grad_norm": 0.5695027547319075,
+      "learning_rate": 6.411042944785276e-05,
+      "loss": 1.5314,
+      "step": 209
+    },
+    {
+      "epoch": 0.009674298613350531,
+      "grad_norm": 0.5657802154412087,
+      "learning_rate": 6.441717791411042e-05,
+      "loss": 1.3973,
+      "step": 210
+    },
+    {
+      "epoch": 0.009720366701985535,
+      "grad_norm": 0.7188241508937754,
+      "learning_rate": 6.472392638036811e-05,
+      "loss": 1.4455,
+      "step": 211
+    },
+    {
+      "epoch": 0.009766434790620537,
+      "grad_norm": 0.6946776996318826,
+      "learning_rate": 6.503067484662577e-05,
+      "loss": 1.3466,
+      "step": 212
+    },
+    {
+      "epoch": 0.009812502879255539,
+      "grad_norm": 0.7693078156017218,
+      "learning_rate": 6.533742331288345e-05,
+      "loss": 1.2718,
+      "step": 213
+    },
+    {
+      "epoch": 0.009858570967890543,
+      "grad_norm": 0.5421221079388185,
+      "learning_rate": 6.56441717791411e-05,
+      "loss": 1.2914,
+      "step": 214
+    },
+    {
+      "epoch": 0.009904639056525544,
+      "grad_norm": 0.8079465991172103,
+      "learning_rate": 6.595092024539877e-05,
+      "loss": 1.3057,
+      "step": 215
+    },
+    {
+      "epoch": 0.009950707145160548,
+      "grad_norm": 0.6622768363574356,
+      "learning_rate": 6.625766871165644e-05,
+      "loss": 1.5215,
+      "step": 216
+    },
+    {
+      "epoch": 0.00999677523379555,
+      "grad_norm": 0.5848111219053423,
+      "learning_rate": 6.656441717791412e-05,
+      "loss": 1.5937,
+      "step": 217
+    },
+    {
+      "epoch": 0.010042843322430552,
+      "grad_norm": 0.6417774602409705,
+      "learning_rate": 6.687116564417179e-05,
+      "loss": 0.9814,
+      "step": 218
+    },
+    {
+      "epoch": 0.010088911411065556,
+      "grad_norm": 0.493836989327599,
+      "learning_rate": 6.717791411042945e-05,
+      "loss": 1.3841,
+      "step": 219
+    },
+    {
+      "epoch": 0.010134979499700558,
+      "grad_norm": 1.0360960329697233,
+      "learning_rate": 6.748466257668711e-05,
+      "loss": 1.3655,
+      "step": 220
+    },
+    {
+      "epoch": 0.01018104758833556,
+      "grad_norm": 0.6642187890011997,
+      "learning_rate": 6.779141104294479e-05,
+      "loss": 1.5123,
+      "step": 221
+    },
+    {
+      "epoch": 0.010227115676970563,
+      "grad_norm": 0.6498016451143526,
+      "learning_rate": 6.809815950920245e-05,
+      "loss": 1.1195,
+      "step": 222
+    },
+    {
+      "epoch": 0.010273183765605565,
+      "grad_norm": 0.6864216241822495,
+      "learning_rate": 6.840490797546014e-05,
+      "loss": 1.2282,
+      "step": 223
+    },
+    {
+      "epoch": 0.010319251854240567,
+      "grad_norm": 0.7746512673688876,
+      "learning_rate": 6.87116564417178e-05,
+      "loss": 0.9202,
+      "step": 224
+    },
+    {
+      "epoch": 0.01036531994287557,
+      "grad_norm": 0.4579667196066817,
+      "learning_rate": 6.901840490797547e-05,
+      "loss": 1.0095,
+      "step": 225
+    },
+    {
+      "epoch": 0.010411388031510573,
+      "grad_norm": 0.6839756627026976,
+      "learning_rate": 6.932515337423313e-05,
+      "loss": 1.5126,
+      "step": 226
+    },
+    {
+      "epoch": 0.010457456120145574,
+      "grad_norm": 0.648328434961087,
+      "learning_rate": 6.963190184049079e-05,
+      "loss": 0.8991,
+      "step": 227
+    },
+    {
+      "epoch": 0.010503524208780578,
+      "grad_norm": 0.4688885988352405,
+      "learning_rate": 6.993865030674847e-05,
+      "loss": 1.3857,
+      "step": 228
+    },
+    {
+      "epoch": 0.01054959229741558,
+      "grad_norm": 0.7043025255096071,
+      "learning_rate": 7.024539877300614e-05,
+      "loss": 1.4061,
+      "step": 229
+    },
+    {
+      "epoch": 0.010595660386050582,
+      "grad_norm": 0.6926487992946778,
+      "learning_rate": 7.055214723926382e-05,
+      "loss": 1.2454,
+      "step": 230
+    },
+    {
+      "epoch": 0.010641728474685586,
+      "grad_norm": 0.5859755233399319,
+      "learning_rate": 7.085889570552148e-05,
+      "loss": 1.3148,
+      "step": 231
+    },
+    {
+      "epoch": 0.010687796563320588,
+      "grad_norm": 0.6024634592993393,
+      "learning_rate": 7.116564417177914e-05,
+      "loss": 1.6153,
+      "step": 232
+    },
+    {
+      "epoch": 0.010733864651955591,
+      "grad_norm": 0.636684697003698,
+      "learning_rate": 7.147239263803681e-05,
+      "loss": 1.2142,
+      "step": 233
+    },
+    {
+      "epoch": 0.010779932740590593,
+      "grad_norm": 0.6015863107490005,
+      "learning_rate": 7.177914110429447e-05,
+      "loss": 1.6616,
+      "step": 234
+    },
+    {
+      "epoch": 0.010826000829225595,
+      "grad_norm": 0.5780126176690528,
+      "learning_rate": 7.208588957055215e-05,
+      "loss": 1.3215,
+      "step": 235
+    },
+    {
+      "epoch": 0.010872068917860599,
+      "grad_norm": 0.5306791153943465,
+      "learning_rate": 7.239263803680982e-05,
+      "loss": 1.3127,
+      "step": 236
+    },
+    {
+      "epoch": 0.0109181370064956,
+      "grad_norm": 0.5847066332277462,
+      "learning_rate": 7.26993865030675e-05,
+      "loss": 1.2226,
+      "step": 237
+    },
+    {
+      "epoch": 0.010964205095130602,
+      "grad_norm": 0.8701752014606419,
+      "learning_rate": 7.300613496932516e-05,
+      "loss": 1.309,
+      "step": 238
+    },
+    {
+      "epoch": 0.011010273183765606,
+      "grad_norm": 0.5909261277552953,
+      "learning_rate": 7.331288343558282e-05,
+      "loss": 0.9567,
+      "step": 239
+    },
+    {
+      "epoch": 0.011056341272400608,
+      "grad_norm": 0.5829700872769413,
+      "learning_rate": 7.361963190184049e-05,
+      "loss": 1.2477,
+      "step": 240
+    },
+    {
+      "epoch": 0.01110240936103561,
+      "grad_norm": 0.5932708355554109,
+      "learning_rate": 7.392638036809815e-05,
+      "loss": 1.4423,
+      "step": 241
+    },
+    {
+      "epoch": 0.011148477449670614,
+      "grad_norm": 0.5457385729134859,
+      "learning_rate": 7.423312883435584e-05,
+      "loss": 1.5415,
+      "step": 242
+    },
+    {
+      "epoch": 0.011194545538305616,
+      "grad_norm": 0.6358064596830111,
+      "learning_rate": 7.45398773006135e-05,
+      "loss": 1.6011,
+      "step": 243
+    },
+    {
+      "epoch": 0.011240613626940617,
+      "grad_norm": 0.6803312394482565,
+      "learning_rate": 7.484662576687118e-05,
+      "loss": 1.3774,
+      "step": 244
+    },
+    {
+      "epoch": 0.011286681715575621,
+      "grad_norm": 0.4602041188685654,
+      "learning_rate": 7.515337423312884e-05,
+      "loss": 1.1624,
+      "step": 245
+    },
+    {
+      "epoch": 0.011332749804210623,
+      "grad_norm": 0.5909927219124327,
+      "learning_rate": 7.54601226993865e-05,
+      "loss": 1.3945,
+      "step": 246
+    },
+    {
+      "epoch": 0.011378817892845627,
+      "grad_norm": 0.5936925215472334,
+      "learning_rate": 7.576687116564417e-05,
+      "loss": 1.3375,
+      "step": 247
+    },
+    {
+      "epoch": 0.011424885981480629,
+      "grad_norm": 0.5663883388138602,
+      "learning_rate": 7.607361963190185e-05,
+      "loss": 1.0942,
+      "step": 248
+    },
+    {
+      "epoch": 0.01147095407011563,
+      "grad_norm": 0.6156450631958176,
+      "learning_rate": 7.638036809815952e-05,
+      "loss": 1.7314,
+      "step": 249
+    },
+    {
+      "epoch": 0.011517022158750634,
+      "grad_norm": 0.6545305401204524,
+      "learning_rate": 7.668711656441718e-05,
+      "loss": 1.1574,
+      "step": 250
+    },
+    {
+      "epoch": 0.011563090247385636,
+      "grad_norm": 0.5790910323139054,
+      "learning_rate": 7.699386503067484e-05,
+      "loss": 1.5236,
+      "step": 251
+    },
+    {
+      "epoch": 0.011609158336020638,
+      "grad_norm": 0.5201603570768956,
+      "learning_rate": 7.730061349693252e-05,
+      "loss": 1.1751,
+      "step": 252
+    },
+    {
+      "epoch": 0.011655226424655642,
+      "grad_norm": 0.5221277517724371,
+      "learning_rate": 7.760736196319018e-05,
+      "loss": 1.51,
+      "step": 253
+    },
+    {
+      "epoch": 0.011701294513290644,
+      "grad_norm": 0.5769678151541204,
+      "learning_rate": 7.791411042944787e-05,
+      "loss": 1.4607,
+      "step": 254
+    },
+    {
+      "epoch": 0.011747362601925646,
+      "grad_norm": 0.613711356248652,
+      "learning_rate": 7.822085889570553e-05,
+      "loss": 1.2504,
+      "step": 255
+    },
+    {
+      "epoch": 0.01179343069056065,
+      "grad_norm": 0.7134329385694058,
+      "learning_rate": 7.85276073619632e-05,
+      "loss": 1.2221,
+      "step": 256
+    },
+    {
+      "epoch": 0.011839498779195651,
+      "grad_norm": 0.7383556063925443,
+      "learning_rate": 7.883435582822086e-05,
+      "loss": 1.6076,
+      "step": 257
+    },
+    {
+      "epoch": 0.011885566867830653,
+      "grad_norm": 0.5439017887864668,
+      "learning_rate": 7.914110429447852e-05,
+      "loss": 1.3944,
+      "step": 258
+    },
+    {
+      "epoch": 0.011931634956465657,
+      "grad_norm": 0.6203280999102432,
+      "learning_rate": 7.94478527607362e-05,
+      "loss": 1.3158,
+      "step": 259
+    },
+    {
+      "epoch": 0.011977703045100659,
+      "grad_norm": 0.6174957557793387,
+      "learning_rate": 7.975460122699386e-05,
+      "loss": 1.5428,
+      "step": 260
+    },
+    {
+      "epoch": 0.01202377113373566,
+      "grad_norm": 0.6209270748965676,
+      "learning_rate": 8.006134969325155e-05,
+      "loss": 1.2603,
+      "step": 261
+    },
+    {
+      "epoch": 0.012069839222370664,
+      "grad_norm": 0.533457165818016,
+      "learning_rate": 8.036809815950921e-05,
+      "loss": 1.5201,
+      "step": 262
+    },
+    {
+      "epoch": 0.012115907311005666,
+      "grad_norm": 0.4370650280529949,
+      "learning_rate": 8.067484662576688e-05,
+      "loss": 1.1046,
+      "step": 263
+    },
+    {
+      "epoch": 0.01216197539964067,
+      "grad_norm": 0.5169462821702575,
+      "learning_rate": 8.098159509202454e-05,
+      "loss": 0.9754,
+      "step": 264
+    },
+    {
+      "epoch": 0.012208043488275672,
+      "grad_norm": 0.5394153146945244,
+      "learning_rate": 8.12883435582822e-05,
+      "loss": 1.7742,
+      "step": 265
+    },
+    {
+      "epoch": 0.012254111576910674,
+      "grad_norm": 0.49408064195028667,
+      "learning_rate": 8.159509202453988e-05,
+      "loss": 1.4332,
+      "step": 266
+    },
+    {
+      "epoch": 0.012300179665545677,
+      "grad_norm": 0.5854556737975919,
+      "learning_rate": 8.190184049079755e-05,
+      "loss": 1.1139,
+      "step": 267
+    },
+    {
+      "epoch": 0.01234624775418068,
+      "grad_norm": 0.6123126904661172,
+      "learning_rate": 8.220858895705523e-05,
+      "loss": 1.2235,
+      "step": 268
+    },
+    {
+      "epoch": 0.012392315842815681,
+      "grad_norm": 0.5649851470982081,
+      "learning_rate": 8.251533742331289e-05,
+      "loss": 1.2704,
+      "step": 269
+    },
+    {
+      "epoch": 0.012438383931450685,
+      "grad_norm": 0.5910242557199674,
+      "learning_rate": 8.282208588957055e-05,
+      "loss": 1.4647,
+      "step": 270
+    },
+    {
+      "epoch": 0.012484452020085687,
+      "grad_norm": 0.6954528546276917,
+      "learning_rate": 8.312883435582822e-05,
+      "loss": 1.1476,
+      "step": 271
+    },
+    {
+      "epoch": 0.012530520108720689,
+      "grad_norm": 0.6418600108132034,
+      "learning_rate": 8.343558282208588e-05,
+      "loss": 1.3947,
+      "step": 272
+    },
+    {
+      "epoch": 0.012576588197355692,
+      "grad_norm": 0.7540627710299023,
+      "learning_rate": 8.374233128834357e-05,
+      "loss": 1.1543,
+      "step": 273
+    },
+    {
+      "epoch": 0.012622656285990694,
+      "grad_norm": 0.5150044429334307,
+      "learning_rate": 8.404907975460123e-05,
+      "loss": 1.2328,
+      "step": 274
+    },
+    {
+      "epoch": 0.012668724374625696,
+      "grad_norm": 0.6489896366373333,
+      "learning_rate": 8.435582822085891e-05,
+      "loss": 1.1264,
+      "step": 275
+    },
+    {
+      "epoch": 0.0127147924632607,
+      "grad_norm": 0.7873113257186471,
+      "learning_rate": 8.466257668711657e-05,
+      "loss": 1.4575,
+      "step": 276
+    },
+    {
+      "epoch": 0.012760860551895702,
+      "grad_norm": 0.5366903974845632,
+      "learning_rate": 8.496932515337423e-05,
+      "loss": 1.2169,
+      "step": 277
+    },
+    {
+      "epoch": 0.012806928640530704,
+      "grad_norm": 0.6490574507337623,
+      "learning_rate": 8.52760736196319e-05,
+      "loss": 1.3718,
+      "step": 278
+    },
+    {
+      "epoch": 0.012852996729165707,
+      "grad_norm": 0.559637599173395,
+      "learning_rate": 8.558282208588958e-05,
+      "loss": 1.365,
+      "step": 279
+    },
+    {
+      "epoch": 0.012899064817800709,
+      "grad_norm": 0.5923791488011673,
+      "learning_rate": 8.588957055214725e-05,
+      "loss": 1.4587,
+      "step": 280
+    },
+    {
+      "epoch": 0.012945132906435713,
+      "grad_norm": 0.668471671972123,
+      "learning_rate": 8.619631901840491e-05,
+      "loss": 1.315,
+      "step": 281
+    },
+    {
+      "epoch": 0.012991200995070715,
+      "grad_norm": 0.5190948005330194,
+      "learning_rate": 8.650306748466259e-05,
+      "loss": 1.2144,
+      "step": 282
+    },
+    {
+      "epoch": 0.013037269083705717,
+      "grad_norm": 0.5808879485117334,
+      "learning_rate": 8.680981595092025e-05,
+      "loss": 1.4827,
+      "step": 283
+    },
+    {
+      "epoch": 0.01308333717234072,
+      "grad_norm": 0.6110226586332892,
+      "learning_rate": 8.711656441717791e-05,
+      "loss": 1.2542,
+      "step": 284
+    },
+    {
+      "epoch": 0.013129405260975722,
+      "grad_norm": 0.47057215176619094,
+      "learning_rate": 8.742331288343558e-05,
+      "loss": 1.3362,
+      "step": 285
+    },
+    {
+      "epoch": 0.013175473349610724,
+      "grad_norm": 0.5960677186698032,
+      "learning_rate": 8.773006134969326e-05,
+      "loss": 1.1163,
+      "step": 286
+    },
+    {
+      "epoch": 0.013221541438245728,
+      "grad_norm": 0.695472806337987,
+      "learning_rate": 8.803680981595093e-05,
+      "loss": 1.4038,
+      "step": 287
+    },
+    {
+      "epoch": 0.01326760952688073,
+      "grad_norm": 0.429777989389087,
+      "learning_rate": 8.83435582822086e-05,
+      "loss": 1.0076,
+      "step": 288
+    },
+    {
+      "epoch": 0.013313677615515732,
+      "grad_norm": 0.5429390170419099,
+      "learning_rate": 8.865030674846625e-05,
+      "loss": 1.1622,
+      "step": 289
+    },
+    {
+      "epoch": 0.013359745704150735,
+      "grad_norm": 0.5315874765207482,
+      "learning_rate": 8.895705521472393e-05,
+      "loss": 1.4387,
+      "step": 290
+    },
+    {
+      "epoch": 0.013405813792785737,
+      "grad_norm": 0.5827858688314658,
+      "learning_rate": 8.926380368098159e-05,
+      "loss": 1.5396,
+      "step": 291
+    },
+    {
+      "epoch": 0.013451881881420739,
+      "grad_norm": 0.5899143297447491,
+      "learning_rate": 8.957055214723928e-05,
+      "loss": 1.2019,
+      "step": 292
+    },
+    {
+      "epoch": 0.013497949970055743,
+      "grad_norm": 0.43260205605471685,
+      "learning_rate": 8.987730061349694e-05,
+      "loss": 1.3047,
+      "step": 293
+    },
+    {
+      "epoch": 0.013544018058690745,
+      "grad_norm": 0.5038076337046336,
+      "learning_rate": 9.018404907975461e-05,
+      "loss": 1.3832,
+      "step": 294
+    },
+    {
+      "epoch": 0.013590086147325748,
+      "grad_norm": 0.680672078990959,
+      "learning_rate": 9.049079754601227e-05,
+      "loss": 1.278,
+      "step": 295
+    },
+    {
+      "epoch": 0.01363615423596075,
+      "grad_norm": 0.5602337158662284,
+      "learning_rate": 9.079754601226993e-05,
+      "loss": 1.3997,
+      "step": 296
+    },
+    {
+      "epoch": 0.013682222324595752,
+      "grad_norm": 0.4705083627802981,
+      "learning_rate": 9.110429447852761e-05,
+      "loss": 1.0813,
+      "step": 297
+    },
+    {
+      "epoch": 0.013728290413230756,
+      "grad_norm": 0.5847259908665945,
+      "learning_rate": 9.141104294478528e-05,
+      "loss": 1.4174,
+      "step": 298
+    },
+    {
+      "epoch": 0.013774358501865758,
+      "grad_norm": 0.7015644568000361,
+      "learning_rate": 9.171779141104296e-05,
+      "loss": 0.9638,
+      "step": 299
+    },
+    {
+      "epoch": 0.01382042659050076,
+      "grad_norm": 0.5825827779495659,
+      "learning_rate": 9.202453987730062e-05,
+      "loss": 1.3652,
+      "step": 300
+    },
+    {
+      "epoch": 0.013866494679135763,
+      "grad_norm": 0.46200521419656726,
+      "learning_rate": 9.233128834355828e-05,
+      "loss": 1.6788,
+      "step": 301
+    },
+    {
+      "epoch": 0.013912562767770765,
+      "grad_norm": 0.6028133702463717,
+      "learning_rate": 9.263803680981595e-05,
+      "loss": 1.3199,
+      "step": 302
+    },
+    {
+      "epoch": 0.013958630856405767,
+      "grad_norm": 0.5754695503034892,
+      "learning_rate": 9.294478527607362e-05,
+      "loss": 1.0535,
+      "step": 303
+    },
+    {
+      "epoch": 0.01400469894504077,
+      "grad_norm": 0.7628413406207298,
+      "learning_rate": 9.325153374233129e-05,
+      "loss": 1.0107,
+      "step": 304
+    },
+    {
+      "epoch": 0.014050767033675773,
+      "grad_norm": 0.4529811390176622,
+      "learning_rate": 9.355828220858896e-05,
+      "loss": 1.3591,
+      "step": 305
+    },
+    {
+      "epoch": 0.014096835122310775,
+      "grad_norm": 0.4524530247152597,
+      "learning_rate": 9.386503067484664e-05,
+      "loss": 1.4876,
+      "step": 306
+    },
+    {
+      "epoch": 0.014142903210945778,
+      "grad_norm": 0.7330207177470147,
+      "learning_rate": 9.41717791411043e-05,
+      "loss": 1.2253,
+      "step": 307
+    },
+    {
+      "epoch": 0.01418897129958078,
+      "grad_norm": 0.48780043377762267,
+      "learning_rate": 9.447852760736196e-05,
+      "loss": 1.3369,
+      "step": 308
+    },
+    {
+      "epoch": 0.014235039388215782,
+      "grad_norm": 0.5909507318239964,
+      "learning_rate": 9.478527607361963e-05,
+      "loss": 1.1909,
+      "step": 309
+    },
+    {
+      "epoch": 0.014281107476850786,
+      "grad_norm": 0.4421798844281286,
+      "learning_rate": 9.50920245398773e-05,
+      "loss": 1.1283,
+      "step": 310
+    },
+    {
+      "epoch": 0.014327175565485788,
+      "grad_norm": 0.649949961550812,
+      "learning_rate": 9.539877300613498e-05,
+      "loss": 1.3377,
+      "step": 311
+    },
+    {
+      "epoch": 0.014373243654120791,
+      "grad_norm": 0.6241980145039505,
+      "learning_rate": 9.570552147239264e-05,
+      "loss": 1.3759,
+      "step": 312
+    },
+    {
+      "epoch": 0.014419311742755793,
+      "grad_norm": 0.5276468509723975,
+      "learning_rate": 9.601226993865032e-05,
+      "loss": 1.3023,
+      "step": 313
+    },
+    {
+      "epoch": 0.014465379831390795,
+      "grad_norm": 0.5876136853700037,
+      "learning_rate": 9.631901840490798e-05,
+      "loss": 1.1025,
+      "step": 314
+    },
+    {
+      "epoch": 0.014511447920025799,
+      "grad_norm": 0.543714279727196,
+      "learning_rate": 9.662576687116564e-05,
+      "loss": 1.2811,
+      "step": 315
+    },
+    {
+      "epoch": 0.0145575160086608,
+      "grad_norm": 0.6452513754624033,
+      "learning_rate": 9.693251533742331e-05,
+      "loss": 1.4672,
+      "step": 316
+    },
+    {
+      "epoch": 0.014603584097295803,
+      "grad_norm": 0.5687897371262159,
+      "learning_rate": 9.723926380368099e-05,
+      "loss": 1.312,
+      "step": 317
+    },
+    {
+      "epoch": 0.014649652185930806,
+      "grad_norm": 0.732969751335369,
+      "learning_rate": 9.754601226993866e-05,
+      "loss": 1.1448,
+      "step": 318
+    },
+    {
+      "epoch": 0.014695720274565808,
+      "grad_norm": 0.5009974483490998,
+      "learning_rate": 9.785276073619632e-05,
+      "loss": 1.3573,
+      "step": 319
+    },
+    {
+      "epoch": 0.01474178836320081,
+      "grad_norm": 0.5954441320032897,
+      "learning_rate": 9.815950920245399e-05,
+      "loss": 1.3573,
+      "step": 320
+    },
+    {
+      "epoch": 0.014787856451835814,
+      "grad_norm": 0.6437075230046724,
+      "learning_rate": 9.846625766871166e-05,
+      "loss": 1.3801,
+      "step": 321
+    },
+    {
+      "epoch": 0.014833924540470816,
+      "grad_norm": 0.6151099461767111,
+      "learning_rate": 9.877300613496932e-05,
+      "loss": 1.3791,
+      "step": 322
+    },
+    {
+      "epoch": 0.014879992629105818,
+      "grad_norm": 0.44715121518971396,
+      "learning_rate": 9.907975460122701e-05,
+      "loss": 1.4754,
+      "step": 323
+    },
+    {
+      "epoch": 0.014926060717740821,
+      "grad_norm": 0.4520315941141799,
+      "learning_rate": 9.938650306748467e-05,
+      "loss": 1.1266,
+      "step": 324
+    },
+    {
+      "epoch": 0.014972128806375823,
+      "grad_norm": 0.6722064141618138,
+      "learning_rate": 9.969325153374234e-05,
+      "loss": 1.0059,
+      "step": 325
+    },
+    {
+      "epoch": 0.015018196895010825,
+      "grad_norm": 0.4513495991388799,
+      "learning_rate": 0.0001,
+      "loss": 1.3664,
+      "step": 326
+    },
+    {
+      "epoch": 0.015064264983645829,
+      "grad_norm": 0.5043898225815876,
+      "learning_rate": 0.00010030674846625767,
+      "loss": 1.3885,
+      "step": 327
+    },
+    {
+      "epoch": 0.01511033307228083,
+      "grad_norm": 0.6486835267511067,
+      "learning_rate": 0.00010061349693251534,
+      "loss": 1.6091,
+      "step": 328
+    },
+    {
+      "epoch": 0.015156401160915834,
+      "grad_norm": 0.5428130042749303,
+      "learning_rate": 0.000100920245398773,
+      "loss": 1.4037,
+      "step": 329
+    },
+    {
+      "epoch": 0.015202469249550836,
+      "grad_norm": 0.6627055418456327,
+      "learning_rate": 0.00010122699386503068,
+      "loss": 1.2482,
+      "step": 330
+    },
+    {
+      "epoch": 0.015248537338185838,
+      "grad_norm": 0.4536867501709119,
+      "learning_rate": 0.00010153374233128834,
+      "loss": 1.1102,
+      "step": 331
+    },
+    {
+      "epoch": 0.015294605426820842,
+      "grad_norm": 0.5079176189534133,
+      "learning_rate": 0.00010184049079754601,
+      "loss": 1.2208,
+      "step": 332
+    },
+    {
+      "epoch": 0.015340673515455844,
+      "grad_norm": 0.7701185935841574,
+      "learning_rate": 0.0001021472392638037,
+      "loss": 1.2086,
+      "step": 333
+    },
+    {
+      "epoch": 0.015386741604090846,
+      "grad_norm": 0.6473275439799034,
+      "learning_rate": 0.00010245398773006136,
+      "loss": 1.2806,
+      "step": 334
+    },
+    {
+      "epoch": 0.01543280969272585,
+      "grad_norm": 0.5897089240781636,
+      "learning_rate": 0.00010276073619631903,
+      "loss": 1.2321,
+      "step": 335
+    },
+    {
+      "epoch": 0.015478877781360851,
+      "grad_norm": 0.6195972034201,
+      "learning_rate": 0.0001030674846625767,
+      "loss": 1.5044,
+      "step": 336
+    },
+    {
+      "epoch": 0.015524945869995853,
+      "grad_norm": 0.5419122714667878,
+      "learning_rate": 0.00010337423312883437,
+      "loss": 1.5276,
+      "step": 337
+    },
+    {
+      "epoch": 0.015571013958630857,
+      "grad_norm": 0.5620927675906606,
+      "learning_rate": 0.00010368098159509203,
+      "loss": 1.6092,
+      "step": 338
+    },
+    {
+      "epoch": 0.015617082047265859,
+      "grad_norm": 0.5888368079284326,
+      "learning_rate": 0.00010398773006134969,
+      "loss": 1.3157,
+      "step": 339
+    },
+    {
+      "epoch": 0.01566315013590086,
+      "grad_norm": 0.6287817348643474,
+      "learning_rate": 0.00010429447852760737,
+      "loss": 1.3603,
+      "step": 340
+    },
+    {
+      "epoch": 0.015709218224535863,
+      "grad_norm": 0.47127242093516164,
+      "learning_rate": 0.00010460122699386503,
+      "loss": 1.3268,
+      "step": 341
+    },
+    {
+      "epoch": 0.015755286313170868,
+      "grad_norm": 0.4921019984264473,
+      "learning_rate": 0.0001049079754601227,
+      "loss": 1.2629,
+      "step": 342
+    },
+    {
+      "epoch": 0.01580135440180587,
+      "grad_norm": 0.5356929573955265,
+      "learning_rate": 0.00010521472392638036,
+      "loss": 1.0794,
+      "step": 343
+    },
+    {
+      "epoch": 0.015847422490440872,
+      "grad_norm": 0.49901021037435656,
+      "learning_rate": 0.00010552147239263804,
+      "loss": 1.4221,
+      "step": 344
+    },
+    {
+      "epoch": 0.015893490579075874,
+      "grad_norm": 0.614707074709489,
+      "learning_rate": 0.00010582822085889572,
+      "loss": 1.4859,
+      "step": 345
+    },
+    {
+      "epoch": 0.015939558667710876,
+      "grad_norm": 0.5698164846123067,
+      "learning_rate": 0.00010613496932515338,
+      "loss": 1.2545,
+      "step": 346
+    },
+    {
+      "epoch": 0.015985626756345878,
+      "grad_norm": 0.5078890546703527,
+      "learning_rate": 0.00010644171779141106,
+      "loss": 1.3656,
+      "step": 347
+    },
+    {
+      "epoch": 0.016031694844980883,
+      "grad_norm": 0.7428269518192409,
+      "learning_rate": 0.00010674846625766872,
+      "loss": 1.3282,
+      "step": 348
+    },
+    {
+      "epoch": 0.016077762933615885,
+      "grad_norm": 0.623891017677552,
+      "learning_rate": 0.0001070552147239264,
+      "loss": 1.0652,
+      "step": 349
+    },
+    {
+      "epoch": 0.016123831022250887,
+      "grad_norm": 0.5940134415227926,
+      "learning_rate": 0.00010736196319018406,
+      "loss": 1.5068,
+      "step": 350
+    },
+    {
+      "epoch": 0.01616989911088589,
+      "grad_norm": 0.565941913185679,
+      "learning_rate": 0.00010766871165644173,
+      "loss": 1.3994,
+      "step": 351
+    },
+    {
+      "epoch": 0.01621596719952089,
+      "grad_norm": 0.546924406924968,
+      "learning_rate": 0.00010797546012269939,
+      "loss": 1.1376,
+      "step": 352
+    },
+    {
+      "epoch": 0.016262035288155896,
+      "grad_norm": 0.5931671778043633,
+      "learning_rate": 0.00010828220858895705,
+      "loss": 0.9807,
+      "step": 353
+    },
+    {
+      "epoch": 0.016308103376790898,
+      "grad_norm": 0.5655692349112963,
+      "learning_rate": 0.00010858895705521473,
+      "loss": 1.3018,
+      "step": 354
+    },
+    {
+      "epoch": 0.0163541714654259,
+      "grad_norm": 0.6698653506380735,
+      "learning_rate": 0.00010889570552147239,
+      "loss": 1.2295,
+      "step": 355
+    },
+    {
+      "epoch": 0.016400239554060902,
+      "grad_norm": 0.5237649369516968,
+      "learning_rate": 0.00010920245398773006,
+      "loss": 1.2873,
+      "step": 356
+    },
+    {
+      "epoch": 0.016446307642695904,
+      "grad_norm": 0.5528699443015928,
+      "learning_rate": 0.00010950920245398772,
+      "loss": 1.2285,
+      "step": 357
+    },
+    {
+      "epoch": 0.016492375731330906,
+      "grad_norm": 0.5355446230105199,
+      "learning_rate": 0.00010981595092024541,
+      "loss": 1.1282,
+      "step": 358
+    },
+    {
+      "epoch": 0.01653844381996591,
+      "grad_norm": 0.594117868446221,
+      "learning_rate": 0.00011012269938650308,
+      "loss": 1.0866,
+      "step": 359
+    },
+    {
+      "epoch": 0.016584511908600913,
+      "grad_norm": 0.46102039397411826,
+      "learning_rate": 0.00011042944785276075,
+      "loss": 1.449,
+      "step": 360
+    },
+    {
+      "epoch": 0.016630579997235915,
+      "grad_norm": 0.46725550590683546,
+      "learning_rate": 0.00011073619631901842,
+      "loss": 1.4541,
+      "step": 361
+    },
+    {
+      "epoch": 0.016676648085870917,
+      "grad_norm": 0.5897483588291145,
+      "learning_rate": 0.00011104294478527608,
+      "loss": 1.507,
+      "step": 362
+    },
+    {
+      "epoch": 0.01672271617450592,
+      "grad_norm": 0.5339712366684339,
+      "learning_rate": 0.00011134969325153376,
+      "loss": 1.2442,
+      "step": 363
+    },
+    {
+      "epoch": 0.01676878426314092,
+      "grad_norm": 0.5621555137679937,
+      "learning_rate": 0.00011165644171779142,
+      "loss": 1.283,
+      "step": 364
+    },
+    {
+      "epoch": 0.016814852351775926,
+      "grad_norm": 0.4977881677905226,
+      "learning_rate": 0.00011196319018404908,
+      "loss": 1.3585,
+      "step": 365
+    },
+    {
+      "epoch": 0.016860920440410928,
+      "grad_norm": 0.47930133908235323,
+      "learning_rate": 0.00011226993865030675,
+      "loss": 0.8179,
+      "step": 366
+    },
+    {
+      "epoch": 0.01690698852904593,
+      "grad_norm": 0.5937708880675047,
+      "learning_rate": 0.00011257668711656441,
+      "loss": 0.9537,
+      "step": 367
+    },
+    {
+      "epoch": 0.016953056617680932,
+      "grad_norm": 0.5648788825647348,
+      "learning_rate": 0.00011288343558282209,
+      "loss": 1.189,
+      "step": 368
+    },
+    {
+      "epoch": 0.016999124706315934,
+      "grad_norm": 0.48544741015458037,
+      "learning_rate": 0.00011319018404907975,
+      "loss": 1.3594,
+      "step": 369
+    },
+    {
+      "epoch": 0.01704519279495094,
+      "grad_norm": 0.4688402046973789,
+      "learning_rate": 0.00011349693251533744,
+      "loss": 1.1365,
+      "step": 370
+    },
+    {
+      "epoch": 0.01709126088358594,
+      "grad_norm": 0.5592857507992868,
+      "learning_rate": 0.00011380368098159511,
+      "loss": 1.3284,
+      "step": 371
+    },
+    {
+      "epoch": 0.017137328972220943,
+      "grad_norm": 0.541079124437854,
+      "learning_rate": 0.00011411042944785277,
+      "loss": 1.1447,
+      "step": 372
+    },
+    {
+      "epoch": 0.017183397060855945,
+      "grad_norm": 0.5081237606713878,
+      "learning_rate": 0.00011441717791411045,
+      "loss": 1.3097,
+      "step": 373
+    },
+    {
+      "epoch": 0.017229465149490947,
+      "grad_norm": 0.4955998085907892,
+      "learning_rate": 0.0001147239263803681,
+      "loss": 1.0736,
+      "step": 374
+    },
+    {
+      "epoch": 0.01727553323812595,
+      "grad_norm": 0.48297245444932946,
+      "learning_rate": 0.00011503067484662578,
+      "loss": 1.3304,
+      "step": 375
+    },
+    {
+      "epoch": 0.017321601326760954,
+      "grad_norm": 0.4883700706612392,
+      "learning_rate": 0.00011533742331288344,
+      "loss": 1.1919,
+      "step": 376
+    },
+    {
+      "epoch": 0.017367669415395956,
+      "grad_norm": 0.5468840815999075,
+      "learning_rate": 0.0001156441717791411,
+      "loss": 1.5316,
+      "step": 377
+    },
+    {
+      "epoch": 0.017413737504030958,
+      "grad_norm": 0.455661906258555,
+      "learning_rate": 0.00011595092024539878,
+      "loss": 1.4039,
+      "step": 378
+    },
+    {
+      "epoch": 0.01745980559266596,
+      "grad_norm": 0.49892467793023865,
+      "learning_rate": 0.00011625766871165644,
+      "loss": 1.6184,
+      "step": 379
+    },
+    {
+      "epoch": 0.017505873681300962,
+      "grad_norm": 0.47762331519746554,
+      "learning_rate": 0.00011656441717791411,
+      "loss": 1.415,
+      "step": 380
+    },
+    {
+      "epoch": 0.017551941769935964,
+      "grad_norm": 0.4808004704252827,
+      "learning_rate": 0.00011687116564417177,
+      "loss": 1.2852,
+      "step": 381
+    },
+    {
+      "epoch": 0.01759800985857097,
+      "grad_norm": 0.5942019866724758,
+      "learning_rate": 0.00011717791411042945,
+      "loss": 1.473,
+      "step": 382
+    },
+    {
+      "epoch": 0.01764407794720597,
+      "grad_norm": 0.5064336593788419,
+      "learning_rate": 0.00011748466257668714,
+      "loss": 1.6749,
+      "step": 383
+    },
+    {
+      "epoch": 0.017690146035840973,
+      "grad_norm": 0.4292276823828754,
+      "learning_rate": 0.0001177914110429448,
+      "loss": 0.9927,
+      "step": 384
+    },
+    {
+      "epoch": 0.017736214124475975,
+      "grad_norm": 0.406139969946228,
+      "learning_rate": 0.00011809815950920247,
+      "loss": 1.3483,
+      "step": 385
+    },
+    {
+      "epoch": 0.017782282213110977,
+      "grad_norm": 0.47883536021591844,
+      "learning_rate": 0.00011840490797546013,
+      "loss": 1.4084,
+      "step": 386
+    },
+    {
+      "epoch": 0.017828350301745982,
+      "grad_norm": 0.6907788141128172,
+      "learning_rate": 0.0001187116564417178,
+      "loss": 0.8245,
+      "step": 387
+    },
+    {
+      "epoch": 0.017874418390380984,
+      "grad_norm": 0.5001291659529251,
+      "learning_rate": 0.00011901840490797547,
+      "loss": 1.4046,
+      "step": 388
+    },
+    {
+      "epoch": 0.017920486479015986,
+      "grad_norm": 0.5340480977727978,
+      "learning_rate": 0.00011932515337423313,
+      "loss": 0.8574,
+      "step": 389
+    },
+    {
+      "epoch": 0.017966554567650988,
+      "grad_norm": 0.5408916919641461,
+      "learning_rate": 0.0001196319018404908,
+      "loss": 1.7748,
+      "step": 390
+    },
+    {
+      "epoch": 0.01801262265628599,
+      "grad_norm": 0.7774073055737248,
+      "learning_rate": 0.00011993865030674846,
+      "loss": 1.3585,
+      "step": 391
+    },
+    {
+      "epoch": 0.01805869074492099,
+      "grad_norm": 0.49291174503657,
+      "learning_rate": 0.00012024539877300614,
+      "loss": 0.9907,
+      "step": 392
+    },
+    {
+      "epoch": 0.018104758833555997,
+      "grad_norm": 0.560353417488851,
+      "learning_rate": 0.0001205521472392638,
+      "loss": 0.9073,
+      "step": 393
+    },
+    {
+      "epoch": 0.018150826922191,
+      "grad_norm": 0.5095942391192007,
+      "learning_rate": 0.00012085889570552147,
+      "loss": 1.4139,
+      "step": 394
+    },
+    {
+      "epoch": 0.018196895010826,
+      "grad_norm": 0.47260911543786116,
+      "learning_rate": 0.00012116564417177916,
+      "loss": 1.2201,
+      "step": 395
+    },
+    {
+      "epoch": 0.018242963099461003,
+      "grad_norm": 0.498160374592355,
+      "learning_rate": 0.00012147239263803682,
+      "loss": 1.4466,
+      "step": 396
+    },
+    {
+      "epoch": 0.018289031188096005,
+      "grad_norm": 0.5532877122990209,
+      "learning_rate": 0.0001217791411042945,
+      "loss": 1.5732,
+      "step": 397
+    },
+    {
+      "epoch": 0.018335099276731007,
+      "grad_norm": 0.47972844263597103,
+      "learning_rate": 0.00012208588957055217,
+      "loss": 1.4676,
+      "step": 398
+    },
+    {
+      "epoch": 0.018381167365366012,
+      "grad_norm": 0.647455162188498,
+      "learning_rate": 0.00012239263803680983,
+      "loss": 1.1728,
+      "step": 399
+    },
+    {
+      "epoch": 0.018427235454001014,
+      "grad_norm": 0.5660137474425724,
+      "learning_rate": 0.0001226993865030675,
+      "loss": 0.8883,
+      "step": 400
+    },
+    {
+      "epoch": 0.018473303542636016,
+      "grad_norm": 0.523119840906882,
+      "learning_rate": 0.00012300613496932515,
+      "loss": 1.1676,
+      "step": 401
+    },
+    {
+      "epoch": 0.018519371631271018,
+      "grad_norm": 0.5432445199996667,
+      "learning_rate": 0.00012331288343558281,
+      "loss": 1.1866,
+      "step": 402
+    },
+    {
+      "epoch": 0.01856543971990602,
+      "grad_norm": 0.5469665693297613,
+      "learning_rate": 0.0001236196319018405,
+      "loss": 1.5376,
+      "step": 403
+    },
+    {
+      "epoch": 0.018611507808541025,
+      "grad_norm": 0.5730325478976448,
+      "learning_rate": 0.00012392638036809816,
+      "loss": 1.6227,
+      "step": 404
+    },
+    {
+      "epoch": 0.018657575897176027,
+      "grad_norm": 0.5788806879671275,
+      "learning_rate": 0.00012423312883435582,
+      "loss": 1.2171,
+      "step": 405
+    },
+    {
+      "epoch": 0.01870364398581103,
+      "grad_norm": 0.5501419042883199,
+      "learning_rate": 0.00012453987730061348,
+      "loss": 1.1212,
+      "step": 406
+    },
+    {
+      "epoch": 0.01874971207444603,
+      "grad_norm": 0.5478438752982194,
+      "learning_rate": 0.00012484662576687117,
+      "loss": 0.897,
+      "step": 407
+    },
+    {
+      "epoch": 0.018795780163081033,
+      "grad_norm": 0.8616673315700679,
+      "learning_rate": 0.00012515337423312886,
+      "loss": 1.103,
+      "step": 408
+    },
+    {
+      "epoch": 0.018841848251716035,
+      "grad_norm": 0.6741709117064105,
+      "learning_rate": 0.00012546012269938652,
+      "loss": 1.6527,
+      "step": 409
+    },
+    {
+      "epoch": 0.01888791634035104,
+      "grad_norm": 0.7858041187764832,
+      "learning_rate": 0.00012576687116564418,
+      "loss": 1.5097,
+      "step": 410
+    },
+    {
+      "epoch": 0.018933984428986042,
+      "grad_norm": 0.5265466114050085,
+      "learning_rate": 0.00012607361963190184,
+      "loss": 1.5183,
+      "step": 411
+    },
+    {
+      "epoch": 0.018980052517621044,
+      "grad_norm": 0.8316212973954304,
+      "learning_rate": 0.00012638036809815953,
+      "loss": 1.4251,
+      "step": 412
+    },
+    {
+      "epoch": 0.019026120606256046,
+      "grad_norm": 0.5396255593854093,
+      "learning_rate": 0.0001266871165644172,
+      "loss": 1.4021,
+      "step": 413
+    },
+    {
+      "epoch": 0.019072188694891048,
+      "grad_norm": 0.6029592373912335,
+      "learning_rate": 0.00012699386503067485,
+      "loss": 1.1602,
+      "step": 414
+    },
+    {
+      "epoch": 0.019118256783526053,
+      "grad_norm": 0.7846103860661221,
+      "learning_rate": 0.0001273006134969325,
+      "loss": 1.6786,
+      "step": 415
+    },
+    {
+      "epoch": 0.019164324872161055,
+      "grad_norm": 0.5893195707858045,
+      "learning_rate": 0.00012760736196319017,
+      "loss": 1.2002,
+      "step": 416
+    },
+    {
+      "epoch": 0.019210392960796057,
+      "grad_norm": 0.5898362831810929,
+      "learning_rate": 0.00012791411042944786,
+      "loss": 1.2307,
+      "step": 417
+    },
+    {
+      "epoch": 0.01925646104943106,
+      "grad_norm": 0.6212892270121383,
+      "learning_rate": 0.00012822085889570552,
+      "loss": 1.4081,
+      "step": 418
+    },
+    {
+      "epoch": 0.01930252913806606,
+      "grad_norm": 0.6048611834282475,
+      "learning_rate": 0.00012852760736196318,
+      "loss": 1.2409,
+      "step": 419
+    },
+    {
+      "epoch": 0.019348597226701063,
+      "grad_norm": 0.5423269173954294,
+      "learning_rate": 0.00012883435582822084,
+      "loss": 1.2113,
+      "step": 420
+    },
+    {
+      "epoch": 0.019394665315336068,
+      "grad_norm": 0.551796545235035,
+      "learning_rate": 0.00012914110429447853,
+      "loss": 1.4677,
+      "step": 421
+    },
+    {
+      "epoch": 0.01944073340397107,
+      "grad_norm": 0.6627148743036899,
+      "learning_rate": 0.00012944785276073622,
+      "loss": 1.2918,
+      "step": 422
+    },
+    {
+      "epoch": 0.019486801492606072,
+      "grad_norm": 0.6587350630650232,
+      "learning_rate": 0.00012975460122699388,
+      "loss": 0.9695,
+      "step": 423
+    },
+    {
+      "epoch": 0.019532869581241074,
+      "grad_norm": 0.7378962524388272,
+      "learning_rate": 0.00013006134969325154,
+      "loss": 1.3801,
+      "step": 424
+    },
+    {
+      "epoch": 0.019578937669876076,
+      "grad_norm": 0.5144017717375662,
+      "learning_rate": 0.0001303680981595092,
+      "loss": 1.0715,
+      "step": 425
+    },
+    {
+      "epoch": 0.019625005758511078,
+      "grad_norm": 0.49960006312037947,
+      "learning_rate": 0.0001306748466257669,
+      "loss": 1.3238,
+      "step": 426
+    },
+    {
+      "epoch": 0.019671073847146083,
+      "grad_norm": 0.49837523222795777,
+      "learning_rate": 0.00013098159509202455,
+      "loss": 1.3027,
+      "step": 427
+    },
+    {
+      "epoch": 0.019717141935781085,
+      "grad_norm": 0.5883905165685954,
+      "learning_rate": 0.0001312883435582822,
+      "loss": 1.0487,
+      "step": 428
+    },
+    {
+      "epoch": 0.019763210024416087,
+      "grad_norm": 0.5568973429252154,
+      "learning_rate": 0.00013159509202453987,
+      "loss": 1.3528,
+      "step": 429
+    },
+    {
+      "epoch": 0.01980927811305109,
+      "grad_norm": 0.586363645151154,
+      "learning_rate": 0.00013190184049079753,
+      "loss": 1.1727,
+      "step": 430
+    },
+    {
+      "epoch": 0.01985534620168609,
+      "grad_norm": 0.5622799567581153,
+      "learning_rate": 0.00013220858895705522,
+      "loss": 1.2653,
+      "step": 431
+    },
+    {
+      "epoch": 0.019901414290321096,
+      "grad_norm": 0.9506834728835041,
+      "learning_rate": 0.00013251533742331288,
+      "loss": 1.4175,
+      "step": 432
+    },
+    {
+      "epoch": 0.019947482378956098,
+      "grad_norm": 0.9478357457538287,
+      "learning_rate": 0.00013282208588957057,
+      "loss": 1.1033,
+      "step": 433
+    },
+    {
+      "epoch": 0.0199935504675911,
+      "grad_norm": 0.5925818285274846,
+      "learning_rate": 0.00013312883435582823,
+      "loss": 1.4938,
+      "step": 434
+    },
+    {
+      "epoch": 0.020039618556226102,
+      "grad_norm": 0.5145086120346835,
+      "learning_rate": 0.0001334355828220859,
+      "loss": 1.4962,
+      "step": 435
+    },
+    {
+      "epoch": 0.020085686644861104,
+      "grad_norm": 0.7065431450298147,
+      "learning_rate": 0.00013374233128834358,
+      "loss": 1.2473,
+      "step": 436
+    },
+    {
+      "epoch": 0.020131754733496106,
+      "grad_norm": 0.5770822215493532,
+      "learning_rate": 0.00013404907975460124,
+      "loss": 1.1198,
+      "step": 437
+    },
+    {
+      "epoch": 0.02017782282213111,
+      "grad_norm": 0.612299960361438,
+      "learning_rate": 0.0001343558282208589,
+      "loss": 1.1615,
+      "step": 438
+    },
+    {
+      "epoch": 0.020223890910766113,
+      "grad_norm": 0.4930049286316884,
+      "learning_rate": 0.00013466257668711656,
+      "loss": 1.2963,
+      "step": 439
+    },
+    {
+      "epoch": 0.020269958999401115,
+      "grad_norm": 0.5768052453810412,
+      "learning_rate": 0.00013496932515337422,
+      "loss": 1.2183,
+      "step": 440
+    },
+    {
+      "epoch": 0.020316027088036117,
+      "grad_norm": 0.7983550021283577,
+      "learning_rate": 0.0001352760736196319,
+      "loss": 1.1055,
+      "step": 441
+    },
+    {
+      "epoch": 0.02036209517667112,
+      "grad_norm": 0.6069004530634943,
+      "learning_rate": 0.00013558282208588957,
+      "loss": 1.3551,
+      "step": 442
+    },
+    {
+      "epoch": 0.02040816326530612,
+      "grad_norm": 0.5611069653031893,
+      "learning_rate": 0.00013588957055214723,
+      "loss": 1.1367,
+      "step": 443
+    },
+    {
+      "epoch": 0.020454231353941126,
+      "grad_norm": 0.5197960459125087,
+      "learning_rate": 0.0001361963190184049,
+      "loss": 1.3769,
+      "step": 444
+    },
+    {
+      "epoch": 0.020500299442576128,
+      "grad_norm": 0.48873942469736054,
+      "learning_rate": 0.00013650306748466258,
+      "loss": 1.6251,
+      "step": 445
+    },
+    {
+      "epoch": 0.02054636753121113,
+      "grad_norm": 0.6548122638323524,
+      "learning_rate": 0.00013680981595092027,
+      "loss": 1.2167,
+      "step": 446
+    },
+    {
+      "epoch": 0.020592435619846132,
+      "grad_norm": 0.6635406090493257,
+      "learning_rate": 0.00013711656441717793,
+      "loss": 1.2842,
+      "step": 447
+    },
+    {
+      "epoch": 0.020638503708481134,
+      "grad_norm": 0.5485282085771375,
+      "learning_rate": 0.0001374233128834356,
+      "loss": 1.086,
+      "step": 448
+    },
+    {
+      "epoch": 0.02068457179711614,
+      "grad_norm": 0.492923934998443,
+      "learning_rate": 0.00013773006134969325,
+      "loss": 1.1198,
+      "step": 449
+    },
+    {
+      "epoch": 0.02073063988575114,
+      "grad_norm": 1.1756709577004028,
+      "learning_rate": 0.00013803680981595094,
+      "loss": 1.0827,
+      "step": 450
+    },
+    {
+      "epoch": 0.020776707974386143,
+      "grad_norm": 0.5156488385884216,
+      "learning_rate": 0.0001383435582822086,
+      "loss": 1.4348,
+      "step": 451
+    },
+    {
+      "epoch": 0.020822776063021145,
+      "grad_norm": 0.5028240330756594,
+      "learning_rate": 0.00013865030674846626,
+      "loss": 1.3906,
+      "step": 452
+    },
+    {
+      "epoch": 0.020868844151656147,
+      "grad_norm": 0.49624087276612355,
+      "learning_rate": 0.00013895705521472392,
+      "loss": 1.5199,
+      "step": 453
+    },
+    {
+      "epoch": 0.02091491224029115,
+      "grad_norm": 0.6571441187603715,
+      "learning_rate": 0.00013926380368098159,
+      "loss": 1.2321,
+      "step": 454
+    },
+    {
+      "epoch": 0.020960980328926154,
+      "grad_norm": 0.8475466663026641,
+      "learning_rate": 0.00013957055214723927,
+      "loss": 1.2527,
+      "step": 455
+    },
+    {
+      "epoch": 0.021007048417561156,
+      "grad_norm": 0.6044812808386559,
+      "learning_rate": 0.00013987730061349693,
+      "loss": 1.1742,
+      "step": 456
+    },
+    {
+      "epoch": 0.021053116506196158,
+      "grad_norm": 0.5866732119557103,
+      "learning_rate": 0.0001401840490797546,
+      "loss": 1.513,
+      "step": 457
+    },
+    {
+      "epoch": 0.02109918459483116,
+      "grad_norm": 0.7279710599378134,
+      "learning_rate": 0.00014049079754601228,
+      "loss": 1.1082,
+      "step": 458
+    },
+    {
+      "epoch": 0.021145252683466162,
+      "grad_norm": 0.5393823733668173,
+      "learning_rate": 0.00014079754601226994,
+      "loss": 1.3346,
+      "step": 459
+    },
+    {
+      "epoch": 0.021191320772101164,
+      "grad_norm": 0.454695178171063,
+      "learning_rate": 0.00014110429447852763,
+      "loss": 0.9804,
+      "step": 460
+    },
+    {
+      "epoch": 0.02123738886073617,
+      "grad_norm": 0.6583580656674628,
+      "learning_rate": 0.0001414110429447853,
+      "loss": 0.984,
+      "step": 461
+    },
+    {
+      "epoch": 0.02128345694937117,
+      "grad_norm": 0.5581037149737764,
+      "learning_rate": 0.00014171779141104295,
+      "loss": 1.1177,
+      "step": 462
+    },
+    {
+      "epoch": 0.021329525038006173,
+      "grad_norm": 0.5215677153896846,
+      "learning_rate": 0.00014202453987730061,
+      "loss": 1.5203,
+      "step": 463
+    },
+    {
+      "epoch": 0.021375593126641175,
+      "grad_norm": 0.7313954741875062,
+      "learning_rate": 0.00014233128834355828,
+      "loss": 1.5379,
+      "step": 464
+    },
+    {
+      "epoch": 0.021421661215276177,
+      "grad_norm": 0.4826304596816824,
+      "learning_rate": 0.00014263803680981596,
+      "loss": 1.5775,
+      "step": 465
+    },
+    {
+      "epoch": 0.021467729303911182,
+      "grad_norm": 0.5599715923312995,
+      "learning_rate": 0.00014294478527607362,
+      "loss": 1.476,
+      "step": 466
+    },
+    {
+      "epoch": 0.021513797392546184,
+      "grad_norm": 0.6530185752014686,
+      "learning_rate": 0.00014325153374233129,
+      "loss": 1.5155,
+      "step": 467
+    },
+    {
+      "epoch": 0.021559865481181186,
+      "grad_norm": 0.5965787367097369,
+      "learning_rate": 0.00014355828220858895,
+      "loss": 1.3472,
+      "step": 468
+    },
+    {
+      "epoch": 0.021605933569816188,
+      "grad_norm": 0.6011438567177697,
+      "learning_rate": 0.00014386503067484663,
+      "loss": 1.1834,
+      "step": 469
+    },
+    {
+      "epoch": 0.02165200165845119,
+      "grad_norm": 0.6205948761299666,
+      "learning_rate": 0.0001441717791411043,
+      "loss": 1.199,
+      "step": 470
+    },
+    {
+      "epoch": 0.021698069747086192,
+      "grad_norm": 0.5573016838720064,
+      "learning_rate": 0.00014447852760736198,
+      "loss": 1.5011,
+      "step": 471
+    },
+    {
+      "epoch": 0.021744137835721197,
+      "grad_norm": 0.5550097054034908,
+      "learning_rate": 0.00014478527607361964,
+      "loss": 1.4969,
+      "step": 472
+    },
+    {
+      "epoch": 0.0217902059243562,
+      "grad_norm": 0.6514367069325447,
+      "learning_rate": 0.0001450920245398773,
+      "loss": 0.9744,
+      "step": 473
+    },
+    {
+      "epoch": 0.0218362740129912,
+      "grad_norm": 0.5735666326486276,
+      "learning_rate": 0.000145398773006135,
+      "loss": 1.4821,
+      "step": 474
+    },
+    {
+      "epoch": 0.021882342101626203,
+      "grad_norm": 0.48878082809421597,
+      "learning_rate": 0.00014570552147239265,
+      "loss": 1.3591,
+      "step": 475
+    },
+    {
+      "epoch": 0.021928410190261205,
+      "grad_norm": 0.49504886426318473,
+      "learning_rate": 0.00014601226993865031,
+      "loss": 1.1424,
+      "step": 476
+    },
+    {
+      "epoch": 0.021974478278896207,
+      "grad_norm": 0.4458121310084816,
+      "learning_rate": 0.00014631901840490798,
+      "loss": 1.1871,
+      "step": 477
+    },
+    {
+      "epoch": 0.022020546367531212,
+      "grad_norm": 0.5518888125506198,
+      "learning_rate": 0.00014662576687116564,
+      "loss": 0.8545,
+      "step": 478
+    },
+    {
+      "epoch": 0.022066614456166214,
+      "grad_norm": 0.5230084321726499,
+      "learning_rate": 0.00014693251533742332,
+      "loss": 1.1481,
+      "step": 479
+    },
+    {
+      "epoch": 0.022112682544801216,
+      "grad_norm": 0.5790547256649033,
+      "learning_rate": 0.00014723926380368098,
+      "loss": 1.2946,
+      "step": 480
+    },
+    {
+      "epoch": 0.022158750633436218,
+      "grad_norm": 0.623239713187122,
+      "learning_rate": 0.00014754601226993865,
+      "loss": 1.3974,
+      "step": 481
+    },
+    {
+      "epoch": 0.02220481872207122,
+      "grad_norm": 0.5659107069650724,
+      "learning_rate": 0.0001478527607361963,
+      "loss": 1.2571,
+      "step": 482
+    },
+    {
+      "epoch": 0.022250886810706225,
+      "grad_norm": 0.4852225235960646,
+      "learning_rate": 0.000148159509202454,
+      "loss": 1.2644,
+      "step": 483
+    },
+    {
+      "epoch": 0.022296954899341227,
+      "grad_norm": 0.5507259519534681,
+      "learning_rate": 0.00014846625766871168,
+      "loss": 1.3336,
+      "step": 484
+    },
+    {
+      "epoch": 0.02234302298797623,
+      "grad_norm": 0.4715813417900144,
+      "learning_rate": 0.00014877300613496934,
+      "loss": 1.4194,
+      "step": 485
+    },
+    {
+      "epoch": 0.02238909107661123,
+      "grad_norm": 0.49426489695751696,
+      "learning_rate": 0.000149079754601227,
+      "loss": 1.3544,
+      "step": 486
+    },
+    {
+      "epoch": 0.022435159165246233,
+      "grad_norm": 0.643707633171356,
+      "learning_rate": 0.00014938650306748467,
+      "loss": 1.417,
+      "step": 487
+    },
+    {
+      "epoch": 0.022481227253881235,
+      "grad_norm": 0.5260491937057732,
+      "learning_rate": 0.00014969325153374235,
+      "loss": 1.3407,
+      "step": 488
+    },
+    {
+      "epoch": 0.02252729534251624,
+      "grad_norm": 0.5570324753135713,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.2827,
+      "step": 489
+    },
+    {
+      "epoch": 0.022573363431151242,
+      "grad_norm": 0.6578102469051025,
+      "learning_rate": 0.00015030674846625767,
+      "loss": 1.2917,
+      "step": 490
+    },
+    {
+      "epoch": 0.022619431519786244,
+      "grad_norm": 0.5642713026289865,
+      "learning_rate": 0.00015061349693251534,
+      "loss": 1.3826,
+      "step": 491
+    },
+    {
+      "epoch": 0.022665499608421246,
+      "grad_norm": 0.5648018141847541,
+      "learning_rate": 0.000150920245398773,
+      "loss": 1.0238,
+      "step": 492
+    },
+    {
+      "epoch": 0.022711567697056248,
+      "grad_norm": 0.531519564603025,
+      "learning_rate": 0.00015122699386503068,
+      "loss": 0.8601,
+      "step": 493
+    },
+    {
+      "epoch": 0.022757635785691253,
+      "grad_norm": 0.4816909421652289,
+      "learning_rate": 0.00015153374233128835,
+      "loss": 0.6733,
+      "step": 494
+    },
+    {
+      "epoch": 0.022803703874326255,
+      "grad_norm": 0.9239009942040249,
+      "learning_rate": 0.000151840490797546,
+      "loss": 1.2017,
+      "step": 495
+    },
+    {
+      "epoch": 0.022849771962961257,
+      "grad_norm": 0.6688991099357369,
+      "learning_rate": 0.0001521472392638037,
+      "loss": 0.8676,
+      "step": 496
+    },
+    {
+      "epoch": 0.02289584005159626,
+      "grad_norm": 0.6537740469506915,
+      "learning_rate": 0.00015245398773006136,
+      "loss": 1.3801,
+      "step": 497
+    },
+    {
+      "epoch": 0.02294190814023126,
+      "grad_norm": 0.5010366746570297,
+      "learning_rate": 0.00015276073619631904,
+      "loss": 1.3659,
+      "step": 498
+    },
+    {
+      "epoch": 0.022987976228866263,
+      "grad_norm": 0.46266047954586986,
+      "learning_rate": 0.0001530674846625767,
+      "loss": 0.9304,
+      "step": 499
+    },
+    {
+      "epoch": 0.02303404431750127,
+      "grad_norm": 0.51947535593283,
+      "learning_rate": 0.00015337423312883436,
+      "loss": 1.5958,
+      "step": 500
+    },
+    {
+      "epoch": 0.02308011240613627,
+      "grad_norm": 0.4897907558235972,
+      "learning_rate": 0.00015368098159509203,
+      "loss": 1.406,
+      "step": 501
+    },
+    {
+      "epoch": 0.023126180494771272,
+      "grad_norm": 0.8258426235632195,
+      "learning_rate": 0.0001539877300613497,
+      "loss": 1.3216,
+      "step": 502
+    },
+    {
+      "epoch": 0.023172248583406274,
+      "grad_norm": 0.5186157842346172,
+      "learning_rate": 0.00015429447852760737,
+      "loss": 1.5257,
+      "step": 503
+    },
+    {
+      "epoch": 0.023218316672041276,
+      "grad_norm": 0.5049838730873579,
+      "learning_rate": 0.00015460122699386504,
+      "loss": 1.3679,
+      "step": 504
+    },
+    {
+      "epoch": 0.023264384760676278,
+      "grad_norm": 0.5743767913232409,
+      "learning_rate": 0.0001549079754601227,
+      "loss": 0.9566,
+      "step": 505
+    },
+    {
+      "epoch": 0.023310452849311283,
+      "grad_norm": 0.5147579979008217,
+      "learning_rate": 0.00015521472392638036,
+      "loss": 1.1672,
+      "step": 506
+    },
+    {
+      "epoch": 0.023356520937946285,
+      "grad_norm": 0.5925525597414084,
+      "learning_rate": 0.00015552147239263805,
+      "loss": 1.073,
+      "step": 507
+    },
+    {
+      "epoch": 0.023402589026581287,
+      "grad_norm": 0.4876000875970032,
+      "learning_rate": 0.00015582822085889573,
+      "loss": 1.348,
+      "step": 508
+    },
+    {
+      "epoch": 0.02344865711521629,
+      "grad_norm": 0.47355155160895,
+      "learning_rate": 0.0001561349693251534,
+      "loss": 1.2083,
+      "step": 509
+    },
+    {
+      "epoch": 0.02349472520385129,
+      "grad_norm": 0.5712487942624723,
+      "learning_rate": 0.00015644171779141105,
+      "loss": 1.0479,
+      "step": 510
+    },
+    {
+      "epoch": 0.023540793292486296,
+      "grad_norm": 0.7027413442103287,
+      "learning_rate": 0.00015674846625766872,
+      "loss": 1.5632,
+      "step": 511
+    },
+    {
+      "epoch": 0.0235868613811213,
+      "grad_norm": 0.5759093219239868,
+      "learning_rate": 0.0001570552147239264,
+      "loss": 1.4641,
+      "step": 512
+    },
+    {
+      "epoch": 0.0236329294697563,
+      "grad_norm": 0.6304901031762065,
+      "learning_rate": 0.00015736196319018406,
+      "loss": 0.8773,
+      "step": 513
+    },
+    {
+      "epoch": 0.023678997558391302,
+      "grad_norm": 0.5608412889519707,
+      "learning_rate": 0.00015766871165644173,
+      "loss": 1.3367,
+      "step": 514
+    },
+    {
+      "epoch": 0.023725065647026304,
+      "grad_norm": 0.7516801701987127,
+      "learning_rate": 0.00015797546012269939,
+      "loss": 1.1182,
+      "step": 515
+    },
+    {
+      "epoch": 0.023771133735661306,
+      "grad_norm": 0.5366105724477778,
+      "learning_rate": 0.00015828220858895705,
+      "loss": 1.6397,
+      "step": 516
+    },
+    {
+      "epoch": 0.02381720182429631,
+      "grad_norm": 0.7242857161529414,
+      "learning_rate": 0.00015858895705521474,
+      "loss": 1.4058,
+      "step": 517
+    },
+    {
+      "epoch": 0.023863269912931313,
+      "grad_norm": 0.5564355917515519,
+      "learning_rate": 0.0001588957055214724,
+      "loss": 1.5376,
+      "step": 518
+    },
+    {
+      "epoch": 0.023909338001566315,
+      "grad_norm": 0.6429913413453947,
+      "learning_rate": 0.00015920245398773006,
+      "loss": 1.2992,
+      "step": 519
+    },
+    {
+      "epoch": 0.023955406090201317,
+      "grad_norm": 0.5404695805609155,
+      "learning_rate": 0.00015950920245398772,
+      "loss": 1.1494,
+      "step": 520
+    },
+    {
+      "epoch": 0.02400147417883632,
+      "grad_norm": 0.40748134994781354,
+      "learning_rate": 0.0001598159509202454,
+      "loss": 1.0737,
+      "step": 521
+    },
+    {
+      "epoch": 0.02404754226747132,
+      "grad_norm": 0.6150578813263151,
+      "learning_rate": 0.0001601226993865031,
+      "loss": 1.1734,
+      "step": 522
+    },
+    {
+      "epoch": 0.024093610356106326,
+      "grad_norm": 0.5428918669822446,
+      "learning_rate": 0.00016042944785276075,
+      "loss": 1.0401,
+      "step": 523
+    },
+    {
+      "epoch": 0.02413967844474133,
+      "grad_norm": 0.5605321496435296,
+      "learning_rate": 0.00016073619631901842,
+      "loss": 1.3918,
+      "step": 524
+    },
+    {
+      "epoch": 0.02418574653337633,
+      "grad_norm": 0.5659727111536695,
+      "learning_rate": 0.00016104294478527608,
+      "loss": 1.2091,
+      "step": 525
+    },
+    {
+      "epoch": 0.024231814622011332,
+      "grad_norm": 0.506109381443957,
+      "learning_rate": 0.00016134969325153376,
+      "loss": 1.2051,
+      "step": 526
+    },
+    {
+      "epoch": 0.024277882710646334,
+      "grad_norm": 0.4969779852570272,
+      "learning_rate": 0.00016165644171779143,
+      "loss": 1.3486,
+      "step": 527
+    },
+    {
+      "epoch": 0.02432395079928134,
+      "grad_norm": 0.7205635401180093,
+      "learning_rate": 0.00016196319018404909,
+      "loss": 1.6054,
+      "step": 528
+    },
+    {
+      "epoch": 0.02437001888791634,
+      "grad_norm": 0.46711568917268875,
+      "learning_rate": 0.00016226993865030675,
+      "loss": 1.3053,
+      "step": 529
+    },
+    {
+      "epoch": 0.024416086976551343,
+      "grad_norm": 0.4937685655151686,
+      "learning_rate": 0.0001625766871165644,
+      "loss": 1.1575,
+      "step": 530
+    },
+    {
+      "epoch": 0.024462155065186345,
+      "grad_norm": 0.6340841581657817,
+      "learning_rate": 0.0001628834355828221,
+      "loss": 1.3176,
+      "step": 531
+    },
+    {
+      "epoch": 0.024508223153821347,
+      "grad_norm": 0.5192810795149426,
+      "learning_rate": 0.00016319018404907976,
+      "loss": 1.0112,
+      "step": 532
+    },
+    {
+      "epoch": 0.02455429124245635,
+      "grad_norm": 0.7286344440773026,
+      "learning_rate": 0.00016349693251533744,
+      "loss": 1.0421,
+      "step": 533
+    },
+    {
+      "epoch": 0.024600359331091354,
+      "grad_norm": 0.5008470821755977,
+      "learning_rate": 0.0001638036809815951,
+      "loss": 1.5747,
+      "step": 534
+    },
+    {
+      "epoch": 0.024646427419726356,
+      "grad_norm": 0.6112119658383529,
+      "learning_rate": 0.00016411042944785277,
+      "loss": 1.7977,
+      "step": 535
+    },
+    {
+      "epoch": 0.02469249550836136,
+      "grad_norm": 0.5810484640099962,
+      "learning_rate": 0.00016441717791411045,
+      "loss": 1.5304,
+      "step": 536
+    },
+    {
+      "epoch": 0.02473856359699636,
+      "grad_norm": 0.5805307399266465,
+      "learning_rate": 0.00016472392638036812,
+      "loss": 0.9542,
+      "step": 537
+    },
+    {
+      "epoch": 0.024784631685631362,
+      "grad_norm": 0.5783042816588378,
+      "learning_rate": 0.00016503067484662578,
+      "loss": 0.9286,
+      "step": 538
+    },
+    {
+      "epoch": 0.024830699774266364,
+      "grad_norm": 0.7520191510667058,
+      "learning_rate": 0.00016533742331288344,
+      "loss": 1.2717,
+      "step": 539
+    },
+    {
+      "epoch": 0.02487676786290137,
+      "grad_norm": 0.7059325267096991,
+      "learning_rate": 0.0001656441717791411,
+      "loss": 1.3362,
+      "step": 540
+    },
+    {
+      "epoch": 0.02492283595153637,
+      "grad_norm": 0.673159032016408,
+      "learning_rate": 0.00016595092024539879,
+      "loss": 1.3921,
+      "step": 541
+    },
+    {
+      "epoch": 0.024968904040171373,
+      "grad_norm": 0.5430627888900998,
+      "learning_rate": 0.00016625766871165645,
+      "loss": 1.4012,
+      "step": 542
+    },
+    {
+      "epoch": 0.025014972128806375,
+      "grad_norm": 0.6605265241948239,
+      "learning_rate": 0.0001665644171779141,
+      "loss": 1.141,
+      "step": 543
+    },
+    {
+      "epoch": 0.025061040217441377,
+      "grad_norm": 0.5946019849206994,
+      "learning_rate": 0.00016687116564417177,
+      "loss": 1.2958,
+      "step": 544
+    },
+    {
+      "epoch": 0.025107108306076382,
+      "grad_norm": 0.6545419228345305,
+      "learning_rate": 0.00016717791411042946,
+      "loss": 1.4041,
+      "step": 545
+    },
+    {
+      "epoch": 0.025153176394711384,
+      "grad_norm": 0.5401121678412126,
+      "learning_rate": 0.00016748466257668714,
+      "loss": 1.4697,
+      "step": 546
+    },
+    {
+      "epoch": 0.025199244483346386,
+      "grad_norm": 0.5098814331212207,
+      "learning_rate": 0.0001677914110429448,
+      "loss": 1.3221,
+      "step": 547
+    },
+    {
+      "epoch": 0.025245312571981388,
+      "grad_norm": 0.7256158452072782,
+      "learning_rate": 0.00016809815950920247,
+      "loss": 0.985,
+      "step": 548
+    },
+    {
+      "epoch": 0.02529138066061639,
+      "grad_norm": 0.693224988051577,
+      "learning_rate": 0.00016840490797546013,
+      "loss": 1.0423,
+      "step": 549
+    },
+    {
+      "epoch": 0.025337448749251392,
+      "grad_norm": 0.829921106814423,
+      "learning_rate": 0.00016871165644171781,
+      "loss": 1.3368,
+      "step": 550
+    },
+    {
+      "epoch": 0.025383516837886397,
+      "grad_norm": 0.5286524656148935,
+      "learning_rate": 0.00016901840490797548,
+      "loss": 1.3455,
+      "step": 551
+    },
+    {
+      "epoch": 0.0254295849265214,
+      "grad_norm": 0.7230056486744934,
+      "learning_rate": 0.00016932515337423314,
+      "loss": 1.224,
+      "step": 552
+    },
+    {
+      "epoch": 0.0254756530151564,
+      "grad_norm": 0.6985399438894404,
+      "learning_rate": 0.0001696319018404908,
+      "loss": 1.2639,
+      "step": 553
+    },
+    {
+      "epoch": 0.025521721103791403,
+      "grad_norm": 0.47093547726794266,
+      "learning_rate": 0.00016993865030674846,
+      "loss": 1.169,
+      "step": 554
+    },
+    {
+      "epoch": 0.025567789192426405,
+      "grad_norm": 0.5552578538702152,
+      "learning_rate": 0.00017024539877300615,
+      "loss": 1.48,
+      "step": 555
+    },
+    {
+      "epoch": 0.025613857281061407,
+      "grad_norm": 0.8526518364083183,
+      "learning_rate": 0.0001705521472392638,
+      "loss": 1.1413,
+      "step": 556
+    },
+    {
+      "epoch": 0.025659925369696412,
+      "grad_norm": 0.6986354413643021,
+      "learning_rate": 0.00017085889570552147,
+      "loss": 1.1064,
+      "step": 557
+    },
+    {
+      "epoch": 0.025705993458331414,
+      "grad_norm": 1.0912611249002593,
+      "learning_rate": 0.00017116564417177916,
+      "loss": 1.1539,
+      "step": 558
+    },
+    {
+      "epoch": 0.025752061546966416,
+      "grad_norm": 0.8954379155977537,
+      "learning_rate": 0.00017147239263803682,
+      "loss": 1.0294,
+      "step": 559
+    },
+    {
+      "epoch": 0.025798129635601418,
+      "grad_norm": 0.5957103038272765,
+      "learning_rate": 0.0001717791411042945,
+      "loss": 1.4308,
+      "step": 560
+    },
+    {
+      "epoch": 0.02584419772423642,
+      "grad_norm": 0.6413606816667416,
+      "learning_rate": 0.00017208588957055217,
+      "loss": 1.1957,
+      "step": 561
+    },
+    {
+      "epoch": 0.025890265812871426,
+      "grad_norm": 0.566440314645742,
+      "learning_rate": 0.00017239263803680983,
+      "loss": 0.8633,
+      "step": 562
+    },
+    {
+      "epoch": 0.025936333901506427,
+      "grad_norm": 0.4758150099289959,
+      "learning_rate": 0.0001726993865030675,
+      "loss": 1.4557,
+      "step": 563
+    },
+    {
+      "epoch": 0.02598240199014143,
+      "grad_norm": 0.446652659911936,
+      "learning_rate": 0.00017300613496932518,
+      "loss": 1.0615,
+      "step": 564
+    },
+    {
+      "epoch": 0.02602847007877643,
+      "grad_norm": 0.509123083494041,
+      "learning_rate": 0.00017331288343558284,
+      "loss": 0.8351,
+      "step": 565
+    },
+    {
+      "epoch": 0.026074538167411433,
+      "grad_norm": 0.5643313417735877,
+      "learning_rate": 0.0001736196319018405,
+      "loss": 1.4633,
+      "step": 566
+    },
+    {
+      "epoch": 0.026120606256046435,
+      "grad_norm": 0.5584853140310495,
+      "learning_rate": 0.00017392638036809816,
+      "loss": 1.6614,
+      "step": 567
+    },
+    {
+      "epoch": 0.02616667434468144,
+      "grad_norm": 0.8311904639545469,
+      "learning_rate": 0.00017423312883435582,
+      "loss": 1.3271,
+      "step": 568
+    },
+    {
+      "epoch": 0.026212742433316442,
+      "grad_norm": 0.5445683950539235,
+      "learning_rate": 0.0001745398773006135,
+      "loss": 1.3867,
+      "step": 569
+    },
+    {
+      "epoch": 0.026258810521951444,
+      "grad_norm": 0.6741713591466492,
+      "learning_rate": 0.00017484662576687117,
+      "loss": 1.1748,
+      "step": 570
+    },
+    {
+      "epoch": 0.026304878610586446,
+      "grad_norm": 0.6172589757558803,
+      "learning_rate": 0.00017515337423312886,
+      "loss": 1.2328,
+      "step": 571
+    },
+    {
+      "epoch": 0.026350946699221448,
+      "grad_norm": 0.67437552435688,
+      "learning_rate": 0.00017546012269938652,
+      "loss": 1.5252,
+      "step": 572
+    },
+    {
+      "epoch": 0.02639701478785645,
+      "grad_norm": 0.6436933360683789,
+      "learning_rate": 0.00017576687116564418,
+      "loss": 1.4914,
+      "step": 573
+    },
+    {
+      "epoch": 0.026443082876491456,
+      "grad_norm": 0.774587517817201,
+      "learning_rate": 0.00017607361963190187,
+      "loss": 1.3843,
+      "step": 574
+    },
+    {
+      "epoch": 0.026489150965126457,
+      "grad_norm": 1.2410273155788658,
+      "learning_rate": 0.00017638036809815953,
+      "loss": 1.3294,
+      "step": 575
+    },
+    {
+      "epoch": 0.02653521905376146,
+      "grad_norm": 0.5173536854752888,
+      "learning_rate": 0.0001766871165644172,
+      "loss": 1.2568,
+      "step": 576
+    },
+    {
+      "epoch": 0.02658128714239646,
+      "grad_norm": 0.41034850294080094,
+      "learning_rate": 0.00017699386503067485,
+      "loss": 1.0055,
+      "step": 577
+    },
+    {
+      "epoch": 0.026627355231031463,
+      "grad_norm": 0.6389827399657563,
+      "learning_rate": 0.0001773006134969325,
+      "loss": 1.1786,
+      "step": 578
+    },
+    {
+      "epoch": 0.02667342331966647,
+      "grad_norm": 0.6161535517671173,
+      "learning_rate": 0.0001776073619631902,
+      "loss": 1.5202,
+      "step": 579
+    },
+    {
+      "epoch": 0.02671949140830147,
+      "grad_norm": 0.652174115517866,
+      "learning_rate": 0.00017791411042944786,
+      "loss": 0.9854,
+      "step": 580
+    },
+    {
+      "epoch": 0.026765559496936472,
+      "grad_norm": 0.5407531781442533,
+      "learning_rate": 0.00017822085889570552,
+      "loss": 1.6345,
+      "step": 581
+    },
+    {
+      "epoch": 0.026811627585571474,
+      "grad_norm": 0.646845264875195,
+      "learning_rate": 0.00017852760736196318,
+      "loss": 1.3744,
+      "step": 582
+    },
+    {
+      "epoch": 0.026857695674206476,
+      "grad_norm": 0.5371725410843516,
+      "learning_rate": 0.00017883435582822087,
+      "loss": 0.9787,
+      "step": 583
+    },
+    {
+      "epoch": 0.026903763762841478,
+      "grad_norm": 0.7233038772213499,
+      "learning_rate": 0.00017914110429447856,
+      "loss": 1.181,
+      "step": 584
+    },
+    {
+      "epoch": 0.026949831851476484,
+      "grad_norm": 0.47397975588958136,
+      "learning_rate": 0.00017944785276073622,
+      "loss": 1.5874,
+      "step": 585
+    },
+    {
+      "epoch": 0.026995899940111485,
+      "grad_norm": 0.5008884413856999,
+      "learning_rate": 0.00017975460122699388,
+      "loss": 1.1101,
+      "step": 586
+    },
+    {
+      "epoch": 0.027041968028746487,
+      "grad_norm": 0.6425122660304662,
+      "learning_rate": 0.00018006134969325154,
+      "loss": 0.9281,
+      "step": 587
+    },
+    {
+      "epoch": 0.02708803611738149,
+      "grad_norm": 0.539787805670795,
+      "learning_rate": 0.00018036809815950923,
+      "loss": 1.2316,
+      "step": 588
+    },
+    {
+      "epoch": 0.02713410420601649,
+      "grad_norm": 0.5017590316809062,
+      "learning_rate": 0.0001806748466257669,
+      "loss": 1.2608,
+      "step": 589
+    },
+    {
+      "epoch": 0.027180172294651497,
+      "grad_norm": 0.5327552609039972,
+      "learning_rate": 0.00018098159509202455,
+      "loss": 1.2271,
+      "step": 590
+    },
+    {
+      "epoch": 0.0272262403832865,
+      "grad_norm": 0.5764260433744264,
+      "learning_rate": 0.0001812883435582822,
+      "loss": 1.5241,
+      "step": 591
+    },
+    {
+      "epoch": 0.0272723084719215,
+      "grad_norm": 0.5657281494170161,
+      "learning_rate": 0.00018159509202453987,
+      "loss": 1.2396,
+      "step": 592
+    },
+    {
+      "epoch": 0.027318376560556502,
+      "grad_norm": 0.6015552660471936,
+      "learning_rate": 0.00018190184049079756,
+      "loss": 1.3838,
+      "step": 593
+    },
+    {
+      "epoch": 0.027364444649191504,
+      "grad_norm": 0.5843919842778933,
+      "learning_rate": 0.00018220858895705522,
+      "loss": 1.411,
+      "step": 594
+    },
+    {
+      "epoch": 0.027410512737826506,
+      "grad_norm": 0.9074356635099659,
+      "learning_rate": 0.00018251533742331288,
+      "loss": 1.3186,
+      "step": 595
+    },
+    {
+      "epoch": 0.02745658082646151,
+      "grad_norm": 0.44420978795126836,
+      "learning_rate": 0.00018282208588957057,
+      "loss": 1.1734,
+      "step": 596
+    },
+    {
+      "epoch": 0.027502648915096514,
+      "grad_norm": 0.4588190292194558,
+      "learning_rate": 0.00018312883435582823,
+      "loss": 1.0764,
+      "step": 597
+    },
+    {
+      "epoch": 0.027548717003731515,
+      "grad_norm": 0.5707004582368886,
+      "learning_rate": 0.00018343558282208592,
+      "loss": 1.3397,
+      "step": 598
+    },
+    {
+      "epoch": 0.027594785092366517,
+      "grad_norm": 0.7299022926412564,
+      "learning_rate": 0.00018374233128834358,
+      "loss": 1.3148,
+      "step": 599
+    },
+    {
+      "epoch": 0.02764085318100152,
+      "grad_norm": 0.626282115069609,
+      "learning_rate": 0.00018404907975460124,
+      "loss": 0.9367,
+      "step": 600
+    },
+    {
+      "epoch": 0.02768692126963652,
+      "grad_norm": 0.5771755017632331,
+      "learning_rate": 0.0001843558282208589,
+      "loss": 1.0043,
+      "step": 601
+    },
+    {
+      "epoch": 0.027732989358271527,
+      "grad_norm": 0.6003076597844547,
+      "learning_rate": 0.00018466257668711656,
+      "loss": 1.5912,
+      "step": 602
+    },
+    {
+      "epoch": 0.02777905744690653,
+      "grad_norm": 0.7504726366767099,
+      "learning_rate": 0.00018496932515337425,
+      "loss": 1.218,
+      "step": 603
+    },
+    {
+      "epoch": 0.02782512553554153,
+      "grad_norm": 0.6058321010034448,
+      "learning_rate": 0.0001852760736196319,
+      "loss": 1.26,
+      "step": 604
+    },
+    {
+      "epoch": 0.027871193624176532,
+      "grad_norm": 0.5520444619340273,
+      "learning_rate": 0.00018558282208588957,
+      "loss": 0.8955,
+      "step": 605
+    },
+    {
+      "epoch": 0.027917261712811534,
+      "grad_norm": 0.5531108479431723,
+      "learning_rate": 0.00018588957055214723,
+      "loss": 1.1573,
+      "step": 606
+    },
+    {
+      "epoch": 0.02796332980144654,
+      "grad_norm": 0.6472317150602768,
+      "learning_rate": 0.00018619631901840492,
+      "loss": 0.9275,
+      "step": 607
+    },
+    {
+      "epoch": 0.02800939789008154,
+      "grad_norm": 0.7861989685255991,
+      "learning_rate": 0.00018650306748466258,
+      "loss": 0.9654,
+      "step": 608
+    },
+    {
+      "epoch": 0.028055465978716543,
+      "grad_norm": 0.6173449187161149,
+      "learning_rate": 0.00018680981595092027,
+      "loss": 1.1699,
+      "step": 609
+    },
+    {
+      "epoch": 0.028101534067351545,
+      "grad_norm": 0.6019912463716274,
+      "learning_rate": 0.00018711656441717793,
+      "loss": 1.2086,
+      "step": 610
+    },
+    {
+      "epoch": 0.028147602155986547,
+      "grad_norm": 0.5543489588411858,
+      "learning_rate": 0.0001874233128834356,
+      "loss": 1.1504,
+      "step": 611
+    },
+    {
+      "epoch": 0.02819367024462155,
+      "grad_norm": 0.5708758106719565,
+      "learning_rate": 0.00018773006134969328,
+      "loss": 1.6596,
+      "step": 612
+    },
+    {
+      "epoch": 0.028239738333256555,
+      "grad_norm": 0.5370439707666345,
+      "learning_rate": 0.00018803680981595094,
+      "loss": 1.7904,
+      "step": 613
+    },
+    {
+      "epoch": 0.028285806421891557,
+      "grad_norm": 0.591019334685924,
+      "learning_rate": 0.0001883435582822086,
+      "loss": 1.3089,
+      "step": 614
+    },
+    {
+      "epoch": 0.02833187451052656,
+      "grad_norm": 0.5383582565134855,
+      "learning_rate": 0.00018865030674846626,
+      "loss": 1.2538,
+      "step": 615
+    },
+    {
+      "epoch": 0.02837794259916156,
+      "grad_norm": 0.52475583144224,
+      "learning_rate": 0.00018895705521472392,
+      "loss": 1.0875,
+      "step": 616
+    },
+    {
+      "epoch": 0.028424010687796562,
+      "grad_norm": 0.46497317998460846,
+      "learning_rate": 0.0001892638036809816,
+      "loss": 1.3686,
+      "step": 617
+    },
+    {
+      "epoch": 0.028470078776431564,
+      "grad_norm": 0.6506601800460828,
+      "learning_rate": 0.00018957055214723927,
+      "loss": 1.3524,
+      "step": 618
+    },
+    {
+      "epoch": 0.02851614686506657,
+      "grad_norm": 0.45958625909057665,
+      "learning_rate": 0.00018987730061349693,
+      "loss": 1.3473,
+      "step": 619
+    },
+    {
+      "epoch": 0.02856221495370157,
+      "grad_norm": 0.6233618988020021,
+      "learning_rate": 0.0001901840490797546,
+      "loss": 1.4012,
+      "step": 620
+    },
+    {
+      "epoch": 0.028608283042336573,
+      "grad_norm": 0.5800510796372809,
+      "learning_rate": 0.00019049079754601228,
+      "loss": 1.4217,
+      "step": 621
+    },
+    {
+      "epoch": 0.028654351130971575,
+      "grad_norm": 0.5393220004840451,
+      "learning_rate": 0.00019079754601226997,
+      "loss": 1.2899,
+      "step": 622
+    },
+    {
+      "epoch": 0.028700419219606577,
+      "grad_norm": 0.46430595812931785,
+      "learning_rate": 0.00019110429447852763,
+      "loss": 1.2388,
+      "step": 623
+    },
+    {
+      "epoch": 0.028746487308241583,
+      "grad_norm": 0.5442511366180596,
+      "learning_rate": 0.0001914110429447853,
+      "loss": 1.4996,
+      "step": 624
+    },
+    {
+      "epoch": 0.028792555396876585,
+      "grad_norm": 0.5926996842630151,
+      "learning_rate": 0.00019171779141104295,
+      "loss": 1.173,
+      "step": 625
+    },
+    {
+      "epoch": 0.028838623485511587,
+      "grad_norm": 0.5204976874125498,
+      "learning_rate": 0.00019202453987730064,
+      "loss": 1.1804,
+      "step": 626
+    },
+    {
+      "epoch": 0.02888469157414659,
+      "grad_norm": 0.5197748792900281,
+      "learning_rate": 0.0001923312883435583,
+      "loss": 1.4822,
+      "step": 627
+    },
+    {
+      "epoch": 0.02893075966278159,
+      "grad_norm": 0.5076100296395281,
+      "learning_rate": 0.00019263803680981596,
+      "loss": 1.3399,
+      "step": 628
+    },
+    {
+      "epoch": 0.028976827751416592,
+      "grad_norm": 0.6164358707512612,
+      "learning_rate": 0.00019294478527607362,
+      "loss": 0.9972,
+      "step": 629
+    },
+    {
+      "epoch": 0.029022895840051598,
+      "grad_norm": 0.7879387475573963,
+      "learning_rate": 0.00019325153374233128,
+      "loss": 0.7786,
+      "step": 630
+    },
+    {
+      "epoch": 0.0290689639286866,
+      "grad_norm": 0.9440641824366766,
+      "learning_rate": 0.00019355828220858897,
+      "loss": 1.2949,
+      "step": 631
+    },
+    {
+      "epoch": 0.0291150320173216,
+      "grad_norm": 0.5186419109090085,
+      "learning_rate": 0.00019386503067484663,
+      "loss": 1.2875,
+      "step": 632
+    },
+    {
+      "epoch": 0.029161100105956603,
+      "grad_norm": 0.5754572098026057,
+      "learning_rate": 0.0001941717791411043,
+      "loss": 1.2353,
+      "step": 633
+    },
+    {
+      "epoch": 0.029207168194591605,
+      "grad_norm": 0.43534398626170495,
+      "learning_rate": 0.00019447852760736198,
+      "loss": 1.1993,
+      "step": 634
+    },
+    {
+      "epoch": 0.029253236283226607,
+      "grad_norm": 0.6588323642044535,
+      "learning_rate": 0.00019478527607361964,
+      "loss": 1.4526,
+      "step": 635
+    },
+    {
+      "epoch": 0.029299304371861613,
+      "grad_norm": 0.7199370151903504,
+      "learning_rate": 0.00019509202453987733,
+      "loss": 1.0257,
+      "step": 636
+    },
+    {
+      "epoch": 0.029345372460496615,
+      "grad_norm": 0.581501285876184,
+      "learning_rate": 0.000195398773006135,
+      "loss": 1.3592,
+      "step": 637
+    },
+    {
+      "epoch": 0.029391440549131616,
+      "grad_norm": 0.6578896488592731,
+      "learning_rate": 0.00019570552147239265,
+      "loss": 1.2616,
+      "step": 638
+    },
+    {
+      "epoch": 0.02943750863776662,
+      "grad_norm": 0.6259063029061305,
+      "learning_rate": 0.0001960122699386503,
+      "loss": 1.6049,
+      "step": 639
+    },
+    {
+      "epoch": 0.02948357672640162,
+      "grad_norm": 0.6192620947364964,
+      "learning_rate": 0.00019631901840490797,
+      "loss": 0.8595,
+      "step": 640
+    },
+    {
+      "epoch": 0.029529644815036626,
+      "grad_norm": 0.8767215638935036,
+      "learning_rate": 0.00019662576687116566,
+      "loss": 1.1011,
+      "step": 641
+    },
+    {
+      "epoch": 0.029575712903671628,
+      "grad_norm": 0.8631983283816006,
+      "learning_rate": 0.00019693251533742332,
+      "loss": 1.2616,
+      "step": 642
+    },
+    {
+      "epoch": 0.02962178099230663,
+      "grad_norm": 0.8954761377471316,
+      "learning_rate": 0.00019723926380368098,
+      "loss": 1.4888,
+      "step": 643
+    },
+    {
+      "epoch": 0.02966784908094163,
+      "grad_norm": 0.5981570714335578,
+      "learning_rate": 0.00019754601226993864,
+      "loss": 1.4755,
+      "step": 644
+    },
+    {
+      "epoch": 0.029713917169576633,
+      "grad_norm": 0.6481458613948916,
+      "learning_rate": 0.00019785276073619633,
+      "loss": 1.3779,
+      "step": 645
+    },
+    {
+      "epoch": 0.029759985258211635,
+      "grad_norm": 0.6621582950456114,
+      "learning_rate": 0.00019815950920245402,
+      "loss": 1.2063,
+      "step": 646
+    },
+    {
+      "epoch": 0.02980605334684664,
+      "grad_norm": 0.6559457343094043,
+      "learning_rate": 0.00019846625766871168,
+      "loss": 1.1292,
+      "step": 647
+    },
+    {
+      "epoch": 0.029852121435481643,
+      "grad_norm": 0.6669813774278682,
+      "learning_rate": 0.00019877300613496934,
+      "loss": 1.0937,
+      "step": 648
+    },
+    {
+      "epoch": 0.029898189524116645,
+      "grad_norm": 0.687670166354479,
+      "learning_rate": 0.000199079754601227,
+      "loss": 1.2676,
+      "step": 649
+    },
+    {
+      "epoch": 0.029944257612751646,
+      "grad_norm": 0.6368194243980484,
+      "learning_rate": 0.0001993865030674847,
+      "loss": 1.1796,
+      "step": 650
+    },
+    {
+      "epoch": 0.02999032570138665,
+      "grad_norm": 0.7715393181892464,
+      "learning_rate": 0.00019969325153374235,
+      "loss": 1.1986,
+      "step": 651
+    },
+    {
+      "epoch": 0.03003639379002165,
+      "grad_norm": 0.6208503453901123,
+      "learning_rate": 0.0002,
+      "loss": 1.5081,
+      "step": 652
+    },
+    {
+      "epoch": 0.030082461878656656,
+      "grad_norm": 0.598121490259459,
+      "learning_rate": 0.00019999999888683573,
+      "loss": 1.3757,
+      "step": 653
+    },
+    {
+      "epoch": 0.030128529967291658,
+      "grad_norm": 0.5359147313823284,
+      "learning_rate": 0.00019999999554734296,
+      "loss": 1.1118,
+      "step": 654
+    },
+    {
+      "epoch": 0.03017459805592666,
+      "grad_norm": 0.9402336634064428,
+      "learning_rate": 0.00019999998998152175,
+      "loss": 1.0626,
+      "step": 655
+    },
+    {
+      "epoch": 0.03022066614456166,
+      "grad_norm": 0.6474272450532342,
+      "learning_rate": 0.00019999998218937223,
+      "loss": 1.443,
+      "step": 656
+    },
+    {
+      "epoch": 0.030266734233196663,
+      "grad_norm": 0.7169300411339717,
+      "learning_rate": 0.00019999997217089453,
+      "loss": 1.2486,
+      "step": 657
+    },
+    {
+      "epoch": 0.03031280232183167,
+      "grad_norm": 0.9078278165392585,
+      "learning_rate": 0.00019999995992608894,
+      "loss": 1.0015,
+      "step": 658
+    },
+    {
+      "epoch": 0.03035887041046667,
+      "grad_norm": 0.7696108224521298,
+      "learning_rate": 0.0001999999454549557,
+      "loss": 1.2933,
+      "step": 659
+    },
+    {
+      "epoch": 0.030404938499101673,
+      "grad_norm": 0.7036348700945554,
+      "learning_rate": 0.00019999992875749517,
+      "loss": 1.3059,
+      "step": 660
+    },
+    {
+      "epoch": 0.030451006587736675,
+      "grad_norm": 0.5344020254634791,
+      "learning_rate": 0.00019999990983370766,
+      "loss": 1.211,
+      "step": 661
+    },
+    {
+      "epoch": 0.030497074676371676,
+      "grad_norm": 0.6248605652801618,
+      "learning_rate": 0.0001999998886835936,
+      "loss": 1.5563,
+      "step": 662
+    },
+    {
+      "epoch": 0.03054314276500668,
+      "grad_norm": 0.6634321886029204,
+      "learning_rate": 0.00019999986530715353,
+      "loss": 0.9544,
+      "step": 663
+    },
+    {
+      "epoch": 0.030589210853641684,
+      "grad_norm": 0.7846429800067349,
+      "learning_rate": 0.00019999983970438789,
+      "loss": 1.3104,
+      "step": 664
+    },
+    {
+      "epoch": 0.030635278942276686,
+      "grad_norm": 0.6849901727460651,
+      "learning_rate": 0.00019999981187529732,
+      "loss": 1.1589,
+      "step": 665
+    },
+    {
+      "epoch": 0.030681347030911688,
+      "grad_norm": 0.7427586007604318,
+      "learning_rate": 0.00019999978181988235,
+      "loss": 1.0837,
+      "step": 666
+    },
+    {
+      "epoch": 0.03072741511954669,
+      "grad_norm": 0.6756382076515431,
+      "learning_rate": 0.0001999997495381437,
+      "loss": 1.1606,
+      "step": 667
+    },
+    {
+      "epoch": 0.03077348320818169,
+      "grad_norm": 0.6377652057718626,
+      "learning_rate": 0.00019999971503008212,
+      "loss": 1.3123,
+      "step": 668
+    },
+    {
+      "epoch": 0.030819551296816697,
+      "grad_norm": 0.6585170461382758,
+      "learning_rate": 0.00019999967829569836,
+      "loss": 1.6919,
+      "step": 669
+    },
+    {
+      "epoch": 0.0308656193854517,
+      "grad_norm": 0.6830534030487112,
+      "learning_rate": 0.00019999963933499317,
+      "loss": 0.9656,
+      "step": 670
+    },
+    {
+      "epoch": 0.0309116874740867,
+      "grad_norm": 0.679713197683336,
+      "learning_rate": 0.0001999995981479675,
+      "loss": 1.1908,
+      "step": 671
+    },
+    {
+      "epoch": 0.030957755562721703,
+      "grad_norm": 0.7180364062019403,
+      "learning_rate": 0.00019999955473462227,
+      "loss": 1.2708,
+      "step": 672
+    },
+    {
+      "epoch": 0.031003823651356704,
+      "grad_norm": 0.576614286620956,
+      "learning_rate": 0.0001999995090949584,
+      "loss": 1.4507,
+      "step": 673
+    },
+    {
+      "epoch": 0.031049891739991706,
+      "grad_norm": 0.9022909626374622,
+      "learning_rate": 0.0001999994612289769,
+      "loss": 1.3268,
+      "step": 674
+    },
+    {
+      "epoch": 0.031095959828626712,
+      "grad_norm": 0.537755219289058,
+      "learning_rate": 0.0001999994111366789,
+      "loss": 1.4925,
+      "step": 675
+    },
+    {
+      "epoch": 0.031142027917261714,
+      "grad_norm": 0.6751983834039169,
+      "learning_rate": 0.00019999935881806542,
+      "loss": 1.4247,
+      "step": 676
+    },
+    {
+      "epoch": 0.031188096005896716,
+      "grad_norm": 0.7605896922780829,
+      "learning_rate": 0.0001999993042731377,
+      "loss": 1.0547,
+      "step": 677
+    },
+    {
+      "epoch": 0.031234164094531718,
+      "grad_norm": 0.7691995768126737,
+      "learning_rate": 0.00019999924750189693,
+      "loss": 1.4735,
+      "step": 678
+    },
+    {
+      "epoch": 0.03128023218316672,
+      "grad_norm": 0.6839397004434498,
+      "learning_rate": 0.0001999991885043444,
+      "loss": 1.0645,
+      "step": 679
+    },
+    {
+      "epoch": 0.03132630027180172,
+      "grad_norm": 0.7512844618207417,
+      "learning_rate": 0.00019999912728048137,
+      "loss": 1.2662,
+      "step": 680
+    },
+    {
+      "epoch": 0.03137236836043673,
+      "grad_norm": 0.6233731485919635,
+      "learning_rate": 0.00019999906383030925,
+      "loss": 1.3632,
+      "step": 681
+    },
+    {
+      "epoch": 0.031418436449071725,
+      "grad_norm": 0.6344269709698929,
+      "learning_rate": 0.00019999899815382942,
+      "loss": 1.5483,
+      "step": 682
+    },
+    {
+      "epoch": 0.03146450453770673,
+      "grad_norm": 0.972029352056632,
+      "learning_rate": 0.00019999893025104338,
+      "loss": 1.157,
+      "step": 683
+    },
+    {
+      "epoch": 0.031510572626341736,
+      "grad_norm": 0.619336062062577,
+      "learning_rate": 0.00019999886012195262,
+      "loss": 1.3899,
+      "step": 684
+    },
+    {
+      "epoch": 0.031556640714976734,
+      "grad_norm": 0.628995341206648,
+      "learning_rate": 0.0001999987877665587,
+      "loss": 1.5231,
+      "step": 685
+    },
+    {
+      "epoch": 0.03160270880361174,
+      "grad_norm": 0.6045591447016283,
+      "learning_rate": 0.0001999987131848632,
+      "loss": 1.2858,
+      "step": 686
+    },
+    {
+      "epoch": 0.03164877689224674,
+      "grad_norm": 0.6062926502835354,
+      "learning_rate": 0.00019999863637686782,
+      "loss": 1.2245,
+      "step": 687
+    },
+    {
+      "epoch": 0.031694844980881744,
+      "grad_norm": 0.6363456266579071,
+      "learning_rate": 0.00019999855734257428,
+      "loss": 1.4204,
+      "step": 688
+    },
+    {
+      "epoch": 0.03174091306951675,
+      "grad_norm": 0.49273973959145084,
+      "learning_rate": 0.00019999847608198433,
+      "loss": 1.3463,
+      "step": 689
+    },
+    {
+      "epoch": 0.03178698115815175,
+      "grad_norm": 0.586015028971623,
+      "learning_rate": 0.00019999839259509977,
+      "loss": 1.2408,
+      "step": 690
+    },
+    {
+      "epoch": 0.03183304924678675,
+      "grad_norm": 0.487779823659222,
+      "learning_rate": 0.0001999983068819224,
+      "loss": 1.3969,
+      "step": 691
+    },
+    {
+      "epoch": 0.03187911733542175,
+      "grad_norm": 0.8816794325237612,
+      "learning_rate": 0.00019999821894245426,
+      "loss": 1.2652,
+      "step": 692
+    },
+    {
+      "epoch": 0.03192518542405676,
+      "grad_norm": 0.5874465994710768,
+      "learning_rate": 0.0001999981287766972,
+      "loss": 1.3377,
+      "step": 693
+    },
+    {
+      "epoch": 0.031971253512691755,
+      "grad_norm": 0.5188167937208316,
+      "learning_rate": 0.0001999980363846533,
+      "loss": 0.8847,
+      "step": 694
+    },
+    {
+      "epoch": 0.03201732160132676,
+      "grad_norm": 0.6589478178212679,
+      "learning_rate": 0.00019999794176632453,
+      "loss": 1.4093,
+      "step": 695
+    },
+    {
+      "epoch": 0.032063389689961766,
+      "grad_norm": 0.522557270456699,
+      "learning_rate": 0.00019999784492171307,
+      "loss": 1.4919,
+      "step": 696
+    },
+    {
+      "epoch": 0.032109457778596764,
+      "grad_norm": 0.6725443795630767,
+      "learning_rate": 0.00019999774585082106,
+      "loss": 1.2566,
+      "step": 697
+    },
+    {
+      "epoch": 0.03215552586723177,
+      "grad_norm": 0.5849732752240484,
+      "learning_rate": 0.00019999764455365068,
+      "loss": 1.1073,
+      "step": 698
+    },
+    {
+      "epoch": 0.03220159395586677,
+      "grad_norm": 0.48866905501052404,
+      "learning_rate": 0.00019999754103020422,
+      "loss": 1.4789,
+      "step": 699
+    },
+    {
+      "epoch": 0.032247662044501774,
+      "grad_norm": 0.5525945564821426,
+      "learning_rate": 0.00019999743528048396,
+      "loss": 1.5065,
+      "step": 700
+    },
+    {
+      "epoch": 0.03229373013313678,
+      "grad_norm": 0.8924169080798068,
+      "learning_rate": 0.00019999732730449226,
+      "loss": 1.0873,
+      "step": 701
+    },
+    {
+      "epoch": 0.03233979822177178,
+      "grad_norm": 0.583562492911861,
+      "learning_rate": 0.00019999721710223153,
+      "loss": 1.5278,
+      "step": 702
+    },
+    {
+      "epoch": 0.03238586631040678,
+      "grad_norm": 0.5649358656832306,
+      "learning_rate": 0.0001999971046737042,
+      "loss": 1.4543,
+      "step": 703
+    },
+    {
+      "epoch": 0.03243193439904178,
+      "grad_norm": 1.1753394183685637,
+      "learning_rate": 0.0001999969900189128,
+      "loss": 1.3722,
+      "step": 704
+    },
+    {
+      "epoch": 0.03247800248767679,
+      "grad_norm": 0.5603789304775639,
+      "learning_rate": 0.00019999687313785992,
+      "loss": 0.9093,
+      "step": 705
+    },
+    {
+      "epoch": 0.03252407057631179,
+      "grad_norm": 0.4876250500177426,
+      "learning_rate": 0.00019999675403054807,
+      "loss": 1.0595,
+      "step": 706
+    },
+    {
+      "epoch": 0.03257013866494679,
+      "grad_norm": 0.6746886368517234,
+      "learning_rate": 0.00019999663269697996,
+      "loss": 1.4415,
+      "step": 707
+    },
+    {
+      "epoch": 0.032616206753581796,
+      "grad_norm": 0.625273923644332,
+      "learning_rate": 0.0001999965091371583,
+      "loss": 1.4908,
+      "step": 708
+    },
+    {
+      "epoch": 0.032662274842216794,
+      "grad_norm": 0.5780468671299953,
+      "learning_rate": 0.0001999963833510858,
+      "loss": 1.143,
+      "step": 709
+    },
+    {
+      "epoch": 0.0327083429308518,
+      "grad_norm": 1.6331092360957078,
+      "learning_rate": 0.00019999625533876527,
+      "loss": 1.2914,
+      "step": 710
+    },
+    {
+      "epoch": 0.0327544110194868,
+      "grad_norm": 0.5893461879109866,
+      "learning_rate": 0.0001999961251001996,
+      "loss": 1.2171,
+      "step": 711
+    },
+    {
+      "epoch": 0.032800479108121804,
+      "grad_norm": 0.613120143083469,
+      "learning_rate": 0.00019999599263539165,
+      "loss": 0.832,
+      "step": 712
+    },
+    {
+      "epoch": 0.03284654719675681,
+      "grad_norm": 0.6347411152124582,
+      "learning_rate": 0.00019999585794434438,
+      "loss": 1.3998,
+      "step": 713
+    },
+    {
+      "epoch": 0.03289261528539181,
+      "grad_norm": 0.6322396584374399,
+      "learning_rate": 0.00019999572102706077,
+      "loss": 1.267,
+      "step": 714
+    },
+    {
+      "epoch": 0.03293868337402681,
+      "grad_norm": 0.7976763142726667,
+      "learning_rate": 0.0001999955818835439,
+      "loss": 1.0641,
+      "step": 715
+    },
+    {
+      "epoch": 0.03298475146266181,
+      "grad_norm": 0.7001687178176297,
+      "learning_rate": 0.00019999544051379687,
+      "loss": 1.6854,
+      "step": 716
+    },
+    {
+      "epoch": 0.03303081955129682,
+      "grad_norm": 0.7402876184835178,
+      "learning_rate": 0.0001999952969178228,
+      "loss": 1.2174,
+      "step": 717
+    },
+    {
+      "epoch": 0.03307688763993182,
+      "grad_norm": 0.7774108129696701,
+      "learning_rate": 0.0001999951510956249,
+      "loss": 1.2028,
+      "step": 718
+    },
+    {
+      "epoch": 0.03312295572856682,
+      "grad_norm": 0.851706714954694,
+      "learning_rate": 0.0001999950030472064,
+      "loss": 1.1806,
+      "step": 719
+    },
+    {
+      "epoch": 0.033169023817201826,
+      "grad_norm": 0.5922954734648771,
+      "learning_rate": 0.00019999485277257063,
+      "loss": 1.4433,
+      "step": 720
+    },
+    {
+      "epoch": 0.033215091905836824,
+      "grad_norm": 0.6019992000288725,
+      "learning_rate": 0.00019999470027172092,
+      "loss": 1.4736,
+      "step": 721
+    },
+    {
+      "epoch": 0.03326115999447183,
+      "grad_norm": 0.7236679092630774,
+      "learning_rate": 0.00019999454554466066,
+      "loss": 1.2769,
+      "step": 722
+    },
+    {
+      "epoch": 0.033307228083106835,
+      "grad_norm": 0.771847645654837,
+      "learning_rate": 0.0001999943885913933,
+      "loss": 1.2642,
+      "step": 723
+    },
+    {
+      "epoch": 0.033353296171741834,
+      "grad_norm": 0.8631720042912306,
+      "learning_rate": 0.00019999422941192234,
+      "loss": 1.4897,
+      "step": 724
+    },
+    {
+      "epoch": 0.03339936426037684,
+      "grad_norm": 0.607524947694161,
+      "learning_rate": 0.00019999406800625132,
+      "loss": 0.8538,
+      "step": 725
+    },
+    {
+      "epoch": 0.03344543234901184,
+      "grad_norm": 0.6681233762874429,
+      "learning_rate": 0.0001999939043743838,
+      "loss": 1.6086,
+      "step": 726
+    },
+    {
+      "epoch": 0.03349150043764684,
+      "grad_norm": 0.7397616613344812,
+      "learning_rate": 0.00019999373851632346,
+      "loss": 1.256,
+      "step": 727
+    },
+    {
+      "epoch": 0.03353756852628184,
+      "grad_norm": 0.8015837444519698,
+      "learning_rate": 0.00019999357043207398,
+      "loss": 0.9359,
+      "step": 728
+    },
+    {
+      "epoch": 0.03358363661491685,
+      "grad_norm": 0.7467410184682428,
+      "learning_rate": 0.00019999340012163912,
+      "loss": 1.5214,
+      "step": 729
+    },
+    {
+      "epoch": 0.03362970470355185,
+      "grad_norm": 0.7169050157027841,
+      "learning_rate": 0.00019999322758502267,
+      "loss": 1.076,
+      "step": 730
+    },
+    {
+      "epoch": 0.03367577279218685,
+      "grad_norm": 1.027706315243458,
+      "learning_rate": 0.00019999305282222842,
+      "loss": 1.4685,
+      "step": 731
+    },
+    {
+      "epoch": 0.033721840880821856,
+      "grad_norm": 0.9156675718808592,
+      "learning_rate": 0.00019999287583326032,
+      "loss": 1.1666,
+      "step": 732
+    },
+    {
+      "epoch": 0.033767908969456854,
+      "grad_norm": 0.9684739412326399,
+      "learning_rate": 0.0001999926966181223,
+      "loss": 1.4559,
+      "step": 733
+    },
+    {
+      "epoch": 0.03381397705809186,
+      "grad_norm": 0.7365319804050164,
+      "learning_rate": 0.00019999251517681835,
+      "loss": 1.0138,
+      "step": 734
+    },
+    {
+      "epoch": 0.033860045146726865,
+      "grad_norm": 0.9185621052071146,
+      "learning_rate": 0.00019999233150935246,
+      "loss": 1.4441,
+      "step": 735
+    },
+    {
+      "epoch": 0.033906113235361864,
+      "grad_norm": 0.7580779094928168,
+      "learning_rate": 0.0001999921456157288,
+      "loss": 1.6798,
+      "step": 736
+    },
+    {
+      "epoch": 0.03395218132399687,
+      "grad_norm": 0.6632878987867218,
+      "learning_rate": 0.00019999195749595145,
+      "loss": 1.0934,
+      "step": 737
+    },
+    {
+      "epoch": 0.03399824941263187,
+      "grad_norm": 0.5336200374913568,
+      "learning_rate": 0.0001999917671500246,
+      "loss": 1.354,
+      "step": 738
+    },
+    {
+      "epoch": 0.03404431750126687,
+      "grad_norm": 0.795193704707052,
+      "learning_rate": 0.00019999157457795255,
+      "loss": 0.9122,
+      "step": 739
+    },
+    {
+      "epoch": 0.03409038558990188,
+      "grad_norm": 0.6310255437639123,
+      "learning_rate": 0.0001999913797797395,
+      "loss": 1.3001,
+      "step": 740
+    },
+    {
+      "epoch": 0.03413645367853688,
+      "grad_norm": 1.1799248885538585,
+      "learning_rate": 0.00019999118275538985,
+      "loss": 0.8611,
+      "step": 741
+    },
+    {
+      "epoch": 0.03418252176717188,
+      "grad_norm": 0.7513564270437892,
+      "learning_rate": 0.00019999098350490795,
+      "loss": 1.0822,
+      "step": 742
+    },
+    {
+      "epoch": 0.03422858985580688,
+      "grad_norm": 0.6140373530559594,
+      "learning_rate": 0.00019999078202829825,
+      "loss": 1.1022,
+      "step": 743
+    },
+    {
+      "epoch": 0.034274657944441886,
+      "grad_norm": 0.8292053989251373,
+      "learning_rate": 0.00019999057832556524,
+      "loss": 1.4266,
+      "step": 744
+    },
+    {
+      "epoch": 0.034320726033076884,
+      "grad_norm": 0.6513572660057803,
+      "learning_rate": 0.00019999037239671344,
+      "loss": 1.1932,
+      "step": 745
+    },
+    {
+      "epoch": 0.03436679412171189,
+      "grad_norm": 0.7986009849851512,
+      "learning_rate": 0.00019999016424174747,
+      "loss": 1.2964,
+      "step": 746
+    },
+    {
+      "epoch": 0.034412862210346895,
+      "grad_norm": 0.682828781079068,
+      "learning_rate": 0.0001999899538606719,
+      "loss": 1.5802,
+      "step": 747
+    },
+    {
+      "epoch": 0.034458930298981894,
+      "grad_norm": 0.6071765970821968,
+      "learning_rate": 0.0001999897412534915,
+      "loss": 0.9658,
+      "step": 748
+    },
+    {
+      "epoch": 0.0345049983876169,
+      "grad_norm": 0.6734155867264336,
+      "learning_rate": 0.00019998952642021092,
+      "loss": 1.4006,
+      "step": 749
+    },
+    {
+      "epoch": 0.0345510664762519,
+      "grad_norm": 0.5622431717759693,
+      "learning_rate": 0.000199989309360835,
+      "loss": 1.1053,
+      "step": 750
+    },
+    {
+      "epoch": 0.0345971345648869,
+      "grad_norm": 0.7905420683866211,
+      "learning_rate": 0.00019998909007536855,
+      "loss": 1.212,
+      "step": 751
+    },
+    {
+      "epoch": 0.03464320265352191,
+      "grad_norm": 0.662041528993047,
+      "learning_rate": 0.00019998886856381645,
+      "loss": 1.2618,
+      "step": 752
+    },
+    {
+      "epoch": 0.03468927074215691,
+      "grad_norm": 0.693965844607136,
+      "learning_rate": 0.00019998864482618362,
+      "loss": 1.0735,
+      "step": 753
+    },
+    {
+      "epoch": 0.03473533883079191,
+      "grad_norm": 0.5744448346463257,
+      "learning_rate": 0.00019998841886247507,
+      "loss": 1.1426,
+      "step": 754
+    },
+    {
+      "epoch": 0.03478140691942691,
+      "grad_norm": 0.6375887966693726,
+      "learning_rate": 0.00019998819067269582,
+      "loss": 1.4317,
+      "step": 755
+    },
+    {
+      "epoch": 0.034827475008061916,
+      "grad_norm": 0.47550361508861755,
+      "learning_rate": 0.00019998796025685094,
+      "loss": 1.4807,
+      "step": 756
+    },
+    {
+      "epoch": 0.03487354309669692,
+      "grad_norm": 0.7060925682733605,
+      "learning_rate": 0.00019998772761494558,
+      "loss": 1.1817,
+      "step": 757
+    },
+    {
+      "epoch": 0.03491961118533192,
+      "grad_norm": 0.4954245007732764,
+      "learning_rate": 0.00019998749274698492,
+      "loss": 1.3145,
+      "step": 758
+    },
+    {
+      "epoch": 0.034965679273966925,
+      "grad_norm": 0.5938223062538304,
+      "learning_rate": 0.00019998725565297413,
+      "loss": 1.2948,
+      "step": 759
+    },
+    {
+      "epoch": 0.035011747362601923,
+      "grad_norm": 0.7361951207894772,
+      "learning_rate": 0.00019998701633291857,
+      "loss": 1.3644,
+      "step": 760
+    },
+    {
+      "epoch": 0.03505781545123693,
+      "grad_norm": 0.5620264046029205,
+      "learning_rate": 0.00019998677478682352,
+      "loss": 0.9454,
+      "step": 761
+    },
+    {
+      "epoch": 0.03510388353987193,
+      "grad_norm": 0.6452558837403971,
+      "learning_rate": 0.0001999865310146944,
+      "loss": 1.4636,
+      "step": 762
+    },
+    {
+      "epoch": 0.03514995162850693,
+      "grad_norm": 1.008664266569099,
+      "learning_rate": 0.00019998628501653656,
+      "loss": 1.4261,
+      "step": 763
+    },
+    {
+      "epoch": 0.03519601971714194,
+      "grad_norm": 0.6757901744702265,
+      "learning_rate": 0.00019998603679235555,
+      "loss": 1.1455,
+      "step": 764
+    },
+    {
+      "epoch": 0.03524208780577694,
+      "grad_norm": 0.6770272050666675,
+      "learning_rate": 0.00019998578634215685,
+      "loss": 1.5836,
+      "step": 765
+    },
+    {
+      "epoch": 0.03528815589441194,
+      "grad_norm": 0.6433776502384918,
+      "learning_rate": 0.00019998553366594611,
+      "loss": 1.7044,
+      "step": 766
+    },
+    {
+      "epoch": 0.03533422398304694,
+      "grad_norm": 0.6550390971669064,
+      "learning_rate": 0.00019998527876372883,
+      "loss": 1.3525,
+      "step": 767
+    },
+    {
+      "epoch": 0.035380292071681946,
+      "grad_norm": 0.5491419059711385,
+      "learning_rate": 0.0001999850216355108,
+      "loss": 1.5315,
+      "step": 768
+    },
+    {
+      "epoch": 0.03542636016031695,
+      "grad_norm": 0.6999214282570793,
+      "learning_rate": 0.00019998476228129768,
+      "loss": 1.4273,
+      "step": 769
+    },
+    {
+      "epoch": 0.03547242824895195,
+      "grad_norm": 0.7293673285223797,
+      "learning_rate": 0.00019998450070109528,
+      "loss": 1.4524,
+      "step": 770
+    },
+    {
+      "epoch": 0.035518496337586955,
+      "grad_norm": 0.9823926127509955,
+      "learning_rate": 0.00019998423689490943,
+      "loss": 0.7932,
+      "step": 771
+    },
+    {
+      "epoch": 0.03556456442622195,
+      "grad_norm": 0.6297575127153403,
+      "learning_rate": 0.00019998397086274592,
+      "loss": 1.5032,
+      "step": 772
+    },
+    {
+      "epoch": 0.03561063251485696,
+      "grad_norm": 0.8937655824555283,
+      "learning_rate": 0.00019998370260461076,
+      "loss": 1.0178,
+      "step": 773
+    },
+    {
+      "epoch": 0.035656700603491964,
+      "grad_norm": 0.6598361675521991,
+      "learning_rate": 0.0001999834321205099,
+      "loss": 1.4459,
+      "step": 774
+    },
+    {
+      "epoch": 0.03570276869212696,
+      "grad_norm": 0.7670179002144609,
+      "learning_rate": 0.00019998315941044937,
+      "loss": 1.2421,
+      "step": 775
+    },
+    {
+      "epoch": 0.03574883678076197,
+      "grad_norm": 0.983101114462648,
+      "learning_rate": 0.00019998288447443523,
+      "loss": 1.2574,
+      "step": 776
+    },
+    {
+      "epoch": 0.035794904869396967,
+      "grad_norm": 9.832919647422212,
+      "learning_rate": 0.0001999826073124736,
+      "loss": 1.3229,
+      "step": 777
+    },
+    {
+      "epoch": 0.03584097295803197,
+      "grad_norm": 0.550698744497293,
+      "learning_rate": 0.00019998232792457063,
+      "loss": 0.8829,
+      "step": 778
+    },
+    {
+      "epoch": 0.03588704104666697,
+      "grad_norm": 0.6721372186849814,
+      "learning_rate": 0.00019998204631073262,
+      "loss": 1.0494,
+      "step": 779
+    },
+    {
+      "epoch": 0.035933109135301976,
+      "grad_norm": 0.4875883314603458,
+      "learning_rate": 0.00019998176247096571,
+      "loss": 1.1069,
+      "step": 780
+    },
+    {
+      "epoch": 0.03597917722393698,
+      "grad_norm": 0.7261714694204979,
+      "learning_rate": 0.00019998147640527632,
+      "loss": 1.109,
+      "step": 781
+    },
+    {
+      "epoch": 0.03602524531257198,
+      "grad_norm": 0.6272316541430056,
+      "learning_rate": 0.00019998118811367078,
+      "loss": 1.6238,
+      "step": 782
+    },
+    {
+      "epoch": 0.036071313401206985,
+      "grad_norm": 0.8066799548213075,
+      "learning_rate": 0.00019998089759615554,
+      "loss": 0.9591,
+      "step": 783
+    },
+    {
+      "epoch": 0.03611738148984198,
+      "grad_norm": 0.610154732523135,
+      "learning_rate": 0.00019998060485273704,
+      "loss": 1.0525,
+      "step": 784
+    },
+    {
+      "epoch": 0.03616344957847699,
+      "grad_norm": 0.5468115755334664,
+      "learning_rate": 0.00019998030988342179,
+      "loss": 1.4152,
+      "step": 785
+    },
+    {
+      "epoch": 0.036209517667111994,
+      "grad_norm": 0.6359844539171537,
+      "learning_rate": 0.00019998001268821638,
+      "loss": 1.4253,
+      "step": 786
+    },
+    {
+      "epoch": 0.03625558575574699,
+      "grad_norm": 0.8110809596074529,
+      "learning_rate": 0.0001999797132671274,
+      "loss": 0.9732,
+      "step": 787
+    },
+    {
+      "epoch": 0.036301653844382,
+      "grad_norm": 0.7440506347547245,
+      "learning_rate": 0.00019997941162016157,
+      "loss": 1.3647,
+      "step": 788
+    },
+    {
+      "epoch": 0.036347721933016996,
+      "grad_norm": 0.632918743186613,
+      "learning_rate": 0.00019997910774732552,
+      "loss": 1.4082,
+      "step": 789
+    },
+    {
+      "epoch": 0.036393790021652,
+      "grad_norm": 1.6847853276124363,
+      "learning_rate": 0.0001999788016486261,
+      "loss": 1.2134,
+      "step": 790
+    },
+    {
+      "epoch": 0.03643985811028701,
+      "grad_norm": 0.8053750445868314,
+      "learning_rate": 0.0001999784933240701,
+      "loss": 1.4028,
+      "step": 791
+    },
+    {
+      "epoch": 0.036485926198922006,
+      "grad_norm": 0.825370462823432,
+      "learning_rate": 0.0001999781827736643,
+      "loss": 1.4516,
+      "step": 792
+    },
+    {
+      "epoch": 0.03653199428755701,
+      "grad_norm": 0.7494624947609396,
+      "learning_rate": 0.00019997786999741574,
+      "loss": 1.4357,
+      "step": 793
+    },
+    {
+      "epoch": 0.03657806237619201,
+      "grad_norm": 0.6532399582937396,
+      "learning_rate": 0.00019997755499533132,
+      "loss": 1.3404,
+      "step": 794
+    },
+    {
+      "epoch": 0.036624130464827015,
+      "grad_norm": 0.7513408468258225,
+      "learning_rate": 0.00019997723776741809,
+      "loss": 1.6232,
+      "step": 795
+    },
+    {
+      "epoch": 0.03667019855346201,
+      "grad_norm": 0.7725331636252956,
+      "learning_rate": 0.00019997691831368303,
+      "loss": 0.8572,
+      "step": 796
+    },
+    {
+      "epoch": 0.03671626664209702,
+      "grad_norm": 0.7457384715148381,
+      "learning_rate": 0.00019997659663413339,
+      "loss": 1.2668,
+      "step": 797
+    },
+    {
+      "epoch": 0.036762334730732024,
+      "grad_norm": 0.9220809916967369,
+      "learning_rate": 0.00019997627272877618,
+      "loss": 1.2777,
+      "step": 798
+    },
+    {
+      "epoch": 0.03680840281936702,
+      "grad_norm": 0.8218232764587173,
+      "learning_rate": 0.0001999759465976187,
+      "loss": 1.2311,
+      "step": 799
+    },
+    {
+      "epoch": 0.03685447090800203,
+      "grad_norm": 0.6295768832826928,
+      "learning_rate": 0.00019997561824066818,
+      "loss": 1.3724,
+      "step": 800
+    },
+    {
+      "epoch": 0.036900538996637026,
+      "grad_norm": 0.7024199200587403,
+      "learning_rate": 0.00019997528765793195,
+      "loss": 1.5667,
+      "step": 801
+    },
+    {
+      "epoch": 0.03694660708527203,
+      "grad_norm": 0.7558503013048128,
+      "learning_rate": 0.00019997495484941737,
+      "loss": 1.3408,
+      "step": 802
+    },
+    {
+      "epoch": 0.03699267517390704,
+      "grad_norm": 0.7330979626272579,
+      "learning_rate": 0.00019997461981513182,
+      "loss": 1.609,
+      "step": 803
+    },
+    {
+      "epoch": 0.037038743262542036,
+      "grad_norm": 0.7281260612894225,
+      "learning_rate": 0.0001999742825550828,
+      "loss": 1.2497,
+      "step": 804
+    },
+    {
+      "epoch": 0.03708481135117704,
+      "grad_norm": 0.6096537516159937,
+      "learning_rate": 0.0001999739430692778,
+      "loss": 1.1065,
+      "step": 805
+    },
+    {
+      "epoch": 0.03713087943981204,
+      "grad_norm": 0.6939512174212273,
+      "learning_rate": 0.00019997360135772435,
+      "loss": 1.4042,
+      "step": 806
+    },
+    {
+      "epoch": 0.037176947528447045,
+      "grad_norm": 0.7793569468361096,
+      "learning_rate": 0.0001999732574204301,
+      "loss": 1.0818,
+      "step": 807
+    },
+    {
+      "epoch": 0.03722301561708205,
+      "grad_norm": 0.6804390191535439,
+      "learning_rate": 0.00019997291125740268,
+      "loss": 1.2772,
+      "step": 808
+    },
+    {
+      "epoch": 0.03726908370571705,
+      "grad_norm": 0.6347526957974149,
+      "learning_rate": 0.00019997256286864983,
+      "loss": 1.2608,
+      "step": 809
+    },
+    {
+      "epoch": 0.037315151794352054,
+      "grad_norm": 0.9265232259313526,
+      "learning_rate": 0.00019997221225417927,
+      "loss": 1.4218,
+      "step": 810
+    },
+    {
+      "epoch": 0.03736121988298705,
+      "grad_norm": 0.7860786100499514,
+      "learning_rate": 0.0001999718594139988,
+      "loss": 1.1293,
+      "step": 811
+    },
+    {
+      "epoch": 0.03740728797162206,
+      "grad_norm": 0.7382928888767862,
+      "learning_rate": 0.00019997150434811634,
+      "loss": 1.4893,
+      "step": 812
+    },
+    {
+      "epoch": 0.03745335606025706,
+      "grad_norm": 0.5582734954821489,
+      "learning_rate": 0.00019997114705653973,
+      "loss": 1.2191,
+      "step": 813
+    },
+    {
+      "epoch": 0.03749942414889206,
+      "grad_norm": 0.6474075007852846,
+      "learning_rate": 0.00019997078753927693,
+      "loss": 1.0599,
+      "step": 814
+    },
+    {
+      "epoch": 0.03754549223752707,
+      "grad_norm": 0.6828185101590261,
+      "learning_rate": 0.00019997042579633596,
+      "loss": 1.5283,
+      "step": 815
+    },
+    {
+      "epoch": 0.037591560326162066,
+      "grad_norm": 0.7447613786538658,
+      "learning_rate": 0.00019997006182772488,
+      "loss": 1.4077,
+      "step": 816
+    },
+    {
+      "epoch": 0.03763762841479707,
+      "grad_norm": 2.066233767933994,
+      "learning_rate": 0.0001999696956334518,
+      "loss": 1.0741,
+      "step": 817
+    },
+    {
+      "epoch": 0.03768369650343207,
+      "grad_norm": 0.6811955106664297,
+      "learning_rate": 0.00019996932721352484,
+      "loss": 1.6299,
+      "step": 818
+    },
+    {
+      "epoch": 0.037729764592067075,
+      "grad_norm": 0.9231482986960212,
+      "learning_rate": 0.00019996895656795226,
+      "loss": 0.9597,
+      "step": 819
+    },
+    {
+      "epoch": 0.03777583268070208,
+      "grad_norm": 0.6951977387872277,
+      "learning_rate": 0.00019996858369674222,
+      "loss": 1.1889,
+      "step": 820
+    },
+    {
+      "epoch": 0.03782190076933708,
+      "grad_norm": 0.827191480787346,
+      "learning_rate": 0.00019996820859990308,
+      "loss": 1.2709,
+      "step": 821
+    },
+    {
+      "epoch": 0.037867968857972084,
+      "grad_norm": 0.6416901819921167,
+      "learning_rate": 0.0001999678312774432,
+      "loss": 1.1396,
+      "step": 822
+    },
+    {
+      "epoch": 0.03791403694660708,
+      "grad_norm": 0.6844321952721149,
+      "learning_rate": 0.00019996745172937096,
+      "loss": 1.1717,
+      "step": 823
+    },
+    {
+      "epoch": 0.03796010503524209,
+      "grad_norm": 0.5511040258130717,
+      "learning_rate": 0.0001999670699556948,
+      "loss": 1.2664,
+      "step": 824
+    },
+    {
+      "epoch": 0.03800617312387709,
+      "grad_norm": 0.7952780735671331,
+      "learning_rate": 0.00019996668595642327,
+      "loss": 1.1358,
+      "step": 825
+    },
+    {
+      "epoch": 0.03805224121251209,
+      "grad_norm": 0.6119942703948295,
+      "learning_rate": 0.00019996629973156486,
+      "loss": 1.2908,
+      "step": 826
+    },
+    {
+      "epoch": 0.0380983093011471,
+      "grad_norm": 0.6845232244329493,
+      "learning_rate": 0.0001999659112811282,
+      "loss": 1.2198,
+      "step": 827
+    },
+    {
+      "epoch": 0.038144377389782096,
+      "grad_norm": 0.7571279776989157,
+      "learning_rate": 0.0001999655206051219,
+      "loss": 1.6072,
+      "step": 828
+    },
+    {
+      "epoch": 0.0381904454784171,
+      "grad_norm": 0.6094332646134862,
+      "learning_rate": 0.0001999651277035547,
+      "loss": 1.2714,
+      "step": 829
+    },
+    {
+      "epoch": 0.038236513567052106,
+      "grad_norm": 0.6882641522918492,
+      "learning_rate": 0.00019996473257643537,
+      "loss": 1.1686,
+      "step": 830
+    },
+    {
+      "epoch": 0.038282581655687105,
+      "grad_norm": 0.8166407469784844,
+      "learning_rate": 0.00019996433522377263,
+      "loss": 1.4533,
+      "step": 831
+    },
+    {
+      "epoch": 0.03832864974432211,
+      "grad_norm": 0.6982663992156749,
+      "learning_rate": 0.00019996393564557542,
+      "loss": 1.1358,
+      "step": 832
+    },
+    {
+      "epoch": 0.03837471783295711,
+      "grad_norm": 0.9404680975011672,
+      "learning_rate": 0.00019996353384185253,
+      "loss": 1.1944,
+      "step": 833
+    },
+    {
+      "epoch": 0.038420785921592114,
+      "grad_norm": 0.7113174500744389,
+      "learning_rate": 0.000199963129812613,
+      "loss": 1.2065,
+      "step": 834
+    },
+    {
+      "epoch": 0.03846685401022711,
+      "grad_norm": 0.6865618473490898,
+      "learning_rate": 0.00019996272355786576,
+      "loss": 1.2543,
+      "step": 835
+    },
+    {
+      "epoch": 0.03851292209886212,
+      "grad_norm": 0.8787730030286536,
+      "learning_rate": 0.0001999623150776199,
+      "loss": 1.3808,
+      "step": 836
+    },
+    {
+      "epoch": 0.03855899018749712,
+      "grad_norm": 0.5806099833551731,
+      "learning_rate": 0.00019996190437188447,
+      "loss": 1.1264,
+      "step": 837
+    },
+    {
+      "epoch": 0.03860505827613212,
+      "grad_norm": 0.7833889462208117,
+      "learning_rate": 0.00019996149144066861,
+      "loss": 1.2341,
+      "step": 838
+    },
+    {
+      "epoch": 0.03865112636476713,
+      "grad_norm": 0.5296681165035761,
+      "learning_rate": 0.0001999610762839816,
+      "loss": 1.1487,
+      "step": 839
+    },
+    {
+      "epoch": 0.038697194453402126,
+      "grad_norm": 0.5567708947781922,
+      "learning_rate": 0.0001999606589018326,
+      "loss": 1.0688,
+      "step": 840
+    },
+    {
+      "epoch": 0.03874326254203713,
+      "grad_norm": 0.48294003406941177,
+      "learning_rate": 0.00019996023929423095,
+      "loss": 1.3285,
+      "step": 841
+    },
+    {
+      "epoch": 0.038789330630672136,
+      "grad_norm": 0.6030326943412234,
+      "learning_rate": 0.00019995981746118592,
+      "loss": 1.267,
+      "step": 842
+    },
+    {
+      "epoch": 0.038835398719307135,
+      "grad_norm": 0.5589778232142922,
+      "learning_rate": 0.00019995939340270696,
+      "loss": 1.1137,
+      "step": 843
+    },
+    {
+      "epoch": 0.03888146680794214,
+      "grad_norm": 0.706320457386479,
+      "learning_rate": 0.00019995896711880352,
+      "loss": 1.4129,
+      "step": 844
+    },
+    {
+      "epoch": 0.03892753489657714,
+      "grad_norm": 0.616735077247593,
+      "learning_rate": 0.00019995853860948505,
+      "loss": 1.1464,
+      "step": 845
+    },
+    {
+      "epoch": 0.038973602985212144,
+      "grad_norm": 0.6172901755808243,
+      "learning_rate": 0.00019995810787476112,
+      "loss": 0.9994,
+      "step": 846
+    },
+    {
+      "epoch": 0.03901967107384715,
+      "grad_norm": 0.8025209141167776,
+      "learning_rate": 0.00019995767491464133,
+      "loss": 1.337,
+      "step": 847
+    },
+    {
+      "epoch": 0.03906573916248215,
+      "grad_norm": 0.6355881854433494,
+      "learning_rate": 0.00019995723972913526,
+      "loss": 1.185,
+      "step": 848
+    },
+    {
+      "epoch": 0.03911180725111715,
+      "grad_norm": 0.7337559645687916,
+      "learning_rate": 0.00019995680231825266,
+      "loss": 1.1091,
+      "step": 849
+    },
+    {
+      "epoch": 0.03915787533975215,
+      "grad_norm": 0.8764070130593173,
+      "learning_rate": 0.00019995636268200324,
+      "loss": 1.2063,
+      "step": 850
+    },
+    {
+      "epoch": 0.03920394342838716,
+      "grad_norm": 0.891546153639371,
+      "learning_rate": 0.0001999559208203968,
+      "loss": 1.4297,
+      "step": 851
+    },
+    {
+      "epoch": 0.039250011517022156,
+      "grad_norm": 0.6892503439885626,
+      "learning_rate": 0.00019995547673344315,
+      "loss": 1.169,
+      "step": 852
+    },
+    {
+      "epoch": 0.03929607960565716,
+      "grad_norm": 0.8343790265852353,
+      "learning_rate": 0.00019995503042115223,
+      "loss": 1.2316,
+      "step": 853
+    },
+    {
+      "epoch": 0.039342147694292166,
+      "grad_norm": 0.5555741955148837,
+      "learning_rate": 0.00019995458188353391,
+      "loss": 1.5199,
+      "step": 854
+    },
+    {
+      "epoch": 0.039388215782927165,
+      "grad_norm": 0.552794315571056,
+      "learning_rate": 0.00019995413112059823,
+      "loss": 1.5238,
+      "step": 855
+    },
+    {
+      "epoch": 0.03943428387156217,
+      "grad_norm": 0.6628098499160139,
+      "learning_rate": 0.0001999536781323552,
+      "loss": 1.0657,
+      "step": 856
+    },
+    {
+      "epoch": 0.03948035196019717,
+      "grad_norm": 0.6113426482329752,
+      "learning_rate": 0.00019995322291881492,
+      "loss": 1.1589,
+      "step": 857
+    },
+    {
+      "epoch": 0.039526420048832174,
+      "grad_norm": 0.737853252217511,
+      "learning_rate": 0.0001999527654799875,
+      "loss": 1.2358,
+      "step": 858
+    },
+    {
+      "epoch": 0.03957248813746718,
+      "grad_norm": 0.7708704869502153,
+      "learning_rate": 0.00019995230581588315,
+      "loss": 1.3103,
+      "step": 859
+    },
+    {
+      "epoch": 0.03961855622610218,
+      "grad_norm": 0.588876402511875,
+      "learning_rate": 0.00019995184392651208,
+      "loss": 1.2196,
+      "step": 860
+    },
+    {
+      "epoch": 0.03966462431473718,
+      "grad_norm": 0.6898873600761574,
+      "learning_rate": 0.0001999513798118846,
+      "loss": 0.6875,
+      "step": 861
+    },
+    {
+      "epoch": 0.03971069240337218,
+      "grad_norm": 0.5978098523386988,
+      "learning_rate": 0.000199950913472011,
+      "loss": 1.2766,
+      "step": 862
+    },
+    {
+      "epoch": 0.03975676049200719,
+      "grad_norm": 0.5112648218838269,
+      "learning_rate": 0.00019995044490690175,
+      "loss": 1.4154,
+      "step": 863
+    },
+    {
+      "epoch": 0.03980282858064219,
+      "grad_norm": 0.5349620048569007,
+      "learning_rate": 0.00019994997411656716,
+      "loss": 1.1626,
+      "step": 864
+    },
+    {
+      "epoch": 0.03984889666927719,
+      "grad_norm": 0.5900252143451371,
+      "learning_rate": 0.0001999495011010178,
+      "loss": 1.5115,
+      "step": 865
+    },
+    {
+      "epoch": 0.039894964757912196,
+      "grad_norm": 0.6212284907559034,
+      "learning_rate": 0.00019994902586026417,
+      "loss": 1.2121,
+      "step": 866
+    },
+    {
+      "epoch": 0.039941032846547195,
+      "grad_norm": 0.5839212850312744,
+      "learning_rate": 0.00019994854839431686,
+      "loss": 1.0049,
+      "step": 867
+    },
+    {
+      "epoch": 0.0399871009351822,
+      "grad_norm": 6.689056170957794,
+      "learning_rate": 0.0001999480687031865,
+      "loss": 1.5668,
+      "step": 868
+    },
+    {
+      "epoch": 0.0400331690238172,
+      "grad_norm": 21.602430792130164,
+      "learning_rate": 0.00019994758678688377,
+      "loss": 2.8208,
+      "step": 869
+    },
+    {
+      "epoch": 0.040079237112452204,
+      "grad_norm": 1.2852759771107944,
+      "learning_rate": 0.0001999471026454194,
+      "loss": 1.1723,
+      "step": 870
+    },
+    {
+      "epoch": 0.04012530520108721,
+      "grad_norm": 0.8017930694270718,
+      "learning_rate": 0.0001999466162788041,
+      "loss": 1.7066,
+      "step": 871
+    },
+    {
+      "epoch": 0.04017137328972221,
+      "grad_norm": 0.8086808091161106,
+      "learning_rate": 0.00019994612768704881,
+      "loss": 1.5776,
+      "step": 872
+    },
+    {
+      "epoch": 0.04021744137835721,
+      "grad_norm": 0.7754224310198025,
+      "learning_rate": 0.00019994563687016434,
+      "loss": 1.7572,
+      "step": 873
+    },
+    {
+      "epoch": 0.04026350946699221,
+      "grad_norm": 2.5410971589819247,
+      "learning_rate": 0.00019994514382816166,
+      "loss": 1.2534,
+      "step": 874
+    },
+    {
+      "epoch": 0.04030957755562722,
+      "grad_norm": 4.494698047800217,
+      "learning_rate": 0.0001999446485610517,
+      "loss": 1.3271,
+      "step": 875
+    },
+    {
+      "epoch": 0.04035564564426222,
+      "grad_norm": 3.3621084872865823,
+      "learning_rate": 0.00019994415106884553,
+      "loss": 1.4214,
+      "step": 876
+    },
+    {
+      "epoch": 0.04040171373289722,
+      "grad_norm": 1.7633110617805348,
+      "learning_rate": 0.0001999436513515542,
+      "loss": 1.8694,
+      "step": 877
+    },
+    {
+      "epoch": 0.040447781821532226,
+      "grad_norm": 1.3535288437362292,
+      "learning_rate": 0.0001999431494091888,
+      "loss": 1.2437,
+      "step": 878
+    },
+    {
+      "epoch": 0.040493849910167225,
+      "grad_norm": 1.0557024833617352,
+      "learning_rate": 0.0001999426452417606,
+      "loss": 1.2683,
+      "step": 879
+    },
+    {
+      "epoch": 0.04053991799880223,
+      "grad_norm": 0.9903560045807442,
+      "learning_rate": 0.00019994213884928072,
+      "loss": 1.6412,
+      "step": 880
+    },
+    {
+      "epoch": 0.040585986087437236,
+      "grad_norm": 0.9523705068978497,
+      "learning_rate": 0.00019994163023176053,
+      "loss": 1.293,
+      "step": 881
+    },
+    {
+      "epoch": 0.040632054176072234,
+      "grad_norm": 0.7617882119106388,
+      "learning_rate": 0.00019994111938921125,
+      "loss": 1.4002,
+      "step": 882
+    },
+    {
+      "epoch": 0.04067812226470724,
+      "grad_norm": 0.9927094488868946,
+      "learning_rate": 0.00019994060632164435,
+      "loss": 1.0398,
+      "step": 883
+    },
+    {
+      "epoch": 0.04072419035334224,
+      "grad_norm": 1.544314057939408,
+      "learning_rate": 0.0001999400910290712,
+      "loss": 1.4309,
+      "step": 884
+    },
+    {
+      "epoch": 0.04077025844197724,
+      "grad_norm": 1.0269355682373524,
+      "learning_rate": 0.0001999395735115033,
+      "loss": 1.2147,
+      "step": 885
+    },
+    {
+      "epoch": 0.04081632653061224,
+      "grad_norm": 0.9818192272550752,
+      "learning_rate": 0.00019993905376895215,
+      "loss": 1.5189,
+      "step": 886
+    },
+    {
+      "epoch": 0.04086239461924725,
+      "grad_norm": 0.742314887199572,
+      "learning_rate": 0.0001999385318014293,
+      "loss": 1.6677,
+      "step": 887
+    },
+    {
+      "epoch": 0.04090846270788225,
+      "grad_norm": 0.717459092286135,
+      "learning_rate": 0.00019993800760894641,
+      "loss": 1.3217,
+      "step": 888
+    },
+    {
+      "epoch": 0.04095453079651725,
+      "grad_norm": 0.8525536031904805,
+      "learning_rate": 0.00019993748119151515,
+      "loss": 1.4812,
+      "step": 889
+    },
+    {
+      "epoch": 0.041000598885152256,
+      "grad_norm": 0.8774728129869073,
+      "learning_rate": 0.0001999369525491472,
+      "loss": 1.2452,
+      "step": 890
+    },
+    {
+      "epoch": 0.041046666973787255,
+      "grad_norm": 1.3837825880676224,
+      "learning_rate": 0.0001999364216818544,
+      "loss": 1.6745,
+      "step": 891
+    },
+    {
+      "epoch": 0.04109273506242226,
+      "grad_norm": 0.8583390333973715,
+      "learning_rate": 0.0001999358885896485,
+      "loss": 1.2053,
+      "step": 892
+    },
+    {
+      "epoch": 0.041138803151057265,
+      "grad_norm": 0.6973587456541566,
+      "learning_rate": 0.0001999353532725414,
+      "loss": 1.2315,
+      "step": 893
+    },
+    {
+      "epoch": 0.041184871239692264,
+      "grad_norm": 1.2561519168275228,
+      "learning_rate": 0.000199934815730545,
+      "loss": 1.502,
+      "step": 894
+    },
+    {
+      "epoch": 0.04123093932832727,
+      "grad_norm": 0.9616590107568807,
+      "learning_rate": 0.00019993427596367127,
+      "loss": 1.5342,
+      "step": 895
+    },
+    {
+      "epoch": 0.04127700741696227,
+      "grad_norm": 0.752458413287621,
+      "learning_rate": 0.00019993373397193226,
+      "loss": 0.9417,
+      "step": 896
+    },
+    {
+      "epoch": 0.04132307550559727,
+      "grad_norm": 0.9338486957563036,
+      "learning_rate": 0.00019993318975533998,
+      "loss": 1.2386,
+      "step": 897
+    },
+    {
+      "epoch": 0.04136914359423228,
+      "grad_norm": 0.7780546397527475,
+      "learning_rate": 0.00019993264331390664,
+      "loss": 1.73,
+      "step": 898
+    },
+    {
+      "epoch": 0.04141521168286728,
+      "grad_norm": 1.2100430913643594,
+      "learning_rate": 0.00019993209464764432,
+      "loss": 1.3684,
+      "step": 899
+    },
+    {
+      "epoch": 0.04146127977150228,
+      "grad_norm": 0.6862428369258369,
+      "learning_rate": 0.00019993154375656523,
+      "loss": 1.4097,
+      "step": 900
+    },
+    {
+      "epoch": 0.04150734786013728,
+      "grad_norm": 0.9144141234008903,
+      "learning_rate": 0.00019993099064068167,
+      "loss": 1.5148,
+      "step": 901
+    },
+    {
+      "epoch": 0.041553415948772286,
+      "grad_norm": 0.6493682700598005,
+      "learning_rate": 0.00019993043530000597,
+      "loss": 1.0542,
+      "step": 902
+    },
+    {
+      "epoch": 0.041599484037407285,
+      "grad_norm": 0.7280443130510269,
+      "learning_rate": 0.00019992987773455046,
+      "loss": 1.3155,
+      "step": 903
+    },
+    {
+      "epoch": 0.04164555212604229,
+      "grad_norm": 0.7099836254342033,
+      "learning_rate": 0.00019992931794432755,
+      "loss": 1.1706,
+      "step": 904
+    },
+    {
+      "epoch": 0.041691620214677295,
+      "grad_norm": 0.6747806807086286,
+      "learning_rate": 0.00019992875592934976,
+      "loss": 1.5038,
+      "step": 905
+    },
+    {
+      "epoch": 0.041737688303312294,
+      "grad_norm": 0.6211794233865656,
+      "learning_rate": 0.00019992819168962952,
+      "loss": 1.2342,
+      "step": 906
+    },
+    {
+      "epoch": 0.0417837563919473,
+      "grad_norm": 0.9500185950208052,
+      "learning_rate": 0.00019992762522517946,
+      "loss": 1.1002,
+      "step": 907
+    },
+    {
+      "epoch": 0.0418298244805823,
+      "grad_norm": 0.869051955053573,
+      "learning_rate": 0.00019992705653601217,
+      "loss": 1.2641,
+      "step": 908
+    },
+    {
+      "epoch": 0.0418758925692173,
+      "grad_norm": 0.682942156536145,
+      "learning_rate": 0.00019992648562214025,
+      "loss": 1.1427,
+      "step": 909
+    },
+    {
+      "epoch": 0.04192196065785231,
+      "grad_norm": 0.804667159917586,
+      "learning_rate": 0.00019992591248357654,
+      "loss": 1.3868,
+      "step": 910
+    },
+    {
+      "epoch": 0.04196802874648731,
+      "grad_norm": 0.6794615918215001,
+      "learning_rate": 0.00019992533712033365,
+      "loss": 1.464,
+      "step": 911
+    },
+    {
+      "epoch": 0.04201409683512231,
+      "grad_norm": 0.6854301552754094,
+      "learning_rate": 0.0001999247595324245,
+      "loss": 1.7221,
+      "step": 912
+    },
+    {
+      "epoch": 0.04206016492375731,
+      "grad_norm": 0.5393944649149385,
+      "learning_rate": 0.00019992417971986194,
+      "loss": 1.2243,
+      "step": 913
+    },
+    {
+      "epoch": 0.042106233012392316,
+      "grad_norm": 0.5820414254592631,
+      "learning_rate": 0.0001999235976826588,
+      "loss": 1.12,
+      "step": 914
+    },
+    {
+      "epoch": 0.04215230110102732,
+      "grad_norm": 0.8000282037410144,
+      "learning_rate": 0.00019992301342082813,
+      "loss": 1.3584,
+      "step": 915
+    },
+    {
+      "epoch": 0.04219836918966232,
+      "grad_norm": 0.6100675660676814,
+      "learning_rate": 0.00019992242693438287,
+      "loss": 1.2094,
+      "step": 916
+    },
+    {
+      "epoch": 0.042244437278297325,
+      "grad_norm": 0.8020490481743079,
+      "learning_rate": 0.0001999218382233361,
+      "loss": 1.2586,
+      "step": 917
+    },
+    {
+      "epoch": 0.042290505366932324,
+      "grad_norm": 0.9824298503367929,
+      "learning_rate": 0.00019992124728770096,
+      "loss": 1.5837,
+      "step": 918
+    },
+    {
+      "epoch": 0.04233657345556733,
+      "grad_norm": 0.7365143792606387,
+      "learning_rate": 0.00019992065412749056,
+      "loss": 1.4739,
+      "step": 919
+    },
+    {
+      "epoch": 0.04238264154420233,
+      "grad_norm": 0.616656600039256,
+      "learning_rate": 0.00019992005874271812,
+      "loss": 1.0333,
+      "step": 920
+    },
+    {
+      "epoch": 0.04242870963283733,
+      "grad_norm": 0.9038758325846982,
+      "learning_rate": 0.0001999194611333969,
+      "loss": 1.1742,
+      "step": 921
+    },
+    {
+      "epoch": 0.04247477772147234,
+      "grad_norm": 0.5638428543944468,
+      "learning_rate": 0.00019991886129954023,
+      "loss": 1.4219,
+      "step": 922
+    },
+    {
+      "epoch": 0.04252084581010734,
+      "grad_norm": 0.713550018365798,
+      "learning_rate": 0.00019991825924116143,
+      "loss": 1.262,
+      "step": 923
+    },
+    {
+      "epoch": 0.04256691389874234,
+      "grad_norm": 0.8470602581764024,
+      "learning_rate": 0.00019991765495827387,
+      "loss": 1.1358,
+      "step": 924
+    },
+    {
+      "epoch": 0.04261298198737734,
+      "grad_norm": 0.536024368599159,
+      "learning_rate": 0.00019991704845089106,
+      "loss": 1.2041,
+      "step": 925
+    },
+    {
+      "epoch": 0.042659050076012346,
+      "grad_norm": 0.6070506667972803,
+      "learning_rate": 0.0001999164397190265,
+      "loss": 1.2587,
+      "step": 926
+    },
+    {
+      "epoch": 0.04270511816464735,
+      "grad_norm": 0.706548932653231,
+      "learning_rate": 0.00019991582876269375,
+      "loss": 1.3827,
+      "step": 927
+    },
+    {
+      "epoch": 0.04275118625328235,
+      "grad_norm": 0.6818257304550918,
+      "learning_rate": 0.00019991521558190638,
+      "loss": 0.786,
+      "step": 928
+    },
+    {
+      "epoch": 0.042797254341917355,
+      "grad_norm": 0.5339431877879769,
+      "learning_rate": 0.00019991460017667803,
+      "loss": 1.1559,
+      "step": 929
+    },
+    {
+      "epoch": 0.042843322430552354,
+      "grad_norm": 0.8999298739791001,
+      "learning_rate": 0.00019991398254702244,
+      "loss": 1.0938,
+      "step": 930
+    },
+    {
+      "epoch": 0.04288939051918736,
+      "grad_norm": 0.5396113183780239,
+      "learning_rate": 0.00019991336269295333,
+      "loss": 1.376,
+      "step": 931
+    },
+    {
+      "epoch": 0.042935458607822365,
+      "grad_norm": 0.5396283106466846,
+      "learning_rate": 0.00019991274061448454,
+      "loss": 1.1513,
+      "step": 932
+    },
+    {
+      "epoch": 0.04298152669645736,
+      "grad_norm": 0.5558206126064201,
+      "learning_rate": 0.0001999121163116299,
+      "loss": 1.4766,
+      "step": 933
+    },
+    {
+      "epoch": 0.04302759478509237,
+      "grad_norm": 0.5855332052948019,
+      "learning_rate": 0.00019991148978440327,
+      "loss": 1.0055,
+      "step": 934
+    },
+    {
+      "epoch": 0.04307366287372737,
+      "grad_norm": 0.5673351479637195,
+      "learning_rate": 0.00019991086103281864,
+      "loss": 0.8871,
+      "step": 935
+    },
+    {
+      "epoch": 0.04311973096236237,
+      "grad_norm": 0.6590266544149581,
+      "learning_rate": 0.00019991023005689,
+      "loss": 1.104,
+      "step": 936
+    },
+    {
+      "epoch": 0.04316579905099737,
+      "grad_norm": 0.68125590554682,
+      "learning_rate": 0.0001999095968566314,
+      "loss": 1.4884,
+      "step": 937
+    },
+    {
+      "epoch": 0.043211867139632376,
+      "grad_norm": 0.6681685576631653,
+      "learning_rate": 0.00019990896143205693,
+      "loss": 1.3862,
+      "step": 938
+    },
+    {
+      "epoch": 0.04325793522826738,
+      "grad_norm": 1.1631510913533707,
+      "learning_rate": 0.00019990832378318074,
+      "loss": 1.0876,
+      "step": 939
+    },
+    {
+      "epoch": 0.04330400331690238,
+      "grad_norm": 0.7919120045754611,
+      "learning_rate": 0.00019990768391001703,
+      "loss": 1.3203,
+      "step": 940
+    },
+    {
+      "epoch": 0.043350071405537385,
+      "grad_norm": 0.8080599964366804,
+      "learning_rate": 0.00019990704181258005,
+      "loss": 1.3584,
+      "step": 941
+    },
+    {
+      "epoch": 0.043396139494172384,
+      "grad_norm": 0.5626600222741492,
+      "learning_rate": 0.00019990639749088409,
+      "loss": 1.1858,
+      "step": 942
+    },
+    {
+      "epoch": 0.04344220758280739,
+      "grad_norm": 0.5887427216474369,
+      "learning_rate": 0.0001999057509449435,
+      "loss": 1.2287,
+      "step": 943
+    },
+    {
+      "epoch": 0.043488275671442395,
+      "grad_norm": 0.649785748532214,
+      "learning_rate": 0.00019990510217477266,
+      "loss": 1.2638,
+      "step": 944
+    },
+    {
+      "epoch": 0.04353434376007739,
+      "grad_norm": 0.7186889478185815,
+      "learning_rate": 0.00019990445118038598,
+      "loss": 1.1217,
+      "step": 945
+    },
+    {
+      "epoch": 0.0435804118487124,
+      "grad_norm": 1.1623506001152983,
+      "learning_rate": 0.00019990379796179804,
+      "loss": 1.1208,
+      "step": 946
+    },
+    {
+      "epoch": 0.0436264799373474,
+      "grad_norm": 0.6900923123093545,
+      "learning_rate": 0.00019990314251902333,
+      "loss": 1.2255,
+      "step": 947
+    },
+    {
+      "epoch": 0.0436725480259824,
+      "grad_norm": 0.6517472108957243,
+      "learning_rate": 0.00019990248485207644,
+      "loss": 1.1934,
+      "step": 948
+    },
+    {
+      "epoch": 0.04371861611461741,
+      "grad_norm": 1.0421230690117615,
+      "learning_rate": 0.000199901824960972,
+      "loss": 1.0491,
+      "step": 949
+    },
+    {
+      "epoch": 0.043764684203252406,
+      "grad_norm": 0.7528581649288901,
+      "learning_rate": 0.00019990116284572475,
+      "loss": 1.2419,
+      "step": 950
+    },
+    {
+      "epoch": 0.04381075229188741,
+      "grad_norm": 0.5894864532258188,
+      "learning_rate": 0.00019990049850634943,
+      "loss": 1.5062,
+      "step": 951
+    },
+    {
+      "epoch": 0.04385682038052241,
+      "grad_norm": 0.5829778802593605,
+      "learning_rate": 0.00019989983194286076,
+      "loss": 1.4378,
+      "step": 952
+    },
+    {
+      "epoch": 0.043902888469157415,
+      "grad_norm": 0.6069593095523182,
+      "learning_rate": 0.00019989916315527363,
+      "loss": 1.2473,
+      "step": 953
+    },
+    {
+      "epoch": 0.043948956557792414,
+      "grad_norm": 0.8423931999085206,
+      "learning_rate": 0.00019989849214360292,
+      "loss": 1.3495,
+      "step": 954
+    },
+    {
+      "epoch": 0.04399502464642742,
+      "grad_norm": 0.5821854564588982,
+      "learning_rate": 0.0001998978189078636,
+      "loss": 1.4772,
+      "step": 955
+    },
+    {
+      "epoch": 0.044041092735062425,
+      "grad_norm": 0.6106529485493869,
+      "learning_rate": 0.00019989714344807057,
+      "loss": 1.5692,
+      "step": 956
+    },
+    {
+      "epoch": 0.04408716082369742,
+      "grad_norm": 0.5446023403564123,
+      "learning_rate": 0.000199896465764239,
+      "loss": 1.5399,
+      "step": 957
+    },
+    {
+      "epoch": 0.04413322891233243,
+      "grad_norm": 0.6106715870399914,
+      "learning_rate": 0.00019989578585638386,
+      "loss": 1.4008,
+      "step": 958
+    },
+    {
+      "epoch": 0.04417929700096743,
+      "grad_norm": 0.594463554406938,
+      "learning_rate": 0.00019989510372452033,
+      "loss": 1.3292,
+      "step": 959
+    },
+    {
+      "epoch": 0.04422536508960243,
+      "grad_norm": 0.5502990772827704,
+      "learning_rate": 0.00019989441936866365,
+      "loss": 1.0499,
+      "step": 960
+    },
+    {
+      "epoch": 0.04427143317823744,
+      "grad_norm": 0.7278469764533988,
+      "learning_rate": 0.00019989373278882897,
+      "loss": 1.5443,
+      "step": 961
+    },
+    {
+      "epoch": 0.044317501266872436,
+      "grad_norm": 0.5323888373739546,
+      "learning_rate": 0.00019989304398503158,
+      "loss": 1.6249,
+      "step": 962
+    },
+    {
+      "epoch": 0.04436356935550744,
+      "grad_norm": 0.7159045681172708,
+      "learning_rate": 0.00019989235295728687,
+      "loss": 1.3031,
+      "step": 963
+    },
+    {
+      "epoch": 0.04440963744414244,
+      "grad_norm": 0.6237541622245248,
+      "learning_rate": 0.00019989165970561024,
+      "loss": 0.8793,
+      "step": 964
+    },
+    {
+      "epoch": 0.044455705532777445,
+      "grad_norm": 0.6157132519803996,
+      "learning_rate": 0.00019989096423001702,
+      "loss": 1.5868,
+      "step": 965
+    },
+    {
+      "epoch": 0.04450177362141245,
+      "grad_norm": 0.4748844338804353,
+      "learning_rate": 0.00019989026653052282,
+      "loss": 1.2268,
+      "step": 966
+    },
+    {
+      "epoch": 0.04454784171004745,
+      "grad_norm": 0.6306497915212245,
+      "learning_rate": 0.00019988956660714306,
+      "loss": 1.38,
+      "step": 967
+    },
+    {
+      "epoch": 0.044593909798682455,
+      "grad_norm": 0.536347627918671,
+      "learning_rate": 0.00019988886445989338,
+      "loss": 1.0087,
+      "step": 968
+    },
+    {
+      "epoch": 0.04463997788731745,
+      "grad_norm": 0.6965884733362043,
+      "learning_rate": 0.0001998881600887894,
+      "loss": 1.2493,
+      "step": 969
+    },
+    {
+      "epoch": 0.04468604597595246,
+      "grad_norm": 0.49723860168859996,
+      "learning_rate": 0.00019988745349384685,
+      "loss": 1.1507,
+      "step": 970
+    },
+    {
+      "epoch": 0.04473211406458746,
+      "grad_norm": 0.7145686613460618,
+      "learning_rate": 0.0001998867446750814,
+      "loss": 1.4192,
+      "step": 971
+    },
+    {
+      "epoch": 0.04477818215322246,
+      "grad_norm": 0.6098593191399865,
+      "learning_rate": 0.00019988603363250884,
+      "loss": 1.1426,
+      "step": 972
+    },
+    {
+      "epoch": 0.04482425024185747,
+      "grad_norm": 0.6245282110479744,
+      "learning_rate": 0.000199885320366145,
+      "loss": 1.3869,
+      "step": 973
+    },
+    {
+      "epoch": 0.044870318330492466,
+      "grad_norm": 0.7739541013483536,
+      "learning_rate": 0.00019988460487600578,
+      "loss": 1.5319,
+      "step": 974
+    },
+    {
+      "epoch": 0.04491638641912747,
+      "grad_norm": 0.5435005864249199,
+      "learning_rate": 0.0001998838871621071,
+      "loss": 1.4515,
+      "step": 975
+    },
+    {
+      "epoch": 0.04496245450776247,
+      "grad_norm": 0.5940973537873893,
+      "learning_rate": 0.00019988316722446494,
+      "loss": 1.8218,
+      "step": 976
+    },
+    {
+      "epoch": 0.045008522596397475,
+      "grad_norm": 0.5983772939799272,
+      "learning_rate": 0.00019988244506309531,
+      "loss": 1.3472,
+      "step": 977
+    },
+    {
+      "epoch": 0.04505459068503248,
+      "grad_norm": 0.5278409086420037,
+      "learning_rate": 0.00019988172067801431,
+      "loss": 1.1108,
+      "step": 978
+    },
+    {
+      "epoch": 0.04510065877366748,
+      "grad_norm": 0.7932339361726242,
+      "learning_rate": 0.00019988099406923805,
+      "loss": 1.1013,
+      "step": 979
+    },
+    {
+      "epoch": 0.045146726862302484,
+      "grad_norm": 0.8870474542730035,
+      "learning_rate": 0.00019988026523678275,
+      "loss": 1.3292,
+      "step": 980
+    },
+    {
+      "epoch": 0.04519279495093748,
+      "grad_norm": 0.6168689947610755,
+      "learning_rate": 0.00019987953418066457,
+      "loss": 1.3146,
+      "step": 981
+    },
+    {
+      "epoch": 0.04523886303957249,
+      "grad_norm": 0.8469615565779184,
+      "learning_rate": 0.00019987880090089984,
+      "loss": 1.3654,
+      "step": 982
+    },
+    {
+      "epoch": 0.045284931128207494,
+      "grad_norm": 0.4329655819626444,
+      "learning_rate": 0.00019987806539750483,
+      "loss": 0.9436,
+      "step": 983
+    },
+    {
+      "epoch": 0.04533099921684249,
+      "grad_norm": 0.6505974172738337,
+      "learning_rate": 0.00019987732767049596,
+      "loss": 1.1545,
+      "step": 984
+    },
+    {
+      "epoch": 0.0453770673054775,
+      "grad_norm": 0.9194169728929895,
+      "learning_rate": 0.00019987658771988964,
+      "loss": 1.0681,
+      "step": 985
+    },
+    {
+      "epoch": 0.045423135394112496,
+      "grad_norm": 0.5698718652379011,
+      "learning_rate": 0.00019987584554570238,
+      "loss": 1.2777,
+      "step": 986
+    },
+    {
+      "epoch": 0.0454692034827475,
+      "grad_norm": 0.4423280154781949,
+      "learning_rate": 0.00019987510114795063,
+      "loss": 1.1383,
+      "step": 987
+    },
+    {
+      "epoch": 0.04551527157138251,
+      "grad_norm": 0.5389242588373032,
+      "learning_rate": 0.00019987435452665101,
+      "loss": 0.9907,
+      "step": 988
+    },
+    {
+      "epoch": 0.045561339660017505,
+      "grad_norm": 0.8373945964412691,
+      "learning_rate": 0.00019987360568182014,
+      "loss": 1.2258,
+      "step": 989
+    },
+    {
+      "epoch": 0.04560740774865251,
+      "grad_norm": 0.615566577030046,
+      "learning_rate": 0.00019987285461347471,
+      "loss": 1.5548,
+      "step": 990
+    },
+    {
+      "epoch": 0.04565347583728751,
+      "grad_norm": 0.7328177075784345,
+      "learning_rate": 0.0001998721013216314,
+      "loss": 1.3032,
+      "step": 991
+    },
+    {
+      "epoch": 0.045699543925922514,
+      "grad_norm": 0.6354387223727682,
+      "learning_rate": 0.00019987134580630697,
+      "loss": 1.1246,
+      "step": 992
+    },
+    {
+      "epoch": 0.04574561201455751,
+      "grad_norm": 0.656739850221389,
+      "learning_rate": 0.00019987058806751832,
+      "loss": 1.2122,
+      "step": 993
+    },
+    {
+      "epoch": 0.04579168010319252,
+      "grad_norm": 1.2584050987458166,
+      "learning_rate": 0.00019986982810528222,
+      "loss": 1.0505,
+      "step": 994
+    },
+    {
+      "epoch": 0.045837748191827524,
+      "grad_norm": 0.5618283446025399,
+      "learning_rate": 0.00019986906591961567,
+      "loss": 1.2564,
+      "step": 995
+    },
+    {
+      "epoch": 0.04588381628046252,
+      "grad_norm": 0.7308549760342561,
+      "learning_rate": 0.0001998683015105356,
+      "loss": 1.5165,
+      "step": 996
+    },
+    {
+      "epoch": 0.04592988436909753,
+      "grad_norm": 0.7904105524675977,
+      "learning_rate": 0.000199867534878059,
+      "loss": 1.469,
+      "step": 997
+    },
+    {
+      "epoch": 0.045975952457732526,
+      "grad_norm": 0.659263297916959,
+      "learning_rate": 0.000199866766022203,
+      "loss": 1.3616,
+      "step": 998
+    },
+    {
+      "epoch": 0.04602202054636753,
+      "grad_norm": 0.5477107315309211,
+      "learning_rate": 0.0001998659949429847,
+      "loss": 1.4988,
+      "step": 999
+    },
+    {
+      "epoch": 0.04606808863500254,
+      "grad_norm": 0.8749849885830427,
+      "learning_rate": 0.00019986522164042124,
+      "loss": 1.3023,
+      "step": 1000
+    },
+    {
+      "epoch": 0.046114156723637535,
+      "grad_norm": 0.8853357326929552,
+      "learning_rate": 0.00019986444611452985,
+      "loss": 1.2937,
+      "step": 1001
+    },
+    {
+      "epoch": 0.04616022481227254,
+      "grad_norm": 0.5925189726457153,
+      "learning_rate": 0.00019986366836532778,
+      "loss": 1.2608,
+      "step": 1002
+    },
+    {
+      "epoch": 0.04620629290090754,
+      "grad_norm": 0.5374792626013507,
+      "learning_rate": 0.0001998628883928324,
+      "loss": 1.3475,
+      "step": 1003
+    },
+    {
+      "epoch": 0.046252360989542544,
+      "grad_norm": 0.570932363721894,
+      "learning_rate": 0.00019986210619706105,
+      "loss": 1.3966,
+      "step": 1004
+    },
+    {
+      "epoch": 0.04629842907817755,
+      "grad_norm": 0.7148721284169033,
+      "learning_rate": 0.0001998613217780311,
+      "loss": 1.2608,
+      "step": 1005
+    },
+    {
+      "epoch": 0.04634449716681255,
+      "grad_norm": 0.5388875510326074,
+      "learning_rate": 0.00019986053513576003,
+      "loss": 1.3667,
+      "step": 1006
+    },
+    {
+      "epoch": 0.046390565255447554,
+      "grad_norm": 0.7552319194453468,
+      "learning_rate": 0.00019985974627026543,
+      "loss": 1.1251,
+      "step": 1007
+    },
+    {
+      "epoch": 0.04643663334408255,
+      "grad_norm": 1.2075053321321616,
+      "learning_rate": 0.00019985895518156475,
+      "loss": 1.2138,
+      "step": 1008
+    },
+    {
+      "epoch": 0.04648270143271756,
+      "grad_norm": 0.6178631818085668,
+      "learning_rate": 0.00019985816186967567,
+      "loss": 1.3016,
+      "step": 1009
+    },
+    {
+      "epoch": 0.046528769521352556,
+      "grad_norm": 0.6802379036074904,
+      "learning_rate": 0.00019985736633461583,
+      "loss": 1.1884,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04657483760998756,
+      "grad_norm": 0.6455949622057531,
+      "learning_rate": 0.00019985656857640297,
+      "loss": 1.3886,
+      "step": 1011
+    },
+    {
+      "epoch": 0.04662090569862257,
+      "grad_norm": 0.721961207843662,
+      "learning_rate": 0.0001998557685950548,
+      "loss": 1.289,
+      "step": 1012
+    },
+    {
+      "epoch": 0.046666973787257565,
+      "grad_norm": 0.5484959260096598,
+      "learning_rate": 0.00019985496639058918,
+      "loss": 1.281,
+      "step": 1013
+    },
+    {
+      "epoch": 0.04671304187589257,
+      "grad_norm": 0.5582826073347394,
+      "learning_rate": 0.00019985416196302393,
+      "loss": 1.2444,
+      "step": 1014
+    },
+    {
+      "epoch": 0.04675910996452757,
+      "grad_norm": 0.5303341428171512,
+      "learning_rate": 0.00019985335531237699,
+      "loss": 1.2301,
+      "step": 1015
+    },
+    {
+      "epoch": 0.046805178053162574,
+      "grad_norm": 0.5932500357628904,
+      "learning_rate": 0.0001998525464386663,
+      "loss": 1.3706,
+      "step": 1016
+    },
+    {
+      "epoch": 0.04685124614179758,
+      "grad_norm": 0.6357986321170452,
+      "learning_rate": 0.00019985173534190986,
+      "loss": 1.3442,
+      "step": 1017
+    },
+    {
+      "epoch": 0.04689731423043258,
+      "grad_norm": 0.6213537704935964,
+      "learning_rate": 0.00019985092202212576,
+      "loss": 1.3739,
+      "step": 1018
+    },
+    {
+      "epoch": 0.046943382319067584,
+      "grad_norm": 0.6730762374680893,
+      "learning_rate": 0.00019985010647933208,
+      "loss": 1.0558,
+      "step": 1019
+    },
+    {
+      "epoch": 0.04698945040770258,
+      "grad_norm": 0.5909626319293223,
+      "learning_rate": 0.000199849288713547,
+      "loss": 1.297,
+      "step": 1020
+    },
+    {
+      "epoch": 0.04703551849633759,
+      "grad_norm": 0.9796384129840167,
+      "learning_rate": 0.0001998484687247887,
+      "loss": 1.1997,
+      "step": 1021
+    },
+    {
+      "epoch": 0.04708158658497259,
+      "grad_norm": 0.6797742267779576,
+      "learning_rate": 0.00019984764651307546,
+      "loss": 0.8644,
+      "step": 1022
+    },
+    {
+      "epoch": 0.04712765467360759,
+      "grad_norm": 0.47429946805765805,
+      "learning_rate": 0.00019984682207842554,
+      "loss": 1.0011,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0471737227622426,
+      "grad_norm": 0.5673155037238143,
+      "learning_rate": 0.00019984599542085735,
+      "loss": 1.167,
+      "step": 1024
+    },
+    {
+      "epoch": 0.047219790850877595,
+      "grad_norm": 0.6951992835734307,
+      "learning_rate": 0.00019984516654038926,
+      "loss": 1.2868,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0472658589395126,
+      "grad_norm": 0.5369023499615226,
+      "learning_rate": 0.00019984433543703976,
+      "loss": 1.7268,
+      "step": 1026
+    },
+    {
+      "epoch": 0.0473119270281476,
+      "grad_norm": 0.4740945819968527,
+      "learning_rate": 0.0001998435021108273,
+      "loss": 1.4405,
+      "step": 1027
+    },
+    {
+      "epoch": 0.047357995116782604,
+      "grad_norm": 0.523095530700902,
+      "learning_rate": 0.00019984266656177048,
+      "loss": 0.8465,
+      "step": 1028
+    },
+    {
+      "epoch": 0.04740406320541761,
+      "grad_norm": 0.6008755070742452,
+      "learning_rate": 0.00019984182878988787,
+      "loss": 0.9166,
+      "step": 1029
+    },
+    {
+      "epoch": 0.04745013129405261,
+      "grad_norm": 0.4642335009250461,
+      "learning_rate": 0.00019984098879519816,
+      "loss": 1.1791,
+      "step": 1030
+    },
+    {
+      "epoch": 0.047496199382687614,
+      "grad_norm": 0.6184730857348495,
+      "learning_rate": 0.00019984014657772,
+      "loss": 1.0062,
+      "step": 1031
+    },
+    {
+      "epoch": 0.04754226747132261,
+      "grad_norm": 0.5444354181480332,
+      "learning_rate": 0.00019983930213747218,
+      "loss": 1.3345,
+      "step": 1032
+    },
+    {
+      "epoch": 0.04758833555995762,
+      "grad_norm": 0.5646781778036669,
+      "learning_rate": 0.00019983845547447344,
+      "loss": 1.4619,
+      "step": 1033
+    },
+    {
+      "epoch": 0.04763440364859262,
+      "grad_norm": 0.5937164681230938,
+      "learning_rate": 0.00019983760658874271,
+      "loss": 1.0687,
+      "step": 1034
+    },
+    {
+      "epoch": 0.04768047173722762,
+      "grad_norm": 0.5517747825875219,
+      "learning_rate": 0.00019983675548029885,
+      "loss": 1.0872,
+      "step": 1035
+    },
+    {
+      "epoch": 0.04772653982586263,
+      "grad_norm": 0.9104728155211729,
+      "learning_rate": 0.00019983590214916084,
+      "loss": 1.348,
+      "step": 1036
+    },
+    {
+      "epoch": 0.047772607914497625,
+      "grad_norm": 0.6699916880261403,
+      "learning_rate": 0.00019983504659534763,
+      "loss": 1.1627,
+      "step": 1037
+    },
+    {
+      "epoch": 0.04781867600313263,
+      "grad_norm": 0.49452569921371076,
+      "learning_rate": 0.00019983418881887828,
+      "loss": 1.5311,
+      "step": 1038
+    },
+    {
+      "epoch": 0.047864744091767636,
+      "grad_norm": 0.5196838005696169,
+      "learning_rate": 0.0001998333288197719,
+      "loss": 1.2367,
+      "step": 1039
+    },
+    {
+      "epoch": 0.047910812180402634,
+      "grad_norm": 0.5103980658161174,
+      "learning_rate": 0.00019983246659804763,
+      "loss": 0.9402,
+      "step": 1040
+    },
+    {
+      "epoch": 0.04795688026903764,
+      "grad_norm": 0.5780115495088558,
+      "learning_rate": 0.00019983160215372467,
+      "loss": 1.3236,
+      "step": 1041
+    },
+    {
+      "epoch": 0.04800294835767264,
+      "grad_norm": 0.49938426185351104,
+      "learning_rate": 0.00019983073548682226,
+      "loss": 1.2965,
+      "step": 1042
+    },
+    {
+      "epoch": 0.048049016446307644,
+      "grad_norm": 0.628999843502964,
+      "learning_rate": 0.00019982986659735968,
+      "loss": 1.2661,
+      "step": 1043
+    },
+    {
+      "epoch": 0.04809508453494264,
+      "grad_norm": 0.46523275817344684,
+      "learning_rate": 0.00019982899548535632,
+      "loss": 1.4403,
+      "step": 1044
+    },
+    {
+      "epoch": 0.04814115262357765,
+      "grad_norm": 0.5965552268277005,
+      "learning_rate": 0.00019982812215083154,
+      "loss": 1.454,
+      "step": 1045
+    },
+    {
+      "epoch": 0.04818722071221265,
+      "grad_norm": 0.585076949859705,
+      "learning_rate": 0.00019982724659380478,
+      "loss": 1.3378,
+      "step": 1046
+    },
+    {
+      "epoch": 0.04823328880084765,
+      "grad_norm": 0.5001518691752529,
+      "learning_rate": 0.00019982636881429554,
+      "loss": 1.3617,
+      "step": 1047
+    },
+    {
+      "epoch": 0.04827935688948266,
+      "grad_norm": 0.5050396533400963,
+      "learning_rate": 0.0001998254888123234,
+      "loss": 1.1994,
+      "step": 1048
+    },
+    {
+      "epoch": 0.048325424978117655,
+      "grad_norm": 0.6305722334679681,
+      "learning_rate": 0.00019982460658790786,
+      "loss": 1.2842,
+      "step": 1049
+    },
+    {
+      "epoch": 0.04837149306675266,
+      "grad_norm": 0.5497156087330919,
+      "learning_rate": 0.00019982372214106864,
+      "loss": 0.9283,
+      "step": 1050
+    },
+    {
+      "epoch": 0.048417561155387666,
+      "grad_norm": 0.6579672384662608,
+      "learning_rate": 0.0001998228354718254,
+      "loss": 1.1937,
+      "step": 1051
+    },
+    {
+      "epoch": 0.048463629244022664,
+      "grad_norm": 0.569880084284879,
+      "learning_rate": 0.00019982194658019792,
+      "loss": 1.2648,
+      "step": 1052
+    },
+    {
+      "epoch": 0.04850969733265767,
+      "grad_norm": 0.5762419811268731,
+      "learning_rate": 0.00019982105546620594,
+      "loss": 1.1991,
+      "step": 1053
+    },
+    {
+      "epoch": 0.04855576542129267,
+      "grad_norm": 0.6902666093724373,
+      "learning_rate": 0.00019982016212986928,
+      "loss": 1.0313,
+      "step": 1054
+    },
+    {
+      "epoch": 0.048601833509927674,
+      "grad_norm": 0.6336928441467731,
+      "learning_rate": 0.00019981926657120793,
+      "loss": 1.2079,
+      "step": 1055
+    },
+    {
+      "epoch": 0.04864790159856268,
+      "grad_norm": 0.8859854448207812,
+      "learning_rate": 0.00019981836879024172,
+      "loss": 1.3856,
+      "step": 1056
+    },
+    {
+      "epoch": 0.04869396968719768,
+      "grad_norm": 0.6178334302079344,
+      "learning_rate": 0.0001998174687869907,
+      "loss": 1.1843,
+      "step": 1057
+    },
+    {
+      "epoch": 0.04874003777583268,
+      "grad_norm": 0.573645457619554,
+      "learning_rate": 0.00019981656656147486,
+      "loss": 1.7136,
+      "step": 1058
+    },
+    {
+      "epoch": 0.04878610586446768,
+      "grad_norm": 0.5755469911223867,
+      "learning_rate": 0.00019981566211371433,
+      "loss": 1.3132,
+      "step": 1059
+    },
+    {
+      "epoch": 0.04883217395310269,
+      "grad_norm": 0.7354905764452407,
+      "learning_rate": 0.00019981475544372925,
+      "loss": 1.2626,
+      "step": 1060
+    },
+    {
+      "epoch": 0.048878242041737685,
+      "grad_norm": 0.6764950953987539,
+      "learning_rate": 0.00019981384655153977,
+      "loss": 0.9607,
+      "step": 1061
+    },
+    {
+      "epoch": 0.04892431013037269,
+      "grad_norm": 0.4512763778926945,
+      "learning_rate": 0.00019981293543716613,
+      "loss": 1.1782,
+      "step": 1062
+    },
+    {
+      "epoch": 0.048970378219007696,
+      "grad_norm": 0.5611314671434139,
+      "learning_rate": 0.00019981202210062865,
+      "loss": 1.6192,
+      "step": 1063
+    },
+    {
+      "epoch": 0.049016446307642694,
+      "grad_norm": 0.5438066535697416,
+      "learning_rate": 0.0001998111065419476,
+      "loss": 1.5207,
+      "step": 1064
+    },
+    {
+      "epoch": 0.0490625143962777,
+      "grad_norm": 0.5678928011402665,
+      "learning_rate": 0.00019981018876114343,
+      "loss": 1.4767,
+      "step": 1065
+    },
+    {
+      "epoch": 0.0491085824849127,
+      "grad_norm": 0.47945916433610863,
+      "learning_rate": 0.00019980926875823657,
+      "loss": 1.3177,
+      "step": 1066
+    },
+    {
+      "epoch": 0.049154650573547704,
+      "grad_norm": 0.5220392449261454,
+      "learning_rate": 0.00019980834653324743,
+      "loss": 1.4271,
+      "step": 1067
+    },
+    {
+      "epoch": 0.04920071866218271,
+      "grad_norm": 0.6943592575548178,
+      "learning_rate": 0.0001998074220861966,
+      "loss": 1.2293,
+      "step": 1068
+    },
+    {
+      "epoch": 0.04924678675081771,
+      "grad_norm": 0.5198054452512066,
+      "learning_rate": 0.00019980649541710466,
+      "loss": 1.223,
+      "step": 1069
+    },
+    {
+      "epoch": 0.04929285483945271,
+      "grad_norm": 0.6229773640113815,
+      "learning_rate": 0.00019980556652599222,
+      "loss": 1.1468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.04933892292808771,
+      "grad_norm": 0.5923889870333745,
+      "learning_rate": 0.00019980463541287996,
+      "loss": 0.7847,
+      "step": 1071
+    },
+    {
+      "epoch": 0.04938499101672272,
+      "grad_norm": 0.5781452725211468,
+      "learning_rate": 0.00019980370207778867,
+      "loss": 1.4899,
+      "step": 1072
+    },
+    {
+      "epoch": 0.04943105910535772,
+      "grad_norm": 0.49584750866995675,
+      "learning_rate": 0.00019980276652073903,
+      "loss": 0.9298,
+      "step": 1073
+    },
+    {
+      "epoch": 0.04947712719399272,
+      "grad_norm": 0.7478270680167264,
+      "learning_rate": 0.00019980182874175192,
+      "loss": 1.4405,
+      "step": 1074
+    },
+    {
+      "epoch": 0.049523195282627726,
+      "grad_norm": 0.6567185144570408,
+      "learning_rate": 0.00019980088874084825,
+      "loss": 1.0708,
+      "step": 1075
+    },
+    {
+      "epoch": 0.049569263371262724,
+      "grad_norm": 0.6102088919945634,
+      "learning_rate": 0.0001997999465180489,
+      "loss": 1.5859,
+      "step": 1076
+    },
+    {
+      "epoch": 0.04961533145989773,
+      "grad_norm": 0.5044531384214817,
+      "learning_rate": 0.00019979900207337488,
+      "loss": 1.2795,
+      "step": 1077
+    },
+    {
+      "epoch": 0.04966139954853273,
+      "grad_norm": 0.6254378325232903,
+      "learning_rate": 0.00019979805540684716,
+      "loss": 0.9344,
+      "step": 1078
+    },
+    {
+      "epoch": 0.049707467637167733,
+      "grad_norm": 0.6252196744883611,
+      "learning_rate": 0.00019979710651848686,
+      "loss": 1.234,
+      "step": 1079
+    },
+    {
+      "epoch": 0.04975353572580274,
+      "grad_norm": 0.5392844444860645,
+      "learning_rate": 0.00019979615540831512,
+      "loss": 1.194,
+      "step": 1080
+    },
+    {
+      "epoch": 0.04979960381443774,
+      "grad_norm": 0.5085023257172269,
+      "learning_rate": 0.00019979520207635308,
+      "loss": 1.5865,
+      "step": 1081
+    },
+    {
+      "epoch": 0.04984567190307274,
+      "grad_norm": 0.5785545792443088,
+      "learning_rate": 0.00019979424652262197,
+      "loss": 1.22,
+      "step": 1082
+    },
+    {
+      "epoch": 0.04989173999170774,
+      "grad_norm": 0.5294010719566948,
+      "learning_rate": 0.0001997932887471431,
+      "loss": 1.1429,
+      "step": 1083
+    },
+    {
+      "epoch": 0.04993780808034275,
+      "grad_norm": 0.5812168130563498,
+      "learning_rate": 0.00019979232874993774,
+      "loss": 1.2077,
+      "step": 1084
+    },
+    {
+      "epoch": 0.04998387616897775,
+      "grad_norm": 0.7501369677770233,
+      "learning_rate": 0.0001997913665310273,
+      "loss": 1.3647,
+      "step": 1085
+    },
+    {
+      "epoch": 0.05002994425761275,
+      "grad_norm": 0.5180791082417556,
+      "learning_rate": 0.00019979040209043317,
+      "loss": 1.0225,
+      "step": 1086
+    },
+    {
+      "epoch": 0.050076012346247756,
+      "grad_norm": 0.7532802940181363,
+      "learning_rate": 0.00019978943542817685,
+      "loss": 1.4086,
+      "step": 1087
+    },
+    {
+      "epoch": 0.050122080434882754,
+      "grad_norm": 0.44886866321796304,
+      "learning_rate": 0.00019978846654427989,
+      "loss": 0.9454,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05016814852351776,
+      "grad_norm": 0.6905484567814839,
+      "learning_rate": 0.00019978749543876376,
+      "loss": 1.0663,
+      "step": 1089
+    },
+    {
+      "epoch": 0.050214216612152765,
+      "grad_norm": 0.5756967644275213,
+      "learning_rate": 0.00019978652211165017,
+      "loss": 1.1376,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05026028470078776,
+      "grad_norm": 0.7403269280052732,
+      "learning_rate": 0.00019978554656296078,
+      "loss": 1.2681,
+      "step": 1091
+    },
+    {
+      "epoch": 0.05030635278942277,
+      "grad_norm": 0.674441163362688,
+      "learning_rate": 0.00019978456879271727,
+      "loss": 1.421,
+      "step": 1092
+    },
+    {
+      "epoch": 0.05035242087805777,
+      "grad_norm": 0.5855688728393856,
+      "learning_rate": 0.00019978358880094144,
+      "loss": 1.5934,
+      "step": 1093
+    },
+    {
+      "epoch": 0.05039848896669277,
+      "grad_norm": 0.5877200128892792,
+      "learning_rate": 0.0001997826065876551,
+      "loss": 1.6072,
+      "step": 1094
+    },
+    {
+      "epoch": 0.05044455705532777,
+      "grad_norm": 0.5504804837157322,
+      "learning_rate": 0.0001997816221528801,
+      "loss": 1.391,
+      "step": 1095
+    },
+    {
+      "epoch": 0.050490625143962777,
+      "grad_norm": 0.5210810962579345,
+      "learning_rate": 0.00019978063549663837,
+      "loss": 1.3858,
+      "step": 1096
+    },
+    {
+      "epoch": 0.05053669323259778,
+      "grad_norm": 0.6122379917902507,
+      "learning_rate": 0.0001997796466189519,
+      "loss": 1.1295,
+      "step": 1097
+    },
+    {
+      "epoch": 0.05058276132123278,
+      "grad_norm": 0.6618998539952142,
+      "learning_rate": 0.00019977865551984268,
+      "loss": 0.9643,
+      "step": 1098
+    },
+    {
+      "epoch": 0.050628829409867786,
+      "grad_norm": 0.5291709506853526,
+      "learning_rate": 0.00019977766219933277,
+      "loss": 0.9311,
+      "step": 1099
+    },
+    {
+      "epoch": 0.050674897498502784,
+      "grad_norm": 0.5008920061449028,
+      "learning_rate": 0.0001997766666574443,
+      "loss": 1.4268,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05072096558713779,
+      "grad_norm": 0.6515597983998577,
+      "learning_rate": 0.0001997756688941994,
+      "loss": 1.2978,
+      "step": 1101
+    },
+    {
+      "epoch": 0.050767033675772795,
+      "grad_norm": 0.5454464810449218,
+      "learning_rate": 0.00019977466890962032,
+      "loss": 1.5496,
+      "step": 1102
+    },
+    {
+      "epoch": 0.05081310176440779,
+      "grad_norm": 0.6239357383983921,
+      "learning_rate": 0.00019977366670372935,
+      "loss": 1.1284,
+      "step": 1103
+    },
+    {
+      "epoch": 0.0508591698530428,
+      "grad_norm": 0.6533834309535551,
+      "learning_rate": 0.00019977266227654873,
+      "loss": 0.9898,
+      "step": 1104
+    },
+    {
+      "epoch": 0.0509052379416778,
+      "grad_norm": 0.5402503397237264,
+      "learning_rate": 0.00019977165562810086,
+      "loss": 1.3687,
+      "step": 1105
+    },
+    {
+      "epoch": 0.0509513060303128,
+      "grad_norm": 0.4630316092995732,
+      "learning_rate": 0.0001997706467584082,
+      "loss": 1.3644,
+      "step": 1106
+    },
+    {
+      "epoch": 0.05099737411894781,
+      "grad_norm": 0.47872987380540394,
+      "learning_rate": 0.0001997696356674931,
+      "loss": 1.1167,
+      "step": 1107
+    },
+    {
+      "epoch": 0.051043442207582806,
+      "grad_norm": 0.6123625028662817,
+      "learning_rate": 0.00019976862235537813,
+      "loss": 1.3023,
+      "step": 1108
+    },
+    {
+      "epoch": 0.05108951029621781,
+      "grad_norm": 0.4630112899610767,
+      "learning_rate": 0.00019976760682208585,
+      "loss": 0.9971,
+      "step": 1109
+    },
+    {
+      "epoch": 0.05113557838485281,
+      "grad_norm": 0.7536543562031227,
+      "learning_rate": 0.00019976658906763892,
+      "loss": 1.2439,
+      "step": 1110
+    },
+    {
+      "epoch": 0.051181646473487816,
+      "grad_norm": 0.5117598309490256,
+      "learning_rate": 0.0001997655690920599,
+      "loss": 1.4431,
+      "step": 1111
+    },
+    {
+      "epoch": 0.051227714562122814,
+      "grad_norm": 0.5821664315312356,
+      "learning_rate": 0.00019976454689537155,
+      "loss": 1.3951,
+      "step": 1112
+    },
+    {
+      "epoch": 0.05127378265075782,
+      "grad_norm": 0.7426101677438536,
+      "learning_rate": 0.00019976352247759663,
+      "loss": 0.9786,
+      "step": 1113
+    },
+    {
+      "epoch": 0.051319850739392825,
+      "grad_norm": 0.5483133140353958,
+      "learning_rate": 0.00019976249583875797,
+      "loss": 1.1002,
+      "step": 1114
+    },
+    {
+      "epoch": 0.05136591882802782,
+      "grad_norm": 0.5388651390324848,
+      "learning_rate": 0.00019976146697887838,
+      "loss": 1.4769,
+      "step": 1115
+    },
+    {
+      "epoch": 0.05141198691666283,
+      "grad_norm": 0.4894023433632612,
+      "learning_rate": 0.00019976043589798074,
+      "loss": 1.1717,
+      "step": 1116
+    },
+    {
+      "epoch": 0.05145805500529783,
+      "grad_norm": 0.44201276122319005,
+      "learning_rate": 0.00019975940259608807,
+      "loss": 1.2709,
+      "step": 1117
+    },
+    {
+      "epoch": 0.05150412309393283,
+      "grad_norm": 0.4107785100115193,
+      "learning_rate": 0.00019975836707322337,
+      "loss": 1.5375,
+      "step": 1118
+    },
+    {
+      "epoch": 0.05155019118256784,
+      "grad_norm": 0.5422803375905418,
+      "learning_rate": 0.00019975732932940967,
+      "loss": 1.1056,
+      "step": 1119
+    },
+    {
+      "epoch": 0.051596259271202836,
+      "grad_norm": 0.8080023887410953,
+      "learning_rate": 0.00019975628936467005,
+      "loss": 0.7682,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05164232735983784,
+      "grad_norm": 0.6374289517864679,
+      "learning_rate": 0.00019975524717902774,
+      "loss": 1.2434,
+      "step": 1121
+    },
+    {
+      "epoch": 0.05168839544847284,
+      "grad_norm": 0.6332750578625612,
+      "learning_rate": 0.00019975420277250588,
+      "loss": 1.642,
+      "step": 1122
+    },
+    {
+      "epoch": 0.051734463537107846,
+      "grad_norm": 0.5125066859728729,
+      "learning_rate": 0.0001997531561451277,
+      "loss": 1.4123,
+      "step": 1123
+    },
+    {
+      "epoch": 0.05178053162574285,
+      "grad_norm": 0.45760752993494397,
+      "learning_rate": 0.00019975210729691659,
+      "loss": 0.9742,
+      "step": 1124
+    },
+    {
+      "epoch": 0.05182659971437785,
+      "grad_norm": 0.5518302372712427,
+      "learning_rate": 0.00019975105622789584,
+      "loss": 1.3665,
+      "step": 1125
+    },
+    {
+      "epoch": 0.051872667803012855,
+      "grad_norm": 0.5255370294915069,
+      "learning_rate": 0.00019975000293808882,
+      "loss": 1.5463,
+      "step": 1126
+    },
+    {
+      "epoch": 0.05191873589164785,
+      "grad_norm": 0.6418277754707972,
+      "learning_rate": 0.00019974894742751906,
+      "loss": 1.187,
+      "step": 1127
+    },
+    {
+      "epoch": 0.05196480398028286,
+      "grad_norm": 0.6093312182534766,
+      "learning_rate": 0.00019974788969621,
+      "loss": 1.3494,
+      "step": 1128
+    },
+    {
+      "epoch": 0.05201087206891786,
+      "grad_norm": 0.5241790371294103,
+      "learning_rate": 0.00019974682974418522,
+      "loss": 1.0891,
+      "step": 1129
+    },
+    {
+      "epoch": 0.05205694015755286,
+      "grad_norm": 0.5722032385831046,
+      "learning_rate": 0.00019974576757146828,
+      "loss": 0.7791,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05210300824618787,
+      "grad_norm": 0.6666545339067546,
+      "learning_rate": 0.00019974470317808286,
+      "loss": 1.4092,
+      "step": 1131
+    },
+    {
+      "epoch": 0.052149076334822866,
+      "grad_norm": 0.6424166579816437,
+      "learning_rate": 0.0001997436365640527,
+      "loss": 1.123,
+      "step": 1132
+    },
+    {
+      "epoch": 0.05219514442345787,
+      "grad_norm": 0.5251600590530996,
+      "learning_rate": 0.00019974256772940143,
+      "loss": 1.3163,
+      "step": 1133
+    },
+    {
+      "epoch": 0.05224121251209287,
+      "grad_norm": 0.503026488486457,
+      "learning_rate": 0.00019974149667415292,
+      "loss": 1.1454,
+      "step": 1134
+    },
+    {
+      "epoch": 0.052287280600727876,
+      "grad_norm": 0.8548904999775788,
+      "learning_rate": 0.00019974042339833103,
+      "loss": 0.9679,
+      "step": 1135
+    },
+    {
+      "epoch": 0.05233334868936288,
+      "grad_norm": 0.828877117433858,
+      "learning_rate": 0.00019973934790195962,
+      "loss": 1.0598,
+      "step": 1136
+    },
+    {
+      "epoch": 0.05237941677799788,
+      "grad_norm": 0.7658744282670162,
+      "learning_rate": 0.00019973827018506267,
+      "loss": 0.8161,
+      "step": 1137
+    },
+    {
+      "epoch": 0.052425484866632885,
+      "grad_norm": 0.8648436378523874,
+      "learning_rate": 0.0001997371902476641,
+      "loss": 1.4674,
+      "step": 1138
+    },
+    {
+      "epoch": 0.05247155295526788,
+      "grad_norm": 0.8457774264308354,
+      "learning_rate": 0.00019973610808978802,
+      "loss": 1.2719,
+      "step": 1139
+    },
+    {
+      "epoch": 0.05251762104390289,
+      "grad_norm": 0.5471796643326545,
+      "learning_rate": 0.00019973502371145853,
+      "loss": 0.9736,
+      "step": 1140
+    },
+    {
+      "epoch": 0.052563689132537894,
+      "grad_norm": 1.0094677990444747,
+      "learning_rate": 0.00019973393711269973,
+      "loss": 1.2199,
+      "step": 1141
+    },
+    {
+      "epoch": 0.05260975722117289,
+      "grad_norm": 0.5686421974979499,
+      "learning_rate": 0.00019973284829353583,
+      "loss": 1.0986,
+      "step": 1142
+    },
+    {
+      "epoch": 0.0526558253098079,
+      "grad_norm": 0.5122735392934946,
+      "learning_rate": 0.00019973175725399104,
+      "loss": 1.2967,
+      "step": 1143
+    },
+    {
+      "epoch": 0.052701893398442896,
+      "grad_norm": 0.6351875609794835,
+      "learning_rate": 0.00019973066399408973,
+      "loss": 1.459,
+      "step": 1144
+    },
+    {
+      "epoch": 0.0527479614870779,
+      "grad_norm": 0.43049458214744535,
+      "learning_rate": 0.00019972956851385613,
+      "loss": 1.054,
+      "step": 1145
+    },
+    {
+      "epoch": 0.0527940295757129,
+      "grad_norm": 0.5294339096718074,
+      "learning_rate": 0.00019972847081331475,
+      "loss": 1.1314,
+      "step": 1146
+    },
+    {
+      "epoch": 0.052840097664347906,
+      "grad_norm": 0.715189822860296,
+      "learning_rate": 0.00019972737089248992,
+      "loss": 0.9504,
+      "step": 1147
+    },
+    {
+      "epoch": 0.05288616575298291,
+      "grad_norm": 0.4918336775508509,
+      "learning_rate": 0.0001997262687514062,
+      "loss": 1.2692,
+      "step": 1148
+    },
+    {
+      "epoch": 0.05293223384161791,
+      "grad_norm": 0.5204228266826297,
+      "learning_rate": 0.00019972516439008807,
+      "loss": 1.0521,
+      "step": 1149
+    },
+    {
+      "epoch": 0.052978301930252915,
+      "grad_norm": 0.6781463358909255,
+      "learning_rate": 0.00019972405780856013,
+      "loss": 1.371,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05302437001888791,
+      "grad_norm": 0.675198466242574,
+      "learning_rate": 0.00019972294900684706,
+      "loss": 1.0941,
+      "step": 1151
+    },
+    {
+      "epoch": 0.05307043810752292,
+      "grad_norm": 0.49235089437515595,
+      "learning_rate": 0.00019972183798497355,
+      "loss": 1.315,
+      "step": 1152
+    },
+    {
+      "epoch": 0.053116506196157924,
+      "grad_norm": 0.5523880527329074,
+      "learning_rate": 0.00019972072474296426,
+      "loss": 1.1688,
+      "step": 1153
+    },
+    {
+      "epoch": 0.05316257428479292,
+      "grad_norm": 0.6339329819112213,
+      "learning_rate": 0.00019971960928084404,
+      "loss": 1.2238,
+      "step": 1154
+    },
+    {
+      "epoch": 0.05320864237342793,
+      "grad_norm": 0.7283595414867862,
+      "learning_rate": 0.00019971849159863768,
+      "loss": 1.3752,
+      "step": 1155
+    },
+    {
+      "epoch": 0.053254710462062926,
+      "grad_norm": 0.5810555097543709,
+      "learning_rate": 0.0001997173716963701,
+      "loss": 1.5693,
+      "step": 1156
+    },
+    {
+      "epoch": 0.05330077855069793,
+      "grad_norm": 0.5166434727398246,
+      "learning_rate": 0.00019971624957406623,
+      "loss": 1.2886,
+      "step": 1157
+    },
+    {
+      "epoch": 0.05334684663933294,
+      "grad_norm": 0.5619763019946507,
+      "learning_rate": 0.00019971512523175102,
+      "loss": 1.4919,
+      "step": 1158
+    },
+    {
+      "epoch": 0.053392914727967936,
+      "grad_norm": 0.5125390529592223,
+      "learning_rate": 0.00019971399866944954,
+      "loss": 1.4915,
+      "step": 1159
+    },
+    {
+      "epoch": 0.05343898281660294,
+      "grad_norm": 0.4995087826302848,
+      "learning_rate": 0.00019971286988718686,
+      "loss": 1.2831,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05348505090523794,
+      "grad_norm": 0.5220935175251712,
+      "learning_rate": 0.00019971173888498806,
+      "loss": 1.1891,
+      "step": 1161
+    },
+    {
+      "epoch": 0.053531118993872945,
+      "grad_norm": 0.7171516273714272,
+      "learning_rate": 0.0001997106056628784,
+      "loss": 1.6514,
+      "step": 1162
+    },
+    {
+      "epoch": 0.05357718708250795,
+      "grad_norm": 0.46145888954289677,
+      "learning_rate": 0.00019970947022088305,
+      "loss": 1.2881,
+      "step": 1163
+    },
+    {
+      "epoch": 0.05362325517114295,
+      "grad_norm": 0.541417582973903,
+      "learning_rate": 0.00019970833255902732,
+      "loss": 0.9459,
+      "step": 1164
+    },
+    {
+      "epoch": 0.053669323259777954,
+      "grad_norm": 0.5010417593451444,
+      "learning_rate": 0.00019970719267733652,
+      "loss": 1.2098,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05371539134841295,
+      "grad_norm": 0.5601504585398086,
+      "learning_rate": 0.00019970605057583604,
+      "loss": 0.8922,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05376145943704796,
+      "grad_norm": 0.8062817559460469,
+      "learning_rate": 0.00019970490625455134,
+      "loss": 1.5766,
+      "step": 1167
+    },
+    {
+      "epoch": 0.053807527525682956,
+      "grad_norm": 0.5214212648410568,
+      "learning_rate": 0.0001997037597135078,
+      "loss": 1.4604,
+      "step": 1168
+    },
+    {
+      "epoch": 0.05385359561431796,
+      "grad_norm": 0.578051591599674,
+      "learning_rate": 0.00019970261095273104,
+      "loss": 1.4614,
+      "step": 1169
+    },
+    {
+      "epoch": 0.05389966370295297,
+      "grad_norm": 0.5088884238036683,
+      "learning_rate": 0.00019970145997224662,
+      "loss": 1.2428,
+      "step": 1170
+    },
+    {
+      "epoch": 0.053945731791587966,
+      "grad_norm": 0.504957061381783,
+      "learning_rate": 0.0001997003067720801,
+      "loss": 1.2572,
+      "step": 1171
+    },
+    {
+      "epoch": 0.05399179988022297,
+      "grad_norm": 0.6584574411068344,
+      "learning_rate": 0.0001996991513522572,
+      "loss": 1.3134,
+      "step": 1172
+    },
+    {
+      "epoch": 0.05403786796885797,
+      "grad_norm": 0.7913345709511379,
+      "learning_rate": 0.00019969799371280368,
+      "loss": 1.2236,
+      "step": 1173
+    },
+    {
+      "epoch": 0.054083936057492975,
+      "grad_norm": 0.6571166396543401,
+      "learning_rate": 0.00019969683385374524,
+      "loss": 1.0249,
+      "step": 1174
+    },
+    {
+      "epoch": 0.05413000414612798,
+      "grad_norm": 0.5904351255747419,
+      "learning_rate": 0.00019969567177510777,
+      "loss": 1.2671,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05417607223476298,
+      "grad_norm": 0.4555905292391452,
+      "learning_rate": 0.00019969450747691709,
+      "loss": 1.4031,
+      "step": 1176
+    },
+    {
+      "epoch": 0.054222140323397984,
+      "grad_norm": 0.5237489966806943,
+      "learning_rate": 0.00019969334095919913,
+      "loss": 1.5707,
+      "step": 1177
+    },
+    {
+      "epoch": 0.05426820841203298,
+      "grad_norm": 0.6111050611189555,
+      "learning_rate": 0.00019969217222197988,
+      "loss": 1.3279,
+      "step": 1178
+    },
+    {
+      "epoch": 0.05431427650066799,
+      "grad_norm": 0.6246741455306403,
+      "learning_rate": 0.00019969100126528537,
+      "loss": 1.1527,
+      "step": 1179
+    },
+    {
+      "epoch": 0.05436034458930299,
+      "grad_norm": 0.5433751318998169,
+      "learning_rate": 0.00019968982808914163,
+      "loss": 0.9841,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05440641267793799,
+      "grad_norm": 0.6931311562303772,
+      "learning_rate": 0.0001996886526935748,
+      "loss": 0.7747,
+      "step": 1181
+    },
+    {
+      "epoch": 0.054452480766573,
+      "grad_norm": 0.5646361519782326,
+      "learning_rate": 0.00019968747507861107,
+      "loss": 1.3053,
+      "step": 1182
+    },
+    {
+      "epoch": 0.054498548855207996,
+      "grad_norm": 0.4134454411740318,
+      "learning_rate": 0.0001996862952442766,
+      "loss": 1.3591,
+      "step": 1183
+    },
+    {
+      "epoch": 0.054544616943843,
+      "grad_norm": 0.5111251691698583,
+      "learning_rate": 0.00019968511319059772,
+      "loss": 1.1093,
+      "step": 1184
+    },
+    {
+      "epoch": 0.054590685032478,
+      "grad_norm": 0.4905429032119355,
+      "learning_rate": 0.00019968392891760067,
+      "loss": 1.2013,
+      "step": 1185
+    },
+    {
+      "epoch": 0.054636753121113005,
+      "grad_norm": 0.6004673788380027,
+      "learning_rate": 0.00019968274242531192,
+      "loss": 1.099,
+      "step": 1186
+    },
+    {
+      "epoch": 0.05468282120974801,
+      "grad_norm": 0.5249574089971653,
+      "learning_rate": 0.00019968155371375784,
+      "loss": 1.4833,
+      "step": 1187
+    },
+    {
+      "epoch": 0.05472888929838301,
+      "grad_norm": 0.46874964728586166,
+      "learning_rate": 0.00019968036278296481,
+      "loss": 1.4453,
+      "step": 1188
+    },
+    {
+      "epoch": 0.054774957387018014,
+      "grad_norm": 0.47435362040438667,
+      "learning_rate": 0.0001996791696329595,
+      "loss": 0.9838,
+      "step": 1189
+    },
+    {
+      "epoch": 0.05482102547565301,
+      "grad_norm": 0.5022696828254765,
+      "learning_rate": 0.00019967797426376834,
+      "loss": 1.2911,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05486709356428802,
+      "grad_norm": 0.5566778780969025,
+      "learning_rate": 0.000199676776675418,
+      "loss": 1.0819,
+      "step": 1191
+    },
+    {
+      "epoch": 0.05491316165292302,
+      "grad_norm": 0.49158051220692395,
+      "learning_rate": 0.00019967557686793514,
+      "loss": 1.3356,
+      "step": 1192
+    },
+    {
+      "epoch": 0.05495922974155802,
+      "grad_norm": 0.6034274588309018,
+      "learning_rate": 0.00019967437484134647,
+      "loss": 1.4104,
+      "step": 1193
+    },
+    {
+      "epoch": 0.05500529783019303,
+      "grad_norm": 0.5182807043958582,
+      "learning_rate": 0.00019967317059567873,
+      "loss": 1.2301,
+      "step": 1194
+    },
+    {
+      "epoch": 0.055051365918828025,
+      "grad_norm": 0.377859482018772,
+      "learning_rate": 0.00019967196413095878,
+      "loss": 1.1306,
+      "step": 1195
+    },
+    {
+      "epoch": 0.05509743400746303,
+      "grad_norm": 0.5275422324509547,
+      "learning_rate": 0.00019967075544721343,
+      "loss": 1.4653,
+      "step": 1196
+    },
+    {
+      "epoch": 0.055143502096098036,
+      "grad_norm": 0.4518093989403645,
+      "learning_rate": 0.00019966954454446963,
+      "loss": 1.5689,
+      "step": 1197
+    },
+    {
+      "epoch": 0.055189570184733035,
+      "grad_norm": 0.49754657419355497,
+      "learning_rate": 0.00019966833142275433,
+      "loss": 1.374,
+      "step": 1198
+    },
+    {
+      "epoch": 0.05523563827336804,
+      "grad_norm": 0.46523741548372244,
+      "learning_rate": 0.00019966711608209448,
+      "loss": 1.3848,
+      "step": 1199
+    },
+    {
+      "epoch": 0.05528170636200304,
+      "grad_norm": 0.6220211863541389,
+      "learning_rate": 0.0001996658985225172,
+      "loss": 1.0142,
+      "step": 1200
+    },
+    {
+      "epoch": 0.055327774450638044,
+      "grad_norm": 0.4396856297951668,
+      "learning_rate": 0.00019966467874404958,
+      "loss": 1.3967,
+      "step": 1201
+    },
+    {
+      "epoch": 0.05537384253927304,
+      "grad_norm": 0.5274010107741036,
+      "learning_rate": 0.00019966345674671878,
+      "loss": 0.9288,
+      "step": 1202
+    },
+    {
+      "epoch": 0.05541991062790805,
+      "grad_norm": 0.6067442789130588,
+      "learning_rate": 0.000199662232530552,
+      "loss": 1.3088,
+      "step": 1203
+    },
+    {
+      "epoch": 0.05546597871654305,
+      "grad_norm": 0.444799746919273,
+      "learning_rate": 0.0001996610060955765,
+      "loss": 1.6128,
+      "step": 1204
+    },
+    {
+      "epoch": 0.05551204680517805,
+      "grad_norm": 0.47384252048981873,
+      "learning_rate": 0.0001996597774418196,
+      "loss": 1.2821,
+      "step": 1205
+    },
+    {
+      "epoch": 0.05555811489381306,
+      "grad_norm": 0.7503503445897636,
+      "learning_rate": 0.0001996585465693086,
+      "loss": 1.0814,
+      "step": 1206
+    },
+    {
+      "epoch": 0.055604182982448055,
+      "grad_norm": 0.5230609640483945,
+      "learning_rate": 0.00019965731347807095,
+      "loss": 1.0497,
+      "step": 1207
+    },
+    {
+      "epoch": 0.05565025107108306,
+      "grad_norm": 0.4659961803119404,
+      "learning_rate": 0.00019965607816813406,
+      "loss": 1.3905,
+      "step": 1208
+    },
+    {
+      "epoch": 0.055696319159718066,
+      "grad_norm": 1.9077672980413145,
+      "learning_rate": 0.0001996548406395255,
+      "loss": 1.4451,
+      "step": 1209
+    },
+    {
+      "epoch": 0.055742387248353065,
+      "grad_norm": 0.5906488809846984,
+      "learning_rate": 0.00019965360089227278,
+      "loss": 0.9278,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05578845533698807,
+      "grad_norm": 0.5204643452608303,
+      "learning_rate": 0.00019965235892640352,
+      "loss": 1.3882,
+      "step": 1211
+    },
+    {
+      "epoch": 0.05583452342562307,
+      "grad_norm": 0.5108795359106992,
+      "learning_rate": 0.00019965111474194535,
+      "loss": 1.4379,
+      "step": 1212
+    },
+    {
+      "epoch": 0.055880591514258074,
+      "grad_norm": 0.48619866867451583,
+      "learning_rate": 0.00019964986833892595,
+      "loss": 1.3763,
+      "step": 1213
+    },
+    {
+      "epoch": 0.05592665960289308,
+      "grad_norm": 0.6106104443802233,
+      "learning_rate": 0.00019964861971737312,
+      "loss": 1.0415,
+      "step": 1214
+    },
+    {
+      "epoch": 0.05597272769152808,
+      "grad_norm": 0.4358299460627517,
+      "learning_rate": 0.0001996473688773146,
+      "loss": 1.2042,
+      "step": 1215
+    },
+    {
+      "epoch": 0.05601879578016308,
+      "grad_norm": 0.5867159691618329,
+      "learning_rate": 0.00019964611581877833,
+      "loss": 1.0817,
+      "step": 1216
+    },
+    {
+      "epoch": 0.05606486386879808,
+      "grad_norm": 0.6591502125009914,
+      "learning_rate": 0.0001996448605417921,
+      "loss": 1.0104,
+      "step": 1217
+    },
+    {
+      "epoch": 0.05611093195743309,
+      "grad_norm": 0.6088733700994733,
+      "learning_rate": 0.00019964360304638393,
+      "loss": 1.3959,
+      "step": 1218
+    },
+    {
+      "epoch": 0.056157000046068085,
+      "grad_norm": 0.5611587082728617,
+      "learning_rate": 0.00019964234333258175,
+      "loss": 1.4024,
+      "step": 1219
+    },
+    {
+      "epoch": 0.05620306813470309,
+      "grad_norm": 0.479283156324924,
+      "learning_rate": 0.0001996410814004137,
+      "loss": 1.3259,
+      "step": 1220
+    },
+    {
+      "epoch": 0.056249136223338096,
+      "grad_norm": 0.5910017479619479,
+      "learning_rate": 0.0001996398172499078,
+      "loss": 1.1815,
+      "step": 1221
+    },
+    {
+      "epoch": 0.056295204311973095,
+      "grad_norm": 0.4925575111847972,
+      "learning_rate": 0.0001996385508810922,
+      "loss": 1.223,
+      "step": 1222
+    },
+    {
+      "epoch": 0.0563412724006081,
+      "grad_norm": 0.5866230078312131,
+      "learning_rate": 0.0001996372822939951,
+      "loss": 1.4213,
+      "step": 1223
+    },
+    {
+      "epoch": 0.0563873404892431,
+      "grad_norm": 0.5269646093816357,
+      "learning_rate": 0.00019963601148864478,
+      "loss": 1.4837,
+      "step": 1224
+    },
+    {
+      "epoch": 0.056433408577878104,
+      "grad_norm": 0.5360171065898665,
+      "learning_rate": 0.00019963473846506947,
+      "loss": 1.1479,
+      "step": 1225
+    },
+    {
+      "epoch": 0.05647947666651311,
+      "grad_norm": 0.5345038844587552,
+      "learning_rate": 0.0001996334632232976,
+      "loss": 1.1816,
+      "step": 1226
+    },
+    {
+      "epoch": 0.05652554475514811,
+      "grad_norm": 0.5596399394236241,
+      "learning_rate": 0.00019963218576335746,
+      "loss": 1.422,
+      "step": 1227
+    },
+    {
+      "epoch": 0.05657161284378311,
+      "grad_norm": 0.5956284578385869,
+      "learning_rate": 0.00019963090608527754,
+      "loss": 1.0694,
+      "step": 1228
+    },
+    {
+      "epoch": 0.05661768093241811,
+      "grad_norm": 0.5541239090771908,
+      "learning_rate": 0.00019962962418908632,
+      "loss": 0.8854,
+      "step": 1229
+    },
+    {
+      "epoch": 0.05666374902105312,
+      "grad_norm": 0.5125401330915287,
+      "learning_rate": 0.00019962834007481237,
+      "loss": 1.6568,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05670981710968812,
+      "grad_norm": 0.5441769645194086,
+      "learning_rate": 0.00019962705374248425,
+      "loss": 1.2174,
+      "step": 1231
+    },
+    {
+      "epoch": 0.05675588519832312,
+      "grad_norm": 0.5938945179071604,
+      "learning_rate": 0.0001996257651921306,
+      "loss": 1.4851,
+      "step": 1232
+    },
+    {
+      "epoch": 0.056801953286958126,
+      "grad_norm": 0.5229464729894249,
+      "learning_rate": 0.0001996244744237801,
+      "loss": 1.1371,
+      "step": 1233
+    },
+    {
+      "epoch": 0.056848021375593125,
+      "grad_norm": 0.4257926776279077,
+      "learning_rate": 0.00019962318143746147,
+      "loss": 1.1149,
+      "step": 1234
+    },
+    {
+      "epoch": 0.05689408946422813,
+      "grad_norm": 0.5918479072562083,
+      "learning_rate": 0.00019962188623320358,
+      "loss": 1.381,
+      "step": 1235
+    },
+    {
+      "epoch": 0.05694015755286313,
+      "grad_norm": 0.40360008868884756,
+      "learning_rate": 0.00019962058881103518,
+      "loss": 1.1747,
+      "step": 1236
+    },
+    {
+      "epoch": 0.056986225641498134,
+      "grad_norm": 0.51095061735553,
+      "learning_rate": 0.0001996192891709852,
+      "loss": 1.0624,
+      "step": 1237
+    },
+    {
+      "epoch": 0.05703229373013314,
+      "grad_norm": 0.49726699648697453,
+      "learning_rate": 0.00019961798731308252,
+      "loss": 1.0613,
+      "step": 1238
+    },
+    {
+      "epoch": 0.05707836181876814,
+      "grad_norm": 0.6701956889846986,
+      "learning_rate": 0.00019961668323735619,
+      "loss": 1.0776,
+      "step": 1239
+    },
+    {
+      "epoch": 0.05712442990740314,
+      "grad_norm": 0.509620973264032,
+      "learning_rate": 0.00019961537694383521,
+      "loss": 1.049,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05717049799603814,
+      "grad_norm": 0.44419571110641587,
+      "learning_rate": 0.00019961406843254867,
+      "loss": 1.3714,
+      "step": 1241
+    },
+    {
+      "epoch": 0.05721656608467315,
+      "grad_norm": 0.6102398764494618,
+      "learning_rate": 0.00019961275770352568,
+      "loss": 1.2997,
+      "step": 1242
+    },
+    {
+      "epoch": 0.05726263417330815,
+      "grad_norm": 0.7479482257125691,
+      "learning_rate": 0.00019961144475679545,
+      "loss": 1.3098,
+      "step": 1243
+    },
+    {
+      "epoch": 0.05730870226194315,
+      "grad_norm": 0.5353598513762766,
+      "learning_rate": 0.00019961012959238722,
+      "loss": 1.01,
+      "step": 1244
+    },
+    {
+      "epoch": 0.057354770350578156,
+      "grad_norm": 0.5573743473080479,
+      "learning_rate": 0.0001996088122103302,
+      "loss": 1.2224,
+      "step": 1245
+    },
+    {
+      "epoch": 0.057400838439213155,
+      "grad_norm": 0.9895097826830096,
+      "learning_rate": 0.0001996074926106538,
+      "loss": 1.2748,
+      "step": 1246
+    },
+    {
+      "epoch": 0.05744690652784816,
+      "grad_norm": 0.5801771020398605,
+      "learning_rate": 0.00019960617079338737,
+      "loss": 1.261,
+      "step": 1247
+    },
+    {
+      "epoch": 0.057492974616483165,
+      "grad_norm": 0.791492015194891,
+      "learning_rate": 0.00019960484675856033,
+      "loss": 1.6117,
+      "step": 1248
+    },
+    {
+      "epoch": 0.057539042705118164,
+      "grad_norm": 0.5572781859814486,
+      "learning_rate": 0.00019960352050620218,
+      "loss": 1.1052,
+      "step": 1249
+    },
+    {
+      "epoch": 0.05758511079375317,
+      "grad_norm": 0.7060207626868199,
+      "learning_rate": 0.0001996021920363424,
+      "loss": 1.0904,
+      "step": 1250
+    },
+    {
+      "epoch": 0.05763117888238817,
+      "grad_norm": 0.4717123319893251,
+      "learning_rate": 0.0001996008613490106,
+      "loss": 1.4151,
+      "step": 1251
+    },
+    {
+      "epoch": 0.05767724697102317,
+      "grad_norm": 0.6289783266561775,
+      "learning_rate": 0.0001995995284442364,
+      "loss": 1.389,
+      "step": 1252
+    },
+    {
+      "epoch": 0.05772331505965817,
+      "grad_norm": 0.757821271299619,
+      "learning_rate": 0.00019959819332204947,
+      "loss": 1.2544,
+      "step": 1253
+    },
+    {
+      "epoch": 0.05776938314829318,
+      "grad_norm": 0.6267938431831968,
+      "learning_rate": 0.00019959685598247958,
+      "loss": 1.4822,
+      "step": 1254
+    },
+    {
+      "epoch": 0.05781545123692818,
+      "grad_norm": 0.4287407825687138,
+      "learning_rate": 0.00019959551642555643,
+      "loss": 1.6132,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05786151932556318,
+      "grad_norm": 0.4807088709830865,
+      "learning_rate": 0.00019959417465130987,
+      "loss": 1.2665,
+      "step": 1256
+    },
+    {
+      "epoch": 0.057907587414198186,
+      "grad_norm": 0.44287254542013094,
+      "learning_rate": 0.00019959283065976982,
+      "loss": 1.07,
+      "step": 1257
+    },
+    {
+      "epoch": 0.057953655502833185,
+      "grad_norm": 0.6429565138185297,
+      "learning_rate": 0.00019959148445096613,
+      "loss": 1.2809,
+      "step": 1258
+    },
+    {
+      "epoch": 0.05799972359146819,
+      "grad_norm": 0.6815577419311504,
+      "learning_rate": 0.00019959013602492877,
+      "loss": 1.0076,
+      "step": 1259
+    },
+    {
+      "epoch": 0.058045791680103195,
+      "grad_norm": 0.5398380139773848,
+      "learning_rate": 0.00019958878538168783,
+      "loss": 1.671,
+      "step": 1260
+    },
+    {
+      "epoch": 0.058091859768738194,
+      "grad_norm": 0.5572191452661106,
+      "learning_rate": 0.00019958743252127334,
+      "loss": 1.4082,
+      "step": 1261
+    },
+    {
+      "epoch": 0.0581379278573732,
+      "grad_norm": 0.5863574917586308,
+      "learning_rate": 0.00019958607744371545,
+      "loss": 1.1746,
+      "step": 1262
+    },
+    {
+      "epoch": 0.0581839959460082,
+      "grad_norm": 0.8689582487470542,
+      "learning_rate": 0.00019958472014904426,
+      "loss": 1.1992,
+      "step": 1263
+    },
+    {
+      "epoch": 0.0582300640346432,
+      "grad_norm": 0.5255540476037924,
+      "learning_rate": 0.00019958336063729003,
+      "loss": 1.3473,
+      "step": 1264
+    },
+    {
+      "epoch": 0.05827613212327821,
+      "grad_norm": 0.7670699882765076,
+      "learning_rate": 0.000199581998908483,
+      "loss": 0.8586,
+      "step": 1265
+    },
+    {
+      "epoch": 0.05832220021191321,
+      "grad_norm": 0.6354927102447859,
+      "learning_rate": 0.00019958063496265357,
+      "loss": 1.2858,
+      "step": 1266
+    },
+    {
+      "epoch": 0.05836826830054821,
+      "grad_norm": 0.6092469233807053,
+      "learning_rate": 0.000199579268799832,
+      "loss": 1.2715,
+      "step": 1267
+    },
+    {
+      "epoch": 0.05841433638918321,
+      "grad_norm": 0.6161638212382382,
+      "learning_rate": 0.0001995779004200488,
+      "loss": 1.1691,
+      "step": 1268
+    },
+    {
+      "epoch": 0.058460404477818216,
+      "grad_norm": 0.45538518136565986,
+      "learning_rate": 0.00019957652982333435,
+      "loss": 1.0359,
+      "step": 1269
+    },
+    {
+      "epoch": 0.058506472566453215,
+      "grad_norm": 0.5851129425791407,
+      "learning_rate": 0.0001995751570097192,
+      "loss": 1.2105,
+      "step": 1270
+    },
+    {
+      "epoch": 0.05855254065508822,
+      "grad_norm": 0.6293223466426111,
+      "learning_rate": 0.0001995737819792339,
+      "loss": 1.2723,
+      "step": 1271
+    },
+    {
+      "epoch": 0.058598608743723225,
+      "grad_norm": 0.5435984299425763,
+      "learning_rate": 0.0001995724047319091,
+      "loss": 1.1913,
+      "step": 1272
+    },
+    {
+      "epoch": 0.058644676832358224,
+      "grad_norm": 0.5260783710711554,
+      "learning_rate": 0.00019957102526777542,
+      "loss": 1.2111,
+      "step": 1273
+    },
+    {
+      "epoch": 0.05869074492099323,
+      "grad_norm": 0.6867179305749185,
+      "learning_rate": 0.00019956964358686363,
+      "loss": 1.0712,
+      "step": 1274
+    },
+    {
+      "epoch": 0.05873681300962823,
+      "grad_norm": 0.46949157851483586,
+      "learning_rate": 0.00019956825968920442,
+      "loss": 1.1973,
+      "step": 1275
+    },
+    {
+      "epoch": 0.05878288109826323,
+      "grad_norm": 0.4911819454684082,
+      "learning_rate": 0.00019956687357482864,
+      "loss": 1.4318,
+      "step": 1276
+    },
+    {
+      "epoch": 0.05882894918689824,
+      "grad_norm": 0.5541587722985081,
+      "learning_rate": 0.00019956548524376711,
+      "loss": 1.2794,
+      "step": 1277
+    },
+    {
+      "epoch": 0.05887501727553324,
+      "grad_norm": 0.7050407895864879,
+      "learning_rate": 0.0001995640946960508,
+      "loss": 1.0372,
+      "step": 1278
+    },
+    {
+      "epoch": 0.05892108536416824,
+      "grad_norm": 0.43682446898437444,
+      "learning_rate": 0.00019956270193171064,
+      "loss": 1.0246,
+      "step": 1279
+    },
+    {
+      "epoch": 0.05896715345280324,
+      "grad_norm": 0.5174563820744016,
+      "learning_rate": 0.00019956130695077766,
+      "loss": 1.4657,
+      "step": 1280
+    },
+    {
+      "epoch": 0.059013221541438246,
+      "grad_norm": 0.46761799433717344,
+      "learning_rate": 0.00019955990975328286,
+      "loss": 1.0646,
+      "step": 1281
+    },
+    {
+      "epoch": 0.05905928963007325,
+      "grad_norm": 0.5133873352526669,
+      "learning_rate": 0.00019955851033925736,
+      "loss": 1.1917,
+      "step": 1282
+    },
+    {
+      "epoch": 0.05910535771870825,
+      "grad_norm": 0.49590111540242765,
+      "learning_rate": 0.00019955710870873236,
+      "loss": 1.6089,
+      "step": 1283
+    },
+    {
+      "epoch": 0.059151425807343255,
+      "grad_norm": 0.47028426475516555,
+      "learning_rate": 0.000199555704861739,
+      "loss": 1.4346,
+      "step": 1284
+    },
+    {
+      "epoch": 0.059197493895978254,
+      "grad_norm": 0.49837388248065817,
+      "learning_rate": 0.00019955429879830864,
+      "loss": 1.4785,
+      "step": 1285
+    },
+    {
+      "epoch": 0.05924356198461326,
+      "grad_norm": 0.4845491029846576,
+      "learning_rate": 0.00019955289051847247,
+      "loss": 1.2591,
+      "step": 1286
+    },
+    {
+      "epoch": 0.05928963007324826,
+      "grad_norm": 0.46500364565433927,
+      "learning_rate": 0.0001995514800222619,
+      "loss": 1.287,
+      "step": 1287
+    },
+    {
+      "epoch": 0.05933569816188326,
+      "grad_norm": 0.5273609590969057,
+      "learning_rate": 0.00019955006730970834,
+      "loss": 1.4519,
+      "step": 1288
+    },
+    {
+      "epoch": 0.05938176625051827,
+      "grad_norm": 1.024260329420094,
+      "learning_rate": 0.0001995486523808432,
+      "loss": 1.1499,
+      "step": 1289
+    },
+    {
+      "epoch": 0.05942783433915327,
+      "grad_norm": 0.6515198980140812,
+      "learning_rate": 0.000199547235235698,
+      "loss": 1.2513,
+      "step": 1290
+    },
+    {
+      "epoch": 0.05947390242778827,
+      "grad_norm": 0.45957696528166486,
+      "learning_rate": 0.0001995458158743043,
+      "loss": 1.3258,
+      "step": 1291
+    },
+    {
+      "epoch": 0.05951997051642327,
+      "grad_norm": 0.460522578522121,
+      "learning_rate": 0.0001995443942966937,
+      "loss": 1.1411,
+      "step": 1292
+    },
+    {
+      "epoch": 0.059566038605058276,
+      "grad_norm": 0.5129848900817557,
+      "learning_rate": 0.00019954297050289783,
+      "loss": 0.8968,
+      "step": 1293
+    },
+    {
+      "epoch": 0.05961210669369328,
+      "grad_norm": 0.5879679602983002,
+      "learning_rate": 0.0001995415444929484,
+      "loss": 1.325,
+      "step": 1294
+    },
+    {
+      "epoch": 0.05965817478232828,
+      "grad_norm": 0.47530595105795576,
+      "learning_rate": 0.00019954011626687715,
+      "loss": 1.1542,
+      "step": 1295
+    },
+    {
+      "epoch": 0.059704242870963285,
+      "grad_norm": 0.4931516825303222,
+      "learning_rate": 0.00019953868582471592,
+      "loss": 1.206,
+      "step": 1296
+    },
+    {
+      "epoch": 0.059750310959598284,
+      "grad_norm": 0.5450533595359576,
+      "learning_rate": 0.0001995372531664965,
+      "loss": 1.1955,
+      "step": 1297
+    },
+    {
+      "epoch": 0.05979637904823329,
+      "grad_norm": 0.6416090634606617,
+      "learning_rate": 0.00019953581829225083,
+      "loss": 1.3917,
+      "step": 1298
+    },
+    {
+      "epoch": 0.059842447136868294,
+      "grad_norm": 0.5587821472379922,
+      "learning_rate": 0.0001995343812020108,
+      "loss": 1.1473,
+      "step": 1299
+    },
+    {
+      "epoch": 0.05988851522550329,
+      "grad_norm": 0.5328602482588007,
+      "learning_rate": 0.00019953294189580846,
+      "loss": 1.2796,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0599345833141383,
+      "grad_norm": 0.4532115494774216,
+      "learning_rate": 0.00019953150037367584,
+      "loss": 1.4908,
+      "step": 1301
+    },
+    {
+      "epoch": 0.0599806514027733,
+      "grad_norm": 0.5411681940574616,
+      "learning_rate": 0.00019953005663564502,
+      "loss": 1.1804,
+      "step": 1302
+    },
+    {
+      "epoch": 0.0600267194914083,
+      "grad_norm": 0.5864726602081987,
+      "learning_rate": 0.00019952861068174812,
+      "loss": 1.2625,
+      "step": 1303
+    },
+    {
+      "epoch": 0.0600727875800433,
+      "grad_norm": 0.5605786049723013,
+      "learning_rate": 0.0001995271625120174,
+      "loss": 1.1572,
+      "step": 1304
+    },
+    {
+      "epoch": 0.060118855668678306,
+      "grad_norm": 0.6070508616696795,
+      "learning_rate": 0.00019952571212648504,
+      "loss": 1.1047,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06016492375731331,
+      "grad_norm": 0.5512185333010151,
+      "learning_rate": 0.00019952425952518335,
+      "loss": 1.4282,
+      "step": 1306
+    },
+    {
+      "epoch": 0.06021099184594831,
+      "grad_norm": 0.4755976995214113,
+      "learning_rate": 0.00019952280470814465,
+      "loss": 1.0758,
+      "step": 1307
+    },
+    {
+      "epoch": 0.060257059934583315,
+      "grad_norm": 0.4760608266413914,
+      "learning_rate": 0.00019952134767540136,
+      "loss": 1.3842,
+      "step": 1308
+    },
+    {
+      "epoch": 0.060303128023218314,
+      "grad_norm": 0.5280011588504248,
+      "learning_rate": 0.0001995198884269859,
+      "loss": 1.3474,
+      "step": 1309
+    },
+    {
+      "epoch": 0.06034919611185332,
+      "grad_norm": 0.5326697441087564,
+      "learning_rate": 0.0001995184269629308,
+      "loss": 1.1814,
+      "step": 1310
+    },
+    {
+      "epoch": 0.060395264200488324,
+      "grad_norm": 0.7110215201797097,
+      "learning_rate": 0.00019951696328326853,
+      "loss": 1.0684,
+      "step": 1311
+    },
+    {
+      "epoch": 0.06044133228912332,
+      "grad_norm": 0.5050347362517732,
+      "learning_rate": 0.0001995154973880317,
+      "loss": 1.3423,
+      "step": 1312
+    },
+    {
+      "epoch": 0.06048740037775833,
+      "grad_norm": 0.5793894930355048,
+      "learning_rate": 0.00019951402927725296,
+      "loss": 1.4424,
+      "step": 1313
+    },
+    {
+      "epoch": 0.06053346846639333,
+      "grad_norm": 0.5044990841406225,
+      "learning_rate": 0.000199512558950965,
+      "loss": 1.1616,
+      "step": 1314
+    },
+    {
+      "epoch": 0.06057953655502833,
+      "grad_norm": 0.4927499310170278,
+      "learning_rate": 0.00019951108640920052,
+      "loss": 1.5602,
+      "step": 1315
+    },
+    {
+      "epoch": 0.06062560464366334,
+      "grad_norm": 0.46265472395867113,
+      "learning_rate": 0.00019950961165199235,
+      "loss": 1.601,
+      "step": 1316
+    },
+    {
+      "epoch": 0.060671672732298336,
+      "grad_norm": 0.45880579044726405,
+      "learning_rate": 0.00019950813467937327,
+      "loss": 1.1013,
+      "step": 1317
+    },
+    {
+      "epoch": 0.06071774082093334,
+      "grad_norm": 0.4513469882613257,
+      "learning_rate": 0.00019950665549137625,
+      "loss": 1.4325,
+      "step": 1318
+    },
+    {
+      "epoch": 0.06076380890956834,
+      "grad_norm": 0.44945499187507204,
+      "learning_rate": 0.0001995051740880341,
+      "loss": 1.2731,
+      "step": 1319
+    },
+    {
+      "epoch": 0.060809876998203345,
+      "grad_norm": 0.6205643722585431,
+      "learning_rate": 0.00019950369046937988,
+      "loss": 1.0498,
+      "step": 1320
+    },
+    {
+      "epoch": 0.060855945086838344,
+      "grad_norm": 0.7362813013764526,
+      "learning_rate": 0.0001995022046354466,
+      "loss": 0.8249,
+      "step": 1321
+    },
+    {
+      "epoch": 0.06090201317547335,
+      "grad_norm": 0.6184642877333666,
+      "learning_rate": 0.0001995007165862674,
+      "loss": 1.0746,
+      "step": 1322
+    },
+    {
+      "epoch": 0.060948081264108354,
+      "grad_norm": 0.4987796247608053,
+      "learning_rate": 0.0001994992263218753,
+      "loss": 1.1324,
+      "step": 1323
+    },
+    {
+      "epoch": 0.06099414935274335,
+      "grad_norm": 0.4893855949202129,
+      "learning_rate": 0.00019949773384230355,
+      "loss": 1.5746,
+      "step": 1324
+    },
+    {
+      "epoch": 0.06104021744137836,
+      "grad_norm": 0.6119223266502348,
+      "learning_rate": 0.00019949623914758532,
+      "loss": 1.2522,
+      "step": 1325
+    },
+    {
+      "epoch": 0.06108628553001336,
+      "grad_norm": 0.5423667614965221,
+      "learning_rate": 0.00019949474223775396,
+      "loss": 1.2579,
+      "step": 1326
+    },
+    {
+      "epoch": 0.06113235361864836,
+      "grad_norm": 0.5074176816813126,
+      "learning_rate": 0.00019949324311284276,
+      "loss": 1.4141,
+      "step": 1327
+    },
+    {
+      "epoch": 0.06117842170728337,
+      "grad_norm": 0.4532242162723257,
+      "learning_rate": 0.00019949174177288508,
+      "loss": 1.31,
+      "step": 1328
+    },
+    {
+      "epoch": 0.061224489795918366,
+      "grad_norm": 0.6730783162530916,
+      "learning_rate": 0.00019949023821791436,
+      "loss": 1.1958,
+      "step": 1329
+    },
+    {
+      "epoch": 0.06127055788455337,
+      "grad_norm": 0.5020549022233975,
+      "learning_rate": 0.0001994887324479641,
+      "loss": 1.078,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06131662597318837,
+      "grad_norm": 0.4662152129466529,
+      "learning_rate": 0.00019948722446306777,
+      "loss": 1.4879,
+      "step": 1331
+    },
+    {
+      "epoch": 0.061362694061823375,
+      "grad_norm": 0.45845735680932176,
+      "learning_rate": 0.000199485714263259,
+      "loss": 1.2346,
+      "step": 1332
+    },
+    {
+      "epoch": 0.06140876215045838,
+      "grad_norm": 0.5543024504655271,
+      "learning_rate": 0.00019948420184857138,
+      "loss": 1.0316,
+      "step": 1333
+    },
+    {
+      "epoch": 0.06145483023909338,
+      "grad_norm": 0.4495489904882786,
+      "learning_rate": 0.00019948268721903855,
+      "loss": 1.1705,
+      "step": 1334
+    },
+    {
+      "epoch": 0.061500898327728384,
+      "grad_norm": 0.5347366508177754,
+      "learning_rate": 0.0001994811703746943,
+      "loss": 1.3854,
+      "step": 1335
+    },
+    {
+      "epoch": 0.06154696641636338,
+      "grad_norm": 0.5028390878472736,
+      "learning_rate": 0.00019947965131557232,
+      "loss": 1.072,
+      "step": 1336
+    },
+    {
+      "epoch": 0.06159303450499839,
+      "grad_norm": 0.5538282155622763,
+      "learning_rate": 0.00019947813004170652,
+      "loss": 1.2352,
+      "step": 1337
+    },
+    {
+      "epoch": 0.061639102593633394,
+      "grad_norm": 0.68458879510938,
+      "learning_rate": 0.00019947660655313075,
+      "loss": 1.1322,
+      "step": 1338
+    },
+    {
+      "epoch": 0.06168517068226839,
+      "grad_norm": 0.3804694594787416,
+      "learning_rate": 0.00019947508084987883,
+      "loss": 1.0827,
+      "step": 1339
+    },
+    {
+      "epoch": 0.0617312387709034,
+      "grad_norm": 0.4476550530070771,
+      "learning_rate": 0.00019947355293198485,
+      "loss": 1.5049,
+      "step": 1340
+    },
+    {
+      "epoch": 0.061777306859538396,
+      "grad_norm": 0.5148141986874867,
+      "learning_rate": 0.00019947202279948275,
+      "loss": 1.5143,
+      "step": 1341
+    },
+    {
+      "epoch": 0.0618233749481734,
+      "grad_norm": 0.5120477559915403,
+      "learning_rate": 0.00019947049045240662,
+      "loss": 1.3374,
+      "step": 1342
+    },
+    {
+      "epoch": 0.0618694430368084,
+      "grad_norm": 0.4776529731627882,
+      "learning_rate": 0.00019946895589079058,
+      "loss": 0.7182,
+      "step": 1343
+    },
+    {
+      "epoch": 0.061915511125443405,
+      "grad_norm": 0.45672316916811023,
+      "learning_rate": 0.00019946741911466877,
+      "loss": 1.1556,
+      "step": 1344
+    },
+    {
+      "epoch": 0.06196157921407841,
+      "grad_norm": 0.554828534833876,
+      "learning_rate": 0.00019946588012407544,
+      "loss": 1.2342,
+      "step": 1345
+    },
+    {
+      "epoch": 0.06200764730271341,
+      "grad_norm": 0.4305724291421675,
+      "learning_rate": 0.00019946433891904485,
+      "loss": 0.7903,
+      "step": 1346
+    },
+    {
+      "epoch": 0.062053715391348414,
+      "grad_norm": 0.49364032911700106,
+      "learning_rate": 0.00019946279549961127,
+      "loss": 1.0294,
+      "step": 1347
+    },
+    {
+      "epoch": 0.06209978347998341,
+      "grad_norm": 0.5708667910782352,
+      "learning_rate": 0.00019946124986580915,
+      "loss": 1.24,
+      "step": 1348
+    },
+    {
+      "epoch": 0.06214585156861842,
+      "grad_norm": 0.5399866784861911,
+      "learning_rate": 0.00019945970201767278,
+      "loss": 1.4918,
+      "step": 1349
+    },
+    {
+      "epoch": 0.062191919657253424,
+      "grad_norm": 0.6016950282826361,
+      "learning_rate": 0.0001994581519552367,
+      "loss": 1.2855,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06223798774588842,
+      "grad_norm": 0.5177084871596485,
+      "learning_rate": 0.0001994565996785354,
+      "loss": 1.5203,
+      "step": 1351
+    },
+    {
+      "epoch": 0.06228405583452343,
+      "grad_norm": 0.5639367127626707,
+      "learning_rate": 0.00019945504518760346,
+      "loss": 1.1491,
+      "step": 1352
+    },
+    {
+      "epoch": 0.062330123923158426,
+      "grad_norm": 0.6552030304641707,
+      "learning_rate": 0.00019945348848247546,
+      "loss": 1.5928,
+      "step": 1353
+    },
+    {
+      "epoch": 0.06237619201179343,
+      "grad_norm": 0.523732090837565,
+      "learning_rate": 0.00019945192956318604,
+      "loss": 1.3971,
+      "step": 1354
+    },
+    {
+      "epoch": 0.06242226010042844,
+      "grad_norm": 0.498941564922459,
+      "learning_rate": 0.00019945036842976995,
+      "loss": 1.4185,
+      "step": 1355
+    },
+    {
+      "epoch": 0.062468328189063435,
+      "grad_norm": 0.5198416760889913,
+      "learning_rate": 0.00019944880508226192,
+      "loss": 1.2953,
+      "step": 1356
+    },
+    {
+      "epoch": 0.06251439627769843,
+      "grad_norm": 0.5020123637889172,
+      "learning_rate": 0.00019944723952069676,
+      "loss": 1.3743,
+      "step": 1357
+    },
+    {
+      "epoch": 0.06256046436633345,
+      "grad_norm": 0.6137202927022659,
+      "learning_rate": 0.00019944567174510932,
+      "loss": 1.2777,
+      "step": 1358
+    },
+    {
+      "epoch": 0.06260653245496844,
+      "grad_norm": 0.48466468761056697,
+      "learning_rate": 0.00019944410175553454,
+      "loss": 1.678,
+      "step": 1359
+    },
+    {
+      "epoch": 0.06265260054360344,
+      "grad_norm": 0.5185067014420517,
+      "learning_rate": 0.00019944252955200734,
+      "loss": 1.1626,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06269866863223846,
+      "grad_norm": 0.5558897779144246,
+      "learning_rate": 0.00019944095513456272,
+      "loss": 0.8356,
+      "step": 1361
+    },
+    {
+      "epoch": 0.06274473672087345,
+      "grad_norm": 0.7889041594519091,
+      "learning_rate": 0.00019943937850323577,
+      "loss": 1.1524,
+      "step": 1362
+    },
+    {
+      "epoch": 0.06279080480950845,
+      "grad_norm": 0.421403762129503,
+      "learning_rate": 0.0001994377996580615,
+      "loss": 1.5688,
+      "step": 1363
+    },
+    {
+      "epoch": 0.06283687289814345,
+      "grad_norm": 0.49294709170040485,
+      "learning_rate": 0.00019943621859907514,
+      "loss": 0.9489,
+      "step": 1364
+    },
+    {
+      "epoch": 0.06288294098677846,
+      "grad_norm": 0.5306755485716318,
+      "learning_rate": 0.00019943463532631187,
+      "loss": 1.4445,
+      "step": 1365
+    },
+    {
+      "epoch": 0.06292900907541346,
+      "grad_norm": 0.4831809325567703,
+      "learning_rate": 0.00019943304983980695,
+      "loss": 1.1369,
+      "step": 1366
+    },
+    {
+      "epoch": 0.06297507716404846,
+      "grad_norm": 0.5515436765397258,
+      "learning_rate": 0.00019943146213959568,
+      "loss": 1.5168,
+      "step": 1367
+    },
+    {
+      "epoch": 0.06302114525268347,
+      "grad_norm": 0.476319108876044,
+      "learning_rate": 0.00019942987222571333,
+      "loss": 1.3463,
+      "step": 1368
+    },
+    {
+      "epoch": 0.06306721334131847,
+      "grad_norm": 0.6281735612813424,
+      "learning_rate": 0.00019942828009819543,
+      "loss": 1.3404,
+      "step": 1369
+    },
+    {
+      "epoch": 0.06311328142995347,
+      "grad_norm": 0.6088899960732135,
+      "learning_rate": 0.00019942668575707731,
+      "loss": 1.3595,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06315934951858847,
+      "grad_norm": 0.5447457927575201,
+      "learning_rate": 0.00019942508920239456,
+      "loss": 0.962,
+      "step": 1371
+    },
+    {
+      "epoch": 0.06320541760722348,
+      "grad_norm": 0.5695050936253727,
+      "learning_rate": 0.00019942349043418264,
+      "loss": 0.9703,
+      "step": 1372
+    },
+    {
+      "epoch": 0.06325148569585848,
+      "grad_norm": 0.47144607015099627,
+      "learning_rate": 0.00019942188945247717,
+      "loss": 1.2844,
+      "step": 1373
+    },
+    {
+      "epoch": 0.06329755378449348,
+      "grad_norm": 0.5315238496115361,
+      "learning_rate": 0.00019942028625731384,
+      "loss": 0.9589,
+      "step": 1374
+    },
+    {
+      "epoch": 0.06334362187312849,
+      "grad_norm": 0.6513899759424725,
+      "learning_rate": 0.0001994186808487283,
+      "loss": 1.2335,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06338968996176349,
+      "grad_norm": 0.7342701211574141,
+      "learning_rate": 0.00019941707322675627,
+      "loss": 0.9923,
+      "step": 1376
+    },
+    {
+      "epoch": 0.06343575805039849,
+      "grad_norm": 0.48026853272159115,
+      "learning_rate": 0.00019941546339143358,
+      "loss": 1.4816,
+      "step": 1377
+    },
+    {
+      "epoch": 0.0634818261390335,
+      "grad_norm": 0.5146840313023356,
+      "learning_rate": 0.00019941385134279607,
+      "loss": 0.9123,
+      "step": 1378
+    },
+    {
+      "epoch": 0.0635278942276685,
+      "grad_norm": 0.46370049403088853,
+      "learning_rate": 0.0001994122370808796,
+      "loss": 0.9167,
+      "step": 1379
+    },
+    {
+      "epoch": 0.0635739623163035,
+      "grad_norm": 0.5001578301086121,
+      "learning_rate": 0.00019941062060572015,
+      "loss": 1.3803,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0636200304049385,
+      "grad_norm": 0.6873909093999854,
+      "learning_rate": 0.00019940900191735367,
+      "loss": 1.1105,
+      "step": 1381
+    },
+    {
+      "epoch": 0.0636660984935735,
+      "grad_norm": 0.526176467957318,
+      "learning_rate": 0.00019940738101581623,
+      "loss": 1.3393,
+      "step": 1382
+    },
+    {
+      "epoch": 0.0637121665822085,
+      "grad_norm": 0.5344916627158632,
+      "learning_rate": 0.00019940575790114385,
+      "loss": 1.0203,
+      "step": 1383
+    },
+    {
+      "epoch": 0.0637582346708435,
+      "grad_norm": 0.5717100463188982,
+      "learning_rate": 0.00019940413257337278,
+      "loss": 0.9078,
+      "step": 1384
+    },
+    {
+      "epoch": 0.06380430275947852,
+      "grad_norm": 0.4854115690563426,
+      "learning_rate": 0.0001994025050325391,
+      "loss": 1.1602,
+      "step": 1385
+    },
+    {
+      "epoch": 0.06385037084811351,
+      "grad_norm": 0.4266745606118856,
+      "learning_rate": 0.00019940087527867907,
+      "loss": 1.3413,
+      "step": 1386
+    },
+    {
+      "epoch": 0.06389643893674851,
+      "grad_norm": 0.7583075835897269,
+      "learning_rate": 0.00019939924331182902,
+      "loss": 1.1071,
+      "step": 1387
+    },
+    {
+      "epoch": 0.06394250702538351,
+      "grad_norm": 0.5873107578560618,
+      "learning_rate": 0.00019939760913202526,
+      "loss": 1.2352,
+      "step": 1388
+    },
+    {
+      "epoch": 0.06398857511401852,
+      "grad_norm": 0.6025092970861982,
+      "learning_rate": 0.00019939597273930412,
+      "loss": 1.2206,
+      "step": 1389
+    },
+    {
+      "epoch": 0.06403464320265352,
+      "grad_norm": 0.4449445647117065,
+      "learning_rate": 0.00019939433413370213,
+      "loss": 0.8453,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06408071129128852,
+      "grad_norm": 0.4578471622399329,
+      "learning_rate": 0.00019939269331525566,
+      "loss": 1.351,
+      "step": 1391
+    },
+    {
+      "epoch": 0.06412677937992353,
+      "grad_norm": 0.3955295203524017,
+      "learning_rate": 0.00019939105028400134,
+      "loss": 1.3468,
+      "step": 1392
+    },
+    {
+      "epoch": 0.06417284746855853,
+      "grad_norm": 0.5100900345945101,
+      "learning_rate": 0.00019938940503997566,
+      "loss": 1.3876,
+      "step": 1393
+    },
+    {
+      "epoch": 0.06421891555719353,
+      "grad_norm": 0.49813730811517765,
+      "learning_rate": 0.00019938775758321532,
+      "loss": 1.3205,
+      "step": 1394
+    },
+    {
+      "epoch": 0.06426498364582854,
+      "grad_norm": 0.5178480688130394,
+      "learning_rate": 0.00019938610791375697,
+      "loss": 1.0289,
+      "step": 1395
+    },
+    {
+      "epoch": 0.06431105173446354,
+      "grad_norm": 0.4694198949185241,
+      "learning_rate": 0.00019938445603163733,
+      "loss": 0.6773,
+      "step": 1396
+    },
+    {
+      "epoch": 0.06435711982309854,
+      "grad_norm": 0.5096822363675494,
+      "learning_rate": 0.0001993828019368932,
+      "loss": 1.2156,
+      "step": 1397
+    },
+    {
+      "epoch": 0.06440318791173354,
+      "grad_norm": 0.4616837942020975,
+      "learning_rate": 0.00019938114562956138,
+      "loss": 1.4359,
+      "step": 1398
+    },
+    {
+      "epoch": 0.06444925600036855,
+      "grad_norm": 0.5982708563012709,
+      "learning_rate": 0.00019937948710967875,
+      "loss": 1.4993,
+      "step": 1399
+    },
+    {
+      "epoch": 0.06449532408900355,
+      "grad_norm": 0.541580627761112,
+      "learning_rate": 0.00019937782637728226,
+      "loss": 1.2607,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06454139217763855,
+      "grad_norm": 0.7226262042846165,
+      "learning_rate": 0.00019937616343240882,
+      "loss": 1.3883,
+      "step": 1401
+    },
+    {
+      "epoch": 0.06458746026627356,
+      "grad_norm": 0.6886577384412403,
+      "learning_rate": 0.00019937449827509555,
+      "loss": 1.3927,
+      "step": 1402
+    },
+    {
+      "epoch": 0.06463352835490856,
+      "grad_norm": 0.7030504763687206,
+      "learning_rate": 0.00019937283090537943,
+      "loss": 1.2935,
+      "step": 1403
+    },
+    {
+      "epoch": 0.06467959644354355,
+      "grad_norm": 0.6431938305249112,
+      "learning_rate": 0.0001993711613232976,
+      "loss": 1.2249,
+      "step": 1404
+    },
+    {
+      "epoch": 0.06472566453217855,
+      "grad_norm": 0.4440054611229617,
+      "learning_rate": 0.0001993694895288873,
+      "loss": 1.572,
+      "step": 1405
+    },
+    {
+      "epoch": 0.06477173262081357,
+      "grad_norm": 0.561371104344811,
+      "learning_rate": 0.00019936781552218563,
+      "loss": 0.8569,
+      "step": 1406
+    },
+    {
+      "epoch": 0.06481780070944856,
+      "grad_norm": 0.5289359869187877,
+      "learning_rate": 0.00019936613930323,
+      "loss": 1.138,
+      "step": 1407
+    },
+    {
+      "epoch": 0.06486386879808356,
+      "grad_norm": 0.48282088886326274,
+      "learning_rate": 0.0001993644608720576,
+      "loss": 1.183,
+      "step": 1408
+    },
+    {
+      "epoch": 0.06490993688671857,
+      "grad_norm": 0.5152717893446296,
+      "learning_rate": 0.00019936278022870588,
+      "loss": 1.4601,
+      "step": 1409
+    },
+    {
+      "epoch": 0.06495600497535357,
+      "grad_norm": 0.5835442700514574,
+      "learning_rate": 0.00019936109737321223,
+      "loss": 1.0244,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06500207306398857,
+      "grad_norm": 0.4737015373543085,
+      "learning_rate": 0.0001993594123056141,
+      "loss": 1.2344,
+      "step": 1411
+    },
+    {
+      "epoch": 0.06504814115262358,
+      "grad_norm": 0.47790839263972784,
+      "learning_rate": 0.00019935772502594902,
+      "loss": 1.1767,
+      "step": 1412
+    },
+    {
+      "epoch": 0.06509420924125858,
+      "grad_norm": 0.6348881790366337,
+      "learning_rate": 0.00019935603553425458,
+      "loss": 1.1579,
+      "step": 1413
+    },
+    {
+      "epoch": 0.06514027732989358,
+      "grad_norm": 0.48328136889667034,
+      "learning_rate": 0.00019935434383056835,
+      "loss": 1.3481,
+      "step": 1414
+    },
+    {
+      "epoch": 0.06518634541852858,
+      "grad_norm": 0.5741847912421049,
+      "learning_rate": 0.000199352649914928,
+      "loss": 1.1756,
+      "step": 1415
+    },
+    {
+      "epoch": 0.06523241350716359,
+      "grad_norm": 0.7071320104763494,
+      "learning_rate": 0.0001993509537873713,
+      "loss": 1.1709,
+      "step": 1416
+    },
+    {
+      "epoch": 0.06527848159579859,
+      "grad_norm": 0.5853563971099917,
+      "learning_rate": 0.0001993492554479359,
+      "loss": 1.3243,
+      "step": 1417
+    },
+    {
+      "epoch": 0.06532454968443359,
+      "grad_norm": 0.5130507242073667,
+      "learning_rate": 0.00019934755489665968,
+      "loss": 1.5355,
+      "step": 1418
+    },
+    {
+      "epoch": 0.0653706177730686,
+      "grad_norm": 0.5120972268693834,
+      "learning_rate": 0.00019934585213358052,
+      "loss": 1.2941,
+      "step": 1419
+    },
+    {
+      "epoch": 0.0654166858617036,
+      "grad_norm": 0.6025182334909026,
+      "learning_rate": 0.00019934414715873632,
+      "loss": 1.3236,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0654627539503386,
+      "grad_norm": 0.5144812145951686,
+      "learning_rate": 0.000199342439972165,
+      "loss": 1.0965,
+      "step": 1421
+    },
+    {
+      "epoch": 0.0655088220389736,
+      "grad_norm": 0.5346228903530887,
+      "learning_rate": 0.0001993407305739046,
+      "loss": 1.2301,
+      "step": 1422
+    },
+    {
+      "epoch": 0.06555489012760861,
+      "grad_norm": 0.5190272762381919,
+      "learning_rate": 0.00019933901896399317,
+      "loss": 1.0991,
+      "step": 1423
+    },
+    {
+      "epoch": 0.06560095821624361,
+      "grad_norm": 0.5381477051477465,
+      "learning_rate": 0.00019933730514246878,
+      "loss": 1.0965,
+      "step": 1424
+    },
+    {
+      "epoch": 0.0656470263048786,
+      "grad_norm": 0.5770640675734682,
+      "learning_rate": 0.00019933558910936966,
+      "loss": 1.1832,
+      "step": 1425
+    },
+    {
+      "epoch": 0.06569309439351362,
+      "grad_norm": 0.5671677715358918,
+      "learning_rate": 0.00019933387086473395,
+      "loss": 1.1294,
+      "step": 1426
+    },
+    {
+      "epoch": 0.06573916248214862,
+      "grad_norm": 0.5019769120602479,
+      "learning_rate": 0.00019933215040859992,
+      "loss": 1.1904,
+      "step": 1427
+    },
+    {
+      "epoch": 0.06578523057078361,
+      "grad_norm": 0.6643936996268444,
+      "learning_rate": 0.00019933042774100587,
+      "loss": 1.3169,
+      "step": 1428
+    },
+    {
+      "epoch": 0.06583129865941863,
+      "grad_norm": 0.8925827406707039,
+      "learning_rate": 0.00019932870286199018,
+      "loss": 1.2021,
+      "step": 1429
+    },
+    {
+      "epoch": 0.06587736674805363,
+      "grad_norm": 0.6063830569061394,
+      "learning_rate": 0.00019932697577159126,
+      "loss": 1.3789,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06592343483668862,
+      "grad_norm": 0.5397822873945308,
+      "learning_rate": 0.0001993252464698475,
+      "loss": 1.1465,
+      "step": 1431
+    },
+    {
+      "epoch": 0.06596950292532362,
+      "grad_norm": 0.8376237570468618,
+      "learning_rate": 0.00019932351495679745,
+      "loss": 1.3297,
+      "step": 1432
+    },
+    {
+      "epoch": 0.06601557101395863,
+      "grad_norm": 0.5214308680777722,
+      "learning_rate": 0.0001993217812324796,
+      "loss": 1.1143,
+      "step": 1433
+    },
+    {
+      "epoch": 0.06606163910259363,
+      "grad_norm": 0.4574529620619701,
+      "learning_rate": 0.00019932004529693264,
+      "loss": 1.5405,
+      "step": 1434
+    },
+    {
+      "epoch": 0.06610770719122863,
+      "grad_norm": 0.5030982002932697,
+      "learning_rate": 0.00019931830715019515,
+      "loss": 1.3875,
+      "step": 1435
+    },
+    {
+      "epoch": 0.06615377527986364,
+      "grad_norm": 0.49369804240947424,
+      "learning_rate": 0.00019931656679230585,
+      "loss": 1.1961,
+      "step": 1436
+    },
+    {
+      "epoch": 0.06619984336849864,
+      "grad_norm": 0.4883708315263336,
+      "learning_rate": 0.00019931482422330348,
+      "loss": 1.39,
+      "step": 1437
+    },
+    {
+      "epoch": 0.06624591145713364,
+      "grad_norm": 0.45948385366356753,
+      "learning_rate": 0.00019931307944322683,
+      "loss": 0.998,
+      "step": 1438
+    },
+    {
+      "epoch": 0.06629197954576864,
+      "grad_norm": 0.5967903876653831,
+      "learning_rate": 0.00019931133245211475,
+      "loss": 0.8829,
+      "step": 1439
+    },
+    {
+      "epoch": 0.06633804763440365,
+      "grad_norm": 0.5115710118145218,
+      "learning_rate": 0.00019930958325000618,
+      "loss": 1.4585,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06638411572303865,
+      "grad_norm": 0.516769388614782,
+      "learning_rate": 0.00019930783183693995,
+      "loss": 1.1432,
+      "step": 1441
+    },
+    {
+      "epoch": 0.06643018381167365,
+      "grad_norm": 0.47694525773320773,
+      "learning_rate": 0.00019930607821295516,
+      "loss": 1.2919,
+      "step": 1442
+    },
+    {
+      "epoch": 0.06647625190030866,
+      "grad_norm": 0.5085738154911145,
+      "learning_rate": 0.00019930432237809083,
+      "loss": 1.3158,
+      "step": 1443
+    },
+    {
+      "epoch": 0.06652231998894366,
+      "grad_norm": 0.6351016646083391,
+      "learning_rate": 0.000199302564332386,
+      "loss": 1.1553,
+      "step": 1444
+    },
+    {
+      "epoch": 0.06656838807757866,
+      "grad_norm": 0.756235323226093,
+      "learning_rate": 0.00019930080407587982,
+      "loss": 1.1966,
+      "step": 1445
+    },
+    {
+      "epoch": 0.06661445616621367,
+      "grad_norm": 0.4957037248090258,
+      "learning_rate": 0.00019929904160861155,
+      "loss": 0.8039,
+      "step": 1446
+    },
+    {
+      "epoch": 0.06666052425484867,
+      "grad_norm": 0.606443367437839,
+      "learning_rate": 0.00019929727693062036,
+      "loss": 1.3322,
+      "step": 1447
+    },
+    {
+      "epoch": 0.06670659234348367,
+      "grad_norm": 0.5205424546385261,
+      "learning_rate": 0.00019929551004194558,
+      "loss": 1.1561,
+      "step": 1448
+    },
+    {
+      "epoch": 0.06675266043211867,
+      "grad_norm": 0.5926748483388383,
+      "learning_rate": 0.0001992937409426265,
+      "loss": 1.3874,
+      "step": 1449
+    },
+    {
+      "epoch": 0.06679872852075368,
+      "grad_norm": 0.6074527723089301,
+      "learning_rate": 0.0001992919696327025,
+      "loss": 0.874,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06684479660938868,
+      "grad_norm": 1.4513228881479363,
+      "learning_rate": 0.0001992901961122131,
+      "loss": 1.0913,
+      "step": 1451
+    },
+    {
+      "epoch": 0.06689086469802367,
+      "grad_norm": 0.8504638415436186,
+      "learning_rate": 0.0001992884203811977,
+      "loss": 1.4583,
+      "step": 1452
+    },
+    {
+      "epoch": 0.06693693278665869,
+      "grad_norm": 0.8836375278321694,
+      "learning_rate": 0.00019928664243969589,
+      "loss": 1.3176,
+      "step": 1453
+    },
+    {
+      "epoch": 0.06698300087529369,
+      "grad_norm": 0.9328994796518115,
+      "learning_rate": 0.00019928486228774718,
+      "loss": 1.1923,
+      "step": 1454
+    },
+    {
+      "epoch": 0.06702906896392868,
+      "grad_norm": 0.5536740550668776,
+      "learning_rate": 0.0001992830799253913,
+      "loss": 1.7102,
+      "step": 1455
+    },
+    {
+      "epoch": 0.06707513705256368,
+      "grad_norm": 0.5195608547036519,
+      "learning_rate": 0.00019928129535266785,
+      "loss": 1.3964,
+      "step": 1456
+    },
+    {
+      "epoch": 0.0671212051411987,
+      "grad_norm": 0.5693671938260518,
+      "learning_rate": 0.00019927950856961658,
+      "loss": 1.2054,
+      "step": 1457
+    },
+    {
+      "epoch": 0.0671672732298337,
+      "grad_norm": 0.8349136954388682,
+      "learning_rate": 0.0001992777195762773,
+      "loss": 1.0729,
+      "step": 1458
+    },
+    {
+      "epoch": 0.06721334131846869,
+      "grad_norm": 0.5057918902865298,
+      "learning_rate": 0.00019927592837268982,
+      "loss": 1.1675,
+      "step": 1459
+    },
+    {
+      "epoch": 0.0672594094071037,
+      "grad_norm": 0.5092509641974343,
+      "learning_rate": 0.000199274134958894,
+      "loss": 1.302,
+      "step": 1460
+    },
+    {
+      "epoch": 0.0673054774957387,
+      "grad_norm": 0.5498306026826177,
+      "learning_rate": 0.00019927233933492982,
+      "loss": 1.0598,
+      "step": 1461
+    },
+    {
+      "epoch": 0.0673515455843737,
+      "grad_norm": 0.6703430526335641,
+      "learning_rate": 0.00019927054150083718,
+      "loss": 1.4529,
+      "step": 1462
+    },
+    {
+      "epoch": 0.06739761367300871,
+      "grad_norm": 0.5194773238710436,
+      "learning_rate": 0.00019926874145665617,
+      "loss": 1.3673,
+      "step": 1463
+    },
+    {
+      "epoch": 0.06744368176164371,
+      "grad_norm": 0.6165389290871098,
+      "learning_rate": 0.00019926693920242682,
+      "loss": 0.9153,
+      "step": 1464
+    },
+    {
+      "epoch": 0.06748974985027871,
+      "grad_norm": 0.5152290804328755,
+      "learning_rate": 0.0001992651347381893,
+      "loss": 1.3363,
+      "step": 1465
+    },
+    {
+      "epoch": 0.06753581793891371,
+      "grad_norm": 0.6417893752676315,
+      "learning_rate": 0.00019926332806398373,
+      "loss": 1.1486,
+      "step": 1466
+    },
+    {
+      "epoch": 0.06758188602754872,
+      "grad_norm": 0.6287462559414765,
+      "learning_rate": 0.0001992615191798504,
+      "loss": 1.11,
+      "step": 1467
+    },
+    {
+      "epoch": 0.06762795411618372,
+      "grad_norm": 0.5289241159115603,
+      "learning_rate": 0.0001992597080858295,
+      "loss": 0.9059,
+      "step": 1468
+    },
+    {
+      "epoch": 0.06767402220481872,
+      "grad_norm": 0.7162621472350191,
+      "learning_rate": 0.0001992578947819614,
+      "loss": 1.1726,
+      "step": 1469
+    },
+    {
+      "epoch": 0.06772009029345373,
+      "grad_norm": 0.593637947052347,
+      "learning_rate": 0.0001992560792682865,
+      "loss": 1.4608,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06776615838208873,
+      "grad_norm": 0.8136128972417956,
+      "learning_rate": 0.00019925426154484513,
+      "loss": 1.3184,
+      "step": 1471
+    },
+    {
+      "epoch": 0.06781222647072373,
+      "grad_norm": 0.4813160501650728,
+      "learning_rate": 0.00019925244161167783,
+      "loss": 1.139,
+      "step": 1472
+    },
+    {
+      "epoch": 0.06785829455935873,
+      "grad_norm": 0.596480403116747,
+      "learning_rate": 0.00019925061946882513,
+      "loss": 1.1781,
+      "step": 1473
+    },
+    {
+      "epoch": 0.06790436264799374,
+      "grad_norm": 0.5797619998445428,
+      "learning_rate": 0.00019924879511632756,
+      "loss": 1.4868,
+      "step": 1474
+    },
+    {
+      "epoch": 0.06795043073662874,
+      "grad_norm": 0.491359210266974,
+      "learning_rate": 0.0001992469685542257,
+      "loss": 1.1756,
+      "step": 1475
+    },
+    {
+      "epoch": 0.06799649882526373,
+      "grad_norm": 0.5263306030373043,
+      "learning_rate": 0.0001992451397825603,
+      "loss": 0.9133,
+      "step": 1476
+    },
+    {
+      "epoch": 0.06804256691389875,
+      "grad_norm": 0.5688438669736144,
+      "learning_rate": 0.000199243308801372,
+      "loss": 1.3164,
+      "step": 1477
+    },
+    {
+      "epoch": 0.06808863500253375,
+      "grad_norm": 0.6865909284509103,
+      "learning_rate": 0.00019924147561070163,
+      "loss": 0.9086,
+      "step": 1478
+    },
+    {
+      "epoch": 0.06813470309116874,
+      "grad_norm": 0.5528265424807864,
+      "learning_rate": 0.00019923964021058993,
+      "loss": 1.2564,
+      "step": 1479
+    },
+    {
+      "epoch": 0.06818077117980376,
+      "grad_norm": 0.5382507592245364,
+      "learning_rate": 0.00019923780260107783,
+      "loss": 1.2604,
+      "step": 1480
+    },
+    {
+      "epoch": 0.06822683926843875,
+      "grad_norm": 0.6113350486999999,
+      "learning_rate": 0.0001992359627822062,
+      "loss": 1.372,
+      "step": 1481
+    },
+    {
+      "epoch": 0.06827290735707375,
+      "grad_norm": 0.6974763864070423,
+      "learning_rate": 0.000199234120754016,
+      "loss": 1.1548,
+      "step": 1482
+    },
+    {
+      "epoch": 0.06831897544570875,
+      "grad_norm": 0.491531547884025,
+      "learning_rate": 0.00019923227651654827,
+      "loss": 1.1467,
+      "step": 1483
+    },
+    {
+      "epoch": 0.06836504353434376,
+      "grad_norm": 0.6202377829573263,
+      "learning_rate": 0.00019923043006984402,
+      "loss": 1.1893,
+      "step": 1484
+    },
+    {
+      "epoch": 0.06841111162297876,
+      "grad_norm": 0.5020477386014991,
+      "learning_rate": 0.00019922858141394444,
+      "loss": 1.2357,
+      "step": 1485
+    },
+    {
+      "epoch": 0.06845717971161376,
+      "grad_norm": 0.44081042001016546,
+      "learning_rate": 0.00019922673054889058,
+      "loss": 1.2022,
+      "step": 1486
+    },
+    {
+      "epoch": 0.06850324780024877,
+      "grad_norm": 0.5746387160368152,
+      "learning_rate": 0.00019922487747472376,
+      "loss": 1.314,
+      "step": 1487
+    },
+    {
+      "epoch": 0.06854931588888377,
+      "grad_norm": 0.4740058835341598,
+      "learning_rate": 0.0001992230221914851,
+      "loss": 1.32,
+      "step": 1488
+    },
+    {
+      "epoch": 0.06859538397751877,
+      "grad_norm": 0.5881252992641264,
+      "learning_rate": 0.00019922116469921607,
+      "loss": 1.252,
+      "step": 1489
+    },
+    {
+      "epoch": 0.06864145206615377,
+      "grad_norm": 0.6289344167175068,
+      "learning_rate": 0.00019921930499795787,
+      "loss": 1.0795,
+      "step": 1490
+    },
+    {
+      "epoch": 0.06868752015478878,
+      "grad_norm": 0.6221899067893231,
+      "learning_rate": 0.00019921744308775202,
+      "loss": 0.8097,
+      "step": 1491
+    },
+    {
+      "epoch": 0.06873358824342378,
+      "grad_norm": 0.497437058791759,
+      "learning_rate": 0.00019921557896863986,
+      "loss": 1.0394,
+      "step": 1492
+    },
+    {
+      "epoch": 0.06877965633205878,
+      "grad_norm": 0.5438463620189287,
+      "learning_rate": 0.00019921371264066298,
+      "loss": 1.4294,
+      "step": 1493
+    },
+    {
+      "epoch": 0.06882572442069379,
+      "grad_norm": 0.5010009359721888,
+      "learning_rate": 0.00019921184410386293,
+      "loss": 1.2792,
+      "step": 1494
+    },
+    {
+      "epoch": 0.06887179250932879,
+      "grad_norm": 0.677613898448485,
+      "learning_rate": 0.00019920997335828126,
+      "loss": 1.2712,
+      "step": 1495
+    },
+    {
+      "epoch": 0.06891786059796379,
+      "grad_norm": 0.5449782701954616,
+      "learning_rate": 0.00019920810040395965,
+      "loss": 1.4357,
+      "step": 1496
+    },
+    {
+      "epoch": 0.0689639286865988,
+      "grad_norm": 1.0311342854661498,
+      "learning_rate": 0.00019920622524093978,
+      "loss": 1.1421,
+      "step": 1497
+    },
+    {
+      "epoch": 0.0690099967752338,
+      "grad_norm": 0.4652361136372153,
+      "learning_rate": 0.0001992043478692634,
+      "loss": 1.3203,
+      "step": 1498
+    },
+    {
+      "epoch": 0.0690560648638688,
+      "grad_norm": 0.503193260617631,
+      "learning_rate": 0.00019920246828897235,
+      "loss": 1.4201,
+      "step": 1499
+    },
+    {
+      "epoch": 0.0691021329525038,
+      "grad_norm": 0.45374534547378875,
+      "learning_rate": 0.0001992005865001084,
+      "loss": 1.3788,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06914820104113881,
+      "grad_norm": 0.625482864143711,
+      "learning_rate": 0.0001991987025027135,
+      "loss": 1.2253,
+      "step": 1501
+    },
+    {
+      "epoch": 0.0691942691297738,
+      "grad_norm": 0.6656885487513983,
+      "learning_rate": 0.0001991968162968296,
+      "loss": 1.1015,
+      "step": 1502
+    },
+    {
+      "epoch": 0.0692403372184088,
+      "grad_norm": 0.6542506599081797,
+      "learning_rate": 0.00019919492788249862,
+      "loss": 1.1074,
+      "step": 1503
+    },
+    {
+      "epoch": 0.06928640530704382,
+      "grad_norm": 0.5591655088638409,
+      "learning_rate": 0.00019919303725976267,
+      "loss": 1.2671,
+      "step": 1504
+    },
+    {
+      "epoch": 0.06933247339567881,
+      "grad_norm": 0.6330376789588191,
+      "learning_rate": 0.00019919114442866384,
+      "loss": 1.1664,
+      "step": 1505
+    },
+    {
+      "epoch": 0.06937854148431381,
+      "grad_norm": 0.505853498222261,
+      "learning_rate": 0.00019918924938924426,
+      "loss": 1.2064,
+      "step": 1506
+    },
+    {
+      "epoch": 0.06942460957294881,
+      "grad_norm": 0.5955160687483567,
+      "learning_rate": 0.00019918735214154612,
+      "loss": 1.4136,
+      "step": 1507
+    },
+    {
+      "epoch": 0.06947067766158382,
+      "grad_norm": 0.6354487601660657,
+      "learning_rate": 0.00019918545268561162,
+      "loss": 1.1523,
+      "step": 1508
+    },
+    {
+      "epoch": 0.06951674575021882,
+      "grad_norm": 0.4513489809800152,
+      "learning_rate": 0.0001991835510214831,
+      "loss": 1.339,
+      "step": 1509
+    },
+    {
+      "epoch": 0.06956281383885382,
+      "grad_norm": 0.4332989574352621,
+      "learning_rate": 0.0001991816471492029,
+      "loss": 1.0982,
+      "step": 1510
+    },
+    {
+      "epoch": 0.06960888192748883,
+      "grad_norm": 0.6158053298557835,
+      "learning_rate": 0.00019917974106881334,
+      "loss": 1.4474,
+      "step": 1511
+    },
+    {
+      "epoch": 0.06965495001612383,
+      "grad_norm": 0.527247448718862,
+      "learning_rate": 0.00019917783278035694,
+      "loss": 1.3997,
+      "step": 1512
+    },
+    {
+      "epoch": 0.06970101810475883,
+      "grad_norm": 0.6400791265930922,
+      "learning_rate": 0.00019917592228387615,
+      "loss": 1.164,
+      "step": 1513
+    },
+    {
+      "epoch": 0.06974708619339384,
+      "grad_norm": 0.5643320243792523,
+      "learning_rate": 0.0001991740095794135,
+      "loss": 1.1996,
+      "step": 1514
+    },
+    {
+      "epoch": 0.06979315428202884,
+      "grad_norm": 0.6331083564322803,
+      "learning_rate": 0.00019917209466701154,
+      "loss": 1.1106,
+      "step": 1515
+    },
+    {
+      "epoch": 0.06983922237066384,
+      "grad_norm": 0.5534981938010868,
+      "learning_rate": 0.00019917017754671296,
+      "loss": 1.0408,
+      "step": 1516
+    },
+    {
+      "epoch": 0.06988529045929884,
+      "grad_norm": 0.6915793407083941,
+      "learning_rate": 0.0001991682582185604,
+      "loss": 1.0384,
+      "step": 1517
+    },
+    {
+      "epoch": 0.06993135854793385,
+      "grad_norm": 0.6377067427367398,
+      "learning_rate": 0.00019916633668259665,
+      "loss": 1.4274,
+      "step": 1518
+    },
+    {
+      "epoch": 0.06997742663656885,
+      "grad_norm": 0.5160154005124289,
+      "learning_rate": 0.00019916441293886444,
+      "loss": 1.2824,
+      "step": 1519
+    },
+    {
+      "epoch": 0.07002349472520385,
+      "grad_norm": 0.6282860080229427,
+      "learning_rate": 0.00019916248698740655,
+      "loss": 1.5069,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07006956281383886,
+      "grad_norm": 0.6936431445138972,
+      "learning_rate": 0.00019916055882826595,
+      "loss": 1.1727,
+      "step": 1521
+    },
+    {
+      "epoch": 0.07011563090247386,
+      "grad_norm": 0.6017975215124463,
+      "learning_rate": 0.00019915862846148558,
+      "loss": 1.3299,
+      "step": 1522
+    },
+    {
+      "epoch": 0.07016169899110886,
+      "grad_norm": 0.6414128607297758,
+      "learning_rate": 0.00019915669588710827,
+      "loss": 0.9631,
+      "step": 1523
+    },
+    {
+      "epoch": 0.07020776707974385,
+      "grad_norm": 0.9532458284166294,
+      "learning_rate": 0.00019915476110517723,
+      "loss": 0.9958,
+      "step": 1524
+    },
+    {
+      "epoch": 0.07025383516837887,
+      "grad_norm": 0.49746599380928624,
+      "learning_rate": 0.0001991528241157354,
+      "loss": 1.3268,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07029990325701387,
+      "grad_norm": 0.9665222452781586,
+      "learning_rate": 0.00019915088491882598,
+      "loss": 1.2753,
+      "step": 1526
+    },
+    {
+      "epoch": 0.07034597134564886,
+      "grad_norm": 0.5394021367602727,
+      "learning_rate": 0.00019914894351449208,
+      "loss": 1.1948,
+      "step": 1527
+    },
+    {
+      "epoch": 0.07039203943428388,
+      "grad_norm": 0.5783519103294661,
+      "learning_rate": 0.00019914699990277699,
+      "loss": 1.0218,
+      "step": 1528
+    },
+    {
+      "epoch": 0.07043810752291887,
+      "grad_norm": 0.5977367355052896,
+      "learning_rate": 0.0001991450540837239,
+      "loss": 1.3427,
+      "step": 1529
+    },
+    {
+      "epoch": 0.07048417561155387,
+      "grad_norm": 0.5259134272564822,
+      "learning_rate": 0.00019914310605737622,
+      "loss": 1.1692,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07053024370018889,
+      "grad_norm": 0.6005131607911207,
+      "learning_rate": 0.00019914115582377729,
+      "loss": 1.3863,
+      "step": 1531
+    },
+    {
+      "epoch": 0.07057631178882388,
+      "grad_norm": 0.9663986322885932,
+      "learning_rate": 0.00019913920338297046,
+      "loss": 1.2756,
+      "step": 1532
+    },
+    {
+      "epoch": 0.07062237987745888,
+      "grad_norm": 0.536267916054544,
+      "learning_rate": 0.00019913724873499931,
+      "loss": 1.166,
+      "step": 1533
+    },
+    {
+      "epoch": 0.07066844796609388,
+      "grad_norm": 0.49968514764358535,
+      "learning_rate": 0.00019913529187990728,
+      "loss": 1.5807,
+      "step": 1534
+    },
+    {
+      "epoch": 0.0707145160547289,
+      "grad_norm": 0.5954791271672133,
+      "learning_rate": 0.00019913333281773796,
+      "loss": 1.2243,
+      "step": 1535
+    },
+    {
+      "epoch": 0.07076058414336389,
+      "grad_norm": 0.5381786898088476,
+      "learning_rate": 0.00019913137154853496,
+      "loss": 1.362,
+      "step": 1536
+    },
+    {
+      "epoch": 0.07080665223199889,
+      "grad_norm": 0.674816595304603,
+      "learning_rate": 0.00019912940807234196,
+      "loss": 1.265,
+      "step": 1537
+    },
+    {
+      "epoch": 0.0708527203206339,
+      "grad_norm": 0.5236876484818728,
+      "learning_rate": 0.00019912744238920262,
+      "loss": 1.4134,
+      "step": 1538
+    },
+    {
+      "epoch": 0.0708987884092689,
+      "grad_norm": 0.6135190525974913,
+      "learning_rate": 0.00019912547449916078,
+      "loss": 0.8291,
+      "step": 1539
+    },
+    {
+      "epoch": 0.0709448564979039,
+      "grad_norm": 0.7521843531438677,
+      "learning_rate": 0.0001991235044022602,
+      "loss": 1.1514,
+      "step": 1540
+    },
+    {
+      "epoch": 0.0709909245865389,
+      "grad_norm": 0.48117129425331295,
+      "learning_rate": 0.00019912153209854477,
+      "loss": 1.4144,
+      "step": 1541
+    },
+    {
+      "epoch": 0.07103699267517391,
+      "grad_norm": 0.5397172437957958,
+      "learning_rate": 0.00019911955758805837,
+      "loss": 1.1872,
+      "step": 1542
+    },
+    {
+      "epoch": 0.07108306076380891,
+      "grad_norm": 0.7982306448766041,
+      "learning_rate": 0.00019911758087084498,
+      "loss": 1.2218,
+      "step": 1543
+    },
+    {
+      "epoch": 0.0711291288524439,
+      "grad_norm": 0.788317546846562,
+      "learning_rate": 0.0001991156019469486,
+      "loss": 1.4539,
+      "step": 1544
+    },
+    {
+      "epoch": 0.07117519694107892,
+      "grad_norm": 0.5681960206344908,
+      "learning_rate": 0.00019911362081641328,
+      "loss": 1.025,
+      "step": 1545
+    },
+    {
+      "epoch": 0.07122126502971392,
+      "grad_norm": 0.9319488386986213,
+      "learning_rate": 0.00019911163747928315,
+      "loss": 1.1769,
+      "step": 1546
+    },
+    {
+      "epoch": 0.07126733311834892,
+      "grad_norm": 0.6563523131281714,
+      "learning_rate": 0.00019910965193560234,
+      "loss": 1.1116,
+      "step": 1547
+    },
+    {
+      "epoch": 0.07131340120698393,
+      "grad_norm": 0.882849611283978,
+      "learning_rate": 0.00019910766418541504,
+      "loss": 1.2695,
+      "step": 1548
+    },
+    {
+      "epoch": 0.07135946929561893,
+      "grad_norm": 0.6273723930425973,
+      "learning_rate": 0.00019910567422876558,
+      "loss": 0.9432,
+      "step": 1549
+    },
+    {
+      "epoch": 0.07140553738425393,
+      "grad_norm": 0.6303769868900637,
+      "learning_rate": 0.00019910368206569817,
+      "loss": 1.3335,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07145160547288892,
+      "grad_norm": 0.589146793624472,
+      "learning_rate": 0.00019910168769625726,
+      "loss": 1.7691,
+      "step": 1551
+    },
+    {
+      "epoch": 0.07149767356152394,
+      "grad_norm": 0.4953486909538088,
+      "learning_rate": 0.00019909969112048717,
+      "loss": 1.3903,
+      "step": 1552
+    },
+    {
+      "epoch": 0.07154374165015893,
+      "grad_norm": 1.6293538363913111,
+      "learning_rate": 0.00019909769233843237,
+      "loss": 1.367,
+      "step": 1553
+    },
+    {
+      "epoch": 0.07158980973879393,
+      "grad_norm": 0.4700290500428555,
+      "learning_rate": 0.00019909569135013734,
+      "loss": 1.4002,
+      "step": 1554
+    },
+    {
+      "epoch": 0.07163587782742895,
+      "grad_norm": 0.5045871643859806,
+      "learning_rate": 0.0001990936881556467,
+      "loss": 1.0928,
+      "step": 1555
+    },
+    {
+      "epoch": 0.07168194591606394,
+      "grad_norm": 0.6994694761035938,
+      "learning_rate": 0.000199091682755005,
+      "loss": 1.1669,
+      "step": 1556
+    },
+    {
+      "epoch": 0.07172801400469894,
+      "grad_norm": 0.7631207003414635,
+      "learning_rate": 0.00019908967514825686,
+      "loss": 1.4705,
+      "step": 1557
+    },
+    {
+      "epoch": 0.07177408209333394,
+      "grad_norm": 1.1742615097187108,
+      "learning_rate": 0.00019908766533544703,
+      "loss": 1.3146,
+      "step": 1558
+    },
+    {
+      "epoch": 0.07182015018196895,
+      "grad_norm": 0.6120863451507998,
+      "learning_rate": 0.0001990856533166202,
+      "loss": 1.0689,
+      "step": 1559
+    },
+    {
+      "epoch": 0.07186621827060395,
+      "grad_norm": 0.6151686448861879,
+      "learning_rate": 0.00019908363909182123,
+      "loss": 1.2707,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07191228635923895,
+      "grad_norm": 1.4016163618351345,
+      "learning_rate": 0.0001990816226610949,
+      "loss": 1.2185,
+      "step": 1561
+    },
+    {
+      "epoch": 0.07195835444787396,
+      "grad_norm": 0.8238712883372914,
+      "learning_rate": 0.00019907960402448615,
+      "loss": 1.1534,
+      "step": 1562
+    },
+    {
+      "epoch": 0.07200442253650896,
+      "grad_norm": 7.663667941671084,
+      "learning_rate": 0.00019907758318203988,
+      "loss": 1.2809,
+      "step": 1563
+    },
+    {
+      "epoch": 0.07205049062514396,
+      "grad_norm": 0.7668865811596047,
+      "learning_rate": 0.0001990755601338011,
+      "loss": 1.4043,
+      "step": 1564
+    },
+    {
+      "epoch": 0.07209655871377897,
+      "grad_norm": 1.0462969173660368,
+      "learning_rate": 0.0001990735348798149,
+      "loss": 1.2504,
+      "step": 1565
+    },
+    {
+      "epoch": 0.07214262680241397,
+      "grad_norm": 1.4576812188483093,
+      "learning_rate": 0.00019907150742012628,
+      "loss": 1.4803,
+      "step": 1566
+    },
+    {
+      "epoch": 0.07218869489104897,
+      "grad_norm": 0.8861517427421983,
+      "learning_rate": 0.0001990694777547804,
+      "loss": 1.3068,
+      "step": 1567
+    },
+    {
+      "epoch": 0.07223476297968397,
+      "grad_norm": 0.7172144316546494,
+      "learning_rate": 0.00019906744588382252,
+      "loss": 1.2634,
+      "step": 1568
+    },
+    {
+      "epoch": 0.07228083106831898,
+      "grad_norm": 0.8739017867838057,
+      "learning_rate": 0.0001990654118072978,
+      "loss": 1.524,
+      "step": 1569
+    },
+    {
+      "epoch": 0.07232689915695398,
+      "grad_norm": 0.6885775049235342,
+      "learning_rate": 0.00019906337552525152,
+      "loss": 1.6287,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07237296724558898,
+      "grad_norm": 0.7570577818047025,
+      "learning_rate": 0.00019906133703772908,
+      "loss": 1.1745,
+      "step": 1571
+    },
+    {
+      "epoch": 0.07241903533422399,
+      "grad_norm": 0.9777421305417565,
+      "learning_rate": 0.0001990592963447758,
+      "loss": 1.1918,
+      "step": 1572
+    },
+    {
+      "epoch": 0.07246510342285899,
+      "grad_norm": 0.6256177759967287,
+      "learning_rate": 0.00019905725344643714,
+      "loss": 1.0512,
+      "step": 1573
+    },
+    {
+      "epoch": 0.07251117151149399,
+      "grad_norm": 0.7247589125695432,
+      "learning_rate": 0.00019905520834275858,
+      "loss": 1.1934,
+      "step": 1574
+    },
+    {
+      "epoch": 0.07255723960012898,
+      "grad_norm": 0.7609948394308028,
+      "learning_rate": 0.00019905316103378565,
+      "loss": 1.1817,
+      "step": 1575
+    },
+    {
+      "epoch": 0.072603307688764,
+      "grad_norm": 0.8250463247483187,
+      "learning_rate": 0.00019905111151956396,
+      "loss": 1.1878,
+      "step": 1576
+    },
+    {
+      "epoch": 0.072649375777399,
+      "grad_norm": 0.6396403440572587,
+      "learning_rate": 0.00019904905980013907,
+      "loss": 1.1788,
+      "step": 1577
+    },
+    {
+      "epoch": 0.07269544386603399,
+      "grad_norm": 0.6967530289965397,
+      "learning_rate": 0.00019904700587555668,
+      "loss": 1.3442,
+      "step": 1578
+    },
+    {
+      "epoch": 0.072741511954669,
+      "grad_norm": 0.6111403404378946,
+      "learning_rate": 0.00019904494974586258,
+      "loss": 1.509,
+      "step": 1579
+    },
+    {
+      "epoch": 0.072787580043304,
+      "grad_norm": 0.6370564101606944,
+      "learning_rate": 0.00019904289141110245,
+      "loss": 1.1039,
+      "step": 1580
+    },
+    {
+      "epoch": 0.072833648131939,
+      "grad_norm": 0.8134478490288952,
+      "learning_rate": 0.0001990408308713222,
+      "loss": 1.2672,
+      "step": 1581
+    },
+    {
+      "epoch": 0.07287971622057401,
+      "grad_norm": 0.7188633420449408,
+      "learning_rate": 0.00019903876812656768,
+      "loss": 1.1543,
+      "step": 1582
+    },
+    {
+      "epoch": 0.07292578430920901,
+      "grad_norm": 1.0278623817970844,
+      "learning_rate": 0.00019903670317688476,
+      "loss": 0.9805,
+      "step": 1583
+    },
+    {
+      "epoch": 0.07297185239784401,
+      "grad_norm": 0.6879437386332967,
+      "learning_rate": 0.0001990346360223195,
+      "loss": 1.2798,
+      "step": 1584
+    },
+    {
+      "epoch": 0.07301792048647901,
+      "grad_norm": 1.0260027152373659,
+      "learning_rate": 0.00019903256666291784,
+      "loss": 1.27,
+      "step": 1585
+    },
+    {
+      "epoch": 0.07306398857511402,
+      "grad_norm": 1.284759417475545,
+      "learning_rate": 0.00019903049509872593,
+      "loss": 1.6629,
+      "step": 1586
+    },
+    {
+      "epoch": 0.07311005666374902,
+      "grad_norm": 0.8920465017265549,
+      "learning_rate": 0.00019902842132978982,
+      "loss": 1.098,
+      "step": 1587
+    },
+    {
+      "epoch": 0.07315612475238402,
+      "grad_norm": 0.9069227891479975,
+      "learning_rate": 0.0001990263453561557,
+      "loss": 1.0086,
+      "step": 1588
+    },
+    {
+      "epoch": 0.07320219284101903,
+      "grad_norm": 0.8182613423507055,
+      "learning_rate": 0.0001990242671778698,
+      "loss": 1.0296,
+      "step": 1589
+    },
+    {
+      "epoch": 0.07324826092965403,
+      "grad_norm": 0.8196301478414959,
+      "learning_rate": 0.0001990221867949784,
+      "loss": 1.4242,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07329432901828903,
+      "grad_norm": 0.8681382459872095,
+      "learning_rate": 0.00019902010420752778,
+      "loss": 1.1255,
+      "step": 1591
+    },
+    {
+      "epoch": 0.07334039710692403,
+      "grad_norm": 0.7250389757665758,
+      "learning_rate": 0.00019901801941556431,
+      "loss": 1.3093,
+      "step": 1592
+    },
+    {
+      "epoch": 0.07338646519555904,
+      "grad_norm": 0.7138998834654877,
+      "learning_rate": 0.00019901593241913442,
+      "loss": 1.4574,
+      "step": 1593
+    },
+    {
+      "epoch": 0.07343253328419404,
+      "grad_norm": 1.546980424755877,
+      "learning_rate": 0.0001990138432182846,
+      "loss": 1.4442,
+      "step": 1594
+    },
+    {
+      "epoch": 0.07347860137282904,
+      "grad_norm": 0.8631652921968546,
+      "learning_rate": 0.00019901175181306132,
+      "loss": 1.259,
+      "step": 1595
+    },
+    {
+      "epoch": 0.07352466946146405,
+      "grad_norm": 0.8236185274574178,
+      "learning_rate": 0.00019900965820351116,
+      "loss": 1.0705,
+      "step": 1596
+    },
+    {
+      "epoch": 0.07357073755009905,
+      "grad_norm": 0.7739894554913392,
+      "learning_rate": 0.0001990075623896807,
+      "loss": 1.1676,
+      "step": 1597
+    },
+    {
+      "epoch": 0.07361680563873405,
+      "grad_norm": 2.8237972474300217,
+      "learning_rate": 0.00019900546437161664,
+      "loss": 1.4194,
+      "step": 1598
+    },
+    {
+      "epoch": 0.07366287372736906,
+      "grad_norm": 0.5507366125252139,
+      "learning_rate": 0.0001990033641493657,
+      "loss": 1.449,
+      "step": 1599
+    },
+    {
+      "epoch": 0.07370894181600406,
+      "grad_norm": 0.8944573428162561,
+      "learning_rate": 0.00019900126172297457,
+      "loss": 1.3276,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07375500990463905,
+      "grad_norm": 1.0362400675333954,
+      "learning_rate": 0.0001989991570924901,
+      "loss": 1.1059,
+      "step": 1601
+    },
+    {
+      "epoch": 0.07380107799327405,
+      "grad_norm": 0.5607011612792004,
+      "learning_rate": 0.0001989970502579592,
+      "loss": 1.4757,
+      "step": 1602
+    },
+    {
+      "epoch": 0.07384714608190907,
+      "grad_norm": 0.6165873512340707,
+      "learning_rate": 0.00019899494121942865,
+      "loss": 0.7288,
+      "step": 1603
+    },
+    {
+      "epoch": 0.07389321417054406,
+      "grad_norm": 0.5855170744120322,
+      "learning_rate": 0.0001989928299769455,
+      "loss": 1.4602,
+      "step": 1604
+    },
+    {
+      "epoch": 0.07393928225917906,
+      "grad_norm": 0.6590683924743276,
+      "learning_rate": 0.00019899071653055675,
+      "loss": 1.0607,
+      "step": 1605
+    },
+    {
+      "epoch": 0.07398535034781407,
+      "grad_norm": 2.5871029572432014,
+      "learning_rate": 0.00019898860088030942,
+      "loss": 1.2764,
+      "step": 1606
+    },
+    {
+      "epoch": 0.07403141843644907,
+      "grad_norm": 0.6259621499839694,
+      "learning_rate": 0.00019898648302625057,
+      "loss": 1.1852,
+      "step": 1607
+    },
+    {
+      "epoch": 0.07407748652508407,
+      "grad_norm": 0.7392154961763915,
+      "learning_rate": 0.00019898436296842745,
+      "loss": 1.0191,
+      "step": 1608
+    },
+    {
+      "epoch": 0.07412355461371908,
+      "grad_norm": 0.6606358246882484,
+      "learning_rate": 0.00019898224070688723,
+      "loss": 1.4572,
+      "step": 1609
+    },
+    {
+      "epoch": 0.07416962270235408,
+      "grad_norm": 0.5846932776693576,
+      "learning_rate": 0.0001989801162416771,
+      "loss": 1.3541,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07421569079098908,
+      "grad_norm": 1.108325937609252,
+      "learning_rate": 0.00019897798957284438,
+      "loss": 1.05,
+      "step": 1611
+    },
+    {
+      "epoch": 0.07426175887962408,
+      "grad_norm": 1.8843066347227175,
+      "learning_rate": 0.00019897586070043648,
+      "loss": 1.4555,
+      "step": 1612
+    },
+    {
+      "epoch": 0.07430782696825909,
+      "grad_norm": 1.5797314217556833,
+      "learning_rate": 0.0001989737296245007,
+      "loss": 1.0548,
+      "step": 1613
+    },
+    {
+      "epoch": 0.07435389505689409,
+      "grad_norm": 0.9560319258567592,
+      "learning_rate": 0.00019897159634508458,
+      "loss": 1.2455,
+      "step": 1614
+    },
+    {
+      "epoch": 0.07439996314552909,
+      "grad_norm": 1.8408901095774495,
+      "learning_rate": 0.00019896946086223557,
+      "loss": 0.9371,
+      "step": 1615
+    },
+    {
+      "epoch": 0.0744460312341641,
+      "grad_norm": 0.83625360077157,
+      "learning_rate": 0.00019896732317600119,
+      "loss": 1.4016,
+      "step": 1616
+    },
+    {
+      "epoch": 0.0744920993227991,
+      "grad_norm": 1.0397569605612822,
+      "learning_rate": 0.00019896518328642903,
+      "loss": 1.476,
+      "step": 1617
+    },
+    {
+      "epoch": 0.0745381674114341,
+      "grad_norm": 0.8812887247663329,
+      "learning_rate": 0.00019896304119356679,
+      "loss": 1.2842,
+      "step": 1618
+    },
+    {
+      "epoch": 0.0745842355000691,
+      "grad_norm": 1.6694848711730146,
+      "learning_rate": 0.0001989608968974621,
+      "loss": 1.5902,
+      "step": 1619
+    },
+    {
+      "epoch": 0.07463030358870411,
+      "grad_norm": 0.868854619068696,
+      "learning_rate": 0.00019895875039816272,
+      "loss": 1.2802,
+      "step": 1620
+    },
+    {
+      "epoch": 0.0746763716773391,
+      "grad_norm": 1.0051707968523453,
+      "learning_rate": 0.00019895660169571647,
+      "loss": 1.1086,
+      "step": 1621
+    },
+    {
+      "epoch": 0.0747224397659741,
+      "grad_norm": 5.423765566010615,
+      "learning_rate": 0.0001989544507901711,
+      "loss": 1.2095,
+      "step": 1622
+    },
+    {
+      "epoch": 0.07476850785460912,
+      "grad_norm": 73.24510972470281,
+      "learning_rate": 0.0001989522976815746,
+      "loss": 2.2631,
+      "step": 1623
+    },
+    {
+      "epoch": 0.07481457594324412,
+      "grad_norm": 3.5671253627180297,
+      "learning_rate": 0.00019895014236997484,
+      "loss": 1.4133,
+      "step": 1624
+    },
+    {
+      "epoch": 0.07486064403187911,
+      "grad_norm": 3.849610690524203,
+      "learning_rate": 0.00019894798485541984,
+      "loss": 1.3671,
+      "step": 1625
+    },
+    {
+      "epoch": 0.07490671212051413,
+      "grad_norm": 177.65252515797926,
+      "learning_rate": 0.00019894582513795762,
+      "loss": 2.2473,
+      "step": 1626
+    },
+    {
+      "epoch": 0.07495278020914913,
+      "grad_norm": 51.99282923051461,
+      "learning_rate": 0.00019894366321763626,
+      "loss": 7.9892,
+      "step": 1627
+    },
+    {
+      "epoch": 0.07499884829778412,
+      "grad_norm": 25.748355840263656,
+      "learning_rate": 0.00019894149909450385,
+      "loss": 6.9943,
+      "step": 1628
+    },
+    {
+      "epoch": 0.07504491638641912,
+      "grad_norm": 20.880723504550133,
+      "learning_rate": 0.00019893933276860865,
+      "loss": 6.0211,
+      "step": 1629
+    },
+    {
+      "epoch": 0.07509098447505413,
+      "grad_norm": 14.876623393807826,
+      "learning_rate": 0.00019893716423999885,
+      "loss": 5.9232,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07513705256368913,
+      "grad_norm": 7.877326796907176,
+      "learning_rate": 0.00019893499350872272,
+      "loss": 5.1976,
+      "step": 1631
+    },
+    {
+      "epoch": 0.07518312065232413,
+      "grad_norm": 3.9382462569499896,
+      "learning_rate": 0.00019893282057482857,
+      "loss": 5.1672,
+      "step": 1632
+    },
+    {
+      "epoch": 0.07522918874095914,
+      "grad_norm": 9.105821457079186,
+      "learning_rate": 0.00019893064543836486,
+      "loss": 4.6104,
+      "step": 1633
+    },
+    {
+      "epoch": 0.07527525682959414,
+      "grad_norm": 18.069964915798238,
+      "learning_rate": 0.00019892846809937994,
+      "loss": 5.0237,
+      "step": 1634
+    },
+    {
+      "epoch": 0.07532132491822914,
+      "grad_norm": 14.59440875866924,
+      "learning_rate": 0.00019892628855792225,
+      "loss": 4.4436,
+      "step": 1635
+    },
+    {
+      "epoch": 0.07536739300686414,
+      "grad_norm": 21.72762248041144,
+      "learning_rate": 0.00019892410681404042,
+      "loss": 5.1858,
+      "step": 1636
+    },
+    {
+      "epoch": 0.07541346109549915,
+      "grad_norm": 18.96773547002135,
+      "learning_rate": 0.00019892192286778294,
+      "loss": 5.7422,
+      "step": 1637
+    },
+    {
+      "epoch": 0.07545952918413415,
+      "grad_norm": 9.805167845938533,
+      "learning_rate": 0.0001989197367191985,
+      "loss": 5.4081,
+      "step": 1638
+    },
+    {
+      "epoch": 0.07550559727276915,
+      "grad_norm": 6.652030191170769,
+      "learning_rate": 0.00019891754836833572,
+      "loss": 3.9803,
+      "step": 1639
+    },
+    {
+      "epoch": 0.07555166536140416,
+      "grad_norm": 4.491343234172737,
+      "learning_rate": 0.00019891535781524334,
+      "loss": 2.8616,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07559773345003916,
+      "grad_norm": 4.847454568485587,
+      "learning_rate": 0.00019891316505997008,
+      "loss": 3.2298,
+      "step": 1641
+    },
+    {
+      "epoch": 0.07564380153867416,
+      "grad_norm": 3.9457122173082597,
+      "learning_rate": 0.00019891097010256483,
+      "loss": 2.9871,
+      "step": 1642
+    },
+    {
+      "epoch": 0.07568986962730917,
+      "grad_norm": 6.656686504971577,
+      "learning_rate": 0.00019890877294307643,
+      "loss": 2.6075,
+      "step": 1643
+    },
+    {
+      "epoch": 0.07573593771594417,
+      "grad_norm": 6.895716531950863,
+      "learning_rate": 0.0001989065735815538,
+      "loss": 2.5978,
+      "step": 1644
+    },
+    {
+      "epoch": 0.07578200580457917,
+      "grad_norm": 3.864119794897403,
+      "learning_rate": 0.0001989043720180459,
+      "loss": 2.2986,
+      "step": 1645
+    },
+    {
+      "epoch": 0.07582807389321417,
+      "grad_norm": 3.7891032623417096,
+      "learning_rate": 0.0001989021682526017,
+      "loss": 2.3882,
+      "step": 1646
+    },
+    {
+      "epoch": 0.07587414198184918,
+      "grad_norm": 6.005647443224046,
+      "learning_rate": 0.00019889996228527034,
+      "loss": 2.8256,
+      "step": 1647
+    },
+    {
+      "epoch": 0.07592021007048418,
+      "grad_norm": 4.032391607719696,
+      "learning_rate": 0.00019889775411610088,
+      "loss": 2.6158,
+      "step": 1648
+    },
+    {
+      "epoch": 0.07596627815911917,
+      "grad_norm": 2.2560469986383302,
+      "learning_rate": 0.0001988955437451425,
+      "loss": 2.3011,
+      "step": 1649
+    },
+    {
+      "epoch": 0.07601234624775419,
+      "grad_norm": 3.4747329347048796,
+      "learning_rate": 0.00019889333117244443,
+      "loss": 2.1502,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07605841433638919,
+      "grad_norm": 2.7956599863306786,
+      "learning_rate": 0.00019889111639805588,
+      "loss": 2.0559,
+      "step": 1651
+    },
+    {
+      "epoch": 0.07610448242502418,
+      "grad_norm": 3.410564230216096,
+      "learning_rate": 0.0001988888994220262,
+      "loss": 2.1597,
+      "step": 1652
+    },
+    {
+      "epoch": 0.07615055051365918,
+      "grad_norm": 3.9962475270641553,
+      "learning_rate": 0.00019888668024440473,
+      "loss": 1.9208,
+      "step": 1653
+    },
+    {
+      "epoch": 0.0761966186022942,
+      "grad_norm": 2.5204424885893175,
+      "learning_rate": 0.0001988844588652409,
+      "loss": 1.8214,
+      "step": 1654
+    },
+    {
+      "epoch": 0.07624268669092919,
+      "grad_norm": 2.9587702863924554,
+      "learning_rate": 0.0001988822352845841,
+      "loss": 1.8623,
+      "step": 1655
+    },
+    {
+      "epoch": 0.07628875477956419,
+      "grad_norm": 2.3493798289324377,
+      "learning_rate": 0.00019888000950248392,
+      "loss": 1.983,
+      "step": 1656
+    },
+    {
+      "epoch": 0.0763348228681992,
+      "grad_norm": 1.817170324469439,
+      "learning_rate": 0.00019887778151898985,
+      "loss": 1.8752,
+      "step": 1657
+    },
+    {
+      "epoch": 0.0763808909568342,
+      "grad_norm": 1.6761714583057903,
+      "learning_rate": 0.0001988755513341515,
+      "loss": 1.5285,
+      "step": 1658
+    },
+    {
+      "epoch": 0.0764269590454692,
+      "grad_norm": 2.0712061184424453,
+      "learning_rate": 0.00019887331894801854,
+      "loss": 1.6293,
+      "step": 1659
+    },
+    {
+      "epoch": 0.07647302713410421,
+      "grad_norm": 2.7110596080865523,
+      "learning_rate": 0.0001988710843606407,
+      "loss": 1.7339,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07651909522273921,
+      "grad_norm": 3.0199443679576,
+      "learning_rate": 0.00019886884757206765,
+      "loss": 1.6001,
+      "step": 1661
+    },
+    {
+      "epoch": 0.07656516331137421,
+      "grad_norm": 2.1543859500238227,
+      "learning_rate": 0.00019886660858234927,
+      "loss": 1.7034,
+      "step": 1662
+    },
+    {
+      "epoch": 0.07661123140000921,
+      "grad_norm": 1.8224270117739605,
+      "learning_rate": 0.00019886436739153534,
+      "loss": 1.6263,
+      "step": 1663
+    },
+    {
+      "epoch": 0.07665729948864422,
+      "grad_norm": 2.1641788943324656,
+      "learning_rate": 0.0001988621239996758,
+      "loss": 1.4233,
+      "step": 1664
+    },
+    {
+      "epoch": 0.07670336757727922,
+      "grad_norm": 3.241356040021733,
+      "learning_rate": 0.00019885987840682058,
+      "loss": 1.7621,
+      "step": 1665
+    },
+    {
+      "epoch": 0.07674943566591422,
+      "grad_norm": 1.528288314590827,
+      "learning_rate": 0.00019885763061301966,
+      "loss": 1.5541,
+      "step": 1666
+    },
+    {
+      "epoch": 0.07679550375454923,
+      "grad_norm": 1.4584170896286162,
+      "learning_rate": 0.0001988553806183231,
+      "loss": 1.5406,
+      "step": 1667
+    },
+    {
+      "epoch": 0.07684157184318423,
+      "grad_norm": 1.4317383355848263,
+      "learning_rate": 0.00019885312842278104,
+      "loss": 1.3877,
+      "step": 1668
+    },
+    {
+      "epoch": 0.07688763993181923,
+      "grad_norm": 1.159045472882773,
+      "learning_rate": 0.00019885087402644351,
+      "loss": 1.8253,
+      "step": 1669
+    },
+    {
+      "epoch": 0.07693370802045423,
+      "grad_norm": 1.6997490841688292,
+      "learning_rate": 0.00019884861742936082,
+      "loss": 1.3579,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07697977610908924,
+      "grad_norm": 2.1036202331079057,
+      "learning_rate": 0.0001988463586315831,
+      "loss": 1.4005,
+      "step": 1671
+    },
+    {
+      "epoch": 0.07702584419772424,
+      "grad_norm": 1.288859638255736,
+      "learning_rate": 0.0001988440976331607,
+      "loss": 1.834,
+      "step": 1672
+    },
+    {
+      "epoch": 0.07707191228635923,
+      "grad_norm": 3.2527071966758148,
+      "learning_rate": 0.00019884183443414396,
+      "loss": 1.6275,
+      "step": 1673
+    },
+    {
+      "epoch": 0.07711798037499425,
+      "grad_norm": 2.70167411696288,
+      "learning_rate": 0.00019883956903458329,
+      "loss": 1.4053,
+      "step": 1674
+    },
+    {
+      "epoch": 0.07716404846362925,
+      "grad_norm": 1.1291450277506756,
+      "learning_rate": 0.00019883730143452902,
+      "loss": 1.7361,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07721011655226424,
+      "grad_norm": 1.4914443565690914,
+      "learning_rate": 0.00019883503163403176,
+      "loss": 1.6523,
+      "step": 1676
+    },
+    {
+      "epoch": 0.07725618464089926,
+      "grad_norm": 2.2195128846458214,
+      "learning_rate": 0.00019883275963314195,
+      "loss": 1.6512,
+      "step": 1677
+    },
+    {
+      "epoch": 0.07730225272953425,
+      "grad_norm": 1.8068107970184906,
+      "learning_rate": 0.0001988304854319102,
+      "loss": 1.3289,
+      "step": 1678
+    },
+    {
+      "epoch": 0.07734832081816925,
+      "grad_norm": 2.7955204477351043,
+      "learning_rate": 0.0001988282090303872,
+      "loss": 1.7672,
+      "step": 1679
+    },
+    {
+      "epoch": 0.07739438890680425,
+      "grad_norm": 1.9271088339685976,
+      "learning_rate": 0.00019882593042862352,
+      "loss": 1.2045,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07744045699543926,
+      "grad_norm": 1.9657875045849278,
+      "learning_rate": 0.00019882364962667,
+      "loss": 1.4639,
+      "step": 1681
+    },
+    {
+      "epoch": 0.07748652508407426,
+      "grad_norm": 1.2208561517334964,
+      "learning_rate": 0.00019882136662457734,
+      "loss": 1.4763,
+      "step": 1682
+    },
+    {
+      "epoch": 0.07753259317270926,
+      "grad_norm": 1.3961170673831707,
+      "learning_rate": 0.00019881908142239637,
+      "loss": 1.5899,
+      "step": 1683
+    },
+    {
+      "epoch": 0.07757866126134427,
+      "grad_norm": 1.1969202166609054,
+      "learning_rate": 0.00019881679402017802,
+      "loss": 1.2586,
+      "step": 1684
+    },
+    {
+      "epoch": 0.07762472934997927,
+      "grad_norm": 1.344548340212687,
+      "learning_rate": 0.00019881450441797322,
+      "loss": 1.377,
+      "step": 1685
+    },
+    {
+      "epoch": 0.07767079743861427,
+      "grad_norm": 1.8478336881356046,
+      "learning_rate": 0.00019881221261583288,
+      "loss": 1.4331,
+      "step": 1686
+    },
+    {
+      "epoch": 0.07771686552724927,
+      "grad_norm": 1.4437086541860582,
+      "learning_rate": 0.00019880991861380805,
+      "loss": 1.3197,
+      "step": 1687
+    },
+    {
+      "epoch": 0.07776293361588428,
+      "grad_norm": 1.5560951302473494,
+      "learning_rate": 0.0001988076224119498,
+      "loss": 1.2814,
+      "step": 1688
+    },
+    {
+      "epoch": 0.07780900170451928,
+      "grad_norm": 1.3666216894744456,
+      "learning_rate": 0.00019880532401030926,
+      "loss": 1.9611,
+      "step": 1689
+    },
+    {
+      "epoch": 0.07785506979315428,
+      "grad_norm": 1.0690110925200287,
+      "learning_rate": 0.00019880302340893758,
+      "loss": 1.3194,
+      "step": 1690
+    },
+    {
+      "epoch": 0.07790113788178929,
+      "grad_norm": 1.5315901457012908,
+      "learning_rate": 0.000198800720607886,
+      "loss": 1.1382,
+      "step": 1691
+    },
+    {
+      "epoch": 0.07794720597042429,
+      "grad_norm": 2.567923491897771,
+      "learning_rate": 0.0001987984156072058,
+      "loss": 1.4795,
+      "step": 1692
+    },
+    {
+      "epoch": 0.07799327405905929,
+      "grad_norm": 1.3654326191034072,
+      "learning_rate": 0.0001987961084069483,
+      "loss": 1.466,
+      "step": 1693
+    },
+    {
+      "epoch": 0.0780393421476943,
+      "grad_norm": 1.5935604736294393,
+      "learning_rate": 0.0001987937990071648,
+      "loss": 1.2351,
+      "step": 1694
+    },
+    {
+      "epoch": 0.0780854102363293,
+      "grad_norm": 1.4374542937944519,
+      "learning_rate": 0.0001987914874079068,
+      "loss": 1.0913,
+      "step": 1695
+    },
+    {
+      "epoch": 0.0781314783249643,
+      "grad_norm": 1.2843611055459179,
+      "learning_rate": 0.00019878917360922568,
+      "loss": 1.6518,
+      "step": 1696
+    },
+    {
+      "epoch": 0.0781775464135993,
+      "grad_norm": 1.1885838371202557,
+      "learning_rate": 0.00019878685761117303,
+      "loss": 1.7647,
+      "step": 1697
+    },
+    {
+      "epoch": 0.0782236145022343,
+      "grad_norm": 1.3158755103317425,
+      "learning_rate": 0.00019878453941380036,
+      "loss": 1.4388,
+      "step": 1698
+    },
+    {
+      "epoch": 0.0782696825908693,
+      "grad_norm": 1.1134400333443728,
+      "learning_rate": 0.0001987822190171593,
+      "loss": 1.0658,
+      "step": 1699
+    },
+    {
+      "epoch": 0.0783157506795043,
+      "grad_norm": 1.456596470152914,
+      "learning_rate": 0.0001987798964213015,
+      "loss": 1.2511,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07836181876813932,
+      "grad_norm": 1.251060730411767,
+      "learning_rate": 0.0001987775716262787,
+      "loss": 1.3216,
+      "step": 1701
+    },
+    {
+      "epoch": 0.07840788685677431,
+      "grad_norm": 0.858095391869021,
+      "learning_rate": 0.00019877524463214263,
+      "loss": 1.2821,
+      "step": 1702
+    },
+    {
+      "epoch": 0.07845395494540931,
+      "grad_norm": 0.9746527560102567,
+      "learning_rate": 0.0001987729154389451,
+      "loss": 1.3697,
+      "step": 1703
+    },
+    {
+      "epoch": 0.07850002303404431,
+      "grad_norm": 0.774237046131867,
+      "learning_rate": 0.00019877058404673797,
+      "loss": 1.742,
+      "step": 1704
+    },
+    {
+      "epoch": 0.07854609112267932,
+      "grad_norm": 1.199180446819006,
+      "learning_rate": 0.00019876825045557316,
+      "loss": 1.4594,
+      "step": 1705
+    },
+    {
+      "epoch": 0.07859215921131432,
+      "grad_norm": 0.821490479584148,
+      "learning_rate": 0.00019876591466550254,
+      "loss": 1.2388,
+      "step": 1706
+    },
+    {
+      "epoch": 0.07863822729994932,
+      "grad_norm": 1.1622530140115501,
+      "learning_rate": 0.00019876357667657823,
+      "loss": 1.3112,
+      "step": 1707
+    },
+    {
+      "epoch": 0.07868429538858433,
+      "grad_norm": 1.0651457487593703,
+      "learning_rate": 0.0001987612364888522,
+      "loss": 1.2766,
+      "step": 1708
+    },
+    {
+      "epoch": 0.07873036347721933,
+      "grad_norm": 1.0392464760530833,
+      "learning_rate": 0.00019875889410237663,
+      "loss": 1.4415,
+      "step": 1709
+    },
+    {
+      "epoch": 0.07877643156585433,
+      "grad_norm": 0.821772938046262,
+      "learning_rate": 0.00019875654951720357,
+      "loss": 1.4304,
+      "step": 1710
+    },
+    {
+      "epoch": 0.07882249965448934,
+      "grad_norm": 0.9189591088824,
+      "learning_rate": 0.00019875420273338527,
+      "loss": 1.5965,
+      "step": 1711
+    },
+    {
+      "epoch": 0.07886856774312434,
+      "grad_norm": 1.9719522635386701,
+      "learning_rate": 0.000198751853750974,
+      "loss": 1.4263,
+      "step": 1712
+    },
+    {
+      "epoch": 0.07891463583175934,
+      "grad_norm": 1.2321381571807366,
+      "learning_rate": 0.000198749502570022,
+      "loss": 1.2719,
+      "step": 1713
+    },
+    {
+      "epoch": 0.07896070392039434,
+      "grad_norm": 1.441827073334531,
+      "learning_rate": 0.00019874714919058167,
+      "loss": 1.2942,
+      "step": 1714
+    },
+    {
+      "epoch": 0.07900677200902935,
+      "grad_norm": 1.5975259208955817,
+      "learning_rate": 0.00019874479361270538,
+      "loss": 1.3044,
+      "step": 1715
+    },
+    {
+      "epoch": 0.07905284009766435,
+      "grad_norm": 1.759322252080767,
+      "learning_rate": 0.00019874243583644554,
+      "loss": 1.3474,
+      "step": 1716
+    },
+    {
+      "epoch": 0.07909890818629935,
+      "grad_norm": 0.8331886751451038,
+      "learning_rate": 0.00019874007586185468,
+      "loss": 1.5596,
+      "step": 1717
+    },
+    {
+      "epoch": 0.07914497627493436,
+      "grad_norm": 1.2756808960685897,
+      "learning_rate": 0.00019873771368898534,
+      "loss": 1.4215,
+      "step": 1718
+    },
+    {
+      "epoch": 0.07919104436356936,
+      "grad_norm": 0.7492005243976285,
+      "learning_rate": 0.0001987353493178901,
+      "loss": 1.1892,
+      "step": 1719
+    },
+    {
+      "epoch": 0.07923711245220436,
+      "grad_norm": 1.1273439303379889,
+      "learning_rate": 0.00019873298274862162,
+      "loss": 1.1703,
+      "step": 1720
+    },
+    {
+      "epoch": 0.07928318054083935,
+      "grad_norm": 1.1199586754179665,
+      "learning_rate": 0.00019873061398123257,
+      "loss": 1.5433,
+      "step": 1721
+    },
+    {
+      "epoch": 0.07932924862947437,
+      "grad_norm": 1.1682048045371216,
+      "learning_rate": 0.00019872824301577567,
+      "loss": 1.1006,
+      "step": 1722
+    },
+    {
+      "epoch": 0.07937531671810936,
+      "grad_norm": 0.8810106214376813,
+      "learning_rate": 0.00019872586985230373,
+      "loss": 1.6153,
+      "step": 1723
+    },
+    {
+      "epoch": 0.07942138480674436,
+      "grad_norm": 1.1192454657593047,
+      "learning_rate": 0.00019872349449086957,
+      "loss": 1.3339,
+      "step": 1724
+    },
+    {
+      "epoch": 0.07946745289537938,
+      "grad_norm": 1.0880820490403307,
+      "learning_rate": 0.00019872111693152608,
+      "loss": 1.1033,
+      "step": 1725
+    },
+    {
+      "epoch": 0.07951352098401437,
+      "grad_norm": 1.1730910313558511,
+      "learning_rate": 0.0001987187371743262,
+      "loss": 1.4163,
+      "step": 1726
+    },
+    {
+      "epoch": 0.07955958907264937,
+      "grad_norm": 1.4273296644638,
+      "learning_rate": 0.0001987163552193229,
+      "loss": 1.3224,
+      "step": 1727
+    },
+    {
+      "epoch": 0.07960565716128438,
+      "grad_norm": 0.785802852594324,
+      "learning_rate": 0.00019871397106656922,
+      "loss": 1.6243,
+      "step": 1728
+    },
+    {
+      "epoch": 0.07965172524991938,
+      "grad_norm": 1.1086116740175374,
+      "learning_rate": 0.0001987115847161182,
+      "loss": 1.2321,
+      "step": 1729
+    },
+    {
+      "epoch": 0.07969779333855438,
+      "grad_norm": 0.8396436341076168,
+      "learning_rate": 0.00019870919616802303,
+      "loss": 1.2931,
+      "step": 1730
+    },
+    {
+      "epoch": 0.07974386142718938,
+      "grad_norm": 1.3806814005016843,
+      "learning_rate": 0.00019870680542233686,
+      "loss": 1.4063,
+      "step": 1731
+    },
+    {
+      "epoch": 0.07978992951582439,
+      "grad_norm": 0.979159248137881,
+      "learning_rate": 0.00019870441247911292,
+      "loss": 1.5262,
+      "step": 1732
+    },
+    {
+      "epoch": 0.07983599760445939,
+      "grad_norm": 0.736823015519698,
+      "learning_rate": 0.00019870201733840443,
+      "loss": 1.5661,
+      "step": 1733
+    },
+    {
+      "epoch": 0.07988206569309439,
+      "grad_norm": 1.0309305713947845,
+      "learning_rate": 0.00019869962000026482,
+      "loss": 1.4649,
+      "step": 1734
+    },
+    {
+      "epoch": 0.0799281337817294,
+      "grad_norm": 0.8835126809103617,
+      "learning_rate": 0.00019869722046474738,
+      "loss": 1.2171,
+      "step": 1735
+    },
+    {
+      "epoch": 0.0799742018703644,
+      "grad_norm": 1.0177976419470485,
+      "learning_rate": 0.00019869481873190553,
+      "loss": 1.6,
+      "step": 1736
+    },
+    {
+      "epoch": 0.0800202699589994,
+      "grad_norm": 0.9023869369034814,
+      "learning_rate": 0.00019869241480179278,
+      "loss": 1.3912,
+      "step": 1737
+    },
+    {
+      "epoch": 0.0800663380476344,
+      "grad_norm": 1.1177353919439368,
+      "learning_rate": 0.0001986900086744627,
+      "loss": 1.3583,
+      "step": 1738
+    },
+    {
+      "epoch": 0.08011240613626941,
+      "grad_norm": 1.1006429081588662,
+      "learning_rate": 0.00019868760034996873,
+      "loss": 1.1309,
+      "step": 1739
+    },
+    {
+      "epoch": 0.08015847422490441,
+      "grad_norm": 1.400846003693539,
+      "learning_rate": 0.00019868518982836455,
+      "loss": 0.8481,
+      "step": 1740
+    },
+    {
+      "epoch": 0.0802045423135394,
+      "grad_norm": 1.056006272538941,
+      "learning_rate": 0.0001986827771097038,
+      "loss": 1.3015,
+      "step": 1741
+    },
+    {
+      "epoch": 0.08025061040217442,
+      "grad_norm": 1.113378554931602,
+      "learning_rate": 0.0001986803621940403,
+      "loss": 1.0477,
+      "step": 1742
+    },
+    {
+      "epoch": 0.08029667849080942,
+      "grad_norm": 0.8614396880975569,
+      "learning_rate": 0.0001986779450814277,
+      "loss": 1.3648,
+      "step": 1743
+    },
+    {
+      "epoch": 0.08034274657944442,
+      "grad_norm": 1.0788136950578968,
+      "learning_rate": 0.00019867552577191986,
+      "loss": 1.5927,
+      "step": 1744
+    },
+    {
+      "epoch": 0.08038881466807943,
+      "grad_norm": 1.205871368806219,
+      "learning_rate": 0.0001986731042655706,
+      "loss": 1.225,
+      "step": 1745
+    },
+    {
+      "epoch": 0.08043488275671443,
+      "grad_norm": 0.919835485843781,
+      "learning_rate": 0.00019867068056243388,
+      "loss": 1.4612,
+      "step": 1746
+    },
+    {
+      "epoch": 0.08048095084534942,
+      "grad_norm": 1.5561512597546878,
+      "learning_rate": 0.00019866825466256365,
+      "loss": 1.1923,
+      "step": 1747
+    },
+    {
+      "epoch": 0.08052701893398442,
+      "grad_norm": 1.122751716061074,
+      "learning_rate": 0.0001986658265660139,
+      "loss": 1.4616,
+      "step": 1748
+    },
+    {
+      "epoch": 0.08057308702261944,
+      "grad_norm": 1.0514680717349483,
+      "learning_rate": 0.00019866339627283867,
+      "loss": 1.5611,
+      "step": 1749
+    },
+    {
+      "epoch": 0.08061915511125443,
+      "grad_norm": 0.9519376621328782,
+      "learning_rate": 0.00019866096378309212,
+      "loss": 1.0416,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08066522319988943,
+      "grad_norm": 0.964495748710859,
+      "learning_rate": 0.0001986585290968284,
+      "loss": 1.2532,
+      "step": 1751
+    },
+    {
+      "epoch": 0.08071129128852444,
+      "grad_norm": 1.4247262287244615,
+      "learning_rate": 0.0001986560922141017,
+      "loss": 1.3097,
+      "step": 1752
+    },
+    {
+      "epoch": 0.08075735937715944,
+      "grad_norm": 1.1042692224311288,
+      "learning_rate": 0.00019865365313496622,
+      "loss": 1.2013,
+      "step": 1753
+    },
+    {
+      "epoch": 0.08080342746579444,
+      "grad_norm": 1.126483680007683,
+      "learning_rate": 0.00019865121185947633,
+      "loss": 1.3745,
+      "step": 1754
+    },
+    {
+      "epoch": 0.08084949555442944,
+      "grad_norm": 1.9319155968371176,
+      "learning_rate": 0.00019864876838768632,
+      "loss": 1.5931,
+      "step": 1755
+    },
+    {
+      "epoch": 0.08089556364306445,
+      "grad_norm": 1.4426463120191713,
+      "learning_rate": 0.00019864632271965067,
+      "loss": 1.0021,
+      "step": 1756
+    },
+    {
+      "epoch": 0.08094163173169945,
+      "grad_norm": 1.325800889699604,
+      "learning_rate": 0.00019864387485542378,
+      "loss": 1.4186,
+      "step": 1757
+    },
+    {
+      "epoch": 0.08098769982033445,
+      "grad_norm": 1.2230431547755727,
+      "learning_rate": 0.00019864142479506016,
+      "loss": 1.2661,
+      "step": 1758
+    },
+    {
+      "epoch": 0.08103376790896946,
+      "grad_norm": 1.1301267170478522,
+      "learning_rate": 0.00019863897253861437,
+      "loss": 1.2247,
+      "step": 1759
+    },
+    {
+      "epoch": 0.08107983599760446,
+      "grad_norm": 1.0079058528943252,
+      "learning_rate": 0.00019863651808614094,
+      "loss": 1.7225,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08112590408623946,
+      "grad_norm": 1.0088860561795443,
+      "learning_rate": 0.00019863406143769458,
+      "loss": 1.4185,
+      "step": 1761
+    },
+    {
+      "epoch": 0.08117197217487447,
+      "grad_norm": 1.0547148061498155,
+      "learning_rate": 0.00019863160259332995,
+      "loss": 1.3587,
+      "step": 1762
+    },
+    {
+      "epoch": 0.08121804026350947,
+      "grad_norm": 1.3147774532198075,
+      "learning_rate": 0.00019862914155310184,
+      "loss": 1.8412,
+      "step": 1763
+    },
+    {
+      "epoch": 0.08126410835214447,
+      "grad_norm": 1.068565235386099,
+      "learning_rate": 0.00019862667831706498,
+      "loss": 1.3561,
+      "step": 1764
+    },
+    {
+      "epoch": 0.08131017644077947,
+      "grad_norm": 0.9635687847785795,
+      "learning_rate": 0.00019862421288527424,
+      "loss": 1.4167,
+      "step": 1765
+    },
+    {
+      "epoch": 0.08135624452941448,
+      "grad_norm": 1.0577352720816695,
+      "learning_rate": 0.0001986217452577845,
+      "loss": 1.7481,
+      "step": 1766
+    },
+    {
+      "epoch": 0.08140231261804948,
+      "grad_norm": 1.0485046728351164,
+      "learning_rate": 0.0001986192754346507,
+      "loss": 1.0616,
+      "step": 1767
+    },
+    {
+      "epoch": 0.08144838070668448,
+      "grad_norm": 0.9462305784179671,
+      "learning_rate": 0.00019861680341592784,
+      "loss": 1.4315,
+      "step": 1768
+    },
+    {
+      "epoch": 0.08149444879531949,
+      "grad_norm": 1.0754394024876082,
+      "learning_rate": 0.00019861432920167095,
+      "loss": 1.5051,
+      "step": 1769
+    },
+    {
+      "epoch": 0.08154051688395449,
+      "grad_norm": 0.891068096837454,
+      "learning_rate": 0.0001986118527919351,
+      "loss": 1.4335,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08158658497258948,
+      "grad_norm": 1.684666543739825,
+      "learning_rate": 0.00019860937418677543,
+      "loss": 1.2019,
+      "step": 1771
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 1.251880960242657,
+      "learning_rate": 0.0001986068933862471,
+      "loss": 1.0994,
+      "step": 1772
+    },
+    {
+      "epoch": 0.0816787211498595,
+      "grad_norm": 0.8649020685509695,
+      "learning_rate": 0.0001986044103904054,
+      "loss": 1.3931,
+      "step": 1773
+    },
+    {
+      "epoch": 0.0817247892384945,
+      "grad_norm": 0.9643135802924216,
+      "learning_rate": 0.00019860192519930556,
+      "loss": 1.4061,
+      "step": 1774
+    },
+    {
+      "epoch": 0.08177085732712949,
+      "grad_norm": 1.6129660224952322,
+      "learning_rate": 0.0001985994378130029,
+      "loss": 1.4937,
+      "step": 1775
+    },
+    {
+      "epoch": 0.0818169254157645,
+      "grad_norm": 1.1338793907196802,
+      "learning_rate": 0.00019859694823155284,
+      "loss": 1.5375,
+      "step": 1776
+    },
+    {
+      "epoch": 0.0818629935043995,
+      "grad_norm": 1.7123279232472781,
+      "learning_rate": 0.00019859445645501077,
+      "loss": 1.2887,
+      "step": 1777
+    },
+    {
+      "epoch": 0.0819090615930345,
+      "grad_norm": 1.1831893025802158,
+      "learning_rate": 0.00019859196248343222,
+      "loss": 1.6348,
+      "step": 1778
+    },
+    {
+      "epoch": 0.08195512968166951,
+      "grad_norm": 1.2119891582247708,
+      "learning_rate": 0.00019858946631687264,
+      "loss": 1.0301,
+      "step": 1779
+    },
+    {
+      "epoch": 0.08200119777030451,
+      "grad_norm": 1.450189566869277,
+      "learning_rate": 0.00019858696795538764,
+      "loss": 1.2137,
+      "step": 1780
+    },
+    {
+      "epoch": 0.08204726585893951,
+      "grad_norm": 2.500743385180486,
+      "learning_rate": 0.00019858446739903283,
+      "loss": 1.5226,
+      "step": 1781
+    },
+    {
+      "epoch": 0.08209333394757451,
+      "grad_norm": 1.407221827287739,
+      "learning_rate": 0.0001985819646478639,
+      "loss": 1.1261,
+      "step": 1782
+    },
+    {
+      "epoch": 0.08213940203620952,
+      "grad_norm": 1.611565957703064,
+      "learning_rate": 0.00019857945970193656,
+      "loss": 0.9534,
+      "step": 1783
+    },
+    {
+      "epoch": 0.08218547012484452,
+      "grad_norm": 1.7072113855347768,
+      "learning_rate": 0.00019857695256130658,
+      "loss": 1.4629,
+      "step": 1784
+    },
+    {
+      "epoch": 0.08223153821347952,
+      "grad_norm": 0.8961069402192444,
+      "learning_rate": 0.00019857444322602977,
+      "loss": 1.552,
+      "step": 1785
+    },
+    {
+      "epoch": 0.08227760630211453,
+      "grad_norm": 2.4897720064304147,
+      "learning_rate": 0.00019857193169616201,
+      "loss": 1.1137,
+      "step": 1786
+    },
+    {
+      "epoch": 0.08232367439074953,
+      "grad_norm": 1.939019917607417,
+      "learning_rate": 0.0001985694179717592,
+      "loss": 1.3711,
+      "step": 1787
+    },
+    {
+      "epoch": 0.08236974247938453,
+      "grad_norm": 1.093749017223252,
+      "learning_rate": 0.0001985669020528773,
+      "loss": 1.2121,
+      "step": 1788
+    },
+    {
+      "epoch": 0.08241581056801953,
+      "grad_norm": 0.9033346112923233,
+      "learning_rate": 0.00019856438393957233,
+      "loss": 1.5048,
+      "step": 1789
+    },
+    {
+      "epoch": 0.08246187865665454,
+      "grad_norm": 1.1106335680959576,
+      "learning_rate": 0.00019856186363190038,
+      "loss": 1.6547,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08250794674528954,
+      "grad_norm": 1.397147359306878,
+      "learning_rate": 0.0001985593411299175,
+      "loss": 1.3028,
+      "step": 1791
+    },
+    {
+      "epoch": 0.08255401483392454,
+      "grad_norm": 1.1181921856588817,
+      "learning_rate": 0.00019855681643367993,
+      "loss": 1.3014,
+      "step": 1792
+    },
+    {
+      "epoch": 0.08260008292255955,
+      "grad_norm": 0.9362429526853848,
+      "learning_rate": 0.00019855428954324377,
+      "loss": 1.1447,
+      "step": 1793
+    },
+    {
+      "epoch": 0.08264615101119455,
+      "grad_norm": 1.026138144297491,
+      "learning_rate": 0.0001985517604586654,
+      "loss": 1.1541,
+      "step": 1794
+    },
+    {
+      "epoch": 0.08269221909982954,
+      "grad_norm": 0.7590039549353118,
+      "learning_rate": 0.000198549229180001,
+      "loss": 1.3568,
+      "step": 1795
+    },
+    {
+      "epoch": 0.08273828718846456,
+      "grad_norm": 0.9327953074096784,
+      "learning_rate": 0.00019854669570730705,
+      "loss": 1.4455,
+      "step": 1796
+    },
+    {
+      "epoch": 0.08278435527709956,
+      "grad_norm": 1.6540633288343725,
+      "learning_rate": 0.00019854416004063983,
+      "loss": 1.4049,
+      "step": 1797
+    },
+    {
+      "epoch": 0.08283042336573455,
+      "grad_norm": 1.3601891763243357,
+      "learning_rate": 0.00019854162218005587,
+      "loss": 1.1436,
+      "step": 1798
+    },
+    {
+      "epoch": 0.08287649145436955,
+      "grad_norm": 1.2429085020482962,
+      "learning_rate": 0.00019853908212561167,
+      "loss": 0.9817,
+      "step": 1799
+    },
+    {
+      "epoch": 0.08292255954300456,
+      "grad_norm": 0.8687704829386835,
+      "learning_rate": 0.0001985365398773638,
+      "loss": 1.3744,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08296862763163956,
+      "grad_norm": 1.0231467074930842,
+      "learning_rate": 0.00019853399543536877,
+      "loss": 1.5915,
+      "step": 1801
+    },
+    {
+      "epoch": 0.08301469572027456,
+      "grad_norm": 0.7600505626322085,
+      "learning_rate": 0.00019853144879968332,
+      "loss": 1.4703,
+      "step": 1802
+    },
+    {
+      "epoch": 0.08306076380890957,
+      "grad_norm": 1.4506086020901088,
+      "learning_rate": 0.00019852889997036405,
+      "loss": 1.4846,
+      "step": 1803
+    },
+    {
+      "epoch": 0.08310683189754457,
+      "grad_norm": 2.9669063663858686,
+      "learning_rate": 0.0001985263489474678,
+      "loss": 1.2024,
+      "step": 1804
+    },
+    {
+      "epoch": 0.08315289998617957,
+      "grad_norm": 1.38858198716373,
+      "learning_rate": 0.00019852379573105132,
+      "loss": 1.1106,
+      "step": 1805
+    },
+    {
+      "epoch": 0.08319896807481457,
+      "grad_norm": 1.7593394229818564,
+      "learning_rate": 0.0001985212403211715,
+      "loss": 1.1269,
+      "step": 1806
+    },
+    {
+      "epoch": 0.08324503616344958,
+      "grad_norm": 2.1775910012090867,
+      "learning_rate": 0.00019851868271788514,
+      "loss": 1.5348,
+      "step": 1807
+    },
+    {
+      "epoch": 0.08329110425208458,
+      "grad_norm": 1.569653726609947,
+      "learning_rate": 0.00019851612292124922,
+      "loss": 1.4306,
+      "step": 1808
+    },
+    {
+      "epoch": 0.08333717234071958,
+      "grad_norm": 1.0083318793185734,
+      "learning_rate": 0.0001985135609313208,
+      "loss": 1.4207,
+      "step": 1809
+    },
+    {
+      "epoch": 0.08338324042935459,
+      "grad_norm": 2.134794041882093,
+      "learning_rate": 0.00019851099674815683,
+      "loss": 1.4053,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08342930851798959,
+      "grad_norm": 1.67081770090858,
+      "learning_rate": 0.00019850843037181443,
+      "loss": 1.1256,
+      "step": 1811
+    },
+    {
+      "epoch": 0.08347537660662459,
+      "grad_norm": 1.1539342054905106,
+      "learning_rate": 0.00019850586180235075,
+      "loss": 1.5694,
+      "step": 1812
+    },
+    {
+      "epoch": 0.0835214446952596,
+      "grad_norm": 1.344245922074858,
+      "learning_rate": 0.00019850329103982293,
+      "loss": 1.6928,
+      "step": 1813
+    },
+    {
+      "epoch": 0.0835675127838946,
+      "grad_norm": 1.8628913634494195,
+      "learning_rate": 0.00019850071808428824,
+      "loss": 1.225,
+      "step": 1814
+    },
+    {
+      "epoch": 0.0836135808725296,
+      "grad_norm": 1.9028391722532134,
+      "learning_rate": 0.00019849814293580398,
+      "loss": 1.3801,
+      "step": 1815
+    },
+    {
+      "epoch": 0.0836596489611646,
+      "grad_norm": 2.039903744074717,
+      "learning_rate": 0.00019849556559442744,
+      "loss": 1.1386,
+      "step": 1816
+    },
+    {
+      "epoch": 0.08370571704979961,
+      "grad_norm": 1.5204506327824685,
+      "learning_rate": 0.000198492986060216,
+      "loss": 1.6796,
+      "step": 1817
+    },
+    {
+      "epoch": 0.0837517851384346,
+      "grad_norm": 1.2085964017494772,
+      "learning_rate": 0.0001984904043332271,
+      "loss": 1.4737,
+      "step": 1818
+    },
+    {
+      "epoch": 0.0837978532270696,
+      "grad_norm": 1.4430162810371439,
+      "learning_rate": 0.00019848782041351822,
+      "loss": 1.4084,
+      "step": 1819
+    },
+    {
+      "epoch": 0.08384392131570462,
+      "grad_norm": 2.0135795994395322,
+      "learning_rate": 0.0001984852343011469,
+      "loss": 1.5743,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08388998940433962,
+      "grad_norm": 2.1902604130821035,
+      "learning_rate": 0.00019848264599617074,
+      "loss": 1.3083,
+      "step": 1821
+    },
+    {
+      "epoch": 0.08393605749297461,
+      "grad_norm": 53.39146018951499,
+      "learning_rate": 0.0001984800554986473,
+      "loss": 1.4943,
+      "step": 1822
+    },
+    {
+      "epoch": 0.08398212558160961,
+      "grad_norm": 27.739820191953466,
+      "learning_rate": 0.0001984774628086343,
+      "loss": 9.9152,
+      "step": 1823
+    },
+    {
+      "epoch": 0.08402819367024462,
+      "grad_norm": 10.09906729484748,
+      "learning_rate": 0.0001984748679261894,
+      "loss": 7.4346,
+      "step": 1824
+    },
+    {
+      "epoch": 0.08407426175887962,
+      "grad_norm": 2.3315289884883703,
+      "learning_rate": 0.00019847227085137043,
+      "loss": 6.9353,
+      "step": 1825
+    },
+    {
+      "epoch": 0.08412032984751462,
+      "grad_norm": 2.4856060768968353,
+      "learning_rate": 0.00019846967158423522,
+      "loss": 7.3441,
+      "step": 1826
+    },
+    {
+      "epoch": 0.08416639793614963,
+      "grad_norm": 2.2143625397737057,
+      "learning_rate": 0.00019846707012484158,
+      "loss": 6.7811,
+      "step": 1827
+    },
+    {
+      "epoch": 0.08421246602478463,
+      "grad_norm": 2.2189380989525795,
+      "learning_rate": 0.00019846446647324748,
+      "loss": 7.1356,
+      "step": 1828
+    },
+    {
+      "epoch": 0.08425853411341963,
+      "grad_norm": 2.822225016477036,
+      "learning_rate": 0.00019846186062951084,
+      "loss": 6.9357,
+      "step": 1829
+    },
+    {
+      "epoch": 0.08430460220205464,
+      "grad_norm": 3.4516434057988423,
+      "learning_rate": 0.0001984592525936897,
+      "loss": 6.5872,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08435067029068964,
+      "grad_norm": 1.8093890875851453,
+      "learning_rate": 0.00019845664236584217,
+      "loss": 6.5518,
+      "step": 1831
+    },
+    {
+      "epoch": 0.08439673837932464,
+      "grad_norm": 2.1378897237358316,
+      "learning_rate": 0.00019845402994602627,
+      "loss": 7.1669,
+      "step": 1832
+    },
+    {
+      "epoch": 0.08444280646795964,
+      "grad_norm": 3.6601805126276683,
+      "learning_rate": 0.00019845141533430022,
+      "loss": 7.2364,
+      "step": 1833
+    },
+    {
+      "epoch": 0.08448887455659465,
+      "grad_norm": 1.8480376425393366,
+      "learning_rate": 0.0001984487985307222,
+      "loss": 6.4471,
+      "step": 1834
+    },
+    {
+      "epoch": 0.08453494264522965,
+      "grad_norm": 1.570378606896317,
+      "learning_rate": 0.0001984461795353505,
+      "loss": 6.9459,
+      "step": 1835
+    },
+    {
+      "epoch": 0.08458101073386465,
+      "grad_norm": 1.7185662114223295,
+      "learning_rate": 0.0001984435583482434,
+      "loss": 6.7247,
+      "step": 1836
+    },
+    {
+      "epoch": 0.08462707882249966,
+      "grad_norm": 1.6198281498467753,
+      "learning_rate": 0.00019844093496945928,
+      "loss": 6.2184,
+      "step": 1837
+    },
+    {
+      "epoch": 0.08467314691113466,
+      "grad_norm": 19.92498871411672,
+      "learning_rate": 0.00019843830939905653,
+      "loss": 6.0044,
+      "step": 1838
+    },
+    {
+      "epoch": 0.08471921499976966,
+      "grad_norm": 2.4288733393085895,
+      "learning_rate": 0.00019843568163709362,
+      "loss": 7.1064,
+      "step": 1839
+    },
+    {
+      "epoch": 0.08476528308840466,
+      "grad_norm": 1.7511260579976657,
+      "learning_rate": 0.00019843305168362903,
+      "loss": 6.3744,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08481135117703967,
+      "grad_norm": 2.9061429743662677,
+      "learning_rate": 0.0001984304195387213,
+      "loss": 7.6299,
+      "step": 1841
+    },
+    {
+      "epoch": 0.08485741926567467,
+      "grad_norm": 2.0061878836788534,
+      "learning_rate": 0.00019842778520242905,
+      "loss": 6.698,
+      "step": 1842
+    },
+    {
+      "epoch": 0.08490348735430966,
+      "grad_norm": 0.7461905132416753,
+      "learning_rate": 0.00019842514867481095,
+      "loss": 7.0204,
+      "step": 1843
+    },
+    {
+      "epoch": 0.08494955544294468,
+      "grad_norm": 1.6067112008783362,
+      "learning_rate": 0.00019842250995592568,
+      "loss": 7.218,
+      "step": 1844
+    },
+    {
+      "epoch": 0.08499562353157968,
+      "grad_norm": 2.098349068442601,
+      "learning_rate": 0.00019841986904583199,
+      "loss": 6.3358,
+      "step": 1845
+    },
+    {
+      "epoch": 0.08504169162021467,
+      "grad_norm": 2.822893900641675,
+      "learning_rate": 0.00019841722594458864,
+      "loss": 6.8805,
+      "step": 1846
+    },
+    {
+      "epoch": 0.08508775970884969,
+      "grad_norm": 1.7523786626317828,
+      "learning_rate": 0.00019841458065225452,
+      "loss": 6.3679,
+      "step": 1847
+    },
+    {
+      "epoch": 0.08513382779748468,
+      "grad_norm": 1.8346308190465956,
+      "learning_rate": 0.0001984119331688885,
+      "loss": 6.1252,
+      "step": 1848
+    },
+    {
+      "epoch": 0.08517989588611968,
+      "grad_norm": 1.322888376999675,
+      "learning_rate": 0.00019840928349454954,
+      "loss": 6.309,
+      "step": 1849
+    },
+    {
+      "epoch": 0.08522596397475468,
+      "grad_norm": 1.5643327741847335,
+      "learning_rate": 0.00019840663162929665,
+      "loss": 6.5599,
+      "step": 1850
+    },
+    {
+      "epoch": 0.0852720320633897,
+      "grad_norm": 1.0129118110265742,
+      "learning_rate": 0.00019840397757318878,
+      "loss": 6.9228,
+      "step": 1851
+    },
+    {
+      "epoch": 0.08531810015202469,
+      "grad_norm": 1.3463090534625028,
+      "learning_rate": 0.0001984013213262851,
+      "loss": 6.1018,
+      "step": 1852
+    },
+    {
+      "epoch": 0.08536416824065969,
+      "grad_norm": 4.71491015496975,
+      "learning_rate": 0.00019839866288864474,
+      "loss": 6.1095,
+      "step": 1853
+    },
+    {
+      "epoch": 0.0854102363292947,
+      "grad_norm": 1.4725384452269772,
+      "learning_rate": 0.00019839600226032684,
+      "loss": 6.5837,
+      "step": 1854
+    },
+    {
+      "epoch": 0.0854563044179297,
+      "grad_norm": 1.50617226113335,
+      "learning_rate": 0.00019839333944139068,
+      "loss": 6.4747,
+      "step": 1855
+    },
+    {
+      "epoch": 0.0855023725065647,
+      "grad_norm": 1.1049540300862113,
+      "learning_rate": 0.0001983906744318955,
+      "loss": 6.7534,
+      "step": 1856
+    },
+    {
+      "epoch": 0.0855484405951997,
+      "grad_norm": 1.2541344049281788,
+      "learning_rate": 0.0001983880072319007,
+      "loss": 6.0457,
+      "step": 1857
+    },
+    {
+      "epoch": 0.08559450868383471,
+      "grad_norm": 1.481757565048243,
+      "learning_rate": 0.00019838533784146558,
+      "loss": 5.5565,
+      "step": 1858
+    },
+    {
+      "epoch": 0.08564057677246971,
+      "grad_norm": 1.973596395688852,
+      "learning_rate": 0.0001983826662606496,
+      "loss": 6.4776,
+      "step": 1859
+    },
+    {
+      "epoch": 0.08568664486110471,
+      "grad_norm": 1.1782311405009658,
+      "learning_rate": 0.00019837999248951228,
+      "loss": 6.3838,
+      "step": 1860
+    },
+    {
+      "epoch": 0.08573271294973972,
+      "grad_norm": 0.7820518645740672,
+      "learning_rate": 0.00019837731652811312,
+      "loss": 6.1675,
+      "step": 1861
+    },
+    {
+      "epoch": 0.08577878103837472,
+      "grad_norm": 1.1824795534371004,
+      "learning_rate": 0.00019837463837651167,
+      "loss": 6.9248,
+      "step": 1862
+    },
+    {
+      "epoch": 0.08582484912700972,
+      "grad_norm": 0.8073869050314264,
+      "learning_rate": 0.00019837195803476757,
+      "loss": 6.2156,
+      "step": 1863
+    },
+    {
+      "epoch": 0.08587091721564473,
+      "grad_norm": 1.2387275591550901,
+      "learning_rate": 0.0001983692755029405,
+      "loss": 7.4036,
+      "step": 1864
+    },
+    {
+      "epoch": 0.08591698530427973,
+      "grad_norm": 0.8336415612711319,
+      "learning_rate": 0.00019836659078109018,
+      "loss": 6.6082,
+      "step": 1865
+    },
+    {
+      "epoch": 0.08596305339291473,
+      "grad_norm": 0.989044322224326,
+      "learning_rate": 0.00019836390386927634,
+      "loss": 6.3779,
+      "step": 1866
+    },
+    {
+      "epoch": 0.08600912148154972,
+      "grad_norm": 1.576021379847767,
+      "learning_rate": 0.0001983612147675589,
+      "loss": 5.7671,
+      "step": 1867
+    },
+    {
+      "epoch": 0.08605518957018474,
+      "grad_norm": 1.2583961623050466,
+      "learning_rate": 0.00019835852347599762,
+      "loss": 6.9533,
+      "step": 1868
+    },
+    {
+      "epoch": 0.08610125765881974,
+      "grad_norm": 0.7296584680468745,
+      "learning_rate": 0.0001983558299946525,
+      "loss": 6.4464,
+      "step": 1869
+    },
+    {
+      "epoch": 0.08614732574745473,
+      "grad_norm": 0.616948027565299,
+      "learning_rate": 0.00019835313432358344,
+      "loss": 5.9995,
+      "step": 1870
+    },
+    {
+      "epoch": 0.08619339383608975,
+      "grad_norm": 0.8190929546212198,
+      "learning_rate": 0.0001983504364628505,
+      "loss": 6.3987,
+      "step": 1871
+    },
+    {
+      "epoch": 0.08623946192472474,
+      "grad_norm": 0.7347032676775593,
+      "learning_rate": 0.00019834773641251374,
+      "loss": 5.8083,
+      "step": 1872
+    },
+    {
+      "epoch": 0.08628553001335974,
+      "grad_norm": 0.8317061137472367,
+      "learning_rate": 0.00019834503417263323,
+      "loss": 6.2528,
+      "step": 1873
+    },
+    {
+      "epoch": 0.08633159810199474,
+      "grad_norm": 0.69038604433008,
+      "learning_rate": 0.00019834232974326915,
+      "loss": 6.1318,
+      "step": 1874
+    },
+    {
+      "epoch": 0.08637766619062975,
+      "grad_norm": 2.2001175030025224,
+      "learning_rate": 0.00019833962312448177,
+      "loss": 5.2761,
+      "step": 1875
+    },
+    {
+      "epoch": 0.08642373427926475,
+      "grad_norm": 16.32219052487346,
+      "learning_rate": 0.00019833691431633124,
+      "loss": 6.1463,
+      "step": 1876
+    },
+    {
+      "epoch": 0.08646980236789975,
+      "grad_norm": 1.8501504175934311,
+      "learning_rate": 0.00019833420331887797,
+      "loss": 6.4624,
+      "step": 1877
+    },
+    {
+      "epoch": 0.08651587045653476,
+      "grad_norm": 2.0834036203350452,
+      "learning_rate": 0.00019833149013218226,
+      "loss": 6.0898,
+      "step": 1878
+    },
+    {
+      "epoch": 0.08656193854516976,
+      "grad_norm": 1.0470220434504334,
+      "learning_rate": 0.00019832877475630452,
+      "loss": 6.6819,
+      "step": 1879
+    },
+    {
+      "epoch": 0.08660800663380476,
+      "grad_norm": 1.5261189256135401,
+      "learning_rate": 0.00019832605719130518,
+      "loss": 6.2845,
+      "step": 1880
+    },
+    {
+      "epoch": 0.08665407472243977,
+      "grad_norm": 0.7663361319910675,
+      "learning_rate": 0.00019832333743724482,
+      "loss": 6.6204,
+      "step": 1881
+    },
+    {
+      "epoch": 0.08670014281107477,
+      "grad_norm": 0.8822660514201166,
+      "learning_rate": 0.0001983206154941839,
+      "loss": 6.2489,
+      "step": 1882
+    },
+    {
+      "epoch": 0.08674621089970977,
+      "grad_norm": 1.4312630337763397,
+      "learning_rate": 0.00019831789136218309,
+      "loss": 6.3253,
+      "step": 1883
+    },
+    {
+      "epoch": 0.08679227898834477,
+      "grad_norm": 0.9780987517412438,
+      "learning_rate": 0.00019831516504130299,
+      "loss": 6.4914,
+      "step": 1884
+    },
+    {
+      "epoch": 0.08683834707697978,
+      "grad_norm": 0.7527813160758163,
+      "learning_rate": 0.00019831243653160432,
+      "loss": 6.4829,
+      "step": 1885
+    },
+    {
+      "epoch": 0.08688441516561478,
+      "grad_norm": 0.7786779597163974,
+      "learning_rate": 0.00019830970583314782,
+      "loss": 6.6816,
+      "step": 1886
+    },
+    {
+      "epoch": 0.08693048325424978,
+      "grad_norm": 1.1735907553434826,
+      "learning_rate": 0.0001983069729459943,
+      "loss": 5.8472,
+      "step": 1887
+    },
+    {
+      "epoch": 0.08697655134288479,
+      "grad_norm": 1.0243259882716733,
+      "learning_rate": 0.00019830423787020455,
+      "loss": 5.7709,
+      "step": 1888
+    },
+    {
+      "epoch": 0.08702261943151979,
+      "grad_norm": 1.08213102391939,
+      "learning_rate": 0.0001983015006058395,
+      "loss": 6.0846,
+      "step": 1889
+    },
+    {
+      "epoch": 0.08706868752015479,
+      "grad_norm": 0.7357580009845719,
+      "learning_rate": 0.00019829876115296012,
+      "loss": 6.1095,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08711475560878978,
+      "grad_norm": 0.9476503010464551,
+      "learning_rate": 0.00019829601951162738,
+      "loss": 5.7493,
+      "step": 1891
+    },
+    {
+      "epoch": 0.0871608236974248,
+      "grad_norm": 0.8515646141745008,
+      "learning_rate": 0.00019829327568190227,
+      "loss": 5.6313,
+      "step": 1892
+    },
+    {
+      "epoch": 0.0872068917860598,
+      "grad_norm": 0.8032855854625014,
+      "learning_rate": 0.00019829052966384593,
+      "loss": 6.255,
+      "step": 1893
+    },
+    {
+      "epoch": 0.0872529598746948,
+      "grad_norm": 1.5617693793851615,
+      "learning_rate": 0.00019828778145751947,
+      "loss": 6.8071,
+      "step": 1894
+    },
+    {
+      "epoch": 0.0872990279633298,
+      "grad_norm": 0.9476447773263463,
+      "learning_rate": 0.0001982850310629841,
+      "loss": 6.1201,
+      "step": 1895
+    },
+    {
+      "epoch": 0.0873450960519648,
+      "grad_norm": 0.9348792417751289,
+      "learning_rate": 0.00019828227848030103,
+      "loss": 6.0872,
+      "step": 1896
+    },
+    {
+      "epoch": 0.0873911641405998,
+      "grad_norm": 1.8686318744795882,
+      "learning_rate": 0.00019827952370953154,
+      "loss": 5.3815,
+      "step": 1897
+    },
+    {
+      "epoch": 0.08743723222923482,
+      "grad_norm": 1.4579257262780079,
+      "learning_rate": 0.00019827676675073698,
+      "loss": 5.9293,
+      "step": 1898
+    },
+    {
+      "epoch": 0.08748330031786981,
+      "grad_norm": 0.7514435817269889,
+      "learning_rate": 0.00019827400760397872,
+      "loss": 6.4499,
+      "step": 1899
+    },
+    {
+      "epoch": 0.08752936840650481,
+      "grad_norm": 1.7948381472410022,
+      "learning_rate": 0.00019827124626931814,
+      "loss": 6.5937,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08757543649513981,
+      "grad_norm": 1.7471346957858027,
+      "learning_rate": 0.0001982684827468168,
+      "loss": 5.9909,
+      "step": 1901
+    },
+    {
+      "epoch": 0.08762150458377482,
+      "grad_norm": 1.2288542322074945,
+      "learning_rate": 0.0001982657170365362,
+      "loss": 5.8816,
+      "step": 1902
+    },
+    {
+      "epoch": 0.08766757267240982,
+      "grad_norm": 1.2758349659369526,
+      "learning_rate": 0.00019826294913853789,
+      "loss": 6.0775,
+      "step": 1903
+    },
+    {
+      "epoch": 0.08771364076104482,
+      "grad_norm": 1.7125176753453883,
+      "learning_rate": 0.00019826017905288348,
+      "loss": 5.828,
+      "step": 1904
+    },
+    {
+      "epoch": 0.08775970884967983,
+      "grad_norm": 1.6084210624575939,
+      "learning_rate": 0.00019825740677963468,
+      "loss": 5.7137,
+      "step": 1905
+    },
+    {
+      "epoch": 0.08780577693831483,
+      "grad_norm": 2.3475662101019834,
+      "learning_rate": 0.0001982546323188532,
+      "loss": 5.7184,
+      "step": 1906
+    },
+    {
+      "epoch": 0.08785184502694983,
+      "grad_norm": 1.4332585826869153,
+      "learning_rate": 0.0001982518556706008,
+      "loss": 5.9962,
+      "step": 1907
+    },
+    {
+      "epoch": 0.08789791311558483,
+      "grad_norm": 3.1923378520625296,
+      "learning_rate": 0.00019824907683493927,
+      "loss": 5.7324,
+      "step": 1908
+    },
+    {
+      "epoch": 0.08794398120421984,
+      "grad_norm": 0.9015521641673158,
+      "learning_rate": 0.00019824629581193052,
+      "loss": 6.4468,
+      "step": 1909
+    },
+    {
+      "epoch": 0.08799004929285484,
+      "grad_norm": 1.7354659139408124,
+      "learning_rate": 0.00019824351260163648,
+      "loss": 5.5804,
+      "step": 1910
+    },
+    {
+      "epoch": 0.08803611738148984,
+      "grad_norm": 2.045157399905315,
+      "learning_rate": 0.00019824072720411902,
+      "loss": 5.2036,
+      "step": 1911
+    },
+    {
+      "epoch": 0.08808218547012485,
+      "grad_norm": 2.3205702874349265,
+      "learning_rate": 0.00019823793961944028,
+      "loss": 5.8459,
+      "step": 1912
+    },
+    {
+      "epoch": 0.08812825355875985,
+      "grad_norm": 1.0327789845803264,
+      "learning_rate": 0.00019823514984766222,
+      "loss": 5.1705,
+      "step": 1913
+    },
+    {
+      "epoch": 0.08817432164739485,
+      "grad_norm": 2.1756508070626177,
+      "learning_rate": 0.000198232357888847,
+      "loss": 5.8262,
+      "step": 1914
+    },
+    {
+      "epoch": 0.08822038973602986,
+      "grad_norm": 2.0284071231967564,
+      "learning_rate": 0.00019822956374305676,
+      "loss": 5.9343,
+      "step": 1915
+    },
+    {
+      "epoch": 0.08826645782466486,
+      "grad_norm": 2.135291647175066,
+      "learning_rate": 0.00019822676741035374,
+      "loss": 5.6818,
+      "step": 1916
+    },
+    {
+      "epoch": 0.08831252591329986,
+      "grad_norm": 1.3363985071252056,
+      "learning_rate": 0.00019822396889080014,
+      "loss": 5.1595,
+      "step": 1917
+    },
+    {
+      "epoch": 0.08835859400193485,
+      "grad_norm": 1.6566683252128942,
+      "learning_rate": 0.00019822116818445827,
+      "loss": 5.7272,
+      "step": 1918
+    },
+    {
+      "epoch": 0.08840466209056987,
+      "grad_norm": 2.0121316664416513,
+      "learning_rate": 0.00019821836529139054,
+      "loss": 5.8384,
+      "step": 1919
+    },
+    {
+      "epoch": 0.08845073017920486,
+      "grad_norm": 0.7696111684712333,
+      "learning_rate": 0.00019821556021165927,
+      "loss": 5.7989,
+      "step": 1920
+    },
+    {
+      "epoch": 0.08849679826783986,
+      "grad_norm": 2.0784294094289937,
+      "learning_rate": 0.00019821275294532698,
+      "loss": 6.077,
+      "step": 1921
+    },
+    {
+      "epoch": 0.08854286635647488,
+      "grad_norm": 1.593867822667572,
+      "learning_rate": 0.00019820994349245617,
+      "loss": 6.0584,
+      "step": 1922
+    },
+    {
+      "epoch": 0.08858893444510987,
+      "grad_norm": 2.281649366392757,
+      "learning_rate": 0.00019820713185310934,
+      "loss": 6.3483,
+      "step": 1923
+    },
+    {
+      "epoch": 0.08863500253374487,
+      "grad_norm": 1.5791638840023499,
+      "learning_rate": 0.00019820431802734914,
+      "loss": 4.9682,
+      "step": 1924
+    },
+    {
+      "epoch": 0.08868107062237987,
+      "grad_norm": 2.4364783715924205,
+      "learning_rate": 0.00019820150201523812,
+      "loss": 5.7638,
+      "step": 1925
+    },
+    {
+      "epoch": 0.08872713871101488,
+      "grad_norm": 1.3978690056948349,
+      "learning_rate": 0.00019819868381683908,
+      "loss": 5.1006,
+      "step": 1926
+    },
+    {
+      "epoch": 0.08877320679964988,
+      "grad_norm": 1.432772888052928,
+      "learning_rate": 0.0001981958634322147,
+      "loss": 6.0067,
+      "step": 1927
+    },
+    {
+      "epoch": 0.08881927488828488,
+      "grad_norm": 1.450345795474353,
+      "learning_rate": 0.0001981930408614278,
+      "loss": 5.8999,
+      "step": 1928
+    },
+    {
+      "epoch": 0.08886534297691989,
+      "grad_norm": 1.1288687137877342,
+      "learning_rate": 0.00019819021610454122,
+      "loss": 6.436,
+      "step": 1929
+    },
+    {
+      "epoch": 0.08891141106555489,
+      "grad_norm": 1.3455353655116848,
+      "learning_rate": 0.00019818738916161784,
+      "loss": 5.8533,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08895747915418989,
+      "grad_norm": 1.7749253840691104,
+      "learning_rate": 0.00019818456003272057,
+      "loss": 5.3026,
+      "step": 1931
+    },
+    {
+      "epoch": 0.0890035472428249,
+      "grad_norm": 0.8789731193451478,
+      "learning_rate": 0.00019818172871791245,
+      "loss": 5.8282,
+      "step": 1932
+    },
+    {
+      "epoch": 0.0890496153314599,
+      "grad_norm": 2.3925970363553577,
+      "learning_rate": 0.00019817889521725644,
+      "loss": 5.1232,
+      "step": 1933
+    },
+    {
+      "epoch": 0.0890956834200949,
+      "grad_norm": 1.0159513462554552,
+      "learning_rate": 0.0001981760595308157,
+      "loss": 5.8974,
+      "step": 1934
+    },
+    {
+      "epoch": 0.0891417515087299,
+      "grad_norm": 1.31066235156928,
+      "learning_rate": 0.00019817322165865335,
+      "loss": 5.0594,
+      "step": 1935
+    },
+    {
+      "epoch": 0.08918781959736491,
+      "grad_norm": 0.7718208207141499,
+      "learning_rate": 0.00019817038160083254,
+      "loss": 5.1894,
+      "step": 1936
+    },
+    {
+      "epoch": 0.08923388768599991,
+      "grad_norm": 1.3442404997512818,
+      "learning_rate": 0.0001981675393574165,
+      "loss": 5.946,
+      "step": 1937
+    },
+    {
+      "epoch": 0.0892799557746349,
+      "grad_norm": 1.0494042745560297,
+      "learning_rate": 0.00019816469492846852,
+      "loss": 5.6509,
+      "step": 1938
+    },
+    {
+      "epoch": 0.08932602386326992,
+      "grad_norm": 1.0776664281007569,
+      "learning_rate": 0.0001981618483140519,
+      "loss": 5.3329,
+      "step": 1939
+    },
+    {
+      "epoch": 0.08937209195190492,
+      "grad_norm": 1.410119782930158,
+      "learning_rate": 0.00019815899951423007,
+      "loss": 5.8153,
+      "step": 1940
+    },
+    {
+      "epoch": 0.08941816004053992,
+      "grad_norm": 1.4869561028308156,
+      "learning_rate": 0.0001981561485290664,
+      "loss": 5.4802,
+      "step": 1941
+    },
+    {
+      "epoch": 0.08946422812917491,
+      "grad_norm": 1.0630418984326906,
+      "learning_rate": 0.00019815329535862442,
+      "loss": 5.4799,
+      "step": 1942
+    },
+    {
+      "epoch": 0.08951029621780993,
+      "grad_norm": 1.580869872941248,
+      "learning_rate": 0.0001981504400029676,
+      "loss": 5.2842,
+      "step": 1943
+    },
+    {
+      "epoch": 0.08955636430644492,
+      "grad_norm": 0.966346662134825,
+      "learning_rate": 0.0001981475824621595,
+      "loss": 4.9688,
+      "step": 1944
+    },
+    {
+      "epoch": 0.08960243239507992,
+      "grad_norm": 1.4223773990403032,
+      "learning_rate": 0.0001981447227362638,
+      "loss": 4.7607,
+      "step": 1945
+    },
+    {
+      "epoch": 0.08964850048371494,
+      "grad_norm": 2.7912252694269655,
+      "learning_rate": 0.0001981418608253441,
+      "loss": 4.8776,
+      "step": 1946
+    },
+    {
+      "epoch": 0.08969456857234993,
+      "grad_norm": 1.1774885869584117,
+      "learning_rate": 0.00019813899672946415,
+      "loss": 5.5289,
+      "step": 1947
+    },
+    {
+      "epoch": 0.08974063666098493,
+      "grad_norm": 2.252207774644817,
+      "learning_rate": 0.00019813613044868773,
+      "loss": 4.4838,
+      "step": 1948
+    },
+    {
+      "epoch": 0.08978670474961994,
+      "grad_norm": 0.685710616432632,
+      "learning_rate": 0.0001981332619830786,
+      "loss": 5.7109,
+      "step": 1949
+    },
+    {
+      "epoch": 0.08983277283825494,
+      "grad_norm": 1.7352733700705374,
+      "learning_rate": 0.0001981303913327007,
+      "loss": 4.6267,
+      "step": 1950
+    },
+    {
+      "epoch": 0.08987884092688994,
+      "grad_norm": 0.7697125159261323,
+      "learning_rate": 0.00019812751849761787,
+      "loss": 5.7336,
+      "step": 1951
+    },
+    {
+      "epoch": 0.08992490901552494,
+      "grad_norm": 0.9412360527199489,
+      "learning_rate": 0.0001981246434778941,
+      "loss": 5.3995,
+      "step": 1952
+    },
+    {
+      "epoch": 0.08997097710415995,
+      "grad_norm": 1.4991942149495168,
+      "learning_rate": 0.00019812176627359337,
+      "loss": 5.1048,
+      "step": 1953
+    },
+    {
+      "epoch": 0.09001704519279495,
+      "grad_norm": 0.9636153687241409,
+      "learning_rate": 0.00019811888688477978,
+      "loss": 5.4302,
+      "step": 1954
+    },
+    {
+      "epoch": 0.09006311328142995,
+      "grad_norm": 1.7602682854679346,
+      "learning_rate": 0.0001981160053115174,
+      "loss": 5.514,
+      "step": 1955
+    },
+    {
+      "epoch": 0.09010918137006496,
+      "grad_norm": 1.0571398516956456,
+      "learning_rate": 0.00019811312155387038,
+      "loss": 5.7113,
+      "step": 1956
+    },
+    {
+      "epoch": 0.09015524945869996,
+      "grad_norm": 1.5697669470102937,
+      "learning_rate": 0.00019811023561190297,
+      "loss": 4.5644,
+      "step": 1957
+    },
+    {
+      "epoch": 0.09020131754733496,
+      "grad_norm": 1.0711805142491395,
+      "learning_rate": 0.00019810734748567936,
+      "loss": 5.5241,
+      "step": 1958
+    },
+    {
+      "epoch": 0.09024738563596997,
+      "grad_norm": 1.5088576014046393,
+      "learning_rate": 0.0001981044571752639,
+      "loss": 5.23,
+      "step": 1959
+    },
+    {
+      "epoch": 0.09029345372460497,
+      "grad_norm": 1.4082272074565416,
+      "learning_rate": 0.0001981015646807209,
+      "loss": 5.7817,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09033952181323997,
+      "grad_norm": 1.1661942695216487,
+      "learning_rate": 0.00019809867000211476,
+      "loss": 5.1255,
+      "step": 1961
+    },
+    {
+      "epoch": 0.09038558990187497,
+      "grad_norm": 1.2012014992502587,
+      "learning_rate": 0.00019809577313950995,
+      "loss": 4.3622,
+      "step": 1962
+    },
+    {
+      "epoch": 0.09043165799050998,
+      "grad_norm": 1.4400183294652367,
+      "learning_rate": 0.00019809287409297094,
+      "loss": 5.8124,
+      "step": 1963
+    },
+    {
+      "epoch": 0.09047772607914498,
+      "grad_norm": 0.8563908093527746,
+      "learning_rate": 0.0001980899728625623,
+      "loss": 5.4294,
+      "step": 1964
+    },
+    {
+      "epoch": 0.09052379416777998,
+      "grad_norm": 1.3058295268149913,
+      "learning_rate": 0.00019808706944834858,
+      "loss": 3.8358,
+      "step": 1965
+    },
+    {
+      "epoch": 0.09056986225641499,
+      "grad_norm": 1.3636991793798343,
+      "learning_rate": 0.00019808416385039444,
+      "loss": 5.5975,
+      "step": 1966
+    },
+    {
+      "epoch": 0.09061593034504999,
+      "grad_norm": 2.0005923634241065,
+      "learning_rate": 0.0001980812560687646,
+      "loss": 4.6522,
+      "step": 1967
+    },
+    {
+      "epoch": 0.09066199843368498,
+      "grad_norm": 0.7669230620113273,
+      "learning_rate": 0.00019807834610352374,
+      "loss": 4.7895,
+      "step": 1968
+    },
+    {
+      "epoch": 0.09070806652231998,
+      "grad_norm": 1.1789902320939165,
+      "learning_rate": 0.00019807543395473668,
+      "loss": 5.7803,
+      "step": 1969
+    },
+    {
+      "epoch": 0.090754134610955,
+      "grad_norm": 0.9979783682805027,
+      "learning_rate": 0.00019807251962246825,
+      "loss": 5.2533,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09080020269959,
+      "grad_norm": 1.3979888032487622,
+      "learning_rate": 0.00019806960310678334,
+      "loss": 4.4439,
+      "step": 1971
+    },
+    {
+      "epoch": 0.09084627078822499,
+      "grad_norm": 0.7296314780548099,
+      "learning_rate": 0.00019806668440774686,
+      "loss": 4.9182,
+      "step": 1972
+    },
+    {
+      "epoch": 0.09089233887686,
+      "grad_norm": 1.1841454629928085,
+      "learning_rate": 0.0001980637635254238,
+      "loss": 5.2582,
+      "step": 1973
+    },
+    {
+      "epoch": 0.090938406965495,
+      "grad_norm": 1.1260613371271555,
+      "learning_rate": 0.00019806084045987917,
+      "loss": 5.3814,
+      "step": 1974
+    },
+    {
+      "epoch": 0.09098447505413,
+      "grad_norm": 1.156159900610722,
+      "learning_rate": 0.00019805791521117807,
+      "loss": 4.6977,
+      "step": 1975
+    },
+    {
+      "epoch": 0.09103054314276501,
+      "grad_norm": 1.9215150310716915,
+      "learning_rate": 0.00019805498777938566,
+      "loss": 5.3067,
+      "step": 1976
+    },
+    {
+      "epoch": 0.09107661123140001,
+      "grad_norm": 1.4824647354003653,
+      "learning_rate": 0.00019805205816456701,
+      "loss": 5.7729,
+      "step": 1977
+    },
+    {
+      "epoch": 0.09112267932003501,
+      "grad_norm": 2.0936127399135107,
+      "learning_rate": 0.00019804912636678748,
+      "loss": 5.0443,
+      "step": 1978
+    },
+    {
+      "epoch": 0.09116874740867001,
+      "grad_norm": 1.6020254752866554,
+      "learning_rate": 0.00019804619238611223,
+      "loss": 5.6246,
+      "step": 1979
+    },
+    {
+      "epoch": 0.09121481549730502,
+      "grad_norm": 1.4380423593378715,
+      "learning_rate": 0.00019804325622260662,
+      "loss": 5.303,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09126088358594002,
+      "grad_norm": 3.4847075434648564,
+      "learning_rate": 0.000198040317876336,
+      "loss": 4.6553,
+      "step": 1981
+    },
+    {
+      "epoch": 0.09130695167457502,
+      "grad_norm": 1.413941986464666,
+      "learning_rate": 0.00019803737734736584,
+      "loss": 4.6316,
+      "step": 1982
+    },
+    {
+      "epoch": 0.09135301976321003,
+      "grad_norm": 3.29186119758617,
+      "learning_rate": 0.00019803443463576158,
+      "loss": 4.6889,
+      "step": 1983
+    },
+    {
+      "epoch": 0.09139908785184503,
+      "grad_norm": 1.3747305314969045,
+      "learning_rate": 0.0001980314897415887,
+      "loss": 5.3896,
+      "step": 1984
+    },
+    {
+      "epoch": 0.09144515594048003,
+      "grad_norm": 2.076727884722302,
+      "learning_rate": 0.00019802854266491278,
+      "loss": 5.4702,
+      "step": 1985
+    },
+    {
+      "epoch": 0.09149122402911503,
+      "grad_norm": 1.5905643318684843,
+      "learning_rate": 0.00019802559340579946,
+      "loss": 4.3543,
+      "step": 1986
+    },
+    {
+      "epoch": 0.09153729211775004,
+      "grad_norm": 1.2421852608017592,
+      "learning_rate": 0.0001980226419643144,
+      "loss": 5.0906,
+      "step": 1987
+    },
+    {
+      "epoch": 0.09158336020638504,
+      "grad_norm": 2.0316783296928262,
+      "learning_rate": 0.00019801968834052324,
+      "loss": 4.5076,
+      "step": 1988
+    },
+    {
+      "epoch": 0.09162942829502004,
+      "grad_norm": 1.252910392977897,
+      "learning_rate": 0.00019801673253449183,
+      "loss": 5.885,
+      "step": 1989
+    },
+    {
+      "epoch": 0.09167549638365505,
+      "grad_norm": 1.5916729743186107,
+      "learning_rate": 0.00019801377454628592,
+      "loss": 5.2934,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09172156447229005,
+      "grad_norm": 1.6036667825050401,
+      "learning_rate": 0.00019801081437597135,
+      "loss": 4.7599,
+      "step": 1991
+    },
+    {
+      "epoch": 0.09176763256092504,
+      "grad_norm": 1.0179299576077825,
+      "learning_rate": 0.00019800785202361406,
+      "loss": 4.3306,
+      "step": 1992
+    },
+    {
+      "epoch": 0.09181370064956006,
+      "grad_norm": 1.482388440104453,
+      "learning_rate": 0.00019800488748928,
+      "loss": 5.0727,
+      "step": 1993
+    },
+    {
+      "epoch": 0.09185976873819506,
+      "grad_norm": 1.1145352112289981,
+      "learning_rate": 0.0001980019207730352,
+      "loss": 5.0356,
+      "step": 1994
+    },
+    {
+      "epoch": 0.09190583682683005,
+      "grad_norm": 1.4203675495213468,
+      "learning_rate": 0.0001979989518749456,
+      "loss": 5.0132,
+      "step": 1995
+    },
+    {
+      "epoch": 0.09195190491546505,
+      "grad_norm": 1.1180614090917016,
+      "learning_rate": 0.00019799598079507744,
+      "loss": 4.818,
+      "step": 1996
+    },
+    {
+      "epoch": 0.09199797300410006,
+      "grad_norm": 1.2076677068033288,
+      "learning_rate": 0.00019799300753349676,
+      "loss": 4.8826,
+      "step": 1997
+    },
+    {
+      "epoch": 0.09204404109273506,
+      "grad_norm": 1.0275487166415698,
+      "learning_rate": 0.0001979900320902698,
+      "loss": 4.5914,
+      "step": 1998
+    },
+    {
+      "epoch": 0.09209010918137006,
+      "grad_norm": 0.9660991752529796,
+      "learning_rate": 0.00019798705446546277,
+      "loss": 4.2893,
+      "step": 1999
+    },
+    {
+      "epoch": 0.09213617727000507,
+      "grad_norm": 1.227768293589412,
+      "learning_rate": 0.00019798407465914202,
+      "loss": 5.1321,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09218224535864007,
+      "grad_norm": 1.0377576602463439,
+      "learning_rate": 0.00019798109267137382,
+      "loss": 4.7893,
+      "step": 2001
+    },
+    {
+      "epoch": 0.09222831344727507,
+      "grad_norm": 1.419184933028019,
+      "learning_rate": 0.00019797810850222461,
+      "loss": 4.6748,
+      "step": 2002
+    },
+    {
+      "epoch": 0.09227438153591007,
+      "grad_norm": 1.3370345360031939,
+      "learning_rate": 0.0001979751221517608,
+      "loss": 4.6488,
+      "step": 2003
+    },
+    {
+      "epoch": 0.09232044962454508,
+      "grad_norm": 2.052993228075356,
+      "learning_rate": 0.0001979721336200489,
+      "loss": 4.05,
+      "step": 2004
+    },
+    {
+      "epoch": 0.09236651771318008,
+      "grad_norm": 1.8360523663542236,
+      "learning_rate": 0.00019796914290715542,
+      "loss": 4.8156,
+      "step": 2005
+    },
+    {
+      "epoch": 0.09241258580181508,
+      "grad_norm": 1.9618562021826396,
+      "learning_rate": 0.00019796615001314695,
+      "loss": 4.9355,
+      "step": 2006
+    },
+    {
+      "epoch": 0.09245865389045009,
+      "grad_norm": 1.2015778857902717,
+      "learning_rate": 0.0001979631549380901,
+      "loss": 4.1943,
+      "step": 2007
+    },
+    {
+      "epoch": 0.09250472197908509,
+      "grad_norm": 1.7163773436568333,
+      "learning_rate": 0.00019796015768205164,
+      "loss": 3.8587,
+      "step": 2008
+    },
+    {
+      "epoch": 0.09255079006772009,
+      "grad_norm": 1.5726690259938652,
+      "learning_rate": 0.00019795715824509822,
+      "loss": 4.8901,
+      "step": 2009
+    },
+    {
+      "epoch": 0.0925968581563551,
+      "grad_norm": 1.6661937811555976,
+      "learning_rate": 0.0001979541566272966,
+      "loss": 4.6018,
+      "step": 2010
+    },
+    {
+      "epoch": 0.0926429262449901,
+      "grad_norm": 1.4824985979231304,
+      "learning_rate": 0.00019795115282871366,
+      "loss": 5.003,
+      "step": 2011
+    },
+    {
+      "epoch": 0.0926889943336251,
+      "grad_norm": 2.2941250450660897,
+      "learning_rate": 0.00019794814684941626,
+      "loss": 4.0078,
+      "step": 2012
+    },
+    {
+      "epoch": 0.0927350624222601,
+      "grad_norm": 1.1785118879216232,
+      "learning_rate": 0.0001979451386894713,
+      "loss": 3.9094,
+      "step": 2013
+    },
+    {
+      "epoch": 0.09278113051089511,
+      "grad_norm": 1.696156395741628,
+      "learning_rate": 0.00019794212834894575,
+      "loss": 4.4677,
+      "step": 2014
+    },
+    {
+      "epoch": 0.0928271985995301,
+      "grad_norm": 2.8391165859893777,
+      "learning_rate": 0.00019793911582790668,
+      "loss": 5.4221,
+      "step": 2015
+    },
+    {
+      "epoch": 0.0928732666881651,
+      "grad_norm": 1.1901730100903958,
+      "learning_rate": 0.0001979361011264211,
+      "loss": 4.6422,
+      "step": 2016
+    },
+    {
+      "epoch": 0.09291933477680012,
+      "grad_norm": 2.547431196420195,
+      "learning_rate": 0.0001979330842445562,
+      "loss": 4.4951,
+      "step": 2017
+    },
+    {
+      "epoch": 0.09296540286543511,
+      "grad_norm": 1.692378309998988,
+      "learning_rate": 0.00019793006518237905,
+      "loss": 5.0752,
+      "step": 2018
+    },
+    {
+      "epoch": 0.09301147095407011,
+      "grad_norm": 2.1916559315257897,
+      "learning_rate": 0.0001979270439399569,
+      "loss": 4.3505,
+      "step": 2019
+    },
+    {
+      "epoch": 0.09305753904270511,
+      "grad_norm": 1.423071298466172,
+      "learning_rate": 0.00019792402051735706,
+      "loss": 5.0171,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09310360713134012,
+      "grad_norm": 2.1477864120532715,
+      "learning_rate": 0.00019792099491464682,
+      "loss": 5.5807,
+      "step": 2021
+    },
+    {
+      "epoch": 0.09314967521997512,
+      "grad_norm": 1.5171987817664958,
+      "learning_rate": 0.0001979179671318935,
+      "loss": 4.7528,
+      "step": 2022
+    },
+    {
+      "epoch": 0.09319574330861012,
+      "grad_norm": 1.9719758922550463,
+      "learning_rate": 0.00019791493716916453,
+      "loss": 6.1501,
+      "step": 2023
+    },
+    {
+      "epoch": 0.09324181139724513,
+      "grad_norm": 1.6448808626389395,
+      "learning_rate": 0.00019791190502652738,
+      "loss": 5.1915,
+      "step": 2024
+    },
+    {
+      "epoch": 0.09328787948588013,
+      "grad_norm": 1.5647007003193731,
+      "learning_rate": 0.00019790887070404955,
+      "loss": 4.1555,
+      "step": 2025
+    },
+    {
+      "epoch": 0.09333394757451513,
+      "grad_norm": 1.5081418853366662,
+      "learning_rate": 0.0001979058342017986,
+      "loss": 4.1392,
+      "step": 2026
+    },
+    {
+      "epoch": 0.09338001566315014,
+      "grad_norm": 1.3124060383230378,
+      "learning_rate": 0.00019790279551984213,
+      "loss": 4.1018,
+      "step": 2027
+    },
+    {
+      "epoch": 0.09342608375178514,
+      "grad_norm": 1.8954522774402287,
+      "learning_rate": 0.00019789975465824778,
+      "loss": 4.5538,
+      "step": 2028
+    },
+    {
+      "epoch": 0.09347215184042014,
+      "grad_norm": 2.3591511222313644,
+      "learning_rate": 0.00019789671161708322,
+      "loss": 4.1709,
+      "step": 2029
+    },
+    {
+      "epoch": 0.09351821992905514,
+      "grad_norm": 1.954564657991809,
+      "learning_rate": 0.00019789366639641625,
+      "loss": 3.9186,
+      "step": 2030
+    },
+    {
+      "epoch": 0.09356428801769015,
+      "grad_norm": 1.4920935753794455,
+      "learning_rate": 0.00019789061899631467,
+      "loss": 5.1897,
+      "step": 2031
+    },
+    {
+      "epoch": 0.09361035610632515,
+      "grad_norm": 1.6173788830375928,
+      "learning_rate": 0.00019788756941684627,
+      "loss": 4.7385,
+      "step": 2032
+    },
+    {
+      "epoch": 0.09365642419496015,
+      "grad_norm": 1.6956920758120178,
+      "learning_rate": 0.00019788451765807903,
+      "loss": 5.753,
+      "step": 2033
+    },
+    {
+      "epoch": 0.09370249228359516,
+      "grad_norm": 2.582864054020192,
+      "learning_rate": 0.00019788146372008082,
+      "loss": 4.1878,
+      "step": 2034
+    },
+    {
+      "epoch": 0.09374856037223016,
+      "grad_norm": 1.586967649050745,
+      "learning_rate": 0.00019787840760291962,
+      "loss": 4.6443,
+      "step": 2035
+    },
+    {
+      "epoch": 0.09379462846086516,
+      "grad_norm": 1.318201854389915,
+      "learning_rate": 0.00019787534930666354,
+      "loss": 4.4014,
+      "step": 2036
+    },
+    {
+      "epoch": 0.09384069654950015,
+      "grad_norm": 1.8623296536815985,
+      "learning_rate": 0.0001978722888313806,
+      "loss": 5.422,
+      "step": 2037
+    },
+    {
+      "epoch": 0.09388676463813517,
+      "grad_norm": 1.810809827302522,
+      "learning_rate": 0.00019786922617713896,
+      "loss": 4.7736,
+      "step": 2038
+    },
+    {
+      "epoch": 0.09393283272677017,
+      "grad_norm": 1.641530796226961,
+      "learning_rate": 0.00019786616134400683,
+      "loss": 4.3588,
+      "step": 2039
+    },
+    {
+      "epoch": 0.09397890081540516,
+      "grad_norm": 2.350791308874932,
+      "learning_rate": 0.00019786309433205242,
+      "loss": 4.8436,
+      "step": 2040
+    },
+    {
+      "epoch": 0.09402496890404018,
+      "grad_norm": 2.1652573059071027,
+      "learning_rate": 0.000197860025141344,
+      "loss": 4.1978,
+      "step": 2041
+    },
+    {
+      "epoch": 0.09407103699267517,
+      "grad_norm": 2.533186640758103,
+      "learning_rate": 0.00019785695377194993,
+      "loss": 4.7805,
+      "step": 2042
+    },
+    {
+      "epoch": 0.09411710508131017,
+      "grad_norm": 2.214396882087193,
+      "learning_rate": 0.00019785388022393856,
+      "loss": 4.3475,
+      "step": 2043
+    },
+    {
+      "epoch": 0.09416317316994519,
+      "grad_norm": 1.7107493781629843,
+      "learning_rate": 0.00019785080449737834,
+      "loss": 4.5843,
+      "step": 2044
+    },
+    {
+      "epoch": 0.09420924125858018,
+      "grad_norm": 2.829769765970948,
+      "learning_rate": 0.00019784772659233775,
+      "loss": 5.0036,
+      "step": 2045
+    },
+    {
+      "epoch": 0.09425530934721518,
+      "grad_norm": 4.4794403498859,
+      "learning_rate": 0.00019784464650888527,
+      "loss": 4.9621,
+      "step": 2046
+    },
+    {
+      "epoch": 0.09430137743585018,
+      "grad_norm": 1.8718262041556757,
+      "learning_rate": 0.00019784156424708951,
+      "loss": 3.6855,
+      "step": 2047
+    },
+    {
+      "epoch": 0.0943474455244852,
+      "grad_norm": 4.063337046875937,
+      "learning_rate": 0.00019783847980701908,
+      "loss": 4.4114,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09439351361312019,
+      "grad_norm": 2.8558357532190892,
+      "learning_rate": 0.00019783539318874268,
+      "loss": 4.5984,
+      "step": 2049
+    },
+    {
+      "epoch": 0.09443958170175519,
+      "grad_norm": 2.7493814257810154,
+      "learning_rate": 0.00019783230439232897,
+      "loss": 4.8351,
+      "step": 2050
+    },
+    {
+      "epoch": 0.0944856497903902,
+      "grad_norm": 1.958738700585108,
+      "learning_rate": 0.00019782921341784677,
+      "loss": 4.7107,
+      "step": 2051
+    },
+    {
+      "epoch": 0.0945317178790252,
+      "grad_norm": 2.9976888773291996,
+      "learning_rate": 0.0001978261202653649,
+      "loss": 4.1398,
+      "step": 2052
+    },
+    {
+      "epoch": 0.0945777859676602,
+      "grad_norm": 2.5248327377966957,
+      "learning_rate": 0.00019782302493495215,
+      "loss": 3.6946,
+      "step": 2053
+    },
+    {
+      "epoch": 0.0946238540562952,
+      "grad_norm": 2.264052831759907,
+      "learning_rate": 0.00019781992742667749,
+      "loss": 3.8058,
+      "step": 2054
+    },
+    {
+      "epoch": 0.09466992214493021,
+      "grad_norm": 3.638545851674497,
+      "learning_rate": 0.00019781682774060988,
+      "loss": 4.2309,
+      "step": 2055
+    },
+    {
+      "epoch": 0.09471599023356521,
+      "grad_norm": 4.431341460097037,
+      "learning_rate": 0.0001978137258768183,
+      "loss": 3.9783,
+      "step": 2056
+    },
+    {
+      "epoch": 0.09476205832220021,
+      "grad_norm": 3.657405976746398,
+      "learning_rate": 0.00019781062183537186,
+      "loss": 3.7891,
+      "step": 2057
+    },
+    {
+      "epoch": 0.09480812641083522,
+      "grad_norm": 5.9438335894718195,
+      "learning_rate": 0.00019780751561633962,
+      "loss": 3.596,
+      "step": 2058
+    },
+    {
+      "epoch": 0.09485419449947022,
+      "grad_norm": 4.357679372977613,
+      "learning_rate": 0.00019780440721979076,
+      "loss": 3.4438,
+      "step": 2059
+    },
+    {
+      "epoch": 0.09490026258810522,
+      "grad_norm": 6.6732689592787215,
+      "learning_rate": 0.00019780129664579446,
+      "loss": 2.8027,
+      "step": 2060
+    },
+    {
+      "epoch": 0.09494633067674023,
+      "grad_norm": 5.311310968162958,
+      "learning_rate": 0.00019779818389442,
+      "loss": 3.0827,
+      "step": 2061
+    },
+    {
+      "epoch": 0.09499239876537523,
+      "grad_norm": 8.683602609514473,
+      "learning_rate": 0.00019779506896573662,
+      "loss": 3.1754,
+      "step": 2062
+    },
+    {
+      "epoch": 0.09503846685401023,
+      "grad_norm": 3.615209703789211,
+      "learning_rate": 0.00019779195185981377,
+      "loss": 2.58,
+      "step": 2063
+    },
+    {
+      "epoch": 0.09508453494264522,
+      "grad_norm": 2.933959667266844,
+      "learning_rate": 0.00019778883257672077,
+      "loss": 2.4243,
+      "step": 2064
+    },
+    {
+      "epoch": 0.09513060303128024,
+      "grad_norm": 3.2766818395522646,
+      "learning_rate": 0.00019778571111652706,
+      "loss": 2.3446,
+      "step": 2065
+    },
+    {
+      "epoch": 0.09517667111991523,
+      "grad_norm": 2.885994675585596,
+      "learning_rate": 0.00019778258747930218,
+      "loss": 2.3008,
+      "step": 2066
+    },
+    {
+      "epoch": 0.09522273920855023,
+      "grad_norm": 4.953966591073837,
+      "learning_rate": 0.00019777946166511562,
+      "loss": 2.3926,
+      "step": 2067
+    },
+    {
+      "epoch": 0.09526880729718525,
+      "grad_norm": 4.04564070467968,
+      "learning_rate": 0.00019777633367403703,
+      "loss": 2.2787,
+      "step": 2068
+    },
+    {
+      "epoch": 0.09531487538582024,
+      "grad_norm": 8.54456890061801,
+      "learning_rate": 0.000197773203506136,
+      "loss": 2.0523,
+      "step": 2069
+    },
+    {
+      "epoch": 0.09536094347445524,
+      "grad_norm": 8.110177515454941,
+      "learning_rate": 0.00019777007116148228,
+      "loss": 2.1725,
+      "step": 2070
+    },
+    {
+      "epoch": 0.09540701156309024,
+      "grad_norm": 3.6328876685166196,
+      "learning_rate": 0.00019776693664014556,
+      "loss": 1.9997,
+      "step": 2071
+    },
+    {
+      "epoch": 0.09545307965172525,
+      "grad_norm": 4.764035746159834,
+      "learning_rate": 0.0001977637999421956,
+      "loss": 2.4081,
+      "step": 2072
+    },
+    {
+      "epoch": 0.09549914774036025,
+      "grad_norm": 2.4454962578475765,
+      "learning_rate": 0.00019776066106770226,
+      "loss": 2.0776,
+      "step": 2073
+    },
+    {
+      "epoch": 0.09554521582899525,
+      "grad_norm": 1.7961622831320796,
+      "learning_rate": 0.00019775752001673547,
+      "loss": 1.9115,
+      "step": 2074
+    },
+    {
+      "epoch": 0.09559128391763026,
+      "grad_norm": 2.6063132965871945,
+      "learning_rate": 0.00019775437678936508,
+      "loss": 1.8626,
+      "step": 2075
+    },
+    {
+      "epoch": 0.09563735200626526,
+      "grad_norm": 4.591723232219041,
+      "learning_rate": 0.0001977512313856611,
+      "loss": 1.5734,
+      "step": 2076
+    },
+    {
+      "epoch": 0.09568342009490026,
+      "grad_norm": 1.5597068023656693,
+      "learning_rate": 0.00019774808380569356,
+      "loss": 1.9449,
+      "step": 2077
+    },
+    {
+      "epoch": 0.09572948818353527,
+      "grad_norm": 1.704446234518652,
+      "learning_rate": 0.00019774493404953254,
+      "loss": 1.3979,
+      "step": 2078
+    },
+    {
+      "epoch": 0.09577555627217027,
+      "grad_norm": 4.531537616332274,
+      "learning_rate": 0.00019774178211724817,
+      "loss": 1.9157,
+      "step": 2079
+    },
+    {
+      "epoch": 0.09582162436080527,
+      "grad_norm": 1.8437468847235419,
+      "learning_rate": 0.00019773862800891062,
+      "loss": 1.6144,
+      "step": 2080
+    },
+    {
+      "epoch": 0.09586769244944027,
+      "grad_norm": 1.3723833661661313,
+      "learning_rate": 0.00019773547172459007,
+      "loss": 1.732,
+      "step": 2081
+    },
+    {
+      "epoch": 0.09591376053807528,
+      "grad_norm": 1.533276606060053,
+      "learning_rate": 0.00019773231326435686,
+      "loss": 1.7878,
+      "step": 2082
+    },
+    {
+      "epoch": 0.09595982862671028,
+      "grad_norm": 2.1347168957116467,
+      "learning_rate": 0.00019772915262828127,
+      "loss": 1.6764,
+      "step": 2083
+    },
+    {
+      "epoch": 0.09600589671534528,
+      "grad_norm": 7.923747733919266,
+      "learning_rate": 0.00019772598981643365,
+      "loss": 1.7687,
+      "step": 2084
+    },
+    {
+      "epoch": 0.09605196480398029,
+      "grad_norm": 2.6808607414839085,
+      "learning_rate": 0.0001977228248288844,
+      "loss": 1.9953,
+      "step": 2085
+    },
+    {
+      "epoch": 0.09609803289261529,
+      "grad_norm": 4.4851192751758235,
+      "learning_rate": 0.00019771965766570405,
+      "loss": 1.4062,
+      "step": 2086
+    },
+    {
+      "epoch": 0.09614410098125029,
+      "grad_norm": 1.3669933089136488,
+      "learning_rate": 0.00019771648832696308,
+      "loss": 1.4345,
+      "step": 2087
+    },
+    {
+      "epoch": 0.09619016906988528,
+      "grad_norm": 4.016595570228585,
+      "learning_rate": 0.00019771331681273205,
+      "loss": 1.6796,
+      "step": 2088
+    },
+    {
+      "epoch": 0.0962362371585203,
+      "grad_norm": 1.991584917643696,
+      "learning_rate": 0.00019771014312308153,
+      "loss": 1.6188,
+      "step": 2089
+    },
+    {
+      "epoch": 0.0962823052471553,
+      "grad_norm": 1.3414305287987873,
+      "learning_rate": 0.00019770696725808224,
+      "loss": 1.8262,
+      "step": 2090
+    },
+    {
+      "epoch": 0.0963283733357903,
+      "grad_norm": 2.5251303109380285,
+      "learning_rate": 0.00019770378921780482,
+      "loss": 1.8034,
+      "step": 2091
+    },
+    {
+      "epoch": 0.0963744414244253,
+      "grad_norm": 2.448091825067566,
+      "learning_rate": 0.0001977006090023201,
+      "loss": 1.4525,
+      "step": 2092
+    },
+    {
+      "epoch": 0.0964205095130603,
+      "grad_norm": 2.0920268014433336,
+      "learning_rate": 0.00019769742661169884,
+      "loss": 1.8853,
+      "step": 2093
+    },
+    {
+      "epoch": 0.0964665776016953,
+      "grad_norm": 2.958891914785194,
+      "learning_rate": 0.00019769424204601186,
+      "loss": 1.4844,
+      "step": 2094
+    },
+    {
+      "epoch": 0.09651264569033031,
+      "grad_norm": 2.854718971613699,
+      "learning_rate": 0.00019769105530533011,
+      "loss": 1.4276,
+      "step": 2095
+    },
+    {
+      "epoch": 0.09655871377896531,
+      "grad_norm": 2.164435521861292,
+      "learning_rate": 0.00019768786638972452,
+      "loss": 1.4699,
+      "step": 2096
+    },
+    {
+      "epoch": 0.09660478186760031,
+      "grad_norm": 1.1850927917138834,
+      "learning_rate": 0.00019768467529926608,
+      "loss": 1.4839,
+      "step": 2097
+    },
+    {
+      "epoch": 0.09665084995623531,
+      "grad_norm": 5.524284395610811,
+      "learning_rate": 0.00019768148203402582,
+      "loss": 1.3305,
+      "step": 2098
+    },
+    {
+      "epoch": 0.09669691804487032,
+      "grad_norm": 3.162677265325548,
+      "learning_rate": 0.00019767828659407487,
+      "loss": 1.3107,
+      "step": 2099
+    },
+    {
+      "epoch": 0.09674298613350532,
+      "grad_norm": 1.7949188851730715,
+      "learning_rate": 0.00019767508897948436,
+      "loss": 1.5136,
+      "step": 2100
+    },
+    {
+      "epoch": 0.09678905422214032,
+      "grad_norm": 15.698053865790811,
+      "learning_rate": 0.00019767188919032544,
+      "loss": 1.6062,
+      "step": 2101
+    },
+    {
+      "epoch": 0.09683512231077533,
+      "grad_norm": 2.5242884459327035,
+      "learning_rate": 0.0001976686872266694,
+      "loss": 1.302,
+      "step": 2102
+    },
+    {
+      "epoch": 0.09688119039941033,
+      "grad_norm": 2.457616483052631,
+      "learning_rate": 0.00019766548308858752,
+      "loss": 1.3351,
+      "step": 2103
+    },
+    {
+      "epoch": 0.09692725848804533,
+      "grad_norm": 8.293705644102035,
+      "learning_rate": 0.0001976622767761511,
+      "loss": 1.4356,
+      "step": 2104
+    },
+    {
+      "epoch": 0.09697332657668033,
+      "grad_norm": 14.902352066386335,
+      "learning_rate": 0.00019765906828943155,
+      "loss": 1.7188,
+      "step": 2105
+    },
+    {
+      "epoch": 0.09701939466531534,
+      "grad_norm": 2.0281577755579363,
+      "learning_rate": 0.00019765585762850028,
+      "loss": 1.7241,
+      "step": 2106
+    },
+    {
+      "epoch": 0.09706546275395034,
+      "grad_norm": 2.4760748783620734,
+      "learning_rate": 0.00019765264479342878,
+      "loss": 1.5203,
+      "step": 2107
+    },
+    {
+      "epoch": 0.09711153084258534,
+      "grad_norm": 2.579193000832151,
+      "learning_rate": 0.0001976494297842886,
+      "loss": 1.5808,
+      "step": 2108
+    },
+    {
+      "epoch": 0.09715759893122035,
+      "grad_norm": 1.623866832298656,
+      "learning_rate": 0.0001976462126011513,
+      "loss": 1.4288,
+      "step": 2109
+    },
+    {
+      "epoch": 0.09720366701985535,
+      "grad_norm": 4.0160872421300455,
+      "learning_rate": 0.0001976429932440885,
+      "loss": 1.3781,
+      "step": 2110
+    },
+    {
+      "epoch": 0.09724973510849035,
+      "grad_norm": 5.91087129726305,
+      "learning_rate": 0.00019763977171317188,
+      "loss": 1.7389,
+      "step": 2111
+    },
+    {
+      "epoch": 0.09729580319712536,
+      "grad_norm": 2.2435274750773213,
+      "learning_rate": 0.00019763654800847316,
+      "loss": 1.5642,
+      "step": 2112
+    },
+    {
+      "epoch": 0.09734187128576036,
+      "grad_norm": 3.092103610886108,
+      "learning_rate": 0.0001976333221300641,
+      "loss": 1.4919,
+      "step": 2113
+    },
+    {
+      "epoch": 0.09738793937439535,
+      "grad_norm": 0.9796798404320602,
+      "learning_rate": 0.00019763009407801653,
+      "loss": 1.5379,
+      "step": 2114
+    },
+    {
+      "epoch": 0.09743400746303035,
+      "grad_norm": 1.6324618862423408,
+      "learning_rate": 0.0001976268638524023,
+      "loss": 1.5152,
+      "step": 2115
+    },
+    {
+      "epoch": 0.09748007555166537,
+      "grad_norm": 1.6733811963986245,
+      "learning_rate": 0.00019762363145329337,
+      "loss": 1.6291,
+      "step": 2116
+    },
+    {
+      "epoch": 0.09752614364030036,
+      "grad_norm": 2.14703810253368,
+      "learning_rate": 0.00019762039688076165,
+      "loss": 1.5088,
+      "step": 2117
+    },
+    {
+      "epoch": 0.09757221172893536,
+      "grad_norm": 2.221860874200408,
+      "learning_rate": 0.00019761716013487918,
+      "loss": 1.4936,
+      "step": 2118
+    },
+    {
+      "epoch": 0.09761827981757037,
+      "grad_norm": 1.876891885440355,
+      "learning_rate": 0.00019761392121571801,
+      "loss": 1.3668,
+      "step": 2119
+    },
+    {
+      "epoch": 0.09766434790620537,
+      "grad_norm": 2.0738829710626234,
+      "learning_rate": 0.0001976106801233503,
+      "loss": 1.4838,
+      "step": 2120
+    },
+    {
+      "epoch": 0.09771041599484037,
+      "grad_norm": 1.5409953894532844,
+      "learning_rate": 0.00019760743685784813,
+      "loss": 1.4597,
+      "step": 2121
+    },
+    {
+      "epoch": 0.09775648408347537,
+      "grad_norm": 1.8400067016063397,
+      "learning_rate": 0.00019760419141928376,
+      "loss": 1.5054,
+      "step": 2122
+    },
+    {
+      "epoch": 0.09780255217211038,
+      "grad_norm": 1.6767567449523015,
+      "learning_rate": 0.00019760094380772938,
+      "loss": 1.3819,
+      "step": 2123
+    },
+    {
+      "epoch": 0.09784862026074538,
+      "grad_norm": 10.340894798316757,
+      "learning_rate": 0.00019759769402325736,
+      "loss": 1.2178,
+      "step": 2124
+    },
+    {
+      "epoch": 0.09789468834938038,
+      "grad_norm": 1.2838186071371624,
+      "learning_rate": 0.00019759444206594005,
+      "loss": 1.6314,
+      "step": 2125
+    },
+    {
+      "epoch": 0.09794075643801539,
+      "grad_norm": 1.7351510027729113,
+      "learning_rate": 0.0001975911879358498,
+      "loss": 1.66,
+      "step": 2126
+    },
+    {
+      "epoch": 0.09798682452665039,
+      "grad_norm": 2.4920600655538934,
+      "learning_rate": 0.00019758793163305907,
+      "loss": 1.6712,
+      "step": 2127
+    },
+    {
+      "epoch": 0.09803289261528539,
+      "grad_norm": 4.133659357900631,
+      "learning_rate": 0.00019758467315764039,
+      "loss": 1.5924,
+      "step": 2128
+    },
+    {
+      "epoch": 0.0980789607039204,
+      "grad_norm": 2.501695034074429,
+      "learning_rate": 0.0001975814125096663,
+      "loss": 1.659,
+      "step": 2129
+    },
+    {
+      "epoch": 0.0981250287925554,
+      "grad_norm": 1.4180762788031545,
+      "learning_rate": 0.00019757814968920934,
+      "loss": 1.7224,
+      "step": 2130
+    },
+    {
+      "epoch": 0.0981710968811904,
+      "grad_norm": 1.3496899800781124,
+      "learning_rate": 0.0001975748846963422,
+      "loss": 1.3704,
+      "step": 2131
+    },
+    {
+      "epoch": 0.0982171649698254,
+      "grad_norm": 1.1916468963056543,
+      "learning_rate": 0.00019757161753113756,
+      "loss": 1.4521,
+      "step": 2132
+    },
+    {
+      "epoch": 0.09826323305846041,
+      "grad_norm": 1.536209325264828,
+      "learning_rate": 0.0001975683481936682,
+      "loss": 1.5357,
+      "step": 2133
+    },
+    {
+      "epoch": 0.09830930114709541,
+      "grad_norm": 1.6923947088186806,
+      "learning_rate": 0.0001975650766840068,
+      "loss": 1.3074,
+      "step": 2134
+    },
+    {
+      "epoch": 0.0983553692357304,
+      "grad_norm": 1.3303687468217944,
+      "learning_rate": 0.00019756180300222625,
+      "loss": 1.4172,
+      "step": 2135
+    },
+    {
+      "epoch": 0.09840143732436542,
+      "grad_norm": 1.1985289361527127,
+      "learning_rate": 0.0001975585271483995,
+      "loss": 1.1458,
+      "step": 2136
+    },
+    {
+      "epoch": 0.09844750541300042,
+      "grad_norm": 2.3567234009187255,
+      "learning_rate": 0.00019755524912259937,
+      "loss": 1.2865,
+      "step": 2137
+    },
+    {
+      "epoch": 0.09849357350163541,
+      "grad_norm": 1.9539173049380847,
+      "learning_rate": 0.00019755196892489888,
+      "loss": 1.1154,
+      "step": 2138
+    },
+    {
+      "epoch": 0.09853964159027041,
+      "grad_norm": 1.7856992006575025,
+      "learning_rate": 0.00019754868655537112,
+      "loss": 1.2431,
+      "step": 2139
+    },
+    {
+      "epoch": 0.09858570967890543,
+      "grad_norm": 0.8620279105991182,
+      "learning_rate": 0.0001975454020140891,
+      "loss": 1.4614,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09863177776754042,
+      "grad_norm": 1.5365780509322462,
+      "learning_rate": 0.00019754211530112592,
+      "loss": 1.4011,
+      "step": 2141
+    },
+    {
+      "epoch": 0.09867784585617542,
+      "grad_norm": 1.1824809522142967,
+      "learning_rate": 0.00019753882641655482,
+      "loss": 1.4344,
+      "step": 2142
+    },
+    {
+      "epoch": 0.09872391394481043,
+      "grad_norm": 1.2990205095318732,
+      "learning_rate": 0.000197535535360449,
+      "loss": 1.4756,
+      "step": 2143
+    },
+    {
+      "epoch": 0.09876998203344543,
+      "grad_norm": 1.9974948521391211,
+      "learning_rate": 0.00019753224213288172,
+      "loss": 1.2083,
+      "step": 2144
+    },
+    {
+      "epoch": 0.09881605012208043,
+      "grad_norm": 1.0551666235920378,
+      "learning_rate": 0.0001975289467339263,
+      "loss": 1.5423,
+      "step": 2145
+    },
+    {
+      "epoch": 0.09886211821071544,
+      "grad_norm": 0.9871074557442939,
+      "learning_rate": 0.0001975256491636561,
+      "loss": 1.5071,
+      "step": 2146
+    },
+    {
+      "epoch": 0.09890818629935044,
+      "grad_norm": 2.167457816576732,
+      "learning_rate": 0.00019752234942214454,
+      "loss": 1.3534,
+      "step": 2147
+    },
+    {
+      "epoch": 0.09895425438798544,
+      "grad_norm": 1.211420103292922,
+      "learning_rate": 0.0001975190475094651,
+      "loss": 1.4757,
+      "step": 2148
+    },
+    {
+      "epoch": 0.09900032247662044,
+      "grad_norm": 1.0146206750339017,
+      "learning_rate": 0.00019751574342569123,
+      "loss": 1.4334,
+      "step": 2149
+    },
+    {
+      "epoch": 0.09904639056525545,
+      "grad_norm": 1.2485352143518236,
+      "learning_rate": 0.00019751243717089655,
+      "loss": 1.3925,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09909245865389045,
+      "grad_norm": 2.502288635576063,
+      "learning_rate": 0.0001975091287451547,
+      "loss": 1.3715,
+      "step": 2151
+    },
+    {
+      "epoch": 0.09913852674252545,
+      "grad_norm": 2.5808474123656797,
+      "learning_rate": 0.0001975058181485393,
+      "loss": 1.5914,
+      "step": 2152
+    },
+    {
+      "epoch": 0.09918459483116046,
+      "grad_norm": 1.08558698738046,
+      "learning_rate": 0.00019750250538112395,
+      "loss": 1.37,
+      "step": 2153
+    },
+    {
+      "epoch": 0.09923066291979546,
+      "grad_norm": 1.3314803047503605,
+      "learning_rate": 0.0001974991904429826,
+      "loss": 1.38,
+      "step": 2154
+    },
+    {
+      "epoch": 0.09927673100843046,
+      "grad_norm": 1.8242745451690612,
+      "learning_rate": 0.0001974958733341889,
+      "loss": 1.325,
+      "step": 2155
+    },
+    {
+      "epoch": 0.09932279909706546,
+      "grad_norm": 2.1173497961426735,
+      "learning_rate": 0.00019749255405481674,
+      "loss": 1.5461,
+      "step": 2156
+    },
+    {
+      "epoch": 0.09936886718570047,
+      "grad_norm": 1.24107212874388,
+      "learning_rate": 0.00019748923260494005,
+      "loss": 1.4357,
+      "step": 2157
+    },
+    {
+      "epoch": 0.09941493527433547,
+      "grad_norm": 1.0510317150775828,
+      "learning_rate": 0.00019748590898463277,
+      "loss": 1.76,
+      "step": 2158
+    },
+    {
+      "epoch": 0.09946100336297047,
+      "grad_norm": 1.1170817833047473,
+      "learning_rate": 0.00019748258319396883,
+      "loss": 1.5986,
+      "step": 2159
+    },
+    {
+      "epoch": 0.09950707145160548,
+      "grad_norm": 1.7428441043783693,
+      "learning_rate": 0.00019747925523302238,
+      "loss": 1.2358,
+      "step": 2160
+    },
+    {
+      "epoch": 0.09955313954024048,
+      "grad_norm": 1.2284766081750107,
+      "learning_rate": 0.00019747592510186741,
+      "loss": 1.1286,
+      "step": 2161
+    },
+    {
+      "epoch": 0.09959920762887547,
+      "grad_norm": 1.875849253426686,
+      "learning_rate": 0.00019747259280057813,
+      "loss": 1.1732,
+      "step": 2162
+    },
+    {
+      "epoch": 0.09964527571751049,
+      "grad_norm": 1.2049539759178678,
+      "learning_rate": 0.0001974692583292287,
+      "loss": 1.2817,
+      "step": 2163
+    },
+    {
+      "epoch": 0.09969134380614549,
+      "grad_norm": 1.1707906748179304,
+      "learning_rate": 0.00019746592168789335,
+      "loss": 1.4977,
+      "step": 2164
+    },
+    {
+      "epoch": 0.09973741189478048,
+      "grad_norm": 1.3918304484310067,
+      "learning_rate": 0.00019746258287664636,
+      "loss": 1.0984,
+      "step": 2165
+    },
+    {
+      "epoch": 0.09978347998341548,
+      "grad_norm": 1.610643262536557,
+      "learning_rate": 0.00019745924189556207,
+      "loss": 0.9286,
+      "step": 2166
+    },
+    {
+      "epoch": 0.0998295480720505,
+      "grad_norm": 1.1111235220772728,
+      "learning_rate": 0.00019745589874471488,
+      "loss": 1.4199,
+      "step": 2167
+    },
+    {
+      "epoch": 0.0998756161606855,
+      "grad_norm": 1.2278386193204054,
+      "learning_rate": 0.0001974525534241792,
+      "loss": 1.5359,
+      "step": 2168
+    },
+    {
+      "epoch": 0.09992168424932049,
+      "grad_norm": 1.2894407918397517,
+      "learning_rate": 0.00019744920593402953,
+      "loss": 1.4132,
+      "step": 2169
+    },
+    {
+      "epoch": 0.0999677523379555,
+      "grad_norm": 1.1891176871771978,
+      "learning_rate": 0.0001974458562743404,
+      "loss": 1.6172,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1000138204265905,
+      "grad_norm": 1.7899923949994543,
+      "learning_rate": 0.00019744250444518632,
+      "loss": 1.3407,
+      "step": 2171
+    },
+    {
+      "epoch": 0.1000598885152255,
+      "grad_norm": 1.0606463172177183,
+      "learning_rate": 0.00019743915044664195,
+      "loss": 1.5668,
+      "step": 2172
+    },
+    {
+      "epoch": 0.1001059566038605,
+      "grad_norm": 2.1079176631435432,
+      "learning_rate": 0.00019743579427878197,
+      "loss": 1.4159,
+      "step": 2173
+    },
+    {
+      "epoch": 0.10015202469249551,
+      "grad_norm": 0.9762544402469904,
+      "learning_rate": 0.00019743243594168113,
+      "loss": 1.369,
+      "step": 2174
+    },
+    {
+      "epoch": 0.10019809278113051,
+      "grad_norm": 1.059513669455098,
+      "learning_rate": 0.00019742907543541412,
+      "loss": 1.6908,
+      "step": 2175
+    },
+    {
+      "epoch": 0.10024416086976551,
+      "grad_norm": 1.1841512566476793,
+      "learning_rate": 0.00019742571276005583,
+      "loss": 1.2774,
+      "step": 2176
+    },
+    {
+      "epoch": 0.10029022895840052,
+      "grad_norm": 2.147290176468075,
+      "learning_rate": 0.0001974223479156811,
+      "loss": 1.5787,
+      "step": 2177
+    },
+    {
+      "epoch": 0.10033629704703552,
+      "grad_norm": 3.302322029521436,
+      "learning_rate": 0.00019741898090236485,
+      "loss": 1.2373,
+      "step": 2178
+    },
+    {
+      "epoch": 0.10038236513567052,
+      "grad_norm": 1.1068467960611872,
+      "learning_rate": 0.00019741561172018202,
+      "loss": 1.1582,
+      "step": 2179
+    },
+    {
+      "epoch": 0.10042843322430553,
+      "grad_norm": 1.2354544189036034,
+      "learning_rate": 0.0001974122403692076,
+      "loss": 1.4096,
+      "step": 2180
+    },
+    {
+      "epoch": 0.10047450131294053,
+      "grad_norm": 3.71593368933978,
+      "learning_rate": 0.0001974088668495167,
+      "loss": 1.2173,
+      "step": 2181
+    },
+    {
+      "epoch": 0.10052056940157553,
+      "grad_norm": 1.2200343915729739,
+      "learning_rate": 0.00019740549116118437,
+      "loss": 1.5931,
+      "step": 2182
+    },
+    {
+      "epoch": 0.10056663749021053,
+      "grad_norm": 1.4993790205511008,
+      "learning_rate": 0.00019740211330428583,
+      "loss": 1.3641,
+      "step": 2183
+    },
+    {
+      "epoch": 0.10061270557884554,
+      "grad_norm": 1.3314329282620925,
+      "learning_rate": 0.00019739873327889623,
+      "loss": 1.4788,
+      "step": 2184
+    },
+    {
+      "epoch": 0.10065877366748054,
+      "grad_norm": 0.8986513666582643,
+      "learning_rate": 0.00019739535108509083,
+      "loss": 0.9932,
+      "step": 2185
+    },
+    {
+      "epoch": 0.10070484175611553,
+      "grad_norm": 1.0242792640792286,
+      "learning_rate": 0.00019739196672294493,
+      "loss": 1.4215,
+      "step": 2186
+    },
+    {
+      "epoch": 0.10075090984475055,
+      "grad_norm": 0.8080013935251633,
+      "learning_rate": 0.0001973885801925339,
+      "loss": 1.4048,
+      "step": 2187
+    },
+    {
+      "epoch": 0.10079697793338555,
+      "grad_norm": 0.9409513308574519,
+      "learning_rate": 0.0001973851914939331,
+      "loss": 1.4751,
+      "step": 2188
+    },
+    {
+      "epoch": 0.10084304602202054,
+      "grad_norm": 1.0489376823239824,
+      "learning_rate": 0.000197381800627218,
+      "loss": 1.5616,
+      "step": 2189
+    },
+    {
+      "epoch": 0.10088911411065554,
+      "grad_norm": 1.0765919996036408,
+      "learning_rate": 0.0001973784075924641,
+      "loss": 1.4519,
+      "step": 2190
+    },
+    {
+      "epoch": 0.10093518219929055,
+      "grad_norm": 1.60321185834697,
+      "learning_rate": 0.0001973750123897469,
+      "loss": 1.5396,
+      "step": 2191
+    },
+    {
+      "epoch": 0.10098125028792555,
+      "grad_norm": 1.2298022648614007,
+      "learning_rate": 0.00019737161501914204,
+      "loss": 1.1552,
+      "step": 2192
+    },
+    {
+      "epoch": 0.10102731837656055,
+      "grad_norm": 1.0870406297506587,
+      "learning_rate": 0.00019736821548072507,
+      "loss": 1.2788,
+      "step": 2193
+    },
+    {
+      "epoch": 0.10107338646519556,
+      "grad_norm": 0.9182649766738237,
+      "learning_rate": 0.0001973648137745718,
+      "loss": 1.4597,
+      "step": 2194
+    },
+    {
+      "epoch": 0.10111945455383056,
+      "grad_norm": 1.2776098837625558,
+      "learning_rate": 0.00019736140990075788,
+      "loss": 1.3318,
+      "step": 2195
+    },
+    {
+      "epoch": 0.10116552264246556,
+      "grad_norm": 1.8312722029980855,
+      "learning_rate": 0.00019735800385935908,
+      "loss": 1.2581,
+      "step": 2196
+    },
+    {
+      "epoch": 0.10121159073110057,
+      "grad_norm": 1.094111604262593,
+      "learning_rate": 0.00019735459565045128,
+      "loss": 1.2931,
+      "step": 2197
+    },
+    {
+      "epoch": 0.10125765881973557,
+      "grad_norm": 1.2473515105460555,
+      "learning_rate": 0.0001973511852741103,
+      "loss": 1.2942,
+      "step": 2198
+    },
+    {
+      "epoch": 0.10130372690837057,
+      "grad_norm": 0.9273940250999783,
+      "learning_rate": 0.00019734777273041213,
+      "loss": 1.4643,
+      "step": 2199
+    },
+    {
+      "epoch": 0.10134979499700557,
+      "grad_norm": 1.3631849229707087,
+      "learning_rate": 0.00019734435801943272,
+      "loss": 1.3122,
+      "step": 2200
+    },
+    {
+      "epoch": 0.10139586308564058,
+      "grad_norm": 1.064047366024789,
+      "learning_rate": 0.00019734094114124808,
+      "loss": 0.9992,
+      "step": 2201
+    },
+    {
+      "epoch": 0.10144193117427558,
+      "grad_norm": 0.8787484793300865,
+      "learning_rate": 0.00019733752209593431,
+      "loss": 1.394,
+      "step": 2202
+    },
+    {
+      "epoch": 0.10148799926291058,
+      "grad_norm": 1.0756484621470674,
+      "learning_rate": 0.00019733410088356747,
+      "loss": 0.9925,
+      "step": 2203
+    },
+    {
+      "epoch": 0.10153406735154559,
+      "grad_norm": 2.3694945828361424,
+      "learning_rate": 0.0001973306775042238,
+      "loss": 1.2027,
+      "step": 2204
+    },
+    {
+      "epoch": 0.10158013544018059,
+      "grad_norm": 0.9998404006459086,
+      "learning_rate": 0.00019732725195797945,
+      "loss": 1.5184,
+      "step": 2205
+    },
+    {
+      "epoch": 0.10162620352881559,
+      "grad_norm": 1.775035852443922,
+      "learning_rate": 0.00019732382424491075,
+      "loss": 1.2607,
+      "step": 2206
+    },
+    {
+      "epoch": 0.10167227161745059,
+      "grad_norm": 1.0697511990300126,
+      "learning_rate": 0.00019732039436509395,
+      "loss": 1.2987,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1017183397060856,
+      "grad_norm": 1.1537413506952066,
+      "learning_rate": 0.00019731696231860545,
+      "loss": 1.4247,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1017644077947206,
+      "grad_norm": 1.1330964625977098,
+      "learning_rate": 0.00019731352810552167,
+      "loss": 1.1894,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1018104758833556,
+      "grad_norm": 0.8448669261855868,
+      "learning_rate": 0.000197310091725919,
+      "loss": 1.6694,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1018565439719906,
+      "grad_norm": 1.3542484127999885,
+      "learning_rate": 0.000197306653179874,
+      "loss": 1.4684,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1019026120606256,
+      "grad_norm": 1.0698578025207863,
+      "learning_rate": 0.00019730321246746324,
+      "loss": 1.2175,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1019486801492606,
+      "grad_norm": 1.375360325803395,
+      "learning_rate": 0.00019729976958876328,
+      "loss": 1.4178,
+      "step": 2213
+    },
+    {
+      "epoch": 0.10199474823789562,
+      "grad_norm": 1.2059043203863196,
+      "learning_rate": 0.00019729632454385077,
+      "loss": 1.391,
+      "step": 2214
+    },
+    {
+      "epoch": 0.10204081632653061,
+      "grad_norm": 1.4544379365384774,
+      "learning_rate": 0.00019729287733280244,
+      "loss": 1.5167,
+      "step": 2215
+    },
+    {
+      "epoch": 0.10208688441516561,
+      "grad_norm": 1.2583037429134318,
+      "learning_rate": 0.00019728942795569498,
+      "loss": 1.2022,
+      "step": 2216
+    },
+    {
+      "epoch": 0.10213295250380061,
+      "grad_norm": 1.4043995543732124,
+      "learning_rate": 0.00019728597641260526,
+      "loss": 1.3678,
+      "step": 2217
+    },
+    {
+      "epoch": 0.10217902059243562,
+      "grad_norm": 1.2370869236215207,
+      "learning_rate": 0.0001972825227036101,
+      "loss": 1.3505,
+      "step": 2218
+    },
+    {
+      "epoch": 0.10222508868107062,
+      "grad_norm": 0.8327136195243131,
+      "learning_rate": 0.0001972790668287863,
+      "loss": 1.7754,
+      "step": 2219
+    },
+    {
+      "epoch": 0.10227115676970562,
+      "grad_norm": 0.7748483363617464,
+      "learning_rate": 0.00019727560878821096,
+      "loss": 1.6182,
+      "step": 2220
+    },
+    {
+      "epoch": 0.10231722485834063,
+      "grad_norm": 1.1231825690087336,
+      "learning_rate": 0.00019727214858196093,
+      "loss": 1.8211,
+      "step": 2221
+    },
+    {
+      "epoch": 0.10236329294697563,
+      "grad_norm": 1.0839813916003866,
+      "learning_rate": 0.00019726868621011332,
+      "loss": 1.4378,
+      "step": 2222
+    },
+    {
+      "epoch": 0.10240936103561063,
+      "grad_norm": 1.0145914858795289,
+      "learning_rate": 0.0001972652216727452,
+      "loss": 1.2402,
+      "step": 2223
+    },
+    {
+      "epoch": 0.10245542912424563,
+      "grad_norm": 0.6724317736970482,
+      "learning_rate": 0.0001972617549699337,
+      "loss": 1.542,
+      "step": 2224
+    },
+    {
+      "epoch": 0.10250149721288064,
+      "grad_norm": 0.9920971145986139,
+      "learning_rate": 0.00019725828610175596,
+      "loss": 1.475,
+      "step": 2225
+    },
+    {
+      "epoch": 0.10254756530151564,
+      "grad_norm": 1.3283154473506855,
+      "learning_rate": 0.00019725481506828923,
+      "loss": 1.2061,
+      "step": 2226
+    },
+    {
+      "epoch": 0.10259363339015064,
+      "grad_norm": 1.0711541377508893,
+      "learning_rate": 0.00019725134186961084,
+      "loss": 1.3111,
+      "step": 2227
+    },
+    {
+      "epoch": 0.10263970147878565,
+      "grad_norm": 1.0626051264707876,
+      "learning_rate": 0.00019724786650579805,
+      "loss": 1.3661,
+      "step": 2228
+    },
+    {
+      "epoch": 0.10268576956742065,
+      "grad_norm": 0.8637582216188219,
+      "learning_rate": 0.00019724438897692826,
+      "loss": 1.3209,
+      "step": 2229
+    },
+    {
+      "epoch": 0.10273183765605565,
+      "grad_norm": 1.4478675650641972,
+      "learning_rate": 0.0001972409092830789,
+      "loss": 1.4609,
+      "step": 2230
+    },
+    {
+      "epoch": 0.10277790574469066,
+      "grad_norm": 1.3867964018700845,
+      "learning_rate": 0.0001972374274243274,
+      "loss": 1.2141,
+      "step": 2231
+    },
+    {
+      "epoch": 0.10282397383332566,
+      "grad_norm": 2.192682575001479,
+      "learning_rate": 0.00019723394340075134,
+      "loss": 1.4028,
+      "step": 2232
+    },
+    {
+      "epoch": 0.10287004192196066,
+      "grad_norm": 0.8922506377297964,
+      "learning_rate": 0.00019723045721242824,
+      "loss": 1.4188,
+      "step": 2233
+    },
+    {
+      "epoch": 0.10291611001059565,
+      "grad_norm": 0.8200105487161009,
+      "learning_rate": 0.0001972269688594357,
+      "loss": 1.4307,
+      "step": 2234
+    },
+    {
+      "epoch": 0.10296217809923067,
+      "grad_norm": 0.9211317319380243,
+      "learning_rate": 0.00019722347834185143,
+      "loss": 1.3411,
+      "step": 2235
+    },
+    {
+      "epoch": 0.10300824618786567,
+      "grad_norm": 0.8270020058383895,
+      "learning_rate": 0.00019721998565975314,
+      "loss": 1.2135,
+      "step": 2236
+    },
+    {
+      "epoch": 0.10305431427650066,
+      "grad_norm": 2.081509641772671,
+      "learning_rate": 0.00019721649081321852,
+      "loss": 1.225,
+      "step": 2237
+    },
+    {
+      "epoch": 0.10310038236513568,
+      "grad_norm": 0.9426796398466794,
+      "learning_rate": 0.00019721299380232545,
+      "loss": 1.1302,
+      "step": 2238
+    },
+    {
+      "epoch": 0.10314645045377067,
+      "grad_norm": 0.9178142979864736,
+      "learning_rate": 0.00019720949462715174,
+      "loss": 1.2169,
+      "step": 2239
+    },
+    {
+      "epoch": 0.10319251854240567,
+      "grad_norm": 0.8903316125450154,
+      "learning_rate": 0.00019720599328777535,
+      "loss": 1.3494,
+      "step": 2240
+    },
+    {
+      "epoch": 0.10323858663104067,
+      "grad_norm": 1.5298976360246956,
+      "learning_rate": 0.00019720248978427415,
+      "loss": 1.2467,
+      "step": 2241
+    },
+    {
+      "epoch": 0.10328465471967568,
+      "grad_norm": 1.2518179557056306,
+      "learning_rate": 0.0001971989841167262,
+      "loss": 1.2939,
+      "step": 2242
+    },
+    {
+      "epoch": 0.10333072280831068,
+      "grad_norm": 1.537995747841208,
+      "learning_rate": 0.0001971954762852095,
+      "loss": 1.5581,
+      "step": 2243
+    },
+    {
+      "epoch": 0.10337679089694568,
+      "grad_norm": 1.274111671501962,
+      "learning_rate": 0.0001971919662898022,
+      "loss": 1.2089,
+      "step": 2244
+    },
+    {
+      "epoch": 0.10342285898558069,
+      "grad_norm": 1.1234511014416146,
+      "learning_rate": 0.00019718845413058245,
+      "loss": 1.4045,
+      "step": 2245
+    },
+    {
+      "epoch": 0.10346892707421569,
+      "grad_norm": 1.2820945295078328,
+      "learning_rate": 0.00019718493980762838,
+      "loss": 1.2288,
+      "step": 2246
+    },
+    {
+      "epoch": 0.10351499516285069,
+      "grad_norm": 0.9975648991611429,
+      "learning_rate": 0.00019718142332101826,
+      "loss": 1.5326,
+      "step": 2247
+    },
+    {
+      "epoch": 0.1035610632514857,
+      "grad_norm": 1.285707533002868,
+      "learning_rate": 0.00019717790467083037,
+      "loss": 1.5254,
+      "step": 2248
+    },
+    {
+      "epoch": 0.1036071313401207,
+      "grad_norm": 0.8726292909178311,
+      "learning_rate": 0.00019717438385714306,
+      "loss": 1.4065,
+      "step": 2249
+    },
+    {
+      "epoch": 0.1036531994287557,
+      "grad_norm": 1.2747017755270902,
+      "learning_rate": 0.00019717086088003472,
+      "loss": 1.023,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1036992675173907,
+      "grad_norm": 1.6012616378718922,
+      "learning_rate": 0.0001971673357395838,
+      "loss": 1.2087,
+      "step": 2251
+    },
+    {
+      "epoch": 0.10374533560602571,
+      "grad_norm": 1.9958279826896639,
+      "learning_rate": 0.00019716380843586871,
+      "loss": 1.2871,
+      "step": 2252
+    },
+    {
+      "epoch": 0.10379140369466071,
+      "grad_norm": 1.0108794552081275,
+      "learning_rate": 0.00019716027896896803,
+      "loss": 1.341,
+      "step": 2253
+    },
+    {
+      "epoch": 0.1038374717832957,
+      "grad_norm": 1.6641333678457537,
+      "learning_rate": 0.00019715674733896034,
+      "loss": 1.3639,
+      "step": 2254
+    },
+    {
+      "epoch": 0.10388353987193072,
+      "grad_norm": 3.828074803743119,
+      "learning_rate": 0.00019715321354592429,
+      "loss": 1.0004,
+      "step": 2255
+    },
+    {
+      "epoch": 0.10392960796056572,
+      "grad_norm": 1.2732071635351423,
+      "learning_rate": 0.0001971496775899385,
+      "loss": 1.3984,
+      "step": 2256
+    },
+    {
+      "epoch": 0.10397567604920072,
+      "grad_norm": 1.4331329639260293,
+      "learning_rate": 0.00019714613947108172,
+      "loss": 1.0888,
+      "step": 2257
+    },
+    {
+      "epoch": 0.10402174413783571,
+      "grad_norm": 2.4902518992563727,
+      "learning_rate": 0.0001971425991894327,
+      "loss": 1.401,
+      "step": 2258
+    },
+    {
+      "epoch": 0.10406781222647073,
+      "grad_norm": 1.2933374565584705,
+      "learning_rate": 0.0001971390567450703,
+      "loss": 1.3832,
+      "step": 2259
+    },
+    {
+      "epoch": 0.10411388031510573,
+      "grad_norm": 2.514769754394211,
+      "learning_rate": 0.0001971355121380733,
+      "loss": 0.9244,
+      "step": 2260
+    },
+    {
+      "epoch": 0.10415994840374072,
+      "grad_norm": 2.3943271252801086,
+      "learning_rate": 0.0001971319653685207,
+      "loss": 1.092,
+      "step": 2261
+    },
+    {
+      "epoch": 0.10420601649237574,
+      "grad_norm": 1.3120967412474662,
+      "learning_rate": 0.0001971284164364915,
+      "loss": 1.1375,
+      "step": 2262
+    },
+    {
+      "epoch": 0.10425208458101073,
+      "grad_norm": 2.9018541215726508,
+      "learning_rate": 0.00019712486534206457,
+      "loss": 1.1203,
+      "step": 2263
+    },
+    {
+      "epoch": 0.10429815266964573,
+      "grad_norm": 1.2585506367518569,
+      "learning_rate": 0.00019712131208531907,
+      "loss": 1.5046,
+      "step": 2264
+    },
+    {
+      "epoch": 0.10434422075828075,
+      "grad_norm": 1.09102509333813,
+      "learning_rate": 0.00019711775666633413,
+      "loss": 1.1358,
+      "step": 2265
+    },
+    {
+      "epoch": 0.10439028884691574,
+      "grad_norm": 2.3111696071428085,
+      "learning_rate": 0.0001971141990851888,
+      "loss": 0.9058,
+      "step": 2266
+    },
+    {
+      "epoch": 0.10443635693555074,
+      "grad_norm": 1.1570627108523577,
+      "learning_rate": 0.00019711063934196234,
+      "loss": 1.6189,
+      "step": 2267
+    },
+    {
+      "epoch": 0.10448242502418574,
+      "grad_norm": 1.3108815010762565,
+      "learning_rate": 0.00019710707743673405,
+      "loss": 1.6065,
+      "step": 2268
+    },
+    {
+      "epoch": 0.10452849311282075,
+      "grad_norm": 1.125491608055442,
+      "learning_rate": 0.00019710351336958316,
+      "loss": 1.2462,
+      "step": 2269
+    },
+    {
+      "epoch": 0.10457456120145575,
+      "grad_norm": 1.4061119456287388,
+      "learning_rate": 0.00019709994714058903,
+      "loss": 1.3475,
+      "step": 2270
+    },
+    {
+      "epoch": 0.10462062929009075,
+      "grad_norm": 1.1519048518540052,
+      "learning_rate": 0.0001970963787498311,
+      "loss": 1.4281,
+      "step": 2271
+    },
+    {
+      "epoch": 0.10466669737872576,
+      "grad_norm": 1.5510315493890363,
+      "learning_rate": 0.00019709280819738874,
+      "loss": 1.4936,
+      "step": 2272
+    },
+    {
+      "epoch": 0.10471276546736076,
+      "grad_norm": 1.3911379646153386,
+      "learning_rate": 0.0001970892354833415,
+      "loss": 1.223,
+      "step": 2273
+    },
+    {
+      "epoch": 0.10475883355599576,
+      "grad_norm": 1.1853024967788761,
+      "learning_rate": 0.0001970856606077689,
+      "loss": 1.6184,
+      "step": 2274
+    },
+    {
+      "epoch": 0.10480490164463076,
+      "grad_norm": 1.2829134921247234,
+      "learning_rate": 0.00019708208357075056,
+      "loss": 1.2246,
+      "step": 2275
+    },
+    {
+      "epoch": 0.10485096973326577,
+      "grad_norm": 1.3068233929847832,
+      "learning_rate": 0.00019707850437236606,
+      "loss": 1.4397,
+      "step": 2276
+    },
+    {
+      "epoch": 0.10489703782190077,
+      "grad_norm": 1.0920845355722404,
+      "learning_rate": 0.00019707492301269513,
+      "loss": 1.7886,
+      "step": 2277
+    },
+    {
+      "epoch": 0.10494310591053577,
+      "grad_norm": 2.46192463761185,
+      "learning_rate": 0.00019707133949181747,
+      "loss": 1.4523,
+      "step": 2278
+    },
+    {
+      "epoch": 0.10498917399917078,
+      "grad_norm": 1.2608944093256036,
+      "learning_rate": 0.0001970677538098129,
+      "loss": 1.6494,
+      "step": 2279
+    },
+    {
+      "epoch": 0.10503524208780578,
+      "grad_norm": 1.2387246972871706,
+      "learning_rate": 0.00019706416596676122,
+      "loss": 1.5808,
+      "step": 2280
+    },
+    {
+      "epoch": 0.10508131017644078,
+      "grad_norm": 2.7537211885866366,
+      "learning_rate": 0.0001970605759627423,
+      "loss": 1.236,
+      "step": 2281
+    },
+    {
+      "epoch": 0.10512737826507579,
+      "grad_norm": 1.4993930294596007,
+      "learning_rate": 0.0001970569837978361,
+      "loss": 1.4943,
+      "step": 2282
+    },
+    {
+      "epoch": 0.10517344635371079,
+      "grad_norm": 1.5689662215223479,
+      "learning_rate": 0.00019705338947212256,
+      "loss": 1.2202,
+      "step": 2283
+    },
+    {
+      "epoch": 0.10521951444234579,
+      "grad_norm": 1.7276496352046065,
+      "learning_rate": 0.0001970497929856817,
+      "loss": 1.4903,
+      "step": 2284
+    },
+    {
+      "epoch": 0.10526558253098078,
+      "grad_norm": 1.6280667824233204,
+      "learning_rate": 0.00019704619433859365,
+      "loss": 1.4166,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1053116506196158,
+      "grad_norm": 1.3407593658175903,
+      "learning_rate": 0.00019704259353093843,
+      "loss": 0.9556,
+      "step": 2286
+    },
+    {
+      "epoch": 0.1053577187082508,
+      "grad_norm": 1.8420501909779605,
+      "learning_rate": 0.00019703899056279628,
+      "loss": 1.8964,
+      "step": 2287
+    },
+    {
+      "epoch": 0.10540378679688579,
+      "grad_norm": 1.7990943495933327,
+      "learning_rate": 0.0001970353854342474,
+      "loss": 1.1682,
+      "step": 2288
+    },
+    {
+      "epoch": 0.1054498548855208,
+      "grad_norm": 0.9992146484887442,
+      "learning_rate": 0.00019703177814537202,
+      "loss": 1.3506,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1054959229741558,
+      "grad_norm": 2.5330447563951455,
+      "learning_rate": 0.0001970281686962505,
+      "loss": 1.6097,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1055419910627908,
+      "grad_norm": 1.6552406891897933,
+      "learning_rate": 0.00019702455708696314,
+      "loss": 1.4616,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1055880591514258,
+      "grad_norm": 5.313622557522923,
+      "learning_rate": 0.00019702094331759042,
+      "loss": 1.618,
+      "step": 2292
+    },
+    {
+      "epoch": 0.10563412724006081,
+      "grad_norm": 1.2336574026774172,
+      "learning_rate": 0.00019701732738821272,
+      "loss": 1.2177,
+      "step": 2293
+    },
+    {
+      "epoch": 0.10568019532869581,
+      "grad_norm": 2.0026320763417433,
+      "learning_rate": 0.00019701370929891057,
+      "loss": 1.745,
+      "step": 2294
+    },
+    {
+      "epoch": 0.10572626341733081,
+      "grad_norm": 2.2181233233944133,
+      "learning_rate": 0.00019701008904976454,
+      "loss": 1.641,
+      "step": 2295
+    },
+    {
+      "epoch": 0.10577233150596582,
+      "grad_norm": 1.408475593451812,
+      "learning_rate": 0.00019700646664085522,
+      "loss": 1.3367,
+      "step": 2296
+    },
+    {
+      "epoch": 0.10581839959460082,
+      "grad_norm": 1.2974297265132573,
+      "learning_rate": 0.00019700284207226326,
+      "loss": 1.4144,
+      "step": 2297
+    },
+    {
+      "epoch": 0.10586446768323582,
+      "grad_norm": 1.830569073635215,
+      "learning_rate": 0.00019699921534406932,
+      "loss": 1.4685,
+      "step": 2298
+    },
+    {
+      "epoch": 0.10591053577187083,
+      "grad_norm": 1.1230031905924718,
+      "learning_rate": 0.00019699558645635418,
+      "loss": 1.4025,
+      "step": 2299
+    },
+    {
+      "epoch": 0.10595660386050583,
+      "grad_norm": 2.161594571210428,
+      "learning_rate": 0.00019699195540919863,
+      "loss": 1.3123,
+      "step": 2300
+    },
+    {
+      "epoch": 0.10600267194914083,
+      "grad_norm": 1.0833698835240022,
+      "learning_rate": 0.00019698832220268347,
+      "loss": 1.4542,
+      "step": 2301
+    },
+    {
+      "epoch": 0.10604874003777583,
+      "grad_norm": 1.2789597725055106,
+      "learning_rate": 0.00019698468683688965,
+      "loss": 1.1673,
+      "step": 2302
+    },
+    {
+      "epoch": 0.10609480812641084,
+      "grad_norm": 1.7175851685082257,
+      "learning_rate": 0.00019698104931189803,
+      "loss": 1.1551,
+      "step": 2303
+    },
+    {
+      "epoch": 0.10614087621504584,
+      "grad_norm": 4.28675194381912,
+      "learning_rate": 0.00019697740962778965,
+      "loss": 1.3381,
+      "step": 2304
+    },
+    {
+      "epoch": 0.10618694430368084,
+      "grad_norm": 1.226641182879569,
+      "learning_rate": 0.00019697376778464554,
+      "loss": 1.2921,
+      "step": 2305
+    },
+    {
+      "epoch": 0.10623301239231585,
+      "grad_norm": 1.5851190883794495,
+      "learning_rate": 0.00019697012378254675,
+      "loss": 1.4593,
+      "step": 2306
+    },
+    {
+      "epoch": 0.10627908048095085,
+      "grad_norm": 2.5701107457804477,
+      "learning_rate": 0.00019696647762157442,
+      "loss": 1.3486,
+      "step": 2307
+    },
+    {
+      "epoch": 0.10632514856958585,
+      "grad_norm": 1.0455589156503684,
+      "learning_rate": 0.00019696282930180975,
+      "loss": 1.4969,
+      "step": 2308
+    },
+    {
+      "epoch": 0.10637121665822084,
+      "grad_norm": 1.3500289242329822,
+      "learning_rate": 0.0001969591788233339,
+      "loss": 1.3865,
+      "step": 2309
+    },
+    {
+      "epoch": 0.10641728474685586,
+      "grad_norm": 1.5524930613613657,
+      "learning_rate": 0.00019695552618622823,
+      "loss": 1.3544,
+      "step": 2310
+    },
+    {
+      "epoch": 0.10646335283549085,
+      "grad_norm": 1.0053188521986116,
+      "learning_rate": 0.000196951871390574,
+      "loss": 1.3751,
+      "step": 2311
+    },
+    {
+      "epoch": 0.10650942092412585,
+      "grad_norm": 1.1430848963726001,
+      "learning_rate": 0.00019694821443645258,
+      "loss": 1.2872,
+      "step": 2312
+    },
+    {
+      "epoch": 0.10655548901276087,
+      "grad_norm": 1.4612059093257606,
+      "learning_rate": 0.0001969445553239454,
+      "loss": 1.2156,
+      "step": 2313
+    },
+    {
+      "epoch": 0.10660155710139586,
+      "grad_norm": 1.0682464058665833,
+      "learning_rate": 0.00019694089405313393,
+      "loss": 1.3892,
+      "step": 2314
+    },
+    {
+      "epoch": 0.10664762519003086,
+      "grad_norm": 1.8764043507131496,
+      "learning_rate": 0.00019693723062409963,
+      "loss": 1.4188,
+      "step": 2315
+    },
+    {
+      "epoch": 0.10669369327866587,
+      "grad_norm": 1.7278376699090612,
+      "learning_rate": 0.00019693356503692412,
+      "loss": 1.0673,
+      "step": 2316
+    },
+    {
+      "epoch": 0.10673976136730087,
+      "grad_norm": 1.627063276971387,
+      "learning_rate": 0.00019692989729168901,
+      "loss": 1.6119,
+      "step": 2317
+    },
+    {
+      "epoch": 0.10678582945593587,
+      "grad_norm": 1.343790729547665,
+      "learning_rate": 0.00019692622738847596,
+      "loss": 1.2382,
+      "step": 2318
+    },
+    {
+      "epoch": 0.10683189754457087,
+      "grad_norm": 1.6041472567603134,
+      "learning_rate": 0.0001969225553273666,
+      "loss": 1.357,
+      "step": 2319
+    },
+    {
+      "epoch": 0.10687796563320588,
+      "grad_norm": 1.1785575613943282,
+      "learning_rate": 0.00019691888110844275,
+      "loss": 1.5824,
+      "step": 2320
+    },
+    {
+      "epoch": 0.10692403372184088,
+      "grad_norm": 1.343304997288376,
+      "learning_rate": 0.0001969152047317862,
+      "loss": 1.344,
+      "step": 2321
+    },
+    {
+      "epoch": 0.10697010181047588,
+      "grad_norm": 2.0040619204395265,
+      "learning_rate": 0.00019691152619747876,
+      "loss": 1.1409,
+      "step": 2322
+    },
+    {
+      "epoch": 0.10701616989911089,
+      "grad_norm": 0.9162174745068978,
+      "learning_rate": 0.00019690784550560237,
+      "loss": 1.444,
+      "step": 2323
+    },
+    {
+      "epoch": 0.10706223798774589,
+      "grad_norm": 1.0386073860976481,
+      "learning_rate": 0.00019690416265623898,
+      "loss": 1.583,
+      "step": 2324
+    },
+    {
+      "epoch": 0.10710830607638089,
+      "grad_norm": 1.2267615499644677,
+      "learning_rate": 0.00019690047764947056,
+      "loss": 1.6335,
+      "step": 2325
+    },
+    {
+      "epoch": 0.1071543741650159,
+      "grad_norm": 1.7074311344840964,
+      "learning_rate": 0.00019689679048537915,
+      "loss": 1.3382,
+      "step": 2326
+    },
+    {
+      "epoch": 0.1072004422536509,
+      "grad_norm": 1.8865821691290543,
+      "learning_rate": 0.00019689310116404685,
+      "loss": 1.5353,
+      "step": 2327
+    },
+    {
+      "epoch": 0.1072465103422859,
+      "grad_norm": 1.9477589298478195,
+      "learning_rate": 0.00019688940968555575,
+      "loss": 1.5353,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1072925784309209,
+      "grad_norm": 2.7623851837787123,
+      "learning_rate": 0.00019688571604998813,
+      "loss": 1.3798,
+      "step": 2329
+    },
+    {
+      "epoch": 0.10733864651955591,
+      "grad_norm": 1.3474034484385182,
+      "learning_rate": 0.00019688202025742614,
+      "loss": 1.3807,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1073847146081909,
+      "grad_norm": 1.4394944776116183,
+      "learning_rate": 0.00019687832230795208,
+      "loss": 1.3904,
+      "step": 2331
+    },
+    {
+      "epoch": 0.1074307826968259,
+      "grad_norm": 1.4731768733914146,
+      "learning_rate": 0.00019687462220164826,
+      "loss": 1.3541,
+      "step": 2332
+    },
+    {
+      "epoch": 0.10747685078546092,
+      "grad_norm": 1.2711064227078208,
+      "learning_rate": 0.00019687091993859715,
+      "loss": 1.4577,
+      "step": 2333
+    },
+    {
+      "epoch": 0.10752291887409592,
+      "grad_norm": 1.0342154907554522,
+      "learning_rate": 0.00019686721551888105,
+      "loss": 1.525,
+      "step": 2334
+    },
+    {
+      "epoch": 0.10756898696273091,
+      "grad_norm": 1.9405766974618357,
+      "learning_rate": 0.0001968635089425825,
+      "loss": 1.3786,
+      "step": 2335
+    },
+    {
+      "epoch": 0.10761505505136591,
+      "grad_norm": 1.1431016331440436,
+      "learning_rate": 0.000196859800209784,
+      "loss": 1.2455,
+      "step": 2336
+    },
+    {
+      "epoch": 0.10766112314000092,
+      "grad_norm": 2.9286528469811564,
+      "learning_rate": 0.00019685608932056817,
+      "loss": 1.6136,
+      "step": 2337
+    },
+    {
+      "epoch": 0.10770719122863592,
+      "grad_norm": 2.4998179748352785,
+      "learning_rate": 0.00019685237627501754,
+      "loss": 1.2752,
+      "step": 2338
+    },
+    {
+      "epoch": 0.10775325931727092,
+      "grad_norm": 1.8148856657979153,
+      "learning_rate": 0.0001968486610732148,
+      "loss": 1.0158,
+      "step": 2339
+    },
+    {
+      "epoch": 0.10779932740590593,
+      "grad_norm": 1.548562587383222,
+      "learning_rate": 0.00019684494371524273,
+      "loss": 1.4653,
+      "step": 2340
+    },
+    {
+      "epoch": 0.10784539549454093,
+      "grad_norm": 1.2915963074377879,
+      "learning_rate": 0.00019684122420118402,
+      "loss": 1.3622,
+      "step": 2341
+    },
+    {
+      "epoch": 0.10789146358317593,
+      "grad_norm": 3.256796486725616,
+      "learning_rate": 0.0001968375025311215,
+      "loss": 1.4339,
+      "step": 2342
+    },
+    {
+      "epoch": 0.10793753167181094,
+      "grad_norm": 1.5040975826806244,
+      "learning_rate": 0.00019683377870513804,
+      "loss": 1.4487,
+      "step": 2343
+    },
+    {
+      "epoch": 0.10798359976044594,
+      "grad_norm": 2.1763879171385403,
+      "learning_rate": 0.00019683005272331654,
+      "loss": 1.2786,
+      "step": 2344
+    },
+    {
+      "epoch": 0.10802966784908094,
+      "grad_norm": 1.0332987605270252,
+      "learning_rate": 0.00019682632458573992,
+      "loss": 1.3764,
+      "step": 2345
+    },
+    {
+      "epoch": 0.10807573593771594,
+      "grad_norm": 1.1967371239142777,
+      "learning_rate": 0.0001968225942924912,
+      "loss": 1.1231,
+      "step": 2346
+    },
+    {
+      "epoch": 0.10812180402635095,
+      "grad_norm": 2.005067157261382,
+      "learning_rate": 0.00019681886184365346,
+      "loss": 1.2069,
+      "step": 2347
+    },
+    {
+      "epoch": 0.10816787211498595,
+      "grad_norm": 1.1676446436131191,
+      "learning_rate": 0.00019681512723930972,
+      "loss": 1.4448,
+      "step": 2348
+    },
+    {
+      "epoch": 0.10821394020362095,
+      "grad_norm": 1.4044518065338525,
+      "learning_rate": 0.00019681139047954323,
+      "loss": 1.4835,
+      "step": 2349
+    },
+    {
+      "epoch": 0.10826000829225596,
+      "grad_norm": 1.0999112955179267,
+      "learning_rate": 0.0001968076515644371,
+      "loss": 1.6755,
+      "step": 2350
+    },
+    {
+      "epoch": 0.10830607638089096,
+      "grad_norm": 1.7931077352997746,
+      "learning_rate": 0.00019680391049407462,
+      "loss": 1.1365,
+      "step": 2351
+    },
+    {
+      "epoch": 0.10835214446952596,
+      "grad_norm": 1.9247140308302506,
+      "learning_rate": 0.00019680016726853904,
+      "loss": 1.6454,
+      "step": 2352
+    },
+    {
+      "epoch": 0.10839821255816096,
+      "grad_norm": 2.167583774226976,
+      "learning_rate": 0.0001967964218879137,
+      "loss": 1.4361,
+      "step": 2353
+    },
+    {
+      "epoch": 0.10844428064679597,
+      "grad_norm": 2.2725366009428813,
+      "learning_rate": 0.00019679267435228202,
+      "loss": 1.3661,
+      "step": 2354
+    },
+    {
+      "epoch": 0.10849034873543097,
+      "grad_norm": 1.0226446786199683,
+      "learning_rate": 0.00019678892466172738,
+      "loss": 1.5423,
+      "step": 2355
+    },
+    {
+      "epoch": 0.10853641682406596,
+      "grad_norm": 1.5026781121278352,
+      "learning_rate": 0.0001967851728163333,
+      "loss": 1.4867,
+      "step": 2356
+    },
+    {
+      "epoch": 0.10858248491270098,
+      "grad_norm": 1.5130560092520557,
+      "learning_rate": 0.0001967814188161833,
+      "loss": 1.2936,
+      "step": 2357
+    },
+    {
+      "epoch": 0.10862855300133598,
+      "grad_norm": 1.1365662686312732,
+      "learning_rate": 0.00019677766266136098,
+      "loss": 1.0825,
+      "step": 2358
+    },
+    {
+      "epoch": 0.10867462108997097,
+      "grad_norm": 3.46422744052517,
+      "learning_rate": 0.0001967739043519499,
+      "loss": 1.3843,
+      "step": 2359
+    },
+    {
+      "epoch": 0.10872068917860599,
+      "grad_norm": 1.5480773959194996,
+      "learning_rate": 0.0001967701438880338,
+      "loss": 1.3191,
+      "step": 2360
+    },
+    {
+      "epoch": 0.10876675726724098,
+      "grad_norm": 0.8765932375113897,
+      "learning_rate": 0.00019676638126969636,
+      "loss": 1.6268,
+      "step": 2361
+    },
+    {
+      "epoch": 0.10881282535587598,
+      "grad_norm": 37.68192422225846,
+      "learning_rate": 0.00019676261649702136,
+      "loss": 1.4714,
+      "step": 2362
+    },
+    {
+      "epoch": 0.10885889344451098,
+      "grad_norm": 1.6673378713026372,
+      "learning_rate": 0.00019675884957009264,
+      "loss": 1.3738,
+      "step": 2363
+    },
+    {
+      "epoch": 0.108904961533146,
+      "grad_norm": 10.802880534132111,
+      "learning_rate": 0.000196755080488994,
+      "loss": 1.6278,
+      "step": 2364
+    },
+    {
+      "epoch": 0.10895102962178099,
+      "grad_norm": 2.55488083533551,
+      "learning_rate": 0.0001967513092538094,
+      "loss": 1.6828,
+      "step": 2365
+    },
+    {
+      "epoch": 0.10899709771041599,
+      "grad_norm": 1.3563347267214916,
+      "learning_rate": 0.00019674753586462282,
+      "loss": 1.622,
+      "step": 2366
+    },
+    {
+      "epoch": 0.109043165799051,
+      "grad_norm": 1.581189405865274,
+      "learning_rate": 0.0001967437603215182,
+      "loss": 1.5346,
+      "step": 2367
+    },
+    {
+      "epoch": 0.109089233887686,
+      "grad_norm": 2.5843618335875314,
+      "learning_rate": 0.00019673998262457966,
+      "loss": 1.6882,
+      "step": 2368
+    },
+    {
+      "epoch": 0.109135301976321,
+      "grad_norm": 2.6452921865644643,
+      "learning_rate": 0.0001967362027738913,
+      "loss": 1.3835,
+      "step": 2369
+    },
+    {
+      "epoch": 0.109181370064956,
+      "grad_norm": 1.9863802131552262,
+      "learning_rate": 0.0001967324207695372,
+      "loss": 1.3264,
+      "step": 2370
+    },
+    {
+      "epoch": 0.10922743815359101,
+      "grad_norm": 2.216225575710332,
+      "learning_rate": 0.00019672863661160167,
+      "loss": 1.4384,
+      "step": 2371
+    },
+    {
+      "epoch": 0.10927350624222601,
+      "grad_norm": 3.554742813972804,
+      "learning_rate": 0.00019672485030016888,
+      "loss": 1.6638,
+      "step": 2372
+    },
+    {
+      "epoch": 0.10931957433086101,
+      "grad_norm": 1.1609933808359194,
+      "learning_rate": 0.00019672106183532312,
+      "loss": 1.3955,
+      "step": 2373
+    },
+    {
+      "epoch": 0.10936564241949602,
+      "grad_norm": 1.913700664593023,
+      "learning_rate": 0.0001967172712171488,
+      "loss": 1.8743,
+      "step": 2374
+    },
+    {
+      "epoch": 0.10941171050813102,
+      "grad_norm": 1.7193809690543378,
+      "learning_rate": 0.00019671347844573024,
+      "loss": 1.3669,
+      "step": 2375
+    },
+    {
+      "epoch": 0.10945777859676602,
+      "grad_norm": 1.1832797496526724,
+      "learning_rate": 0.00019670968352115194,
+      "loss": 1.5361,
+      "step": 2376
+    },
+    {
+      "epoch": 0.10950384668540103,
+      "grad_norm": 1.5449301059291622,
+      "learning_rate": 0.0001967058864434983,
+      "loss": 1.149,
+      "step": 2377
+    },
+    {
+      "epoch": 0.10954991477403603,
+      "grad_norm": 1.5046763145008133,
+      "learning_rate": 0.00019670208721285398,
+      "loss": 1.3474,
+      "step": 2378
+    },
+    {
+      "epoch": 0.10959598286267103,
+      "grad_norm": 1.1946638026905259,
+      "learning_rate": 0.00019669828582930348,
+      "loss": 1.4119,
+      "step": 2379
+    },
+    {
+      "epoch": 0.10964205095130602,
+      "grad_norm": 1.2936388251395277,
+      "learning_rate": 0.00019669448229293142,
+      "loss": 1.7158,
+      "step": 2380
+    },
+    {
+      "epoch": 0.10968811903994104,
+      "grad_norm": 1.9294327162394416,
+      "learning_rate": 0.00019669067660382252,
+      "loss": 1.3627,
+      "step": 2381
+    },
+    {
+      "epoch": 0.10973418712857604,
+      "grad_norm": 1.0943531238687383,
+      "learning_rate": 0.0001966868687620615,
+      "loss": 1.4121,
+      "step": 2382
+    },
+    {
+      "epoch": 0.10978025521721103,
+      "grad_norm": 1.2683062625951453,
+      "learning_rate": 0.00019668305876773313,
+      "loss": 1.3752,
+      "step": 2383
+    },
+    {
+      "epoch": 0.10982632330584605,
+      "grad_norm": 1.6018551773389853,
+      "learning_rate": 0.00019667924662092223,
+      "loss": 1.6053,
+      "step": 2384
+    },
+    {
+      "epoch": 0.10987239139448104,
+      "grad_norm": 1.123884556983763,
+      "learning_rate": 0.00019667543232171366,
+      "loss": 1.733,
+      "step": 2385
+    },
+    {
+      "epoch": 0.10991845948311604,
+      "grad_norm": 2.0148381635560573,
+      "learning_rate": 0.00019667161587019237,
+      "loss": 1.4638,
+      "step": 2386
+    },
+    {
+      "epoch": 0.10996452757175104,
+      "grad_norm": 1.3605593539807537,
+      "learning_rate": 0.0001966677972664433,
+      "loss": 1.2296,
+      "step": 2387
+    },
+    {
+      "epoch": 0.11001059566038605,
+      "grad_norm": 1.2255127029109083,
+      "learning_rate": 0.00019666397651055148,
+      "loss": 1.3963,
+      "step": 2388
+    },
+    {
+      "epoch": 0.11005666374902105,
+      "grad_norm": 1.3200980186501612,
+      "learning_rate": 0.00019666015360260197,
+      "loss": 1.3496,
+      "step": 2389
+    },
+    {
+      "epoch": 0.11010273183765605,
+      "grad_norm": 1.7619889484489413,
+      "learning_rate": 0.00019665632854267985,
+      "loss": 1.4613,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11014879992629106,
+      "grad_norm": 1.1477786777552543,
+      "learning_rate": 0.00019665250133087033,
+      "loss": 1.4792,
+      "step": 2391
+    },
+    {
+      "epoch": 0.11019486801492606,
+      "grad_norm": 2.4538345844852683,
+      "learning_rate": 0.0001966486719672586,
+      "loss": 1.3553,
+      "step": 2392
+    },
+    {
+      "epoch": 0.11024093610356106,
+      "grad_norm": 1.5512248053216044,
+      "learning_rate": 0.0001966448404519299,
+      "loss": 1.4772,
+      "step": 2393
+    },
+    {
+      "epoch": 0.11028700419219607,
+      "grad_norm": 1.3284619938427122,
+      "learning_rate": 0.00019664100678496954,
+      "loss": 1.4399,
+      "step": 2394
+    },
+    {
+      "epoch": 0.11033307228083107,
+      "grad_norm": 1.2632157430049815,
+      "learning_rate": 0.00019663717096646285,
+      "loss": 1.4702,
+      "step": 2395
+    },
+    {
+      "epoch": 0.11037914036946607,
+      "grad_norm": 1.137347479635243,
+      "learning_rate": 0.00019663333299649525,
+      "loss": 1.5225,
+      "step": 2396
+    },
+    {
+      "epoch": 0.11042520845810107,
+      "grad_norm": 1.2195227177048236,
+      "learning_rate": 0.0001966294928751522,
+      "loss": 1.2384,
+      "step": 2397
+    },
+    {
+      "epoch": 0.11047127654673608,
+      "grad_norm": 4.445155289392034,
+      "learning_rate": 0.00019662565060251915,
+      "loss": 1.3145,
+      "step": 2398
+    },
+    {
+      "epoch": 0.11051734463537108,
+      "grad_norm": 1.6705270299536032,
+      "learning_rate": 0.00019662180617868166,
+      "loss": 1.5228,
+      "step": 2399
+    },
+    {
+      "epoch": 0.11056341272400608,
+      "grad_norm": 2.0897381151657575,
+      "learning_rate": 0.00019661795960372534,
+      "loss": 1.2848,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11060948081264109,
+      "grad_norm": 1.8251674285039223,
+      "learning_rate": 0.00019661411087773582,
+      "loss": 0.9773,
+      "step": 2401
+    },
+    {
+      "epoch": 0.11065554890127609,
+      "grad_norm": 2.783434665398051,
+      "learning_rate": 0.0001966102600007988,
+      "loss": 1.8273,
+      "step": 2402
+    },
+    {
+      "epoch": 0.11070161698991109,
+      "grad_norm": 1.7904485111091126,
+      "learning_rate": 0.00019660640697299994,
+      "loss": 1.6736,
+      "step": 2403
+    },
+    {
+      "epoch": 0.11074768507854608,
+      "grad_norm": 1.4455959339549747,
+      "learning_rate": 0.00019660255179442513,
+      "loss": 1.552,
+      "step": 2404
+    },
+    {
+      "epoch": 0.1107937531671811,
+      "grad_norm": 1.5256018200758286,
+      "learning_rate": 0.0001965986944651601,
+      "loss": 1.2738,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1108398212558161,
+      "grad_norm": 2.8259761598925945,
+      "learning_rate": 0.00019659483498529078,
+      "loss": 1.5749,
+      "step": 2406
+    },
+    {
+      "epoch": 0.1108858893444511,
+      "grad_norm": 1.8320750233260517,
+      "learning_rate": 0.0001965909733549031,
+      "loss": 1.6243,
+      "step": 2407
+    },
+    {
+      "epoch": 0.1109319574330861,
+      "grad_norm": 1.9841564841980588,
+      "learning_rate": 0.00019658710957408298,
+      "loss": 1.5284,
+      "step": 2408
+    },
+    {
+      "epoch": 0.1109780255217211,
+      "grad_norm": 1.5495819798820951,
+      "learning_rate": 0.0001965832436429165,
+      "loss": 1.0356,
+      "step": 2409
+    },
+    {
+      "epoch": 0.1110240936103561,
+      "grad_norm": 2.959484582354105,
+      "learning_rate": 0.0001965793755614897,
+      "loss": 1.0331,
+      "step": 2410
+    },
+    {
+      "epoch": 0.11107016169899112,
+      "grad_norm": 1.7864888623994395,
+      "learning_rate": 0.00019657550532988874,
+      "loss": 1.6597,
+      "step": 2411
+    },
+    {
+      "epoch": 0.11111622978762611,
+      "grad_norm": 3.0256641746442603,
+      "learning_rate": 0.0001965716329481997,
+      "loss": 1.4411,
+      "step": 2412
+    },
+    {
+      "epoch": 0.11116229787626111,
+      "grad_norm": 1.7366447667671279,
+      "learning_rate": 0.00019656775841650882,
+      "loss": 1.4659,
+      "step": 2413
+    },
+    {
+      "epoch": 0.11120836596489611,
+      "grad_norm": 2.8443625509734605,
+      "learning_rate": 0.00019656388173490243,
+      "loss": 1.5469,
+      "step": 2414
+    },
+    {
+      "epoch": 0.11125443405353112,
+      "grad_norm": 1.9932833131839862,
+      "learning_rate": 0.00019656000290346675,
+      "loss": 1.3674,
+      "step": 2415
+    },
+    {
+      "epoch": 0.11130050214216612,
+      "grad_norm": 2.2146201620594432,
+      "learning_rate": 0.00019655612192228818,
+      "loss": 1.2748,
+      "step": 2416
+    },
+    {
+      "epoch": 0.11134657023080112,
+      "grad_norm": 2.3706075555981845,
+      "learning_rate": 0.00019655223879145312,
+      "loss": 1.6287,
+      "step": 2417
+    },
+    {
+      "epoch": 0.11139263831943613,
+      "grad_norm": 2.340951307100998,
+      "learning_rate": 0.00019654835351104802,
+      "loss": 1.34,
+      "step": 2418
+    },
+    {
+      "epoch": 0.11143870640807113,
+      "grad_norm": 2.209816930822457,
+      "learning_rate": 0.00019654446608115937,
+      "loss": 1.1311,
+      "step": 2419
+    },
+    {
+      "epoch": 0.11148477449670613,
+      "grad_norm": 2.0227836557973746,
+      "learning_rate": 0.0001965405765018737,
+      "loss": 1.4226,
+      "step": 2420
+    },
+    {
+      "epoch": 0.11153084258534113,
+      "grad_norm": 2.087300735567473,
+      "learning_rate": 0.00019653668477327764,
+      "loss": 1.1696,
+      "step": 2421
+    },
+    {
+      "epoch": 0.11157691067397614,
+      "grad_norm": 1.5082904191082338,
+      "learning_rate": 0.00019653279089545782,
+      "loss": 1.4854,
+      "step": 2422
+    },
+    {
+      "epoch": 0.11162297876261114,
+      "grad_norm": 1.5690341861081198,
+      "learning_rate": 0.00019652889486850093,
+      "loss": 1.3188,
+      "step": 2423
+    },
+    {
+      "epoch": 0.11166904685124614,
+      "grad_norm": 1.9861924053606106,
+      "learning_rate": 0.0001965249966924937,
+      "loss": 1.1113,
+      "step": 2424
+    },
+    {
+      "epoch": 0.11171511493988115,
+      "grad_norm": 1.5369139962161518,
+      "learning_rate": 0.00019652109636752294,
+      "loss": 1.7524,
+      "step": 2425
+    },
+    {
+      "epoch": 0.11176118302851615,
+      "grad_norm": 1.369152915997438,
+      "learning_rate": 0.00019651719389367545,
+      "loss": 1.1834,
+      "step": 2426
+    },
+    {
+      "epoch": 0.11180725111715115,
+      "grad_norm": 1.345365169920432,
+      "learning_rate": 0.00019651328927103813,
+      "loss": 1.2718,
+      "step": 2427
+    },
+    {
+      "epoch": 0.11185331920578616,
+      "grad_norm": 1.176808524236659,
+      "learning_rate": 0.00019650938249969792,
+      "loss": 1.255,
+      "step": 2428
+    },
+    {
+      "epoch": 0.11189938729442116,
+      "grad_norm": 1.2580158622316375,
+      "learning_rate": 0.00019650547357974175,
+      "loss": 1.5046,
+      "step": 2429
+    },
+    {
+      "epoch": 0.11194545538305616,
+      "grad_norm": 1.2228552546931462,
+      "learning_rate": 0.00019650156251125673,
+      "loss": 0.9645,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11199152347169115,
+      "grad_norm": 1.5463641512091006,
+      "learning_rate": 0.00019649764929432988,
+      "loss": 1.0559,
+      "step": 2431
+    },
+    {
+      "epoch": 0.11203759156032617,
+      "grad_norm": 1.1496325562723209,
+      "learning_rate": 0.00019649373392904832,
+      "loss": 1.3866,
+      "step": 2432
+    },
+    {
+      "epoch": 0.11208365964896116,
+      "grad_norm": 1.066212105891343,
+      "learning_rate": 0.00019648981641549922,
+      "loss": 1.6043,
+      "step": 2433
+    },
+    {
+      "epoch": 0.11212972773759616,
+      "grad_norm": 1.0176584498710342,
+      "learning_rate": 0.0001964858967537698,
+      "loss": 1.729,
+      "step": 2434
+    },
+    {
+      "epoch": 0.11217579582623118,
+      "grad_norm": 1.3054142207905268,
+      "learning_rate": 0.0001964819749439473,
+      "loss": 1.5642,
+      "step": 2435
+    },
+    {
+      "epoch": 0.11222186391486617,
+      "grad_norm": 1.4710453337254552,
+      "learning_rate": 0.00019647805098611912,
+      "loss": 1.2065,
+      "step": 2436
+    },
+    {
+      "epoch": 0.11226793200350117,
+      "grad_norm": 1.6507133667162868,
+      "learning_rate": 0.0001964741248803725,
+      "loss": 1.465,
+      "step": 2437
+    },
+    {
+      "epoch": 0.11231400009213617,
+      "grad_norm": 2.711912276940597,
+      "learning_rate": 0.00019647019662679495,
+      "loss": 1.5222,
+      "step": 2438
+    },
+    {
+      "epoch": 0.11236006818077118,
+      "grad_norm": 0.9462737708612213,
+      "learning_rate": 0.0001964662662254739,
+      "loss": 1.552,
+      "step": 2439
+    },
+    {
+      "epoch": 0.11240613626940618,
+      "grad_norm": 1.5029206480252675,
+      "learning_rate": 0.00019646233367649676,
+      "loss": 1.326,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11245220435804118,
+      "grad_norm": 1.413411567375485,
+      "learning_rate": 0.0001964583989799512,
+      "loss": 1.4795,
+      "step": 2441
+    },
+    {
+      "epoch": 0.11249827244667619,
+      "grad_norm": 1.8894243468287575,
+      "learning_rate": 0.00019645446213592478,
+      "loss": 1.4954,
+      "step": 2442
+    },
+    {
+      "epoch": 0.11254434053531119,
+      "grad_norm": 2.573785241738933,
+      "learning_rate": 0.00019645052314450515,
+      "loss": 1.3072,
+      "step": 2443
+    },
+    {
+      "epoch": 0.11259040862394619,
+      "grad_norm": 1.0857723427956016,
+      "learning_rate": 0.00019644658200578002,
+      "loss": 1.4908,
+      "step": 2444
+    },
+    {
+      "epoch": 0.1126364767125812,
+      "grad_norm": 3.161029520035044,
+      "learning_rate": 0.00019644263871983707,
+      "loss": 1.5362,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1126825448012162,
+      "grad_norm": 2.3163827323618364,
+      "learning_rate": 0.00019643869328676413,
+      "loss": 1.1957,
+      "step": 2446
+    },
+    {
+      "epoch": 0.1127286128898512,
+      "grad_norm": 1.118677893865271,
+      "learning_rate": 0.00019643474570664907,
+      "loss": 1.3876,
+      "step": 2447
+    },
+    {
+      "epoch": 0.1127746809784862,
+      "grad_norm": 8.158098485577046,
+      "learning_rate": 0.00019643079597957974,
+      "loss": 1.1841,
+      "step": 2448
+    },
+    {
+      "epoch": 0.11282074906712121,
+      "grad_norm": 6.974100585706861,
+      "learning_rate": 0.00019642684410564406,
+      "loss": 1.8066,
+      "step": 2449
+    },
+    {
+      "epoch": 0.11286681715575621,
+      "grad_norm": 1.7457122277248633,
+      "learning_rate": 0.00019642289008493007,
+      "loss": 1.3138,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1129128852443912,
+      "grad_norm": 19.61754530671276,
+      "learning_rate": 0.00019641893391752574,
+      "loss": 1.4772,
+      "step": 2451
+    },
+    {
+      "epoch": 0.11295895333302622,
+      "grad_norm": 42.920368482267534,
+      "learning_rate": 0.00019641497560351917,
+      "loss": 5.0952,
+      "step": 2452
+    },
+    {
+      "epoch": 0.11300502142166122,
+      "grad_norm": 2.2549164252150757,
+      "learning_rate": 0.0001964110151429985,
+      "loss": 1.7786,
+      "step": 2453
+    },
+    {
+      "epoch": 0.11305108951029622,
+      "grad_norm": 8.429000213898954,
+      "learning_rate": 0.00019640705253605185,
+      "loss": 1.7689,
+      "step": 2454
+    },
+    {
+      "epoch": 0.11309715759893121,
+      "grad_norm": 4.6825937389159655,
+      "learning_rate": 0.00019640308778276752,
+      "loss": 2.0762,
+      "step": 2455
+    },
+    {
+      "epoch": 0.11314322568756623,
+      "grad_norm": 11.593369801051045,
+      "learning_rate": 0.0001963991208832337,
+      "loss": 2.3822,
+      "step": 2456
+    },
+    {
+      "epoch": 0.11318929377620122,
+      "grad_norm": 5.5067035905123705,
+      "learning_rate": 0.00019639515183753875,
+      "loss": 2.2618,
+      "step": 2457
+    },
+    {
+      "epoch": 0.11323536186483622,
+      "grad_norm": 5.832972983333566,
+      "learning_rate": 0.00019639118064577104,
+      "loss": 1.5852,
+      "step": 2458
+    },
+    {
+      "epoch": 0.11328142995347124,
+      "grad_norm": 5.802844958153406,
+      "learning_rate": 0.000196387207308019,
+      "loss": 1.5084,
+      "step": 2459
+    },
+    {
+      "epoch": 0.11332749804210623,
+      "grad_norm": 2.2737293497283932,
+      "learning_rate": 0.00019638323182437099,
+      "loss": 1.5201,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11337356613074123,
+      "grad_norm": 2.763253072084708,
+      "learning_rate": 0.0001963792541949156,
+      "loss": 1.5227,
+      "step": 2461
+    },
+    {
+      "epoch": 0.11341963421937624,
+      "grad_norm": 3.886090143539712,
+      "learning_rate": 0.0001963752744197414,
+      "loss": 1.7835,
+      "step": 2462
+    },
+    {
+      "epoch": 0.11346570230801124,
+      "grad_norm": 4.727982835850657,
+      "learning_rate": 0.00019637129249893694,
+      "loss": 1.7499,
+      "step": 2463
+    },
+    {
+      "epoch": 0.11351177039664624,
+      "grad_norm": 3.8144274318561835,
+      "learning_rate": 0.00019636730843259094,
+      "loss": 1.9332,
+      "step": 2464
+    },
+    {
+      "epoch": 0.11355783848528124,
+      "grad_norm": 3.3685390177600043,
+      "learning_rate": 0.00019636332222079198,
+      "loss": 1.7154,
+      "step": 2465
+    },
+    {
+      "epoch": 0.11360390657391625,
+      "grad_norm": 8.466463965295102,
+      "learning_rate": 0.00019635933386362892,
+      "loss": 1.7772,
+      "step": 2466
+    },
+    {
+      "epoch": 0.11364997466255125,
+      "grad_norm": 3.2007519132044795,
+      "learning_rate": 0.00019635534336119053,
+      "loss": 1.6296,
+      "step": 2467
+    },
+    {
+      "epoch": 0.11369604275118625,
+      "grad_norm": 4.25761233986753,
+      "learning_rate": 0.0001963513507135656,
+      "loss": 2.1526,
+      "step": 2468
+    },
+    {
+      "epoch": 0.11374211083982126,
+      "grad_norm": 3.642115046180779,
+      "learning_rate": 0.00019634735592084307,
+      "loss": 1.7378,
+      "step": 2469
+    },
+    {
+      "epoch": 0.11378817892845626,
+      "grad_norm": 3.006718975296261,
+      "learning_rate": 0.00019634335898311185,
+      "loss": 1.5762,
+      "step": 2470
+    },
+    {
+      "epoch": 0.11383424701709126,
+      "grad_norm": 3.1073631883603197,
+      "learning_rate": 0.00019633935990046096,
+      "loss": 1.6587,
+      "step": 2471
+    },
+    {
+      "epoch": 0.11388031510572626,
+      "grad_norm": 4.246559416170993,
+      "learning_rate": 0.00019633535867297937,
+      "loss": 1.6534,
+      "step": 2472
+    },
+    {
+      "epoch": 0.11392638319436127,
+      "grad_norm": 2.363627397542333,
+      "learning_rate": 0.00019633135530075624,
+      "loss": 1.7232,
+      "step": 2473
+    },
+    {
+      "epoch": 0.11397245128299627,
+      "grad_norm": 2.0536141369367624,
+      "learning_rate": 0.00019632734978388064,
+      "loss": 1.5824,
+      "step": 2474
+    },
+    {
+      "epoch": 0.11401851937163127,
+      "grad_norm": 3.521129014524405,
+      "learning_rate": 0.00019632334212244174,
+      "loss": 1.7895,
+      "step": 2475
+    },
+    {
+      "epoch": 0.11406458746026628,
+      "grad_norm": 2.5827528163816464,
+      "learning_rate": 0.0001963193323165288,
+      "loss": 1.6881,
+      "step": 2476
+    },
+    {
+      "epoch": 0.11411065554890128,
+      "grad_norm": 2.416123347854421,
+      "learning_rate": 0.00019631532036623107,
+      "loss": 1.6997,
+      "step": 2477
+    },
+    {
+      "epoch": 0.11415672363753628,
+      "grad_norm": 4.001096273107987,
+      "learning_rate": 0.00019631130627163792,
+      "loss": 1.3524,
+      "step": 2478
+    },
+    {
+      "epoch": 0.11420279172617129,
+      "grad_norm": 5.0177033177776815,
+      "learning_rate": 0.00019630729003283865,
+      "loss": 1.3886,
+      "step": 2479
+    },
+    {
+      "epoch": 0.11424885981480629,
+      "grad_norm": 4.908217592974506,
+      "learning_rate": 0.00019630327164992268,
+      "loss": 1.443,
+      "step": 2480
+    },
+    {
+      "epoch": 0.11429492790344128,
+      "grad_norm": 14.647405752103438,
+      "learning_rate": 0.00019629925112297952,
+      "loss": 1.5608,
+      "step": 2481
+    },
+    {
+      "epoch": 0.11434099599207628,
+      "grad_norm": 9.19750846696933,
+      "learning_rate": 0.00019629522845209863,
+      "loss": 1.4204,
+      "step": 2482
+    },
+    {
+      "epoch": 0.1143870640807113,
+      "grad_norm": 3.8220775468051467,
+      "learning_rate": 0.0001962912036373696,
+      "loss": 1.5786,
+      "step": 2483
+    },
+    {
+      "epoch": 0.1144331321693463,
+      "grad_norm": 5.854543422580773,
+      "learning_rate": 0.00019628717667888201,
+      "loss": 1.5222,
+      "step": 2484
+    },
+    {
+      "epoch": 0.11447920025798129,
+      "grad_norm": 4.008710929496249,
+      "learning_rate": 0.00019628314757672556,
+      "loss": 1.4195,
+      "step": 2485
+    },
+    {
+      "epoch": 0.1145252683466163,
+      "grad_norm": 4.863330507543313,
+      "learning_rate": 0.00019627911633098987,
+      "loss": 1.7666,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1145713364352513,
+      "grad_norm": 3.8139654737441178,
+      "learning_rate": 0.00019627508294176477,
+      "loss": 1.8653,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1146174045238863,
+      "grad_norm": 4.916938274543771,
+      "learning_rate": 0.00019627104740914004,
+      "loss": 1.3883,
+      "step": 2488
+    },
+    {
+      "epoch": 0.1146634726125213,
+      "grad_norm": 2.1565190006754897,
+      "learning_rate": 0.00019626700973320548,
+      "loss": 1.7166,
+      "step": 2489
+    },
+    {
+      "epoch": 0.11470954070115631,
+      "grad_norm": 2.7151625432047775,
+      "learning_rate": 0.000196262969914051,
+      "loss": 1.6765,
+      "step": 2490
+    },
+    {
+      "epoch": 0.11475560878979131,
+      "grad_norm": 2.7481860504245117,
+      "learning_rate": 0.00019625892795176658,
+      "loss": 1.4389,
+      "step": 2491
+    },
+    {
+      "epoch": 0.11480167687842631,
+      "grad_norm": 15.295508267489303,
+      "learning_rate": 0.0001962548838464422,
+      "loss": 1.2738,
+      "step": 2492
+    },
+    {
+      "epoch": 0.11484774496706132,
+      "grad_norm": 2.1608331251222985,
+      "learning_rate": 0.00019625083759816785,
+      "loss": 1.5578,
+      "step": 2493
+    },
+    {
+      "epoch": 0.11489381305569632,
+      "grad_norm": 2.9753695803959626,
+      "learning_rate": 0.00019624678920703362,
+      "loss": 1.8651,
+      "step": 2494
+    },
+    {
+      "epoch": 0.11493988114433132,
+      "grad_norm": 9.217342998133224,
+      "learning_rate": 0.00019624273867312967,
+      "loss": 1.6464,
+      "step": 2495
+    },
+    {
+      "epoch": 0.11498594923296633,
+      "grad_norm": 1.6690110503823385,
+      "learning_rate": 0.00019623868599654616,
+      "loss": 1.5632,
+      "step": 2496
+    },
+    {
+      "epoch": 0.11503201732160133,
+      "grad_norm": 5.907084206761236,
+      "learning_rate": 0.00019623463117737333,
+      "loss": 1.7937,
+      "step": 2497
+    },
+    {
+      "epoch": 0.11507808541023633,
+      "grad_norm": 3.4442406369780203,
+      "learning_rate": 0.00019623057421570142,
+      "loss": 1.7124,
+      "step": 2498
+    },
+    {
+      "epoch": 0.11512415349887133,
+      "grad_norm": 2.725898783940716,
+      "learning_rate": 0.0001962265151116208,
+      "loss": 1.6562,
+      "step": 2499
+    },
+    {
+      "epoch": 0.11517022158750634,
+      "grad_norm": 2.006923615502276,
+      "learning_rate": 0.0001962224538652218,
+      "loss": 1.7766,
+      "step": 2500
+    },
+    {
+      "epoch": 0.11521628967614134,
+      "grad_norm": 1.654579645963725,
+      "learning_rate": 0.00019621839047659484,
+      "loss": 1.6215,
+      "step": 2501
+    },
+    {
+      "epoch": 0.11526235776477634,
+      "grad_norm": 2.7283978780738516,
+      "learning_rate": 0.00019621432494583042,
+      "loss": 1.4607,
+      "step": 2502
+    },
+    {
+      "epoch": 0.11530842585341135,
+      "grad_norm": 1.7341235285263799,
+      "learning_rate": 0.00019621025727301902,
+      "loss": 1.6569,
+      "step": 2503
+    },
+    {
+      "epoch": 0.11535449394204635,
+      "grad_norm": 2.8546691742681762,
+      "learning_rate": 0.00019620618745825122,
+      "loss": 1.6549,
+      "step": 2504
+    },
+    {
+      "epoch": 0.11540056203068134,
+      "grad_norm": 2.2863222370872633,
+      "learning_rate": 0.00019620211550161758,
+      "loss": 1.7574,
+      "step": 2505
+    },
+    {
+      "epoch": 0.11544663011931634,
+      "grad_norm": 2.6210675860267525,
+      "learning_rate": 0.0001961980414032088,
+      "loss": 1.5049,
+      "step": 2506
+    },
+    {
+      "epoch": 0.11549269820795136,
+      "grad_norm": 1.9090314826542167,
+      "learning_rate": 0.00019619396516311558,
+      "loss": 1.4317,
+      "step": 2507
+    },
+    {
+      "epoch": 0.11553876629658635,
+      "grad_norm": 1.488167630937673,
+      "learning_rate": 0.00019618988678142866,
+      "loss": 1.7748,
+      "step": 2508
+    },
+    {
+      "epoch": 0.11558483438522135,
+      "grad_norm": 1.5819417019390676,
+      "learning_rate": 0.00019618580625823886,
+      "loss": 1.5536,
+      "step": 2509
+    },
+    {
+      "epoch": 0.11563090247385636,
+      "grad_norm": 3.606454717494447,
+      "learning_rate": 0.000196181723593637,
+      "loss": 1.5504,
+      "step": 2510
+    },
+    {
+      "epoch": 0.11567697056249136,
+      "grad_norm": 1.5301934022870494,
+      "learning_rate": 0.00019617763878771393,
+      "loss": 1.3689,
+      "step": 2511
+    },
+    {
+      "epoch": 0.11572303865112636,
+      "grad_norm": 2.612213936308485,
+      "learning_rate": 0.0001961735518405607,
+      "loss": 1.5041,
+      "step": 2512
+    },
+    {
+      "epoch": 0.11576910673976137,
+      "grad_norm": 3.4422512101197755,
+      "learning_rate": 0.00019616946275226825,
+      "loss": 1.3621,
+      "step": 2513
+    },
+    {
+      "epoch": 0.11581517482839637,
+      "grad_norm": 2.587333929860802,
+      "learning_rate": 0.00019616537152292757,
+      "loss": 1.5193,
+      "step": 2514
+    },
+    {
+      "epoch": 0.11586124291703137,
+      "grad_norm": 1.967400816384379,
+      "learning_rate": 0.00019616127815262976,
+      "loss": 1.4779,
+      "step": 2515
+    },
+    {
+      "epoch": 0.11590731100566637,
+      "grad_norm": 2.8326533261478906,
+      "learning_rate": 0.00019615718264146602,
+      "loss": 1.3658,
+      "step": 2516
+    },
+    {
+      "epoch": 0.11595337909430138,
+      "grad_norm": 2.8094027948059948,
+      "learning_rate": 0.00019615308498952748,
+      "loss": 1.7575,
+      "step": 2517
+    },
+    {
+      "epoch": 0.11599944718293638,
+      "grad_norm": 3.3986201917411165,
+      "learning_rate": 0.00019614898519690533,
+      "loss": 1.3201,
+      "step": 2518
+    },
+    {
+      "epoch": 0.11604551527157138,
+      "grad_norm": 4.0609184483867145,
+      "learning_rate": 0.00019614488326369093,
+      "loss": 1.3749,
+      "step": 2519
+    },
+    {
+      "epoch": 0.11609158336020639,
+      "grad_norm": 4.982961044540292,
+      "learning_rate": 0.00019614077918997556,
+      "loss": 1.1958,
+      "step": 2520
+    },
+    {
+      "epoch": 0.11613765144884139,
+      "grad_norm": 3.1077961458284324,
+      "learning_rate": 0.00019613667297585058,
+      "loss": 1.6036,
+      "step": 2521
+    },
+    {
+      "epoch": 0.11618371953747639,
+      "grad_norm": 3.8440686636814894,
+      "learning_rate": 0.0001961325646214074,
+      "loss": 1.6356,
+      "step": 2522
+    },
+    {
+      "epoch": 0.11622978762611139,
+      "grad_norm": 10.720374476930113,
+      "learning_rate": 0.0001961284541267375,
+      "loss": 1.2794,
+      "step": 2523
+    },
+    {
+      "epoch": 0.1162758557147464,
+      "grad_norm": 6.000246272477061,
+      "learning_rate": 0.00019612434149193238,
+      "loss": 1.5342,
+      "step": 2524
+    },
+    {
+      "epoch": 0.1163219238033814,
+      "grad_norm": 34.097854800253465,
+      "learning_rate": 0.00019612022671708362,
+      "loss": 1.3421,
+      "step": 2525
+    },
+    {
+      "epoch": 0.1163679918920164,
+      "grad_norm": 1.6274156227397965,
+      "learning_rate": 0.00019611610980228283,
+      "loss": 1.3908,
+      "step": 2526
+    },
+    {
+      "epoch": 0.11641405998065141,
+      "grad_norm": 3.7355764807139797,
+      "learning_rate": 0.00019611199074762167,
+      "loss": 1.4516,
+      "step": 2527
+    },
+    {
+      "epoch": 0.1164601280692864,
+      "grad_norm": 1.2733711484890184,
+      "learning_rate": 0.00019610786955319185,
+      "loss": 1.431,
+      "step": 2528
+    },
+    {
+      "epoch": 0.1165061961579214,
+      "grad_norm": 1.2155202913865575,
+      "learning_rate": 0.00019610374621908508,
+      "loss": 1.5889,
+      "step": 2529
+    },
+    {
+      "epoch": 0.11655226424655642,
+      "grad_norm": 1.352668659323126,
+      "learning_rate": 0.00019609962074539316,
+      "loss": 1.5759,
+      "step": 2530
+    },
+    {
+      "epoch": 0.11659833233519142,
+      "grad_norm": 0.9738249591868543,
+      "learning_rate": 0.00019609549313220797,
+      "loss": 1.7404,
+      "step": 2531
+    },
+    {
+      "epoch": 0.11664440042382641,
+      "grad_norm": 1.142433334065138,
+      "learning_rate": 0.00019609136337962144,
+      "loss": 1.4295,
+      "step": 2532
+    },
+    {
+      "epoch": 0.11669046851246141,
+      "grad_norm": 2.313172716070439,
+      "learning_rate": 0.00019608723148772543,
+      "loss": 1.4872,
+      "step": 2533
+    },
+    {
+      "epoch": 0.11673653660109642,
+      "grad_norm": 2.224275344610855,
+      "learning_rate": 0.00019608309745661195,
+      "loss": 1.623,
+      "step": 2534
+    },
+    {
+      "epoch": 0.11678260468973142,
+      "grad_norm": 1.6006045291975648,
+      "learning_rate": 0.00019607896128637307,
+      "loss": 1.1621,
+      "step": 2535
+    },
+    {
+      "epoch": 0.11682867277836642,
+      "grad_norm": 1.2118511757445833,
+      "learning_rate": 0.00019607482297710088,
+      "loss": 1.6344,
+      "step": 2536
+    },
+    {
+      "epoch": 0.11687474086700143,
+      "grad_norm": 5.845203525562706,
+      "learning_rate": 0.00019607068252888745,
+      "loss": 1.5858,
+      "step": 2537
+    },
+    {
+      "epoch": 0.11692080895563643,
+      "grad_norm": 1.313470231241952,
+      "learning_rate": 0.00019606653994182503,
+      "loss": 1.399,
+      "step": 2538
+    },
+    {
+      "epoch": 0.11696687704427143,
+      "grad_norm": 2.466940586616623,
+      "learning_rate": 0.0001960623952160058,
+      "loss": 1.5112,
+      "step": 2539
+    },
+    {
+      "epoch": 0.11701294513290643,
+      "grad_norm": 1.507291897321067,
+      "learning_rate": 0.00019605824835152203,
+      "loss": 1.3761,
+      "step": 2540
+    },
+    {
+      "epoch": 0.11705901322154144,
+      "grad_norm": 4.306765531616128,
+      "learning_rate": 0.00019605409934846612,
+      "loss": 1.5432,
+      "step": 2541
+    },
+    {
+      "epoch": 0.11710508131017644,
+      "grad_norm": 8.648621511881935,
+      "learning_rate": 0.00019604994820693035,
+      "loss": 1.3531,
+      "step": 2542
+    },
+    {
+      "epoch": 0.11715114939881144,
+      "grad_norm": 4.232382338196365,
+      "learning_rate": 0.0001960457949270072,
+      "loss": 1.3154,
+      "step": 2543
+    },
+    {
+      "epoch": 0.11719721748744645,
+      "grad_norm": 3.0358607874805035,
+      "learning_rate": 0.00019604163950878907,
+      "loss": 1.6334,
+      "step": 2544
+    },
+    {
+      "epoch": 0.11724328557608145,
+      "grad_norm": 4.4856775272831575,
+      "learning_rate": 0.00019603748195236854,
+      "loss": 1.4026,
+      "step": 2545
+    },
+    {
+      "epoch": 0.11728935366471645,
+      "grad_norm": 4.882983574284761,
+      "learning_rate": 0.00019603332225783812,
+      "loss": 1.3454,
+      "step": 2546
+    },
+    {
+      "epoch": 0.11733542175335146,
+      "grad_norm": 12.452624193260878,
+      "learning_rate": 0.00019602916042529045,
+      "loss": 1.3909,
+      "step": 2547
+    },
+    {
+      "epoch": 0.11738148984198646,
+      "grad_norm": 3.0303108135140238,
+      "learning_rate": 0.0001960249964548182,
+      "loss": 1.4087,
+      "step": 2548
+    },
+    {
+      "epoch": 0.11742755793062146,
+      "grad_norm": 2.729055995704955,
+      "learning_rate": 0.00019602083034651402,
+      "loss": 1.423,
+      "step": 2549
+    },
+    {
+      "epoch": 0.11747362601925646,
+      "grad_norm": 4.396345880780402,
+      "learning_rate": 0.00019601666210047072,
+      "loss": 1.5118,
+      "step": 2550
+    },
+    {
+      "epoch": 0.11751969410789147,
+      "grad_norm": 1.5231634239886997,
+      "learning_rate": 0.00019601249171678105,
+      "loss": 1.8516,
+      "step": 2551
+    },
+    {
+      "epoch": 0.11756576219652647,
+      "grad_norm": 4.221691092295368,
+      "learning_rate": 0.00019600831919553788,
+      "loss": 1.2398,
+      "step": 2552
+    },
+    {
+      "epoch": 0.11761183028516146,
+      "grad_norm": 3.231064652831053,
+      "learning_rate": 0.00019600414453683414,
+      "loss": 1.6611,
+      "step": 2553
+    },
+    {
+      "epoch": 0.11765789837379648,
+      "grad_norm": 1.9203509352206893,
+      "learning_rate": 0.0001959999677407627,
+      "loss": 1.555,
+      "step": 2554
+    },
+    {
+      "epoch": 0.11770396646243148,
+      "grad_norm": 2.725367927259495,
+      "learning_rate": 0.00019599578880741657,
+      "loss": 1.625,
+      "step": 2555
+    },
+    {
+      "epoch": 0.11775003455106647,
+      "grad_norm": 1.762069773109157,
+      "learning_rate": 0.0001959916077368888,
+      "loss": 1.7318,
+      "step": 2556
+    },
+    {
+      "epoch": 0.11779610263970147,
+      "grad_norm": 2.142829761026809,
+      "learning_rate": 0.0001959874245292725,
+      "loss": 1.6588,
+      "step": 2557
+    },
+    {
+      "epoch": 0.11784217072833648,
+      "grad_norm": 1.1985509998822665,
+      "learning_rate": 0.00019598323918466076,
+      "loss": 1.6295,
+      "step": 2558
+    },
+    {
+      "epoch": 0.11788823881697148,
+      "grad_norm": 1.6407889606021226,
+      "learning_rate": 0.0001959790517031468,
+      "loss": 1.4084,
+      "step": 2559
+    },
+    {
+      "epoch": 0.11793430690560648,
+      "grad_norm": 2.5809539196499687,
+      "learning_rate": 0.00019597486208482382,
+      "loss": 1.2984,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1179803749942415,
+      "grad_norm": 1.525848241798921,
+      "learning_rate": 0.00019597067032978504,
+      "loss": 1.5314,
+      "step": 2561
+    },
+    {
+      "epoch": 0.11802644308287649,
+      "grad_norm": 1.6206790308118766,
+      "learning_rate": 0.0001959664764381239,
+      "loss": 1.5471,
+      "step": 2562
+    },
+    {
+      "epoch": 0.11807251117151149,
+      "grad_norm": 2.14409123861202,
+      "learning_rate": 0.0001959622804099337,
+      "loss": 1.5936,
+      "step": 2563
+    },
+    {
+      "epoch": 0.1181185792601465,
+      "grad_norm": 1.6288419563863707,
+      "learning_rate": 0.00019595808224530785,
+      "loss": 1.5976,
+      "step": 2564
+    },
+    {
+      "epoch": 0.1181646473487815,
+      "grad_norm": 1.421478410954294,
+      "learning_rate": 0.00019595388194433984,
+      "loss": 1.6935,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1182107154374165,
+      "grad_norm": 1.110869513523423,
+      "learning_rate": 0.00019594967950712318,
+      "loss": 1.3835,
+      "step": 2566
+    },
+    {
+      "epoch": 0.1182567835260515,
+      "grad_norm": 1.774443121975387,
+      "learning_rate": 0.0001959454749337514,
+      "loss": 1.3109,
+      "step": 2567
+    },
+    {
+      "epoch": 0.11830285161468651,
+      "grad_norm": 1.9561872949786843,
+      "learning_rate": 0.00019594126822431815,
+      "loss": 1.5121,
+      "step": 2568
+    },
+    {
+      "epoch": 0.11834891970332151,
+      "grad_norm": 1.5056415790594766,
+      "learning_rate": 0.00019593705937891707,
+      "loss": 1.6543,
+      "step": 2569
+    },
+    {
+      "epoch": 0.11839498779195651,
+      "grad_norm": 2.2745677560637794,
+      "learning_rate": 0.00019593284839764186,
+      "loss": 1.6946,
+      "step": 2570
+    },
+    {
+      "epoch": 0.11844105588059152,
+      "grad_norm": 1.758750836847383,
+      "learning_rate": 0.00019592863528058625,
+      "loss": 1.4105,
+      "step": 2571
+    },
+    {
+      "epoch": 0.11848712396922652,
+      "grad_norm": 1.7927026858547095,
+      "learning_rate": 0.0001959244200278441,
+      "loss": 1.5659,
+      "step": 2572
+    },
+    {
+      "epoch": 0.11853319205786152,
+      "grad_norm": 2.156211048513434,
+      "learning_rate": 0.00019592020263950918,
+      "loss": 1.4571,
+      "step": 2573
+    },
+    {
+      "epoch": 0.11857926014649652,
+      "grad_norm": 1.671891557823231,
+      "learning_rate": 0.00019591598311567544,
+      "loss": 1.5612,
+      "step": 2574
+    },
+    {
+      "epoch": 0.11862532823513153,
+      "grad_norm": 2.920988844235431,
+      "learning_rate": 0.00019591176145643675,
+      "loss": 1.3445,
+      "step": 2575
+    },
+    {
+      "epoch": 0.11867139632376653,
+      "grad_norm": 3.7108717145898042,
+      "learning_rate": 0.00019590753766188717,
+      "loss": 1.4385,
+      "step": 2576
+    },
+    {
+      "epoch": 0.11871746441240152,
+      "grad_norm": 2.024792917590775,
+      "learning_rate": 0.00019590331173212073,
+      "loss": 1.364,
+      "step": 2577
+    },
+    {
+      "epoch": 0.11876353250103654,
+      "grad_norm": 1.8225787700743306,
+      "learning_rate": 0.00019589908366723146,
+      "loss": 1.3325,
+      "step": 2578
+    },
+    {
+      "epoch": 0.11880960058967154,
+      "grad_norm": 2.2702742571032903,
+      "learning_rate": 0.00019589485346731356,
+      "loss": 1.3571,
+      "step": 2579
+    },
+    {
+      "epoch": 0.11885566867830653,
+      "grad_norm": 3.5518708713738874,
+      "learning_rate": 0.00019589062113246116,
+      "loss": 1.5277,
+      "step": 2580
+    },
+    {
+      "epoch": 0.11890173676694155,
+      "grad_norm": 1.1582766441621255,
+      "learning_rate": 0.0001958863866627685,
+      "loss": 1.2224,
+      "step": 2581
+    },
+    {
+      "epoch": 0.11894780485557654,
+      "grad_norm": 1.4267745430711811,
+      "learning_rate": 0.0001958821500583298,
+      "loss": 1.6528,
+      "step": 2582
+    },
+    {
+      "epoch": 0.11899387294421154,
+      "grad_norm": 2.0847098992645656,
+      "learning_rate": 0.00019587791131923948,
+      "loss": 0.9796,
+      "step": 2583
+    },
+    {
+      "epoch": 0.11903994103284654,
+      "grad_norm": 1.2521648298595784,
+      "learning_rate": 0.00019587367044559187,
+      "loss": 1.6018,
+      "step": 2584
+    },
+    {
+      "epoch": 0.11908600912148155,
+      "grad_norm": 2.990595224738499,
+      "learning_rate": 0.00019586942743748137,
+      "loss": 1.3042,
+      "step": 2585
+    },
+    {
+      "epoch": 0.11913207721011655,
+      "grad_norm": 1.0875836711069482,
+      "learning_rate": 0.00019586518229500243,
+      "loss": 1.5221,
+      "step": 2586
+    },
+    {
+      "epoch": 0.11917814529875155,
+      "grad_norm": 2.795074323029351,
+      "learning_rate": 0.00019586093501824958,
+      "loss": 1.3838,
+      "step": 2587
+    },
+    {
+      "epoch": 0.11922421338738656,
+      "grad_norm": 1.7841332755767172,
+      "learning_rate": 0.00019585668560731737,
+      "loss": 1.4868,
+      "step": 2588
+    },
+    {
+      "epoch": 0.11927028147602156,
+      "grad_norm": 1.0516133959907463,
+      "learning_rate": 0.00019585243406230042,
+      "loss": 1.6935,
+      "step": 2589
+    },
+    {
+      "epoch": 0.11931634956465656,
+      "grad_norm": 1.4540445571175546,
+      "learning_rate": 0.0001958481803832934,
+      "loss": 1.4403,
+      "step": 2590
+    },
+    {
+      "epoch": 0.11936241765329156,
+      "grad_norm": 1.0480077613250958,
+      "learning_rate": 0.000195843924570391,
+      "loss": 1.0877,
+      "step": 2591
+    },
+    {
+      "epoch": 0.11940848574192657,
+      "grad_norm": 1.126139812441366,
+      "learning_rate": 0.00019583966662368792,
+      "loss": 1.5091,
+      "step": 2592
+    },
+    {
+      "epoch": 0.11945455383056157,
+      "grad_norm": 1.9074266855669286,
+      "learning_rate": 0.000195835406543279,
+      "loss": 1.4626,
+      "step": 2593
+    },
+    {
+      "epoch": 0.11950062191919657,
+      "grad_norm": 2.206230676457771,
+      "learning_rate": 0.0001958311443292591,
+      "loss": 1.4052,
+      "step": 2594
+    },
+    {
+      "epoch": 0.11954669000783158,
+      "grad_norm": 1.8434597571246152,
+      "learning_rate": 0.00019582687998172303,
+      "loss": 1.4631,
+      "step": 2595
+    },
+    {
+      "epoch": 0.11959275809646658,
+      "grad_norm": 1.076205421749008,
+      "learning_rate": 0.00019582261350076582,
+      "loss": 1.3748,
+      "step": 2596
+    },
+    {
+      "epoch": 0.11963882618510158,
+      "grad_norm": 1.199171314028877,
+      "learning_rate": 0.00019581834488648247,
+      "loss": 1.3994,
+      "step": 2597
+    },
+    {
+      "epoch": 0.11968489427373659,
+      "grad_norm": 3.6662092005853575,
+      "learning_rate": 0.00019581407413896788,
+      "loss": 1.3928,
+      "step": 2598
+    },
+    {
+      "epoch": 0.11973096236237159,
+      "grad_norm": 2.064409150833345,
+      "learning_rate": 0.00019580980125831726,
+      "loss": 1.6307,
+      "step": 2599
+    },
+    {
+      "epoch": 0.11977703045100659,
+      "grad_norm": 2.101561136570377,
+      "learning_rate": 0.00019580552624462566,
+      "loss": 1.4608,
+      "step": 2600
+    },
+    {
+      "epoch": 0.11982309853964158,
+      "grad_norm": 1.4228776381251576,
+      "learning_rate": 0.00019580124909798833,
+      "loss": 1.5321,
+      "step": 2601
+    },
+    {
+      "epoch": 0.1198691666282766,
+      "grad_norm": 2.271600132237475,
+      "learning_rate": 0.00019579696981850044,
+      "loss": 1.3831,
+      "step": 2602
+    },
+    {
+      "epoch": 0.1199152347169116,
+      "grad_norm": 3.898076253539868,
+      "learning_rate": 0.00019579268840625726,
+      "loss": 1.1775,
+      "step": 2603
+    },
+    {
+      "epoch": 0.1199613028055466,
+      "grad_norm": 3.2947590398542226,
+      "learning_rate": 0.00019578840486135415,
+      "loss": 1.3007,
+      "step": 2604
+    },
+    {
+      "epoch": 0.1200073708941816,
+      "grad_norm": 1.575417164517264,
+      "learning_rate": 0.00019578411918388643,
+      "loss": 1.6824,
+      "step": 2605
+    },
+    {
+      "epoch": 0.1200534389828166,
+      "grad_norm": 1.5625154975012419,
+      "learning_rate": 0.00019577983137394954,
+      "loss": 1.539,
+      "step": 2606
+    },
+    {
+      "epoch": 0.1200995070714516,
+      "grad_norm": 1.1510561713624041,
+      "learning_rate": 0.00019577554143163892,
+      "loss": 1.517,
+      "step": 2607
+    },
+    {
+      "epoch": 0.1201455751600866,
+      "grad_norm": 1.2648139775862806,
+      "learning_rate": 0.00019577124935705008,
+      "loss": 1.5205,
+      "step": 2608
+    },
+    {
+      "epoch": 0.12019164324872161,
+      "grad_norm": 1.5811921341237407,
+      "learning_rate": 0.0001957669551502786,
+      "loss": 1.67,
+      "step": 2609
+    },
+    {
+      "epoch": 0.12023771133735661,
+      "grad_norm": 1.4128600568518537,
+      "learning_rate": 0.0001957626588114201,
+      "loss": 1.3981,
+      "step": 2610
+    },
+    {
+      "epoch": 0.12028377942599161,
+      "grad_norm": 2.1975368676095415,
+      "learning_rate": 0.00019575836034057013,
+      "loss": 1.5253,
+      "step": 2611
+    },
+    {
+      "epoch": 0.12032984751462662,
+      "grad_norm": 2.214724545344047,
+      "learning_rate": 0.0001957540597378245,
+      "loss": 1.4823,
+      "step": 2612
+    },
+    {
+      "epoch": 0.12037591560326162,
+      "grad_norm": 2.1834656333036673,
+      "learning_rate": 0.0001957497570032789,
+      "loss": 1.0914,
+      "step": 2613
+    },
+    {
+      "epoch": 0.12042198369189662,
+      "grad_norm": 1.4409914017429142,
+      "learning_rate": 0.00019574545213702914,
+      "loss": 1.3872,
+      "step": 2614
+    },
+    {
+      "epoch": 0.12046805178053163,
+      "grad_norm": 1.586537456051479,
+      "learning_rate": 0.0001957411451391711,
+      "loss": 1.3843,
+      "step": 2615
+    },
+    {
+      "epoch": 0.12051411986916663,
+      "grad_norm": 1.5617790246825574,
+      "learning_rate": 0.00019573683600980054,
+      "loss": 1.1746,
+      "step": 2616
+    },
+    {
+      "epoch": 0.12056018795780163,
+      "grad_norm": 2.1890283873640466,
+      "learning_rate": 0.00019573252474901353,
+      "loss": 1.1813,
+      "step": 2617
+    },
+    {
+      "epoch": 0.12060625604643663,
+      "grad_norm": 1.7584381961453663,
+      "learning_rate": 0.000195728211356906,
+      "loss": 1.3836,
+      "step": 2618
+    },
+    {
+      "epoch": 0.12065232413507164,
+      "grad_norm": 1.508539234728666,
+      "learning_rate": 0.000195723895833574,
+      "loss": 1.4662,
+      "step": 2619
+    },
+    {
+      "epoch": 0.12069839222370664,
+      "grad_norm": 2.292587331727649,
+      "learning_rate": 0.00019571957817911354,
+      "loss": 1.2405,
+      "step": 2620
+    },
+    {
+      "epoch": 0.12074446031234164,
+      "grad_norm": 1.8909178544858063,
+      "learning_rate": 0.00019571525839362085,
+      "loss": 1.2217,
+      "step": 2621
+    },
+    {
+      "epoch": 0.12079052840097665,
+      "grad_norm": 1.5198846014422225,
+      "learning_rate": 0.000195710936477192,
+      "loss": 1.2531,
+      "step": 2622
+    },
+    {
+      "epoch": 0.12083659648961165,
+      "grad_norm": 2.0529235880959926,
+      "learning_rate": 0.0001957066124299233,
+      "loss": 1.7373,
+      "step": 2623
+    },
+    {
+      "epoch": 0.12088266457824665,
+      "grad_norm": 1.1607753740062103,
+      "learning_rate": 0.00019570228625191095,
+      "loss": 1.5823,
+      "step": 2624
+    },
+    {
+      "epoch": 0.12092873266688164,
+      "grad_norm": 1.1126439705913613,
+      "learning_rate": 0.0001956979579432513,
+      "loss": 1.2367,
+      "step": 2625
+    },
+    {
+      "epoch": 0.12097480075551666,
+      "grad_norm": 1.8935073550826607,
+      "learning_rate": 0.0001956936275040407,
+      "loss": 1.7998,
+      "step": 2626
+    },
+    {
+      "epoch": 0.12102086884415166,
+      "grad_norm": 2.0972569489795023,
+      "learning_rate": 0.00019568929493437556,
+      "loss": 1.3905,
+      "step": 2627
+    },
+    {
+      "epoch": 0.12106693693278665,
+      "grad_norm": 3.135204347544434,
+      "learning_rate": 0.00019568496023435238,
+      "loss": 1.3648,
+      "step": 2628
+    },
+    {
+      "epoch": 0.12111300502142167,
+      "grad_norm": 1.4725242776662628,
+      "learning_rate": 0.00019568062340406757,
+      "loss": 1.5093,
+      "step": 2629
+    },
+    {
+      "epoch": 0.12115907311005666,
+      "grad_norm": 1.3912267440853812,
+      "learning_rate": 0.00019567628444361774,
+      "loss": 1.4984,
+      "step": 2630
+    },
+    {
+      "epoch": 0.12120514119869166,
+      "grad_norm": 1.475879864015986,
+      "learning_rate": 0.0001956719433530995,
+      "loss": 1.5926,
+      "step": 2631
+    },
+    {
+      "epoch": 0.12125120928732668,
+      "grad_norm": 1.648102898075821,
+      "learning_rate": 0.0001956676001326095,
+      "loss": 1.7146,
+      "step": 2632
+    },
+    {
+      "epoch": 0.12129727737596167,
+      "grad_norm": 2.999356650053094,
+      "learning_rate": 0.00019566325478224437,
+      "loss": 1.2583,
+      "step": 2633
+    },
+    {
+      "epoch": 0.12134334546459667,
+      "grad_norm": 3.018524033874648,
+      "learning_rate": 0.00019565890730210097,
+      "loss": 1.5597,
+      "step": 2634
+    },
+    {
+      "epoch": 0.12138941355323167,
+      "grad_norm": 1.726101559316234,
+      "learning_rate": 0.00019565455769227598,
+      "loss": 1.1545,
+      "step": 2635
+    },
+    {
+      "epoch": 0.12143548164186668,
+      "grad_norm": 1.7358747872840814,
+      "learning_rate": 0.0001956502059528663,
+      "loss": 1.4175,
+      "step": 2636
+    },
+    {
+      "epoch": 0.12148154973050168,
+      "grad_norm": 1.1326211068423047,
+      "learning_rate": 0.00019564585208396875,
+      "loss": 1.6493,
+      "step": 2637
+    },
+    {
+      "epoch": 0.12152761781913668,
+      "grad_norm": 2.1068277819902788,
+      "learning_rate": 0.00019564149608568033,
+      "loss": 1.2682,
+      "step": 2638
+    },
+    {
+      "epoch": 0.12157368590777169,
+      "grad_norm": 2.842578077566498,
+      "learning_rate": 0.00019563713795809796,
+      "loss": 1.5307,
+      "step": 2639
+    },
+    {
+      "epoch": 0.12161975399640669,
+      "grad_norm": 1.7731506052148565,
+      "learning_rate": 0.0001956327777013187,
+      "loss": 1.1384,
+      "step": 2640
+    },
+    {
+      "epoch": 0.12166582208504169,
+      "grad_norm": 1.8465828464408098,
+      "learning_rate": 0.00019562841531543966,
+      "loss": 1.4829,
+      "step": 2641
+    },
+    {
+      "epoch": 0.12171189017367669,
+      "grad_norm": 2.65337096115276,
+      "learning_rate": 0.0001956240508005579,
+      "loss": 1.1255,
+      "step": 2642
+    },
+    {
+      "epoch": 0.1217579582623117,
+      "grad_norm": 1.813673624715679,
+      "learning_rate": 0.00019561968415677058,
+      "loss": 1.3856,
+      "step": 2643
+    },
+    {
+      "epoch": 0.1218040263509467,
+      "grad_norm": 1.2280038320010813,
+      "learning_rate": 0.00019561531538417497,
+      "loss": 1.6957,
+      "step": 2644
+    },
+    {
+      "epoch": 0.1218500944395817,
+      "grad_norm": 1.1033582112055929,
+      "learning_rate": 0.0001956109444828683,
+      "loss": 1.7303,
+      "step": 2645
+    },
+    {
+      "epoch": 0.12189616252821671,
+      "grad_norm": 1.2857462656995264,
+      "learning_rate": 0.0001956065714529479,
+      "loss": 1.2366,
+      "step": 2646
+    },
+    {
+      "epoch": 0.12194223061685171,
+      "grad_norm": 2.312279094607007,
+      "learning_rate": 0.0001956021962945111,
+      "loss": 1.4835,
+      "step": 2647
+    },
+    {
+      "epoch": 0.1219882987054867,
+      "grad_norm": 2.436145463853769,
+      "learning_rate": 0.00019559781900765536,
+      "loss": 1.3266,
+      "step": 2648
+    },
+    {
+      "epoch": 0.12203436679412172,
+      "grad_norm": 1.596134336387999,
+      "learning_rate": 0.00019559343959247807,
+      "loss": 1.3912,
+      "step": 2649
+    },
+    {
+      "epoch": 0.12208043488275672,
+      "grad_norm": 3.021642242245169,
+      "learning_rate": 0.00019558905804907675,
+      "loss": 1.2316,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12212650297139171,
+      "grad_norm": 2.229458800438505,
+      "learning_rate": 0.00019558467437754895,
+      "loss": 1.4863,
+      "step": 2651
+    },
+    {
+      "epoch": 0.12217257106002671,
+      "grad_norm": 6.893842951218711,
+      "learning_rate": 0.00019558028857799225,
+      "loss": 1.2301,
+      "step": 2652
+    },
+    {
+      "epoch": 0.12221863914866173,
+      "grad_norm": 1.9800819624641932,
+      "learning_rate": 0.00019557590065050437,
+      "loss": 1.3659,
+      "step": 2653
+    },
+    {
+      "epoch": 0.12226470723729672,
+      "grad_norm": 2.385123103531126,
+      "learning_rate": 0.00019557151059518288,
+      "loss": 1.2023,
+      "step": 2654
+    },
+    {
+      "epoch": 0.12231077532593172,
+      "grad_norm": 1.9783305208601218,
+      "learning_rate": 0.00019556711841212562,
+      "loss": 1.405,
+      "step": 2655
+    },
+    {
+      "epoch": 0.12235684341456673,
+      "grad_norm": 9.789662525489085,
+      "learning_rate": 0.0001955627241014303,
+      "loss": 1.7106,
+      "step": 2656
+    },
+    {
+      "epoch": 0.12240291150320173,
+      "grad_norm": 4.176973067681562,
+      "learning_rate": 0.0001955583276631948,
+      "loss": 1.2466,
+      "step": 2657
+    },
+    {
+      "epoch": 0.12244897959183673,
+      "grad_norm": 2.588221134995953,
+      "learning_rate": 0.00019555392909751697,
+      "loss": 1.383,
+      "step": 2658
+    },
+    {
+      "epoch": 0.12249504768047173,
+      "grad_norm": 1.7836506564679053,
+      "learning_rate": 0.00019554952840449477,
+      "loss": 1.8516,
+      "step": 2659
+    },
+    {
+      "epoch": 0.12254111576910674,
+      "grad_norm": 3.6535679495391733,
+      "learning_rate": 0.00019554512558422614,
+      "loss": 1.2808,
+      "step": 2660
+    },
+    {
+      "epoch": 0.12258718385774174,
+      "grad_norm": 2.2686495609078094,
+      "learning_rate": 0.00019554072063680908,
+      "loss": 1.5519,
+      "step": 2661
+    },
+    {
+      "epoch": 0.12263325194637674,
+      "grad_norm": 3.6329970632349613,
+      "learning_rate": 0.00019553631356234173,
+      "loss": 1.5715,
+      "step": 2662
+    },
+    {
+      "epoch": 0.12267932003501175,
+      "grad_norm": 1.8494645224747603,
+      "learning_rate": 0.00019553190436092218,
+      "loss": 1.6628,
+      "step": 2663
+    },
+    {
+      "epoch": 0.12272538812364675,
+      "grad_norm": 3.8806415329931334,
+      "learning_rate": 0.00019552749303264853,
+      "loss": 1.1147,
+      "step": 2664
+    },
+    {
+      "epoch": 0.12277145621228175,
+      "grad_norm": 2.175878766892258,
+      "learning_rate": 0.00019552307957761912,
+      "loss": 1.3701,
+      "step": 2665
+    },
+    {
+      "epoch": 0.12281752430091676,
+      "grad_norm": 2.163804988724653,
+      "learning_rate": 0.00019551866399593207,
+      "loss": 1.4924,
+      "step": 2666
+    },
+    {
+      "epoch": 0.12286359238955176,
+      "grad_norm": 5.343808035141412,
+      "learning_rate": 0.00019551424628768576,
+      "loss": 1.5163,
+      "step": 2667
+    },
+    {
+      "epoch": 0.12290966047818676,
+      "grad_norm": 1.892362915647032,
+      "learning_rate": 0.00019550982645297852,
+      "loss": 1.632,
+      "step": 2668
+    },
+    {
+      "epoch": 0.12295572856682176,
+      "grad_norm": 2.7291815723892077,
+      "learning_rate": 0.00019550540449190878,
+      "loss": 1.425,
+      "step": 2669
+    },
+    {
+      "epoch": 0.12300179665545677,
+      "grad_norm": 2.7090698939238416,
+      "learning_rate": 0.00019550098040457492,
+      "loss": 1.6086,
+      "step": 2670
+    },
+    {
+      "epoch": 0.12304786474409177,
+      "grad_norm": 1.0772606277176762,
+      "learning_rate": 0.00019549655419107554,
+      "loss": 1.4501,
+      "step": 2671
+    },
+    {
+      "epoch": 0.12309393283272677,
+      "grad_norm": 1.6411819232751261,
+      "learning_rate": 0.00019549212585150908,
+      "loss": 1.4164,
+      "step": 2672
+    },
+    {
+      "epoch": 0.12314000092136178,
+      "grad_norm": 2.235226228644429,
+      "learning_rate": 0.0001954876953859742,
+      "loss": 1.4796,
+      "step": 2673
+    },
+    {
+      "epoch": 0.12318606900999678,
+      "grad_norm": 2.026435263820993,
+      "learning_rate": 0.00019548326279456948,
+      "loss": 1.518,
+      "step": 2674
+    },
+    {
+      "epoch": 0.12323213709863177,
+      "grad_norm": 1.273960687113167,
+      "learning_rate": 0.00019547882807739364,
+      "loss": 1.5373,
+      "step": 2675
+    },
+    {
+      "epoch": 0.12327820518726679,
+      "grad_norm": 1.870558587412234,
+      "learning_rate": 0.00019547439123454542,
+      "loss": 1.4136,
+      "step": 2676
+    },
+    {
+      "epoch": 0.12332427327590179,
+      "grad_norm": 1.5453732025653117,
+      "learning_rate": 0.0001954699522661236,
+      "loss": 1.4234,
+      "step": 2677
+    },
+    {
+      "epoch": 0.12337034136453678,
+      "grad_norm": 1.2196072462834393,
+      "learning_rate": 0.00019546551117222697,
+      "loss": 1.6378,
+      "step": 2678
+    },
+    {
+      "epoch": 0.12341640945317178,
+      "grad_norm": 1.3038129923705206,
+      "learning_rate": 0.0001954610679529544,
+      "loss": 1.5512,
+      "step": 2679
+    },
+    {
+      "epoch": 0.1234624775418068,
+      "grad_norm": 1.2652826944595357,
+      "learning_rate": 0.00019545662260840482,
+      "loss": 1.7583,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1235085456304418,
+      "grad_norm": 1.2475224023504323,
+      "learning_rate": 0.00019545217513867725,
+      "loss": 1.1037,
+      "step": 2681
+    },
+    {
+      "epoch": 0.12355461371907679,
+      "grad_norm": 1.299188840864384,
+      "learning_rate": 0.00019544772554387066,
+      "loss": 1.2076,
+      "step": 2682
+    },
+    {
+      "epoch": 0.1236006818077118,
+      "grad_norm": 2.512494397925874,
+      "learning_rate": 0.00019544327382408413,
+      "loss": 1.2167,
+      "step": 2683
+    },
+    {
+      "epoch": 0.1236467498963468,
+      "grad_norm": 1.9724989424462145,
+      "learning_rate": 0.00019543881997941673,
+      "loss": 1.2467,
+      "step": 2684
+    },
+    {
+      "epoch": 0.1236928179849818,
+      "grad_norm": 1.0927530450213994,
+      "learning_rate": 0.00019543436400996768,
+      "loss": 1.4969,
+      "step": 2685
+    },
+    {
+      "epoch": 0.1237388860736168,
+      "grad_norm": 1.1507868024036645,
+      "learning_rate": 0.0001954299059158361,
+      "loss": 1.5205,
+      "step": 2686
+    },
+    {
+      "epoch": 0.12378495416225181,
+      "grad_norm": 1.2811468153819614,
+      "learning_rate": 0.00019542544569712132,
+      "loss": 1.2832,
+      "step": 2687
+    },
+    {
+      "epoch": 0.12383102225088681,
+      "grad_norm": 2.268748887627783,
+      "learning_rate": 0.0001954209833539226,
+      "loss": 1.2828,
+      "step": 2688
+    },
+    {
+      "epoch": 0.12387709033952181,
+      "grad_norm": 1.381655576996567,
+      "learning_rate": 0.00019541651888633932,
+      "loss": 1.5628,
+      "step": 2689
+    },
+    {
+      "epoch": 0.12392315842815682,
+      "grad_norm": 1.9711368086605066,
+      "learning_rate": 0.00019541205229447082,
+      "loss": 1.6192,
+      "step": 2690
+    },
+    {
+      "epoch": 0.12396922651679182,
+      "grad_norm": 1.2634821952368964,
+      "learning_rate": 0.0001954075835784166,
+      "loss": 1.77,
+      "step": 2691
+    },
+    {
+      "epoch": 0.12401529460542682,
+      "grad_norm": 6.468747899206288,
+      "learning_rate": 0.00019540311273827612,
+      "loss": 1.2802,
+      "step": 2692
+    },
+    {
+      "epoch": 0.12406136269406183,
+      "grad_norm": 1.8265710374399935,
+      "learning_rate": 0.0001953986397741489,
+      "loss": 1.3668,
+      "step": 2693
+    },
+    {
+      "epoch": 0.12410743078269683,
+      "grad_norm": 2.994366770610838,
+      "learning_rate": 0.00019539416468613454,
+      "loss": 1.5709,
+      "step": 2694
+    },
+    {
+      "epoch": 0.12415349887133183,
+      "grad_norm": 1.1786967965318114,
+      "learning_rate": 0.0001953896874743327,
+      "loss": 1.3535,
+      "step": 2695
+    },
+    {
+      "epoch": 0.12419956695996683,
+      "grad_norm": 2.7808770074633893,
+      "learning_rate": 0.00019538520813884298,
+      "loss": 1.0559,
+      "step": 2696
+    },
+    {
+      "epoch": 0.12424563504860184,
+      "grad_norm": 1.2257152542140664,
+      "learning_rate": 0.00019538072667976518,
+      "loss": 1.3583,
+      "step": 2697
+    },
+    {
+      "epoch": 0.12429170313723684,
+      "grad_norm": 1.214048576600122,
+      "learning_rate": 0.00019537624309719904,
+      "loss": 1.4411,
+      "step": 2698
+    },
+    {
+      "epoch": 0.12433777122587183,
+      "grad_norm": 2.321056849531834,
+      "learning_rate": 0.00019537175739124438,
+      "loss": 1.3104,
+      "step": 2699
+    },
+    {
+      "epoch": 0.12438383931450685,
+      "grad_norm": 3.0010901907623078,
+      "learning_rate": 0.00019536726956200106,
+      "loss": 1.2541,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12442990740314185,
+      "grad_norm": 1.2549661067064675,
+      "learning_rate": 0.00019536277960956904,
+      "loss": 1.225,
+      "step": 2701
+    },
+    {
+      "epoch": 0.12447597549177684,
+      "grad_norm": 3.264674099220428,
+      "learning_rate": 0.0001953582875340482,
+      "loss": 1.3182,
+      "step": 2702
+    },
+    {
+      "epoch": 0.12452204358041184,
+      "grad_norm": 1.0993626728120218,
+      "learning_rate": 0.0001953537933355386,
+      "loss": 1.2023,
+      "step": 2703
+    },
+    {
+      "epoch": 0.12456811166904685,
+      "grad_norm": 1.5794043455485764,
+      "learning_rate": 0.00019534929701414032,
+      "loss": 1.5602,
+      "step": 2704
+    },
+    {
+      "epoch": 0.12461417975768185,
+      "grad_norm": 1.118240439882163,
+      "learning_rate": 0.00019534479856995337,
+      "loss": 1.5138,
+      "step": 2705
+    },
+    {
+      "epoch": 0.12466024784631685,
+      "grad_norm": 1.8298481177533095,
+      "learning_rate": 0.00019534029800307803,
+      "loss": 1.6074,
+      "step": 2706
+    },
+    {
+      "epoch": 0.12470631593495186,
+      "grad_norm": 1.6105811303127195,
+      "learning_rate": 0.00019533579531361438,
+      "loss": 1.4253,
+      "step": 2707
+    },
+    {
+      "epoch": 0.12475238402358686,
+      "grad_norm": 4.856308042667574,
+      "learning_rate": 0.00019533129050166274,
+      "loss": 1.64,
+      "step": 2708
+    },
+    {
+      "epoch": 0.12479845211222186,
+      "grad_norm": 1.1288630857555573,
+      "learning_rate": 0.00019532678356732336,
+      "loss": 1.3198,
+      "step": 2709
+    },
+    {
+      "epoch": 0.12484452020085687,
+      "grad_norm": 1.2733071236040843,
+      "learning_rate": 0.0001953222745106966,
+      "loss": 1.4331,
+      "step": 2710
+    },
+    {
+      "epoch": 0.12489058828949187,
+      "grad_norm": 2.803200107788735,
+      "learning_rate": 0.00019531776333188284,
+      "loss": 1.5542,
+      "step": 2711
+    },
+    {
+      "epoch": 0.12493665637812687,
+      "grad_norm": 5.15653437382759,
+      "learning_rate": 0.00019531325003098254,
+      "loss": 1.4276,
+      "step": 2712
+    },
+    {
+      "epoch": 0.12498272446676187,
+      "grad_norm": 1.9490137490543245,
+      "learning_rate": 0.0001953087346080961,
+      "loss": 1.3317,
+      "step": 2713
+    },
+    {
+      "epoch": 0.12502879255539687,
+      "grad_norm": 1.2260127097100533,
+      "learning_rate": 0.00019530421706332414,
+      "loss": 1.3514,
+      "step": 2714
+    },
+    {
+      "epoch": 0.12507486064403187,
+      "grad_norm": 3.4115389826486955,
+      "learning_rate": 0.0001952996973967672,
+      "loss": 1.3684,
+      "step": 2715
+    },
+    {
+      "epoch": 0.1251209287326669,
+      "grad_norm": 1.275279286176182,
+      "learning_rate": 0.00019529517560852588,
+      "loss": 1.5272,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1251669968213019,
+      "grad_norm": 1.9189951075779008,
+      "learning_rate": 0.0001952906516987009,
+      "loss": 1.3959,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1252130649099369,
+      "grad_norm": 1.3477739948028178,
+      "learning_rate": 0.00019528612566739291,
+      "loss": 1.4756,
+      "step": 2718
+    },
+    {
+      "epoch": 0.1252591329985719,
+      "grad_norm": 2.006570865145105,
+      "learning_rate": 0.00019528159751470273,
+      "loss": 1.2349,
+      "step": 2719
+    },
+    {
+      "epoch": 0.12530520108720689,
+      "grad_norm": 1.3682347219635873,
+      "learning_rate": 0.00019527706724073118,
+      "loss": 1.4622,
+      "step": 2720
+    },
+    {
+      "epoch": 0.12535126917584188,
+      "grad_norm": 1.2784861369224212,
+      "learning_rate": 0.00019527253484557906,
+      "loss": 1.4629,
+      "step": 2721
+    },
+    {
+      "epoch": 0.1253973372644769,
+      "grad_norm": 2.1090944963725398,
+      "learning_rate": 0.00019526800032934732,
+      "loss": 1.3182,
+      "step": 2722
+    },
+    {
+      "epoch": 0.1254434053531119,
+      "grad_norm": 1.3942739291114479,
+      "learning_rate": 0.00019526346369213692,
+      "loss": 1.2625,
+      "step": 2723
+    },
+    {
+      "epoch": 0.1254894734417469,
+      "grad_norm": 2.3183492011088473,
+      "learning_rate": 0.00019525892493404883,
+      "loss": 1.5116,
+      "step": 2724
+    },
+    {
+      "epoch": 0.1255355415303819,
+      "grad_norm": 4.634269449673797,
+      "learning_rate": 0.0001952543840551841,
+      "loss": 0.9339,
+      "step": 2725
+    },
+    {
+      "epoch": 0.1255816096190169,
+      "grad_norm": 2.0494085813161216,
+      "learning_rate": 0.00019524984105564384,
+      "loss": 1.4262,
+      "step": 2726
+    },
+    {
+      "epoch": 0.1256276777076519,
+      "grad_norm": 1.219007857367777,
+      "learning_rate": 0.0001952452959355292,
+      "loss": 1.6304,
+      "step": 2727
+    },
+    {
+      "epoch": 0.1256737457962869,
+      "grad_norm": 2.2659300874551818,
+      "learning_rate": 0.00019524074869494134,
+      "loss": 1.1108,
+      "step": 2728
+    },
+    {
+      "epoch": 0.12571981388492193,
+      "grad_norm": 1.2120463255732843,
+      "learning_rate": 0.00019523619933398153,
+      "loss": 1.3108,
+      "step": 2729
+    },
+    {
+      "epoch": 0.12576588197355693,
+      "grad_norm": 2.66976405742326,
+      "learning_rate": 0.00019523164785275102,
+      "loss": 1.4854,
+      "step": 2730
+    },
+    {
+      "epoch": 0.12581195006219192,
+      "grad_norm": 1.4190811469665678,
+      "learning_rate": 0.00019522709425135118,
+      "loss": 1.2495,
+      "step": 2731
+    },
+    {
+      "epoch": 0.12585801815082692,
+      "grad_norm": 1.430966923315359,
+      "learning_rate": 0.00019522253852988333,
+      "loss": 1.2194,
+      "step": 2732
+    },
+    {
+      "epoch": 0.12590408623946192,
+      "grad_norm": 1.3667030959473165,
+      "learning_rate": 0.00019521798068844896,
+      "loss": 1.5295,
+      "step": 2733
+    },
+    {
+      "epoch": 0.12595015432809692,
+      "grad_norm": 7.962366982898507,
+      "learning_rate": 0.00019521342072714953,
+      "loss": 1.2257,
+      "step": 2734
+    },
+    {
+      "epoch": 0.12599622241673192,
+      "grad_norm": 2.2361410160310538,
+      "learning_rate": 0.00019520885864608652,
+      "loss": 1.3662,
+      "step": 2735
+    },
+    {
+      "epoch": 0.12604229050536694,
+      "grad_norm": 0.9915801886034459,
+      "learning_rate": 0.0001952042944453615,
+      "loss": 1.4662,
+      "step": 2736
+    },
+    {
+      "epoch": 0.12608835859400194,
+      "grad_norm": 2.3154636603871506,
+      "learning_rate": 0.00019519972812507611,
+      "loss": 1.2432,
+      "step": 2737
+    },
+    {
+      "epoch": 0.12613442668263694,
+      "grad_norm": 2.3742661509754592,
+      "learning_rate": 0.00019519515968533205,
+      "loss": 1.2235,
+      "step": 2738
+    },
+    {
+      "epoch": 0.12618049477127194,
+      "grad_norm": 6.810354769264509,
+      "learning_rate": 0.00019519058912623093,
+      "loss": 1.1497,
+      "step": 2739
+    },
+    {
+      "epoch": 0.12622656285990694,
+      "grad_norm": 0.9894029165746012,
+      "learning_rate": 0.0001951860164478746,
+      "loss": 1.5476,
+      "step": 2740
+    },
+    {
+      "epoch": 0.12627263094854194,
+      "grad_norm": 1.8280723792515288,
+      "learning_rate": 0.0001951814416503648,
+      "loss": 1.1666,
+      "step": 2741
+    },
+    {
+      "epoch": 0.12631869903717693,
+      "grad_norm": 2.530827007881725,
+      "learning_rate": 0.00019517686473380342,
+      "loss": 1.3937,
+      "step": 2742
+    },
+    {
+      "epoch": 0.12636476712581196,
+      "grad_norm": 1.8487830837168509,
+      "learning_rate": 0.0001951722856982923,
+      "loss": 1.3036,
+      "step": 2743
+    },
+    {
+      "epoch": 0.12641083521444696,
+      "grad_norm": 1.623748665127775,
+      "learning_rate": 0.00019516770454393347,
+      "loss": 1.2101,
+      "step": 2744
+    },
+    {
+      "epoch": 0.12645690330308196,
+      "grad_norm": 1.0039589803423081,
+      "learning_rate": 0.00019516312127082887,
+      "loss": 1.5797,
+      "step": 2745
+    },
+    {
+      "epoch": 0.12650297139171696,
+      "grad_norm": 4.636098966325606,
+      "learning_rate": 0.00019515853587908052,
+      "loss": 1.2828,
+      "step": 2746
+    },
+    {
+      "epoch": 0.12654903948035195,
+      "grad_norm": 1.6090106773783197,
+      "learning_rate": 0.00019515394836879056,
+      "loss": 1.1626,
+      "step": 2747
+    },
+    {
+      "epoch": 0.12659510756898695,
+      "grad_norm": 1.8792733274490647,
+      "learning_rate": 0.00019514935874006105,
+      "loss": 1.2363,
+      "step": 2748
+    },
+    {
+      "epoch": 0.12664117565762195,
+      "grad_norm": 1.8022745761418333,
+      "learning_rate": 0.00019514476699299428,
+      "loss": 1.6689,
+      "step": 2749
+    },
+    {
+      "epoch": 0.12668724374625698,
+      "grad_norm": 1.5301083356061091,
+      "learning_rate": 0.00019514017312769236,
+      "loss": 1.3426,
+      "step": 2750
+    },
+    {
+      "epoch": 0.12673331183489198,
+      "grad_norm": 2.490492610091644,
+      "learning_rate": 0.00019513557714425763,
+      "loss": 1.1097,
+      "step": 2751
+    },
+    {
+      "epoch": 0.12677937992352697,
+      "grad_norm": 1.167324267554098,
+      "learning_rate": 0.0001951309790427924,
+      "loss": 1.4904,
+      "step": 2752
+    },
+    {
+      "epoch": 0.12682544801216197,
+      "grad_norm": 1.2588714705069781,
+      "learning_rate": 0.00019512637882339903,
+      "loss": 1.5328,
+      "step": 2753
+    },
+    {
+      "epoch": 0.12687151610079697,
+      "grad_norm": 1.7365393215457687,
+      "learning_rate": 0.00019512177648617996,
+      "loss": 1.2235,
+      "step": 2754
+    },
+    {
+      "epoch": 0.12691758418943197,
+      "grad_norm": 1.3732927678616842,
+      "learning_rate": 0.0001951171720312376,
+      "loss": 1.3481,
+      "step": 2755
+    },
+    {
+      "epoch": 0.126963652278067,
+      "grad_norm": 1.2290000542314738,
+      "learning_rate": 0.00019511256545867454,
+      "loss": 1.4284,
+      "step": 2756
+    },
+    {
+      "epoch": 0.127009720366702,
+      "grad_norm": 2.1222491160706825,
+      "learning_rate": 0.00019510795676859325,
+      "loss": 1.2717,
+      "step": 2757
+    },
+    {
+      "epoch": 0.127055788455337,
+      "grad_norm": 1.1179259769590595,
+      "learning_rate": 0.0001951033459610964,
+      "loss": 1.1155,
+      "step": 2758
+    },
+    {
+      "epoch": 0.127101856543972,
+      "grad_norm": 0.9723458673747156,
+      "learning_rate": 0.00019509873303628664,
+      "loss": 1.2775,
+      "step": 2759
+    },
+    {
+      "epoch": 0.127147924632607,
+      "grad_norm": 1.5588281399455655,
+      "learning_rate": 0.0001950941179942666,
+      "loss": 1.3831,
+      "step": 2760
+    },
+    {
+      "epoch": 0.127193992721242,
+      "grad_norm": 1.479395070029732,
+      "learning_rate": 0.00019508950083513914,
+      "loss": 1.4229,
+      "step": 2761
+    },
+    {
+      "epoch": 0.127240060809877,
+      "grad_norm": 1.7761505540547964,
+      "learning_rate": 0.0001950848815590069,
+      "loss": 1.3377,
+      "step": 2762
+    },
+    {
+      "epoch": 0.127286128898512,
+      "grad_norm": 1.5403384542252128,
+      "learning_rate": 0.00019508026016597287,
+      "loss": 1.3933,
+      "step": 2763
+    },
+    {
+      "epoch": 0.127332196987147,
+      "grad_norm": 1.0404811149870168,
+      "learning_rate": 0.00019507563665613985,
+      "loss": 1.3702,
+      "step": 2764
+    },
+    {
+      "epoch": 0.127378265075782,
+      "grad_norm": 1.496850616148944,
+      "learning_rate": 0.00019507101102961082,
+      "loss": 1.3634,
+      "step": 2765
+    },
+    {
+      "epoch": 0.127424333164417,
+      "grad_norm": 1.3226731919030443,
+      "learning_rate": 0.00019506638328648873,
+      "loss": 1.5046,
+      "step": 2766
+    },
+    {
+      "epoch": 0.127470401253052,
+      "grad_norm": 1.5681396631994327,
+      "learning_rate": 0.00019506175342687661,
+      "loss": 1.0965,
+      "step": 2767
+    },
+    {
+      "epoch": 0.127516469341687,
+      "grad_norm": 1.3347652343245342,
+      "learning_rate": 0.00019505712145087757,
+      "loss": 1.789,
+      "step": 2768
+    },
+    {
+      "epoch": 0.127562537430322,
+      "grad_norm": 1.8767150644677872,
+      "learning_rate": 0.0001950524873585947,
+      "loss": 1.2411,
+      "step": 2769
+    },
+    {
+      "epoch": 0.12760860551895703,
+      "grad_norm": 0.9846325201519147,
+      "learning_rate": 0.00019504785115013117,
+      "loss": 1.4883,
+      "step": 2770
+    },
+    {
+      "epoch": 0.12765467360759203,
+      "grad_norm": 1.1882458027915848,
+      "learning_rate": 0.0001950432128255902,
+      "loss": 1.4541,
+      "step": 2771
+    },
+    {
+      "epoch": 0.12770074169622703,
+      "grad_norm": 0.9809468073851066,
+      "learning_rate": 0.00019503857238507507,
+      "loss": 1.3147,
+      "step": 2772
+    },
+    {
+      "epoch": 0.12774680978486203,
+      "grad_norm": 1.1275300640328334,
+      "learning_rate": 0.0001950339298286891,
+      "loss": 1.521,
+      "step": 2773
+    },
+    {
+      "epoch": 0.12779287787349702,
+      "grad_norm": 1.4178049585586066,
+      "learning_rate": 0.0001950292851565356,
+      "loss": 1.7669,
+      "step": 2774
+    },
+    {
+      "epoch": 0.12783894596213202,
+      "grad_norm": 1.0025144061665818,
+      "learning_rate": 0.00019502463836871802,
+      "loss": 1.3928,
+      "step": 2775
+    },
+    {
+      "epoch": 0.12788501405076702,
+      "grad_norm": 1.1543757662604566,
+      "learning_rate": 0.00019501998946533982,
+      "loss": 1.2694,
+      "step": 2776
+    },
+    {
+      "epoch": 0.12793108213940205,
+      "grad_norm": 1.139478925975603,
+      "learning_rate": 0.00019501533844650445,
+      "loss": 1.7793,
+      "step": 2777
+    },
+    {
+      "epoch": 0.12797715022803705,
+      "grad_norm": 1.5166145539167524,
+      "learning_rate": 0.00019501068531231549,
+      "loss": 1.4672,
+      "step": 2778
+    },
+    {
+      "epoch": 0.12802321831667204,
+      "grad_norm": 2.9486173956924224,
+      "learning_rate": 0.00019500603006287654,
+      "loss": 1.1672,
+      "step": 2779
+    },
+    {
+      "epoch": 0.12806928640530704,
+      "grad_norm": 2.528321805641383,
+      "learning_rate": 0.00019500137269829124,
+      "loss": 1.4563,
+      "step": 2780
+    },
+    {
+      "epoch": 0.12811535449394204,
+      "grad_norm": 2.341865275986884,
+      "learning_rate": 0.00019499671321866323,
+      "loss": 1.6115,
+      "step": 2781
+    },
+    {
+      "epoch": 0.12816142258257704,
+      "grad_norm": 1.9570019585915779,
+      "learning_rate": 0.00019499205162409633,
+      "loss": 1.0019,
+      "step": 2782
+    },
+    {
+      "epoch": 0.12820749067121204,
+      "grad_norm": 2.284956617718929,
+      "learning_rate": 0.00019498738791469427,
+      "loss": 1.3328,
+      "step": 2783
+    },
+    {
+      "epoch": 0.12825355875984706,
+      "grad_norm": 1.819716489593961,
+      "learning_rate": 0.00019498272209056086,
+      "loss": 1.6319,
+      "step": 2784
+    },
+    {
+      "epoch": 0.12829962684848206,
+      "grad_norm": 3.007974891950235,
+      "learning_rate": 0.00019497805415180002,
+      "loss": 1.4521,
+      "step": 2785
+    },
+    {
+      "epoch": 0.12834569493711706,
+      "grad_norm": 4.586926159258868,
+      "learning_rate": 0.00019497338409851564,
+      "loss": 1.302,
+      "step": 2786
+    },
+    {
+      "epoch": 0.12839176302575206,
+      "grad_norm": 3.8565788336078803,
+      "learning_rate": 0.00019496871193081173,
+      "loss": 1.2243,
+      "step": 2787
+    },
+    {
+      "epoch": 0.12843783111438706,
+      "grad_norm": 3.988183270629042,
+      "learning_rate": 0.00019496403764879229,
+      "loss": 1.387,
+      "step": 2788
+    },
+    {
+      "epoch": 0.12848389920302206,
+      "grad_norm": 1.641640549952054,
+      "learning_rate": 0.00019495936125256137,
+      "loss": 1.1621,
+      "step": 2789
+    },
+    {
+      "epoch": 0.12852996729165708,
+      "grad_norm": 1.236463990858675,
+      "learning_rate": 0.00019495468274222306,
+      "loss": 1.6229,
+      "step": 2790
+    },
+    {
+      "epoch": 0.12857603538029208,
+      "grad_norm": 1.7701990842472146,
+      "learning_rate": 0.00019495000211788158,
+      "loss": 1.3619,
+      "step": 2791
+    },
+    {
+      "epoch": 0.12862210346892708,
+      "grad_norm": 2.888577970851478,
+      "learning_rate": 0.0001949453193796411,
+      "loss": 1.3042,
+      "step": 2792
+    },
+    {
+      "epoch": 0.12866817155756208,
+      "grad_norm": 2.7860528290610973,
+      "learning_rate": 0.0001949406345276059,
+      "loss": 1.5071,
+      "step": 2793
+    },
+    {
+      "epoch": 0.12871423964619708,
+      "grad_norm": 1.0627461637638078,
+      "learning_rate": 0.00019493594756188025,
+      "loss": 1.3303,
+      "step": 2794
+    },
+    {
+      "epoch": 0.12876030773483207,
+      "grad_norm": 1.8471066310337132,
+      "learning_rate": 0.00019493125848256851,
+      "loss": 1.0274,
+      "step": 2795
+    },
+    {
+      "epoch": 0.12880637582346707,
+      "grad_norm": 2.9620330775800294,
+      "learning_rate": 0.00019492656728977505,
+      "loss": 1.415,
+      "step": 2796
+    },
+    {
+      "epoch": 0.1288524439121021,
+      "grad_norm": 2.0861343814176263,
+      "learning_rate": 0.00019492187398360435,
+      "loss": 1.1273,
+      "step": 2797
+    },
+    {
+      "epoch": 0.1288985120007371,
+      "grad_norm": 0.9623320042937237,
+      "learning_rate": 0.0001949171785641609,
+      "loss": 1.2019,
+      "step": 2798
+    },
+    {
+      "epoch": 0.1289445800893721,
+      "grad_norm": 2.4697713913494224,
+      "learning_rate": 0.0001949124810315492,
+      "loss": 1.6128,
+      "step": 2799
+    },
+    {
+      "epoch": 0.1289906481780071,
+      "grad_norm": 1.5458528275243795,
+      "learning_rate": 0.00019490778138587383,
+      "loss": 1.2467,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1290367162666421,
+      "grad_norm": 2.4833653841472216,
+      "learning_rate": 0.0001949030796272395,
+      "loss": 1.2775,
+      "step": 2801
+    },
+    {
+      "epoch": 0.1290827843552771,
+      "grad_norm": 0.9819432483013696,
+      "learning_rate": 0.00019489837575575075,
+      "loss": 1.5671,
+      "step": 2802
+    },
+    {
+      "epoch": 0.1291288524439121,
+      "grad_norm": 1.2769054220859875,
+      "learning_rate": 0.00019489366977151241,
+      "loss": 1.5433,
+      "step": 2803
+    },
+    {
+      "epoch": 0.12917492053254712,
+      "grad_norm": 2.2062836331900693,
+      "learning_rate": 0.00019488896167462922,
+      "loss": 1.4663,
+      "step": 2804
+    },
+    {
+      "epoch": 0.12922098862118211,
+      "grad_norm": 2.54231522136851,
+      "learning_rate": 0.000194884251465206,
+      "loss": 1.4967,
+      "step": 2805
+    },
+    {
+      "epoch": 0.1292670567098171,
+      "grad_norm": 2.58124109283377,
+      "learning_rate": 0.00019487953914334766,
+      "loss": 1.4457,
+      "step": 2806
+    },
+    {
+      "epoch": 0.1293131247984521,
+      "grad_norm": 1.2740618591716422,
+      "learning_rate": 0.00019487482470915897,
+      "loss": 1.4426,
+      "step": 2807
+    },
+    {
+      "epoch": 0.1293591928870871,
+      "grad_norm": 1.1768438446426637,
+      "learning_rate": 0.00019487010816274505,
+      "loss": 1.6277,
+      "step": 2808
+    },
+    {
+      "epoch": 0.1294052609757221,
+      "grad_norm": 1.5684794313345891,
+      "learning_rate": 0.00019486538950421083,
+      "loss": 1.4577,
+      "step": 2809
+    },
+    {
+      "epoch": 0.1294513290643571,
+      "grad_norm": 1.24616674507227,
+      "learning_rate": 0.00019486066873366134,
+      "loss": 1.3678,
+      "step": 2810
+    },
+    {
+      "epoch": 0.12949739715299213,
+      "grad_norm": 3.5352446485610463,
+      "learning_rate": 0.00019485594585120177,
+      "loss": 1.4569,
+      "step": 2811
+    },
+    {
+      "epoch": 0.12954346524162713,
+      "grad_norm": 1.2908696464327338,
+      "learning_rate": 0.00019485122085693716,
+      "loss": 1.4966,
+      "step": 2812
+    },
+    {
+      "epoch": 0.12958953333026213,
+      "grad_norm": 1.114307213355888,
+      "learning_rate": 0.00019484649375097276,
+      "loss": 1.4992,
+      "step": 2813
+    },
+    {
+      "epoch": 0.12963560141889713,
+      "grad_norm": 1.281148993003846,
+      "learning_rate": 0.00019484176453341383,
+      "loss": 1.3962,
+      "step": 2814
+    },
+    {
+      "epoch": 0.12968166950753213,
+      "grad_norm": 0.7992411825698598,
+      "learning_rate": 0.00019483703320436559,
+      "loss": 1.5072,
+      "step": 2815
+    },
+    {
+      "epoch": 0.12972773759616713,
+      "grad_norm": 3.655737422810413,
+      "learning_rate": 0.00019483229976393344,
+      "loss": 1.3583,
+      "step": 2816
+    },
+    {
+      "epoch": 0.12977380568480212,
+      "grad_norm": 1.144825291568825,
+      "learning_rate": 0.00019482756421222274,
+      "loss": 1.6496,
+      "step": 2817
+    },
+    {
+      "epoch": 0.12981987377343715,
+      "grad_norm": 1.5804758561837213,
+      "learning_rate": 0.00019482282654933891,
+      "loss": 1.7703,
+      "step": 2818
+    },
+    {
+      "epoch": 0.12986594186207215,
+      "grad_norm": 1.8303979151941436,
+      "learning_rate": 0.00019481808677538744,
+      "loss": 1.2578,
+      "step": 2819
+    },
+    {
+      "epoch": 0.12991200995070715,
+      "grad_norm": 1.4607203299686644,
+      "learning_rate": 0.00019481334489047388,
+      "loss": 1.3788,
+      "step": 2820
+    },
+    {
+      "epoch": 0.12995807803934215,
+      "grad_norm": 1.3544083047152957,
+      "learning_rate": 0.0001948086008947037,
+      "loss": 1.4681,
+      "step": 2821
+    },
+    {
+      "epoch": 0.13000414612797714,
+      "grad_norm": 1.4859381954578192,
+      "learning_rate": 0.0001948038547881826,
+      "loss": 1.38,
+      "step": 2822
+    },
+    {
+      "epoch": 0.13005021421661214,
+      "grad_norm": 1.3155244578264664,
+      "learning_rate": 0.00019479910657101626,
+      "loss": 1.3336,
+      "step": 2823
+    },
+    {
+      "epoch": 0.13009628230524717,
+      "grad_norm": 3.0447777210447846,
+      "learning_rate": 0.00019479435624331033,
+      "loss": 1.4223,
+      "step": 2824
+    },
+    {
+      "epoch": 0.13014235039388217,
+      "grad_norm": 1.1156861936988303,
+      "learning_rate": 0.00019478960380517058,
+      "loss": 1.5205,
+      "step": 2825
+    },
+    {
+      "epoch": 0.13018841848251717,
+      "grad_norm": 6.101168412958453,
+      "learning_rate": 0.0001947848492567029,
+      "loss": 1.1283,
+      "step": 2826
+    },
+    {
+      "epoch": 0.13023448657115216,
+      "grad_norm": 0.9457635099215048,
+      "learning_rate": 0.000194780092598013,
+      "loss": 1.6469,
+      "step": 2827
+    },
+    {
+      "epoch": 0.13028055465978716,
+      "grad_norm": 1.3641778789002446,
+      "learning_rate": 0.00019477533382920688,
+      "loss": 1.2516,
+      "step": 2828
+    },
+    {
+      "epoch": 0.13032662274842216,
+      "grad_norm": 1.4585401583556012,
+      "learning_rate": 0.00019477057295039045,
+      "loss": 1.4775,
+      "step": 2829
+    },
+    {
+      "epoch": 0.13037269083705716,
+      "grad_norm": 4.21855721308667,
+      "learning_rate": 0.00019476580996166974,
+      "loss": 1.2849,
+      "step": 2830
+    },
+    {
+      "epoch": 0.13041875892569219,
+      "grad_norm": 1.4322566601394748,
+      "learning_rate": 0.0001947610448631507,
+      "loss": 1.5087,
+      "step": 2831
+    },
+    {
+      "epoch": 0.13046482701432718,
+      "grad_norm": 0.8168555821022674,
+      "learning_rate": 0.00019475627765493952,
+      "loss": 1.1701,
+      "step": 2832
+    },
+    {
+      "epoch": 0.13051089510296218,
+      "grad_norm": 1.0156064515577075,
+      "learning_rate": 0.0001947515083371423,
+      "loss": 1.5002,
+      "step": 2833
+    },
+    {
+      "epoch": 0.13055696319159718,
+      "grad_norm": 1.041504194146429,
+      "learning_rate": 0.00019474673690986517,
+      "loss": 1.4361,
+      "step": 2834
+    },
+    {
+      "epoch": 0.13060303128023218,
+      "grad_norm": 1.2555512404993276,
+      "learning_rate": 0.0001947419633732144,
+      "loss": 1.1361,
+      "step": 2835
+    },
+    {
+      "epoch": 0.13064909936886718,
+      "grad_norm": 2.4862918381030537,
+      "learning_rate": 0.00019473718772729633,
+      "loss": 1.2332,
+      "step": 2836
+    },
+    {
+      "epoch": 0.13069516745750218,
+      "grad_norm": 0.8757967721558869,
+      "learning_rate": 0.00019473240997221715,
+      "loss": 1.4858,
+      "step": 2837
+    },
+    {
+      "epoch": 0.1307412355461372,
+      "grad_norm": 1.0429269817541913,
+      "learning_rate": 0.0001947276301080833,
+      "loss": 1.3628,
+      "step": 2838
+    },
+    {
+      "epoch": 0.1307873036347722,
+      "grad_norm": 0.8535887791368458,
+      "learning_rate": 0.0001947228481350012,
+      "loss": 1.5428,
+      "step": 2839
+    },
+    {
+      "epoch": 0.1308333717234072,
+      "grad_norm": 1.322403950134832,
+      "learning_rate": 0.0001947180640530773,
+      "loss": 1.0812,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1308794398120422,
+      "grad_norm": 1.0216690650234361,
+      "learning_rate": 0.00019471327786241814,
+      "loss": 1.6182,
+      "step": 2841
+    },
+    {
+      "epoch": 0.1309255079006772,
+      "grad_norm": 1.3084400721566989,
+      "learning_rate": 0.0001947084895631302,
+      "loss": 1.0499,
+      "step": 2842
+    },
+    {
+      "epoch": 0.1309715759893122,
+      "grad_norm": 1.1856777268605707,
+      "learning_rate": 0.00019470369915532015,
+      "loss": 1.3639,
+      "step": 2843
+    },
+    {
+      "epoch": 0.1310176440779472,
+      "grad_norm": 1.1763474507852394,
+      "learning_rate": 0.00019469890663909462,
+      "loss": 1.2304,
+      "step": 2844
+    },
+    {
+      "epoch": 0.13106371216658222,
+      "grad_norm": 1.2079081515032737,
+      "learning_rate": 0.0001946941120145603,
+      "loss": 1.505,
+      "step": 2845
+    },
+    {
+      "epoch": 0.13110978025521722,
+      "grad_norm": 1.9929950896575792,
+      "learning_rate": 0.00019468931528182398,
+      "loss": 1.3235,
+      "step": 2846
+    },
+    {
+      "epoch": 0.13115584834385222,
+      "grad_norm": 1.7541868877494422,
+      "learning_rate": 0.00019468451644099238,
+      "loss": 1.6254,
+      "step": 2847
+    },
+    {
+      "epoch": 0.13120191643248721,
+      "grad_norm": 1.3449112969320343,
+      "learning_rate": 0.00019467971549217236,
+      "loss": 1.3656,
+      "step": 2848
+    },
+    {
+      "epoch": 0.1312479845211222,
+      "grad_norm": 1.2632617179946044,
+      "learning_rate": 0.0001946749124354708,
+      "loss": 1.458,
+      "step": 2849
+    },
+    {
+      "epoch": 0.1312940526097572,
+      "grad_norm": 1.2089582532430314,
+      "learning_rate": 0.0001946701072709947,
+      "loss": 1.5372,
+      "step": 2850
+    },
+    {
+      "epoch": 0.13134012069839224,
+      "grad_norm": 1.7459739099236908,
+      "learning_rate": 0.00019466529999885094,
+      "loss": 1.3229,
+      "step": 2851
+    },
+    {
+      "epoch": 0.13138618878702724,
+      "grad_norm": 1.7077553509342318,
+      "learning_rate": 0.0001946604906191466,
+      "loss": 1.5658,
+      "step": 2852
+    },
+    {
+      "epoch": 0.13143225687566223,
+      "grad_norm": 4.267165331976592,
+      "learning_rate": 0.00019465567913198878,
+      "loss": 1.4075,
+      "step": 2853
+    },
+    {
+      "epoch": 0.13147832496429723,
+      "grad_norm": 7.7589841271566735,
+      "learning_rate": 0.00019465086553748453,
+      "loss": 1.5689,
+      "step": 2854
+    },
+    {
+      "epoch": 0.13152439305293223,
+      "grad_norm": 1.8173384260037506,
+      "learning_rate": 0.00019464604983574106,
+      "loss": 1.6438,
+      "step": 2855
+    },
+    {
+      "epoch": 0.13157046114156723,
+      "grad_norm": 1.904878455230755,
+      "learning_rate": 0.00019464123202686558,
+      "loss": 1.1441,
+      "step": 2856
+    },
+    {
+      "epoch": 0.13161652923020223,
+      "grad_norm": 2.2011783188471794,
+      "learning_rate": 0.00019463641211096535,
+      "loss": 1.1447,
+      "step": 2857
+    },
+    {
+      "epoch": 0.13166259731883725,
+      "grad_norm": 6.656617890023987,
+      "learning_rate": 0.00019463159008814763,
+      "loss": 1.1765,
+      "step": 2858
+    },
+    {
+      "epoch": 0.13170866540747225,
+      "grad_norm": 1.823443793214553,
+      "learning_rate": 0.00019462676595851988,
+      "loss": 1.5014,
+      "step": 2859
+    },
+    {
+      "epoch": 0.13175473349610725,
+      "grad_norm": 1.7960112792878018,
+      "learning_rate": 0.00019462193972218938,
+      "loss": 1.4143,
+      "step": 2860
+    },
+    {
+      "epoch": 0.13180080158474225,
+      "grad_norm": 2.151029442945084,
+      "learning_rate": 0.00019461711137926365,
+      "loss": 1.4107,
+      "step": 2861
+    },
+    {
+      "epoch": 0.13184686967337725,
+      "grad_norm": 2.2627356668908076,
+      "learning_rate": 0.00019461228092985017,
+      "loss": 1.2442,
+      "step": 2862
+    },
+    {
+      "epoch": 0.13189293776201225,
+      "grad_norm": 4.8357901891646335,
+      "learning_rate": 0.00019460744837405649,
+      "loss": 1.1446,
+      "step": 2863
+    },
+    {
+      "epoch": 0.13193900585064725,
+      "grad_norm": 1.3619835355255512,
+      "learning_rate": 0.00019460261371199018,
+      "loss": 1.7153,
+      "step": 2864
+    },
+    {
+      "epoch": 0.13198507393928227,
+      "grad_norm": 1.4949308813711986,
+      "learning_rate": 0.0001945977769437589,
+      "loss": 1.6059,
+      "step": 2865
+    },
+    {
+      "epoch": 0.13203114202791727,
+      "grad_norm": 1.6209594920852597,
+      "learning_rate": 0.0001945929380694703,
+      "loss": 1.6831,
+      "step": 2866
+    },
+    {
+      "epoch": 0.13207721011655227,
+      "grad_norm": 10.697645598006103,
+      "learning_rate": 0.00019458809708923212,
+      "loss": 1.4833,
+      "step": 2867
+    },
+    {
+      "epoch": 0.13212327820518727,
+      "grad_norm": 2.591573553830692,
+      "learning_rate": 0.00019458325400315216,
+      "loss": 1.6104,
+      "step": 2868
+    },
+    {
+      "epoch": 0.13216934629382227,
+      "grad_norm": 1.4993491417489146,
+      "learning_rate": 0.0001945784088113382,
+      "loss": 1.3048,
+      "step": 2869
+    },
+    {
+      "epoch": 0.13221541438245726,
+      "grad_norm": 1.4427886963852026,
+      "learning_rate": 0.00019457356151389817,
+      "loss": 1.6416,
+      "step": 2870
+    },
+    {
+      "epoch": 0.13226148247109226,
+      "grad_norm": 1.5137766638190917,
+      "learning_rate": 0.0001945687121109399,
+      "loss": 1.3939,
+      "step": 2871
+    },
+    {
+      "epoch": 0.1323075505597273,
+      "grad_norm": 2.3562957467737617,
+      "learning_rate": 0.00019456386060257147,
+      "loss": 1.1496,
+      "step": 2872
+    },
+    {
+      "epoch": 0.1323536186483623,
+      "grad_norm": 1.8363927751496558,
+      "learning_rate": 0.00019455900698890077,
+      "loss": 1.6439,
+      "step": 2873
+    },
+    {
+      "epoch": 0.13239968673699729,
+      "grad_norm": 1.9406520462483314,
+      "learning_rate": 0.00019455415127003596,
+      "loss": 1.4256,
+      "step": 2874
+    },
+    {
+      "epoch": 0.13244575482563228,
+      "grad_norm": 1.5837695897107773,
+      "learning_rate": 0.00019454929344608506,
+      "loss": 1.2485,
+      "step": 2875
+    },
+    {
+      "epoch": 0.13249182291426728,
+      "grad_norm": 4.993328678680114,
+      "learning_rate": 0.00019454443351715628,
+      "loss": 1.3968,
+      "step": 2876
+    },
+    {
+      "epoch": 0.13253789100290228,
+      "grad_norm": 2.1504205448147573,
+      "learning_rate": 0.00019453957148335776,
+      "loss": 1.4437,
+      "step": 2877
+    },
+    {
+      "epoch": 0.13258395909153728,
+      "grad_norm": 6.099947533483714,
+      "learning_rate": 0.00019453470734479784,
+      "loss": 1.5307,
+      "step": 2878
+    },
+    {
+      "epoch": 0.1326300271801723,
+      "grad_norm": 1.5078504012047682,
+      "learning_rate": 0.0001945298411015847,
+      "loss": 1.6207,
+      "step": 2879
+    },
+    {
+      "epoch": 0.1326760952688073,
+      "grad_norm": 1.460591346320933,
+      "learning_rate": 0.0001945249727538268,
+      "loss": 1.6122,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1327221633574423,
+      "grad_norm": 1.7629603766549178,
+      "learning_rate": 0.0001945201023016324,
+      "loss": 1.5594,
+      "step": 2881
+    },
+    {
+      "epoch": 0.1327682314460773,
+      "grad_norm": 3.3848639888032115,
+      "learning_rate": 0.00019451522974511,
+      "loss": 1.6706,
+      "step": 2882
+    },
+    {
+      "epoch": 0.1328142995347123,
+      "grad_norm": 1.1836819114592516,
+      "learning_rate": 0.00019451035508436806,
+      "loss": 1.3867,
+      "step": 2883
+    },
+    {
+      "epoch": 0.1328603676233473,
+      "grad_norm": 1.4585717204324191,
+      "learning_rate": 0.00019450547831951514,
+      "loss": 1.5127,
+      "step": 2884
+    },
+    {
+      "epoch": 0.13290643571198232,
+      "grad_norm": 0.8511633969076341,
+      "learning_rate": 0.00019450059945065978,
+      "loss": 1.469,
+      "step": 2885
+    },
+    {
+      "epoch": 0.13295250380061732,
+      "grad_norm": 1.2722918371462,
+      "learning_rate": 0.0001944957184779106,
+      "loss": 1.3811,
+      "step": 2886
+    },
+    {
+      "epoch": 0.13299857188925232,
+      "grad_norm": 2.362492254890088,
+      "learning_rate": 0.00019449083540137627,
+      "loss": 1.5828,
+      "step": 2887
+    },
+    {
+      "epoch": 0.13304463997788732,
+      "grad_norm": 0.9799260687564674,
+      "learning_rate": 0.00019448595022116552,
+      "loss": 1.4802,
+      "step": 2888
+    },
+    {
+      "epoch": 0.13309070806652232,
+      "grad_norm": 0.9988696284802046,
+      "learning_rate": 0.00019448106293738708,
+      "loss": 1.2813,
+      "step": 2889
+    },
+    {
+      "epoch": 0.13313677615515732,
+      "grad_norm": 1.0054884300544706,
+      "learning_rate": 0.00019447617355014977,
+      "loss": 1.3161,
+      "step": 2890
+    },
+    {
+      "epoch": 0.13318284424379231,
+      "grad_norm": 1.349580178881571,
+      "learning_rate": 0.00019447128205956246,
+      "loss": 1.3084,
+      "step": 2891
+    },
+    {
+      "epoch": 0.13322891233242734,
+      "grad_norm": 1.750134797249221,
+      "learning_rate": 0.00019446638846573407,
+      "loss": 1.3692,
+      "step": 2892
+    },
+    {
+      "epoch": 0.13327498042106234,
+      "grad_norm": 2.6062989661279077,
+      "learning_rate": 0.00019446149276877345,
+      "loss": 1.4301,
+      "step": 2893
+    },
+    {
+      "epoch": 0.13332104850969734,
+      "grad_norm": 1.2745015016110306,
+      "learning_rate": 0.00019445659496878973,
+      "loss": 0.9622,
+      "step": 2894
+    },
+    {
+      "epoch": 0.13336711659833234,
+      "grad_norm": 1.6560011104436976,
+      "learning_rate": 0.00019445169506589183,
+      "loss": 1.0619,
+      "step": 2895
+    },
+    {
+      "epoch": 0.13341318468696733,
+      "grad_norm": 1.1464297204832934,
+      "learning_rate": 0.0001944467930601889,
+      "loss": 1.2864,
+      "step": 2896
+    },
+    {
+      "epoch": 0.13345925277560233,
+      "grad_norm": 0.9985776306419257,
+      "learning_rate": 0.00019444188895179006,
+      "loss": 1.5566,
+      "step": 2897
+    },
+    {
+      "epoch": 0.13350532086423733,
+      "grad_norm": 1.042788423035844,
+      "learning_rate": 0.0001944369827408045,
+      "loss": 1.4223,
+      "step": 2898
+    },
+    {
+      "epoch": 0.13355138895287236,
+      "grad_norm": 1.6187293345531415,
+      "learning_rate": 0.00019443207442734146,
+      "loss": 1.2277,
+      "step": 2899
+    },
+    {
+      "epoch": 0.13359745704150736,
+      "grad_norm": 1.9229312258731486,
+      "learning_rate": 0.0001944271640115102,
+      "loss": 1.624,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13364352513014235,
+      "grad_norm": 1.4138122828962325,
+      "learning_rate": 0.00019442225149342,
+      "loss": 1.1012,
+      "step": 2901
+    },
+    {
+      "epoch": 0.13368959321877735,
+      "grad_norm": 1.3908502555712123,
+      "learning_rate": 0.00019441733687318032,
+      "loss": 1.4332,
+      "step": 2902
+    },
+    {
+      "epoch": 0.13373566130741235,
+      "grad_norm": 1.5431245809850003,
+      "learning_rate": 0.00019441242015090052,
+      "loss": 1.2824,
+      "step": 2903
+    },
+    {
+      "epoch": 0.13378172939604735,
+      "grad_norm": 3.0689572827663474,
+      "learning_rate": 0.00019440750132669002,
+      "loss": 1.2549,
+      "step": 2904
+    },
+    {
+      "epoch": 0.13382779748468235,
+      "grad_norm": 1.4448963805309687,
+      "learning_rate": 0.00019440258040065843,
+      "loss": 1.2719,
+      "step": 2905
+    },
+    {
+      "epoch": 0.13387386557331737,
+      "grad_norm": 2.412597559831858,
+      "learning_rate": 0.00019439765737291523,
+      "loss": 1.0398,
+      "step": 2906
+    },
+    {
+      "epoch": 0.13391993366195237,
+      "grad_norm": 1.372130191685505,
+      "learning_rate": 0.00019439273224357005,
+      "loss": 1.2896,
+      "step": 2907
+    },
+    {
+      "epoch": 0.13396600175058737,
+      "grad_norm": 2.577529727110201,
+      "learning_rate": 0.00019438780501273252,
+      "loss": 0.9746,
+      "step": 2908
+    },
+    {
+      "epoch": 0.13401206983922237,
+      "grad_norm": 1.3315979061841523,
+      "learning_rate": 0.0001943828756805124,
+      "loss": 1.4202,
+      "step": 2909
+    },
+    {
+      "epoch": 0.13405813792785737,
+      "grad_norm": 1.1247547343975512,
+      "learning_rate": 0.0001943779442470193,
+      "loss": 1.4035,
+      "step": 2910
+    },
+    {
+      "epoch": 0.13410420601649237,
+      "grad_norm": 1.251045424686158,
+      "learning_rate": 0.00019437301071236317,
+      "loss": 1.4468,
+      "step": 2911
+    },
+    {
+      "epoch": 0.13415027410512737,
+      "grad_norm": 2.066427601414091,
+      "learning_rate": 0.00019436807507665376,
+      "loss": 1.2774,
+      "step": 2912
+    },
+    {
+      "epoch": 0.1341963421937624,
+      "grad_norm": 1.4571887630081857,
+      "learning_rate": 0.00019436313734000097,
+      "loss": 1.2746,
+      "step": 2913
+    },
+    {
+      "epoch": 0.1342424102823974,
+      "grad_norm": 2.041959254401401,
+      "learning_rate": 0.0001943581975025147,
+      "loss": 1.6934,
+      "step": 2914
+    },
+    {
+      "epoch": 0.1342884783710324,
+      "grad_norm": 1.595560864692847,
+      "learning_rate": 0.00019435325556430494,
+      "loss": 1.1027,
+      "step": 2915
+    },
+    {
+      "epoch": 0.1343345464596674,
+      "grad_norm": 1.4329280949375007,
+      "learning_rate": 0.00019434831152548174,
+      "loss": 1.4586,
+      "step": 2916
+    },
+    {
+      "epoch": 0.13438061454830239,
+      "grad_norm": 1.2465024096900663,
+      "learning_rate": 0.00019434336538615518,
+      "loss": 1.4867,
+      "step": 2917
+    },
+    {
+      "epoch": 0.13442668263693738,
+      "grad_norm": 1.044671692722305,
+      "learning_rate": 0.00019433841714643532,
+      "loss": 1.0589,
+      "step": 2918
+    },
+    {
+      "epoch": 0.1344727507255724,
+      "grad_norm": 1.2928447075682423,
+      "learning_rate": 0.0001943334668064324,
+      "loss": 1.4251,
+      "step": 2919
+    },
+    {
+      "epoch": 0.1345188188142074,
+      "grad_norm": 10.031187972079131,
+      "learning_rate": 0.00019432851436625655,
+      "loss": 1.381,
+      "step": 2920
+    },
+    {
+      "epoch": 0.1345648869028424,
+      "grad_norm": 1.5789389505479414,
+      "learning_rate": 0.00019432355982601805,
+      "loss": 1.2622,
+      "step": 2921
+    },
+    {
+      "epoch": 0.1346109549914774,
+      "grad_norm": 2.0141058587932013,
+      "learning_rate": 0.0001943186031858273,
+      "loss": 1.332,
+      "step": 2922
+    },
+    {
+      "epoch": 0.1346570230801124,
+      "grad_norm": 0.9647920365580146,
+      "learning_rate": 0.0001943136444457945,
+      "loss": 1.6487,
+      "step": 2923
+    },
+    {
+      "epoch": 0.1347030911687474,
+      "grad_norm": 1.6114698520195523,
+      "learning_rate": 0.00019430868360603015,
+      "loss": 1.4982,
+      "step": 2924
+    },
+    {
+      "epoch": 0.1347491592573824,
+      "grad_norm": 1.256017146110911,
+      "learning_rate": 0.00019430372066664463,
+      "loss": 1.7709,
+      "step": 2925
+    },
+    {
+      "epoch": 0.13479522734601743,
+      "grad_norm": 1.2955928976760593,
+      "learning_rate": 0.0001942987556277485,
+      "loss": 1.0098,
+      "step": 2926
+    },
+    {
+      "epoch": 0.13484129543465243,
+      "grad_norm": 2.1209586744182465,
+      "learning_rate": 0.00019429378848945226,
+      "loss": 1.331,
+      "step": 2927
+    },
+    {
+      "epoch": 0.13488736352328742,
+      "grad_norm": 2.8082862620026647,
+      "learning_rate": 0.0001942888192518665,
+      "loss": 1.5207,
+      "step": 2928
+    },
+    {
+      "epoch": 0.13493343161192242,
+      "grad_norm": 1.4630177787943515,
+      "learning_rate": 0.00019428384791510183,
+      "loss": 1.489,
+      "step": 2929
+    },
+    {
+      "epoch": 0.13497949970055742,
+      "grad_norm": 1.4527486333606252,
+      "learning_rate": 0.00019427887447926896,
+      "loss": 1.5137,
+      "step": 2930
+    },
+    {
+      "epoch": 0.13502556778919242,
+      "grad_norm": 1.429150159135778,
+      "learning_rate": 0.0001942738989444786,
+      "loss": 1.0172,
+      "step": 2931
+    },
+    {
+      "epoch": 0.13507163587782742,
+      "grad_norm": 1.2341933094359214,
+      "learning_rate": 0.00019426892131084155,
+      "loss": 1.0715,
+      "step": 2932
+    },
+    {
+      "epoch": 0.13511770396646244,
+      "grad_norm": 1.1042241263486288,
+      "learning_rate": 0.00019426394157846856,
+      "loss": 1.6186,
+      "step": 2933
+    },
+    {
+      "epoch": 0.13516377205509744,
+      "grad_norm": 2.4566014406803416,
+      "learning_rate": 0.00019425895974747057,
+      "loss": 1.4785,
+      "step": 2934
+    },
+    {
+      "epoch": 0.13520984014373244,
+      "grad_norm": 1.2879772179886124,
+      "learning_rate": 0.00019425397581795845,
+      "loss": 1.4176,
+      "step": 2935
+    },
+    {
+      "epoch": 0.13525590823236744,
+      "grad_norm": 3.1131500221739694,
+      "learning_rate": 0.00019424898979004317,
+      "loss": 1.4623,
+      "step": 2936
+    },
+    {
+      "epoch": 0.13530197632100244,
+      "grad_norm": 5.096177919955234,
+      "learning_rate": 0.00019424400166383573,
+      "loss": 1.4543,
+      "step": 2937
+    },
+    {
+      "epoch": 0.13534804440963744,
+      "grad_norm": 2.2624662404242724,
+      "learning_rate": 0.00019423901143944723,
+      "loss": 1.3575,
+      "step": 2938
+    },
+    {
+      "epoch": 0.13539411249827243,
+      "grad_norm": 1.0635695260339963,
+      "learning_rate": 0.0001942340191169887,
+      "loss": 1.4498,
+      "step": 2939
+    },
+    {
+      "epoch": 0.13544018058690746,
+      "grad_norm": 1.7933050290847088,
+      "learning_rate": 0.00019422902469657126,
+      "loss": 1.3995,
+      "step": 2940
+    },
+    {
+      "epoch": 0.13548624867554246,
+      "grad_norm": 2.279062012408367,
+      "learning_rate": 0.0001942240281783062,
+      "loss": 1.2632,
+      "step": 2941
+    },
+    {
+      "epoch": 0.13553231676417746,
+      "grad_norm": 1.3531202670182063,
+      "learning_rate": 0.0001942190295623047,
+      "loss": 1.4555,
+      "step": 2942
+    },
+    {
+      "epoch": 0.13557838485281246,
+      "grad_norm": 1.7394701393271519,
+      "learning_rate": 0.00019421402884867807,
+      "loss": 1.2408,
+      "step": 2943
+    },
+    {
+      "epoch": 0.13562445294144745,
+      "grad_norm": 1.1799204496145075,
+      "learning_rate": 0.00019420902603753764,
+      "loss": 1.6526,
+      "step": 2944
+    },
+    {
+      "epoch": 0.13567052103008245,
+      "grad_norm": 1.5674807809621127,
+      "learning_rate": 0.00019420402112899475,
+      "loss": 1.7543,
+      "step": 2945
+    },
+    {
+      "epoch": 0.13571658911871745,
+      "grad_norm": 1.4947272543610142,
+      "learning_rate": 0.00019419901412316088,
+      "loss": 1.622,
+      "step": 2946
+    },
+    {
+      "epoch": 0.13576265720735248,
+      "grad_norm": 1.401378128722394,
+      "learning_rate": 0.00019419400502014743,
+      "loss": 1.4221,
+      "step": 2947
+    },
+    {
+      "epoch": 0.13580872529598748,
+      "grad_norm": 1.4739877173714273,
+      "learning_rate": 0.00019418899382006601,
+      "loss": 1.705,
+      "step": 2948
+    },
+    {
+      "epoch": 0.13585479338462247,
+      "grad_norm": 4.562058718730405,
+      "learning_rate": 0.00019418398052302815,
+      "loss": 1.4229,
+      "step": 2949
+    },
+    {
+      "epoch": 0.13590086147325747,
+      "grad_norm": 1.2409004504256829,
+      "learning_rate": 0.00019417896512914545,
+      "loss": 1.4219,
+      "step": 2950
+    },
+    {
+      "epoch": 0.13594692956189247,
+      "grad_norm": 1.1964098587981766,
+      "learning_rate": 0.00019417394763852954,
+      "loss": 1.2613,
+      "step": 2951
+    },
+    {
+      "epoch": 0.13599299765052747,
+      "grad_norm": 1.2603820056551853,
+      "learning_rate": 0.0001941689280512922,
+      "loss": 1.4138,
+      "step": 2952
+    },
+    {
+      "epoch": 0.1360390657391625,
+      "grad_norm": 1.3147950771696484,
+      "learning_rate": 0.0001941639063675451,
+      "loss": 1.2745,
+      "step": 2953
+    },
+    {
+      "epoch": 0.1360851338277975,
+      "grad_norm": 2.6072451700853896,
+      "learning_rate": 0.0001941588825874001,
+      "loss": 1.5303,
+      "step": 2954
+    },
+    {
+      "epoch": 0.1361312019164325,
+      "grad_norm": 1.411525923992724,
+      "learning_rate": 0.00019415385671096903,
+      "loss": 1.5477,
+      "step": 2955
+    },
+    {
+      "epoch": 0.1361772700050675,
+      "grad_norm": 1.4511162042600572,
+      "learning_rate": 0.0001941488287383638,
+      "loss": 1.4929,
+      "step": 2956
+    },
+    {
+      "epoch": 0.1362233380937025,
+      "grad_norm": 2.067030088481983,
+      "learning_rate": 0.0001941437986696963,
+      "loss": 1.3834,
+      "step": 2957
+    },
+    {
+      "epoch": 0.1362694061823375,
+      "grad_norm": 2.661383376411514,
+      "learning_rate": 0.00019413876650507857,
+      "loss": 1.5972,
+      "step": 2958
+    },
+    {
+      "epoch": 0.1363154742709725,
+      "grad_norm": 1.5654793288509927,
+      "learning_rate": 0.00019413373224462258,
+      "loss": 1.4321,
+      "step": 2959
+    },
+    {
+      "epoch": 0.1363615423596075,
+      "grad_norm": 1.6411028023966967,
+      "learning_rate": 0.00019412869588844047,
+      "loss": 1.4709,
+      "step": 2960
+    },
+    {
+      "epoch": 0.1364076104482425,
+      "grad_norm": 1.594003820432256,
+      "learning_rate": 0.00019412365743664436,
+      "loss": 1.5011,
+      "step": 2961
+    },
+    {
+      "epoch": 0.1364536785368775,
+      "grad_norm": 1.4414765851057332,
+      "learning_rate": 0.0001941186168893464,
+      "loss": 1.4124,
+      "step": 2962
+    },
+    {
+      "epoch": 0.1364997466255125,
+      "grad_norm": 2.016296742581296,
+      "learning_rate": 0.00019411357424665879,
+      "loss": 1.2252,
+      "step": 2963
+    },
+    {
+      "epoch": 0.1365458147141475,
+      "grad_norm": 1.3446518635622078,
+      "learning_rate": 0.00019410852950869383,
+      "loss": 1.437,
+      "step": 2964
+    },
+    {
+      "epoch": 0.1365918828027825,
+      "grad_norm": 1.6801803531861947,
+      "learning_rate": 0.0001941034826755638,
+      "loss": 1.3595,
+      "step": 2965
+    },
+    {
+      "epoch": 0.1366379508914175,
+      "grad_norm": 1.926991434513644,
+      "learning_rate": 0.0001940984337473811,
+      "loss": 1.3657,
+      "step": 2966
+    },
+    {
+      "epoch": 0.13668401898005253,
+      "grad_norm": 1.7313755923131753,
+      "learning_rate": 0.00019409338272425813,
+      "loss": 1.6035,
+      "step": 2967
+    },
+    {
+      "epoch": 0.13673008706868753,
+      "grad_norm": 1.8204354595044605,
+      "learning_rate": 0.0001940883296063073,
+      "loss": 1.5181,
+      "step": 2968
+    },
+    {
+      "epoch": 0.13677615515732253,
+      "grad_norm": 4.4546009426310045,
+      "learning_rate": 0.00019408327439364117,
+      "loss": 1.3066,
+      "step": 2969
+    },
+    {
+      "epoch": 0.13682222324595752,
+      "grad_norm": 1.6912999353379992,
+      "learning_rate": 0.00019407821708637224,
+      "loss": 1.0717,
+      "step": 2970
+    },
+    {
+      "epoch": 0.13686829133459252,
+      "grad_norm": 1.6546962679385229,
+      "learning_rate": 0.0001940731576846131,
+      "loss": 1.4468,
+      "step": 2971
+    },
+    {
+      "epoch": 0.13691435942322752,
+      "grad_norm": 2.4055357490615776,
+      "learning_rate": 0.00019406809618847644,
+      "loss": 1.3777,
+      "step": 2972
+    },
+    {
+      "epoch": 0.13696042751186252,
+      "grad_norm": 1.7812607495094053,
+      "learning_rate": 0.0001940630325980749,
+      "loss": 1.5187,
+      "step": 2973
+    },
+    {
+      "epoch": 0.13700649560049755,
+      "grad_norm": 2.926899356932234,
+      "learning_rate": 0.00019405796691352118,
+      "loss": 1.1956,
+      "step": 2974
+    },
+    {
+      "epoch": 0.13705256368913254,
+      "grad_norm": 2.310214643717819,
+      "learning_rate": 0.00019405289913492814,
+      "loss": 1.4505,
+      "step": 2975
+    },
+    {
+      "epoch": 0.13709863177776754,
+      "grad_norm": 1.856252930355902,
+      "learning_rate": 0.00019404782926240856,
+      "loss": 1.5818,
+      "step": 2976
+    },
+    {
+      "epoch": 0.13714469986640254,
+      "grad_norm": 2.5723859988717885,
+      "learning_rate": 0.00019404275729607533,
+      "loss": 1.6491,
+      "step": 2977
+    },
+    {
+      "epoch": 0.13719076795503754,
+      "grad_norm": 1.6811102371228581,
+      "learning_rate": 0.00019403768323604138,
+      "loss": 1.5537,
+      "step": 2978
+    },
+    {
+      "epoch": 0.13723683604367254,
+      "grad_norm": 2.547273259486589,
+      "learning_rate": 0.0001940326070824196,
+      "loss": 1.5785,
+      "step": 2979
+    },
+    {
+      "epoch": 0.13728290413230754,
+      "grad_norm": 3.037268527514981,
+      "learning_rate": 0.00019402752883532307,
+      "loss": 1.1778,
+      "step": 2980
+    },
+    {
+      "epoch": 0.13732897222094256,
+      "grad_norm": 3.010361237133217,
+      "learning_rate": 0.00019402244849486483,
+      "loss": 1.1969,
+      "step": 2981
+    },
+    {
+      "epoch": 0.13737504030957756,
+      "grad_norm": 1.1958409882484897,
+      "learning_rate": 0.000194017366061158,
+      "loss": 1.5365,
+      "step": 2982
+    },
+    {
+      "epoch": 0.13742110839821256,
+      "grad_norm": 3.971994847190469,
+      "learning_rate": 0.00019401228153431572,
+      "loss": 1.2508,
+      "step": 2983
+    },
+    {
+      "epoch": 0.13746717648684756,
+      "grad_norm": 1.9419682000775655,
+      "learning_rate": 0.0001940071949144512,
+      "loss": 1.4022,
+      "step": 2984
+    },
+    {
+      "epoch": 0.13751324457548256,
+      "grad_norm": 1.8595382916560683,
+      "learning_rate": 0.00019400210620167764,
+      "loss": 0.9595,
+      "step": 2985
+    },
+    {
+      "epoch": 0.13755931266411756,
+      "grad_norm": 2.441184402051875,
+      "learning_rate": 0.00019399701539610839,
+      "loss": 1.3889,
+      "step": 2986
+    },
+    {
+      "epoch": 0.13760538075275258,
+      "grad_norm": 1.8285289221982175,
+      "learning_rate": 0.00019399192249785677,
+      "loss": 1.4551,
+      "step": 2987
+    },
+    {
+      "epoch": 0.13765144884138758,
+      "grad_norm": 7.131832150920086,
+      "learning_rate": 0.00019398682750703614,
+      "loss": 1.6128,
+      "step": 2988
+    },
+    {
+      "epoch": 0.13769751693002258,
+      "grad_norm": 1.4279705852554323,
+      "learning_rate": 0.00019398173042375994,
+      "loss": 1.465,
+      "step": 2989
+    },
+    {
+      "epoch": 0.13774358501865758,
+      "grad_norm": 2.8905947322895242,
+      "learning_rate": 0.00019397663124814165,
+      "loss": 1.5023,
+      "step": 2990
+    },
+    {
+      "epoch": 0.13778965310729258,
+      "grad_norm": 1.9025330015147854,
+      "learning_rate": 0.00019397152998029484,
+      "loss": 1.4107,
+      "step": 2991
+    },
+    {
+      "epoch": 0.13783572119592757,
+      "grad_norm": 1.6056947003358564,
+      "learning_rate": 0.000193966426620333,
+      "loss": 1.4347,
+      "step": 2992
+    },
+    {
+      "epoch": 0.13788178928456257,
+      "grad_norm": 1.6057167886124772,
+      "learning_rate": 0.00019396132116836982,
+      "loss": 1.4154,
+      "step": 2993
+    },
+    {
+      "epoch": 0.1379278573731976,
+      "grad_norm": 2.1837088203943997,
+      "learning_rate": 0.0001939562136245189,
+      "loss": 1.4148,
+      "step": 2994
+    },
+    {
+      "epoch": 0.1379739254618326,
+      "grad_norm": 1.7858382139262479,
+      "learning_rate": 0.00019395110398889403,
+      "loss": 1.6423,
+      "step": 2995
+    },
+    {
+      "epoch": 0.1380199935504676,
+      "grad_norm": 2.8552064600171887,
+      "learning_rate": 0.0001939459922616089,
+      "loss": 1.5461,
+      "step": 2996
+    },
+    {
+      "epoch": 0.1380660616391026,
+      "grad_norm": 3.5219636627624493,
+      "learning_rate": 0.00019394087844277732,
+      "loss": 1.7434,
+      "step": 2997
+    },
+    {
+      "epoch": 0.1381121297277376,
+      "grad_norm": 1.7588123305409438,
+      "learning_rate": 0.00019393576253251319,
+      "loss": 1.5186,
+      "step": 2998
+    },
+    {
+      "epoch": 0.1381581978163726,
+      "grad_norm": 1.024670647807686,
+      "learning_rate": 0.00019393064453093034,
+      "loss": 1.3919,
+      "step": 2999
+    },
+    {
+      "epoch": 0.1382042659050076,
+      "grad_norm": 1.6739739563851193,
+      "learning_rate": 0.00019392552443814274,
+      "loss": 1.7679,
+      "step": 3000
+    },
+    {
+      "epoch": 0.13825033399364262,
+      "grad_norm": 2.611575033569473,
+      "learning_rate": 0.0001939204022542644,
+      "loss": 1.3416,
+      "step": 3001
+    },
+    {
+      "epoch": 0.13829640208227761,
+      "grad_norm": 1.4152228093874792,
+      "learning_rate": 0.00019391527797940936,
+      "loss": 1.6537,
+      "step": 3002
+    },
+    {
+      "epoch": 0.1383424701709126,
+      "grad_norm": 2.587913513448412,
+      "learning_rate": 0.00019391015161369167,
+      "loss": 1.4085,
+      "step": 3003
+    },
+    {
+      "epoch": 0.1383885382595476,
+      "grad_norm": 1.7153440693266484,
+      "learning_rate": 0.00019390502315722548,
+      "loss": 1.3318,
+      "step": 3004
+    },
+    {
+      "epoch": 0.1384346063481826,
+      "grad_norm": 2.2851062439151932,
+      "learning_rate": 0.00019389989261012496,
+      "loss": 1.3579,
+      "step": 3005
+    },
+    {
+      "epoch": 0.1384806744368176,
+      "grad_norm": 1.5220366898989446,
+      "learning_rate": 0.00019389475997250432,
+      "loss": 1.3899,
+      "step": 3006
+    },
+    {
+      "epoch": 0.1385267425254526,
+      "grad_norm": 1.1869995091987267,
+      "learning_rate": 0.00019388962524447787,
+      "loss": 1.2561,
+      "step": 3007
+    },
+    {
+      "epoch": 0.13857281061408763,
+      "grad_norm": 2.7564989110291935,
+      "learning_rate": 0.00019388448842615988,
+      "loss": 1.1615,
+      "step": 3008
+    },
+    {
+      "epoch": 0.13861887870272263,
+      "grad_norm": 1.5748989669918494,
+      "learning_rate": 0.00019387934951766472,
+      "loss": 1.5867,
+      "step": 3009
+    },
+    {
+      "epoch": 0.13866494679135763,
+      "grad_norm": 1.3816570804567865,
+      "learning_rate": 0.00019387420851910684,
+      "loss": 1.6237,
+      "step": 3010
+    },
+    {
+      "epoch": 0.13871101487999263,
+      "grad_norm": 1.6662491802041348,
+      "learning_rate": 0.00019386906543060066,
+      "loss": 1.4799,
+      "step": 3011
+    },
+    {
+      "epoch": 0.13875708296862763,
+      "grad_norm": 11.991855857779678,
+      "learning_rate": 0.0001938639202522607,
+      "loss": 1.2994,
+      "step": 3012
+    },
+    {
+      "epoch": 0.13880315105726262,
+      "grad_norm": 2.014158392902162,
+      "learning_rate": 0.0001938587729842015,
+      "loss": 1.5651,
+      "step": 3013
+    },
+    {
+      "epoch": 0.13884921914589762,
+      "grad_norm": 1.762899361858972,
+      "learning_rate": 0.00019385362362653763,
+      "loss": 1.9043,
+      "step": 3014
+    },
+    {
+      "epoch": 0.13889528723453265,
+      "grad_norm": 2.9825428824601223,
+      "learning_rate": 0.00019384847217938375,
+      "loss": 1.414,
+      "step": 3015
+    },
+    {
+      "epoch": 0.13894135532316765,
+      "grad_norm": 2.332091315276757,
+      "learning_rate": 0.00019384331864285458,
+      "loss": 1.7363,
+      "step": 3016
+    },
+    {
+      "epoch": 0.13898742341180265,
+      "grad_norm": 2.827801613388813,
+      "learning_rate": 0.00019383816301706482,
+      "loss": 1.3254,
+      "step": 3017
+    },
+    {
+      "epoch": 0.13903349150043764,
+      "grad_norm": 3.9356223740968765,
+      "learning_rate": 0.00019383300530212923,
+      "loss": 1.3101,
+      "step": 3018
+    },
+    {
+      "epoch": 0.13907955958907264,
+      "grad_norm": 14.29651985835451,
+      "learning_rate": 0.0001938278454981627,
+      "loss": 1.547,
+      "step": 3019
+    },
+    {
+      "epoch": 0.13912562767770764,
+      "grad_norm": 6.071513250664917,
+      "learning_rate": 0.00019382268360528006,
+      "loss": 1.3699,
+      "step": 3020
+    },
+    {
+      "epoch": 0.13917169576634267,
+      "grad_norm": 1.7422398281690021,
+      "learning_rate": 0.00019381751962359623,
+      "loss": 1.6346,
+      "step": 3021
+    },
+    {
+      "epoch": 0.13921776385497767,
+      "grad_norm": 1.841721632220204,
+      "learning_rate": 0.0001938123535532262,
+      "loss": 1.7644,
+      "step": 3022
+    },
+    {
+      "epoch": 0.13926383194361266,
+      "grad_norm": 1.7942208488417963,
+      "learning_rate": 0.00019380718539428496,
+      "loss": 1.7229,
+      "step": 3023
+    },
+    {
+      "epoch": 0.13930990003224766,
+      "grad_norm": 1.389850727879691,
+      "learning_rate": 0.0001938020151468876,
+      "loss": 1.4871,
+      "step": 3024
+    },
+    {
+      "epoch": 0.13935596812088266,
+      "grad_norm": 2.641197378165523,
+      "learning_rate": 0.0001937968428111492,
+      "loss": 1.6458,
+      "step": 3025
+    },
+    {
+      "epoch": 0.13940203620951766,
+      "grad_norm": 2.130767368347178,
+      "learning_rate": 0.00019379166838718492,
+      "loss": 1.5975,
+      "step": 3026
+    },
+    {
+      "epoch": 0.13944810429815266,
+      "grad_norm": 1.6876070511236898,
+      "learning_rate": 0.00019378649187510994,
+      "loss": 1.4983,
+      "step": 3027
+    },
+    {
+      "epoch": 0.13949417238678768,
+      "grad_norm": 4.320094145839094,
+      "learning_rate": 0.00019378131327503958,
+      "loss": 1.6433,
+      "step": 3028
+    },
+    {
+      "epoch": 0.13954024047542268,
+      "grad_norm": 1.547665761289758,
+      "learning_rate": 0.00019377613258708904,
+      "loss": 1.3147,
+      "step": 3029
+    },
+    {
+      "epoch": 0.13958630856405768,
+      "grad_norm": 1.6647842871346479,
+      "learning_rate": 0.0001937709498113737,
+      "loss": 1.5234,
+      "step": 3030
+    },
+    {
+      "epoch": 0.13963237665269268,
+      "grad_norm": 2.482192423277411,
+      "learning_rate": 0.00019376576494800897,
+      "loss": 1.5748,
+      "step": 3031
+    },
+    {
+      "epoch": 0.13967844474132768,
+      "grad_norm": 2.463409393238185,
+      "learning_rate": 0.00019376057799711026,
+      "loss": 1.6932,
+      "step": 3032
+    },
+    {
+      "epoch": 0.13972451282996268,
+      "grad_norm": 2.3561905657640962,
+      "learning_rate": 0.000193755388958793,
+      "loss": 1.362,
+      "step": 3033
+    },
+    {
+      "epoch": 0.13977058091859768,
+      "grad_norm": 3.2606500952171147,
+      "learning_rate": 0.0001937501978331728,
+      "loss": 1.7314,
+      "step": 3034
+    },
+    {
+      "epoch": 0.1398166490072327,
+      "grad_norm": 5.08928947313532,
+      "learning_rate": 0.00019374500462036518,
+      "loss": 1.5654,
+      "step": 3035
+    },
+    {
+      "epoch": 0.1398627170958677,
+      "grad_norm": 1.3543617169358804,
+      "learning_rate": 0.00019373980932048577,
+      "loss": 1.3524,
+      "step": 3036
+    },
+    {
+      "epoch": 0.1399087851845027,
+      "grad_norm": 1.545660925852375,
+      "learning_rate": 0.00019373461193365022,
+      "loss": 1.5398,
+      "step": 3037
+    },
+    {
+      "epoch": 0.1399548532731377,
+      "grad_norm": 2.5460126214462675,
+      "learning_rate": 0.0001937294124599743,
+      "loss": 1.306,
+      "step": 3038
+    },
+    {
+      "epoch": 0.1400009213617727,
+      "grad_norm": 1.407238465721734,
+      "learning_rate": 0.0001937242108995737,
+      "loss": 1.6411,
+      "step": 3039
+    },
+    {
+      "epoch": 0.1400469894504077,
+      "grad_norm": 2.5427996260221777,
+      "learning_rate": 0.0001937190072525642,
+      "loss": 1.4721,
+      "step": 3040
+    },
+    {
+      "epoch": 0.1400930575390427,
+      "grad_norm": 4.260029814384888,
+      "learning_rate": 0.00019371380151906177,
+      "loss": 1.4168,
+      "step": 3041
+    },
+    {
+      "epoch": 0.14013912562767772,
+      "grad_norm": 17.561674416855194,
+      "learning_rate": 0.00019370859369918222,
+      "loss": 1.5589,
+      "step": 3042
+    },
+    {
+      "epoch": 0.14018519371631272,
+      "grad_norm": 1.8253682478107174,
+      "learning_rate": 0.0001937033837930415,
+      "loss": 1.1184,
+      "step": 3043
+    },
+    {
+      "epoch": 0.14023126180494772,
+      "grad_norm": 2.0556278845855456,
+      "learning_rate": 0.00019369817180075558,
+      "loss": 1.5787,
+      "step": 3044
+    },
+    {
+      "epoch": 0.14027732989358271,
+      "grad_norm": 1.7896540671513432,
+      "learning_rate": 0.00019369295772244053,
+      "loss": 1.5343,
+      "step": 3045
+    },
+    {
+      "epoch": 0.1403233979822177,
+      "grad_norm": 2.3930762835643202,
+      "learning_rate": 0.00019368774155821244,
+      "loss": 1.7384,
+      "step": 3046
+    },
+    {
+      "epoch": 0.1403694660708527,
+      "grad_norm": 2.841561471493084,
+      "learning_rate": 0.00019368252330818743,
+      "loss": 1.4343,
+      "step": 3047
+    },
+    {
+      "epoch": 0.1404155341594877,
+      "grad_norm": 2.204714642317954,
+      "learning_rate": 0.00019367730297248167,
+      "loss": 1.704,
+      "step": 3048
+    },
+    {
+      "epoch": 0.14046160224812274,
+      "grad_norm": 2.490536841064764,
+      "learning_rate": 0.00019367208055121136,
+      "loss": 1.5617,
+      "step": 3049
+    },
+    {
+      "epoch": 0.14050767033675773,
+      "grad_norm": 1.5122425525551642,
+      "learning_rate": 0.0001936668560444928,
+      "loss": 1.9618,
+      "step": 3050
+    },
+    {
+      "epoch": 0.14055373842539273,
+      "grad_norm": 2.3719454543210117,
+      "learning_rate": 0.00019366162945244232,
+      "loss": 1.6722,
+      "step": 3051
+    },
+    {
+      "epoch": 0.14059980651402773,
+      "grad_norm": 2.5125915854178187,
+      "learning_rate": 0.0001936564007751762,
+      "loss": 1.3974,
+      "step": 3052
+    },
+    {
+      "epoch": 0.14064587460266273,
+      "grad_norm": 3.6128815324314267,
+      "learning_rate": 0.0001936511700128109,
+      "loss": 1.2608,
+      "step": 3053
+    },
+    {
+      "epoch": 0.14069194269129773,
+      "grad_norm": 2.2002772185996715,
+      "learning_rate": 0.00019364593716546296,
+      "loss": 1.6003,
+      "step": 3054
+    },
+    {
+      "epoch": 0.14073801077993275,
+      "grad_norm": 6.985215936969935,
+      "learning_rate": 0.00019364070223324873,
+      "loss": 1.2869,
+      "step": 3055
+    },
+    {
+      "epoch": 0.14078407886856775,
+      "grad_norm": 16.336015233998594,
+      "learning_rate": 0.00019363546521628485,
+      "loss": 1.7349,
+      "step": 3056
+    },
+    {
+      "epoch": 0.14083014695720275,
+      "grad_norm": 3.9979718410440466,
+      "learning_rate": 0.00019363022611468787,
+      "loss": 1.6385,
+      "step": 3057
+    },
+    {
+      "epoch": 0.14087621504583775,
+      "grad_norm": 3.1485200719156143,
+      "learning_rate": 0.00019362498492857445,
+      "loss": 1.5845,
+      "step": 3058
+    },
+    {
+      "epoch": 0.14092228313447275,
+      "grad_norm": 3.6408927775054254,
+      "learning_rate": 0.0001936197416580613,
+      "loss": 1.8157,
+      "step": 3059
+    },
+    {
+      "epoch": 0.14096835122310775,
+      "grad_norm": 1.9674766401406183,
+      "learning_rate": 0.00019361449630326508,
+      "loss": 1.7719,
+      "step": 3060
+    },
+    {
+      "epoch": 0.14101441931174274,
+      "grad_norm": 3.4951285923043867,
+      "learning_rate": 0.00019360924886430264,
+      "loss": 1.4969,
+      "step": 3061
+    },
+    {
+      "epoch": 0.14106048740037777,
+      "grad_norm": 1.990068043589338,
+      "learning_rate": 0.0001936039993412908,
+      "loss": 1.8305,
+      "step": 3062
+    },
+    {
+      "epoch": 0.14110655548901277,
+      "grad_norm": 1.8928590464246775,
+      "learning_rate": 0.00019359874773434638,
+      "loss": 1.4715,
+      "step": 3063
+    },
+    {
+      "epoch": 0.14115262357764777,
+      "grad_norm": 1.7561252473540199,
+      "learning_rate": 0.00019359349404358636,
+      "loss": 1.4604,
+      "step": 3064
+    },
+    {
+      "epoch": 0.14119869166628277,
+      "grad_norm": 2.4116495992392735,
+      "learning_rate": 0.00019358823826912764,
+      "loss": 1.682,
+      "step": 3065
+    },
+    {
+      "epoch": 0.14124475975491776,
+      "grad_norm": 2.067000484872219,
+      "learning_rate": 0.00019358298041108728,
+      "loss": 1.6097,
+      "step": 3066
+    },
+    {
+      "epoch": 0.14129082784355276,
+      "grad_norm": 2.5962308048515474,
+      "learning_rate": 0.0001935777204695823,
+      "loss": 1.601,
+      "step": 3067
+    },
+    {
+      "epoch": 0.14133689593218776,
+      "grad_norm": 2.444401414517089,
+      "learning_rate": 0.00019357245844472986,
+      "loss": 1.5224,
+      "step": 3068
+    },
+    {
+      "epoch": 0.1413829640208228,
+      "grad_norm": 3.0435691630866644,
+      "learning_rate": 0.00019356719433664705,
+      "loss": 1.3252,
+      "step": 3069
+    },
+    {
+      "epoch": 0.1414290321094578,
+      "grad_norm": 3.0850952138177963,
+      "learning_rate": 0.0001935619281454511,
+      "loss": 1.2737,
+      "step": 3070
+    },
+    {
+      "epoch": 0.14147510019809278,
+      "grad_norm": 1.6923524314336729,
+      "learning_rate": 0.00019355665987125923,
+      "loss": 1.7471,
+      "step": 3071
+    },
+    {
+      "epoch": 0.14152116828672778,
+      "grad_norm": 2.884688704932528,
+      "learning_rate": 0.00019355138951418876,
+      "loss": 1.6907,
+      "step": 3072
+    },
+    {
+      "epoch": 0.14156723637536278,
+      "grad_norm": 7.016640869537686,
+      "learning_rate": 0.00019354611707435702,
+      "loss": 1.6019,
+      "step": 3073
+    },
+    {
+      "epoch": 0.14161330446399778,
+      "grad_norm": 2.179812852618325,
+      "learning_rate": 0.00019354084255188137,
+      "loss": 1.4475,
+      "step": 3074
+    },
+    {
+      "epoch": 0.14165937255263278,
+      "grad_norm": 9.856073587459433,
+      "learning_rate": 0.00019353556594687925,
+      "loss": 1.5112,
+      "step": 3075
+    },
+    {
+      "epoch": 0.1417054406412678,
+      "grad_norm": 2.7517185835583753,
+      "learning_rate": 0.0001935302872594681,
+      "loss": 1.5374,
+      "step": 3076
+    },
+    {
+      "epoch": 0.1417515087299028,
+      "grad_norm": 3.156764698832684,
+      "learning_rate": 0.0001935250064897655,
+      "loss": 1.5786,
+      "step": 3077
+    },
+    {
+      "epoch": 0.1417975768185378,
+      "grad_norm": 2.6990465068458964,
+      "learning_rate": 0.00019351972363788902,
+      "loss": 1.8306,
+      "step": 3078
+    },
+    {
+      "epoch": 0.1418436449071728,
+      "grad_norm": 5.445504309259714,
+      "learning_rate": 0.0001935144387039562,
+      "loss": 1.4724,
+      "step": 3079
+    },
+    {
+      "epoch": 0.1418897129958078,
+      "grad_norm": 1.9301484921668324,
+      "learning_rate": 0.0001935091516880848,
+      "loss": 1.6876,
+      "step": 3080
+    },
+    {
+      "epoch": 0.1419357810844428,
+      "grad_norm": 2.686496103854366,
+      "learning_rate": 0.0001935038625903924,
+      "loss": 1.5322,
+      "step": 3081
+    },
+    {
+      "epoch": 0.1419818491730778,
+      "grad_norm": 1.8362010281601797,
+      "learning_rate": 0.00019349857141099687,
+      "loss": 1.5472,
+      "step": 3082
+    },
+    {
+      "epoch": 0.14202791726171282,
+      "grad_norm": 46.41789226228838,
+      "learning_rate": 0.00019349327815001597,
+      "loss": 1.6544,
+      "step": 3083
+    },
+    {
+      "epoch": 0.14207398535034782,
+      "grad_norm": 3.917812265041287,
+      "learning_rate": 0.0001934879828075675,
+      "loss": 1.3378,
+      "step": 3084
+    },
+    {
+      "epoch": 0.14212005343898282,
+      "grad_norm": 2.8316497121665445,
+      "learning_rate": 0.0001934826853837694,
+      "loss": 1.3507,
+      "step": 3085
+    },
+    {
+      "epoch": 0.14216612152761782,
+      "grad_norm": 4.096719085266754,
+      "learning_rate": 0.00019347738587873964,
+      "loss": 2.0304,
+      "step": 3086
+    },
+    {
+      "epoch": 0.14221218961625282,
+      "grad_norm": 3.26831065462915,
+      "learning_rate": 0.0001934720842925961,
+      "loss": 1.9522,
+      "step": 3087
+    },
+    {
+      "epoch": 0.1422582577048878,
+      "grad_norm": 4.610820112766136,
+      "learning_rate": 0.00019346678062545693,
+      "loss": 2.0305,
+      "step": 3088
+    },
+    {
+      "epoch": 0.14230432579352284,
+      "grad_norm": 5.635920084862719,
+      "learning_rate": 0.00019346147487744013,
+      "loss": 1.9884,
+      "step": 3089
+    },
+    {
+      "epoch": 0.14235039388215784,
+      "grad_norm": 6.276529122982152,
+      "learning_rate": 0.00019345616704866383,
+      "loss": 1.6339,
+      "step": 3090
+    },
+    {
+      "epoch": 0.14239646197079284,
+      "grad_norm": 5.590999200062283,
+      "learning_rate": 0.00019345085713924626,
+      "loss": 1.8042,
+      "step": 3091
+    },
+    {
+      "epoch": 0.14244253005942784,
+      "grad_norm": 2.800151474213026,
+      "learning_rate": 0.00019344554514930554,
+      "loss": 1.8487,
+      "step": 3092
+    },
+    {
+      "epoch": 0.14248859814806283,
+      "grad_norm": 4.2884397002659576,
+      "learning_rate": 0.00019344023107896,
+      "loss": 1.5068,
+      "step": 3093
+    },
+    {
+      "epoch": 0.14253466623669783,
+      "grad_norm": 2.1270895563944445,
+      "learning_rate": 0.00019343491492832793,
+      "loss": 1.9714,
+      "step": 3094
+    },
+    {
+      "epoch": 0.14258073432533283,
+      "grad_norm": 4.1278773656801855,
+      "learning_rate": 0.00019342959669752766,
+      "loss": 1.9922,
+      "step": 3095
+    },
+    {
+      "epoch": 0.14262680241396786,
+      "grad_norm": 2.6528410968839675,
+      "learning_rate": 0.00019342427638667763,
+      "loss": 1.8944,
+      "step": 3096
+    },
+    {
+      "epoch": 0.14267287050260286,
+      "grad_norm": 7.03110848742673,
+      "learning_rate": 0.0001934189539958963,
+      "loss": 1.5537,
+      "step": 3097
+    },
+    {
+      "epoch": 0.14271893859123785,
+      "grad_norm": 3.824482877872851,
+      "learning_rate": 0.00019341362952530214,
+      "loss": 1.8166,
+      "step": 3098
+    },
+    {
+      "epoch": 0.14276500667987285,
+      "grad_norm": 3.185025699726533,
+      "learning_rate": 0.00019340830297501365,
+      "loss": 1.8202,
+      "step": 3099
+    },
+    {
+      "epoch": 0.14281107476850785,
+      "grad_norm": 5.494960069075554,
+      "learning_rate": 0.0001934029743451495,
+      "loss": 1.9654,
+      "step": 3100
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 2.6346295247350344,
+      "learning_rate": 0.00019339764363582825,
+      "loss": 1.831,
+      "step": 3101
+    },
+    {
+      "epoch": 0.14290321094577785,
+      "grad_norm": 4.0190406292607985,
+      "learning_rate": 0.0001933923108471686,
+      "loss": 1.6428,
+      "step": 3102
+    },
+    {
+      "epoch": 0.14294927903441287,
+      "grad_norm": 3.1019854232737574,
+      "learning_rate": 0.00019338697597928928,
+      "loss": 1.7445,
+      "step": 3103
+    },
+    {
+      "epoch": 0.14299534712304787,
+      "grad_norm": 3.8490878834398683,
+      "learning_rate": 0.0001933816390323091,
+      "loss": 1.512,
+      "step": 3104
+    },
+    {
+      "epoch": 0.14304141521168287,
+      "grad_norm": 3.8040436397680897,
+      "learning_rate": 0.00019337630000634682,
+      "loss": 1.8926,
+      "step": 3105
+    },
+    {
+      "epoch": 0.14308748330031787,
+      "grad_norm": 4.051946226523484,
+      "learning_rate": 0.0001933709589015213,
+      "loss": 1.5578,
+      "step": 3106
+    },
+    {
+      "epoch": 0.14313355138895287,
+      "grad_norm": 4.286796817620762,
+      "learning_rate": 0.00019336561571795153,
+      "loss": 1.9215,
+      "step": 3107
+    },
+    {
+      "epoch": 0.14317961947758787,
+      "grad_norm": 3.4922451011009334,
+      "learning_rate": 0.00019336027045575637,
+      "loss": 1.6725,
+      "step": 3108
+    },
+    {
+      "epoch": 0.14322568756622286,
+      "grad_norm": 2.824168372386935,
+      "learning_rate": 0.00019335492311505487,
+      "loss": 2.0948,
+      "step": 3109
+    },
+    {
+      "epoch": 0.1432717556548579,
+      "grad_norm": 2.0908978852543023,
+      "learning_rate": 0.0001933495736959661,
+      "loss": 1.635,
+      "step": 3110
+    },
+    {
+      "epoch": 0.1433178237434929,
+      "grad_norm": 2.598062770365577,
+      "learning_rate": 0.00019334422219860912,
+      "loss": 1.6986,
+      "step": 3111
+    },
+    {
+      "epoch": 0.1433638918321279,
+      "grad_norm": 2.67883809431684,
+      "learning_rate": 0.00019333886862310312,
+      "loss": 1.7703,
+      "step": 3112
+    },
+    {
+      "epoch": 0.14340995992076289,
+      "grad_norm": 3.7462509357928644,
+      "learning_rate": 0.0001933335129695672,
+      "loss": 1.5426,
+      "step": 3113
+    },
+    {
+      "epoch": 0.14345602800939788,
+      "grad_norm": 3.8352579037634573,
+      "learning_rate": 0.00019332815523812066,
+      "loss": 1.4536,
+      "step": 3114
+    },
+    {
+      "epoch": 0.14350209609803288,
+      "grad_norm": 1.8409611834881556,
+      "learning_rate": 0.00019332279542888276,
+      "loss": 1.5036,
+      "step": 3115
+    },
+    {
+      "epoch": 0.14354816418666788,
+      "grad_norm": 1.9247591252594718,
+      "learning_rate": 0.00019331743354197286,
+      "loss": 1.382,
+      "step": 3116
+    },
+    {
+      "epoch": 0.1435942322753029,
+      "grad_norm": 2.278881842164614,
+      "learning_rate": 0.0001933120695775103,
+      "loss": 1.5531,
+      "step": 3117
+    },
+    {
+      "epoch": 0.1436403003639379,
+      "grad_norm": 2.2140406974056983,
+      "learning_rate": 0.0001933067035356145,
+      "loss": 1.5517,
+      "step": 3118
+    },
+    {
+      "epoch": 0.1436863684525729,
+      "grad_norm": 2.576751881948214,
+      "learning_rate": 0.0001933013354164049,
+      "loss": 1.2646,
+      "step": 3119
+    },
+    {
+      "epoch": 0.1437324365412079,
+      "grad_norm": 2.3680951457002397,
+      "learning_rate": 0.0001932959652200011,
+      "loss": 1.1958,
+      "step": 3120
+    },
+    {
+      "epoch": 0.1437785046298429,
+      "grad_norm": 1.652163098523988,
+      "learning_rate": 0.00019329059294652257,
+      "loss": 1.8905,
+      "step": 3121
+    },
+    {
+      "epoch": 0.1438245727184779,
+      "grad_norm": 5.155174543028785,
+      "learning_rate": 0.00019328521859608897,
+      "loss": 1.4862,
+      "step": 3122
+    },
+    {
+      "epoch": 0.14387064080711293,
+      "grad_norm": 1.5922463918600565,
+      "learning_rate": 0.0001932798421688199,
+      "loss": 1.5299,
+      "step": 3123
+    },
+    {
+      "epoch": 0.14391670889574792,
+      "grad_norm": 1.851803202480659,
+      "learning_rate": 0.0001932744636648351,
+      "loss": 1.5364,
+      "step": 3124
+    },
+    {
+      "epoch": 0.14396277698438292,
+      "grad_norm": 1.2253635583350844,
+      "learning_rate": 0.0001932690830842543,
+      "loss": 1.5829,
+      "step": 3125
+    },
+    {
+      "epoch": 0.14400884507301792,
+      "grad_norm": 2.0385283631684286,
+      "learning_rate": 0.00019326370042719728,
+      "loss": 1.6058,
+      "step": 3126
+    },
+    {
+      "epoch": 0.14405491316165292,
+      "grad_norm": 2.7117876700959487,
+      "learning_rate": 0.0001932583156937839,
+      "loss": 1.2503,
+      "step": 3127
+    },
+    {
+      "epoch": 0.14410098125028792,
+      "grad_norm": 2.002169862517079,
+      "learning_rate": 0.000193252928884134,
+      "loss": 1.3945,
+      "step": 3128
+    },
+    {
+      "epoch": 0.14414704933892292,
+      "grad_norm": 2.7996505005700967,
+      "learning_rate": 0.00019324753999836754,
+      "loss": 1.635,
+      "step": 3129
+    },
+    {
+      "epoch": 0.14419311742755794,
+      "grad_norm": 2.816799760668648,
+      "learning_rate": 0.0001932421490366045,
+      "loss": 1.4996,
+      "step": 3130
+    },
+    {
+      "epoch": 0.14423918551619294,
+      "grad_norm": 2.2141797553632525,
+      "learning_rate": 0.0001932367559989649,
+      "loss": 1.7774,
+      "step": 3131
+    },
+    {
+      "epoch": 0.14428525360482794,
+      "grad_norm": 2.39444344830856,
+      "learning_rate": 0.00019323136088556875,
+      "loss": 1.4631,
+      "step": 3132
+    },
+    {
+      "epoch": 0.14433132169346294,
+      "grad_norm": 3.721317435450103,
+      "learning_rate": 0.00019322596369653625,
+      "loss": 1.3385,
+      "step": 3133
+    },
+    {
+      "epoch": 0.14437738978209794,
+      "grad_norm": 2.157714692233336,
+      "learning_rate": 0.0001932205644319875,
+      "loss": 1.6321,
+      "step": 3134
+    },
+    {
+      "epoch": 0.14442345787073294,
+      "grad_norm": 2.892531434652211,
+      "learning_rate": 0.00019321516309204272,
+      "loss": 1.496,
+      "step": 3135
+    },
+    {
+      "epoch": 0.14446952595936793,
+      "grad_norm": 3.054795460044118,
+      "learning_rate": 0.00019320975967682217,
+      "loss": 1.6077,
+      "step": 3136
+    },
+    {
+      "epoch": 0.14451559404800296,
+      "grad_norm": 1.7959431254040852,
+      "learning_rate": 0.0001932043541864461,
+      "loss": 1.6206,
+      "step": 3137
+    },
+    {
+      "epoch": 0.14456166213663796,
+      "grad_norm": 2.1341296173126736,
+      "learning_rate": 0.00019319894662103495,
+      "loss": 1.5268,
+      "step": 3138
+    },
+    {
+      "epoch": 0.14460773022527296,
+      "grad_norm": 1.940737577653498,
+      "learning_rate": 0.00019319353698070904,
+      "loss": 1.5323,
+      "step": 3139
+    },
+    {
+      "epoch": 0.14465379831390796,
+      "grad_norm": 1.5433564754353597,
+      "learning_rate": 0.00019318812526558882,
+      "loss": 1.8824,
+      "step": 3140
+    },
+    {
+      "epoch": 0.14469986640254295,
+      "grad_norm": 1.7736221620520918,
+      "learning_rate": 0.00019318271147579478,
+      "loss": 1.4887,
+      "step": 3141
+    },
+    {
+      "epoch": 0.14474593449117795,
+      "grad_norm": 2.5204352747172942,
+      "learning_rate": 0.00019317729561144743,
+      "loss": 1.6667,
+      "step": 3142
+    },
+    {
+      "epoch": 0.14479200257981295,
+      "grad_norm": 1.5365735179639426,
+      "learning_rate": 0.00019317187767266733,
+      "loss": 1.3093,
+      "step": 3143
+    },
+    {
+      "epoch": 0.14483807066844798,
+      "grad_norm": 2.8550499949871955,
+      "learning_rate": 0.00019316645765957518,
+      "loss": 1.4806,
+      "step": 3144
+    },
+    {
+      "epoch": 0.14488413875708298,
+      "grad_norm": 7.540276326683114,
+      "learning_rate": 0.00019316103557229158,
+      "loss": 1.5824,
+      "step": 3145
+    },
+    {
+      "epoch": 0.14493020684571797,
+      "grad_norm": 2.310601327504833,
+      "learning_rate": 0.00019315561141093726,
+      "loss": 1.5239,
+      "step": 3146
+    },
+    {
+      "epoch": 0.14497627493435297,
+      "grad_norm": 2.932606882790063,
+      "learning_rate": 0.00019315018517563297,
+      "loss": 1.4827,
+      "step": 3147
+    },
+    {
+      "epoch": 0.14502234302298797,
+      "grad_norm": 2.6703023474092937,
+      "learning_rate": 0.00019314475686649953,
+      "loss": 1.4506,
+      "step": 3148
+    },
+    {
+      "epoch": 0.14506841111162297,
+      "grad_norm": 3.9791058078408263,
+      "learning_rate": 0.0001931393264836578,
+      "loss": 1.5661,
+      "step": 3149
+    },
+    {
+      "epoch": 0.14511447920025797,
+      "grad_norm": 3.3326770705293964,
+      "learning_rate": 0.00019313389402722867,
+      "loss": 1.4885,
+      "step": 3150
+    },
+    {
+      "epoch": 0.145160547288893,
+      "grad_norm": 3.370889801290939,
+      "learning_rate": 0.00019312845949733304,
+      "loss": 1.6311,
+      "step": 3151
+    },
+    {
+      "epoch": 0.145206615377528,
+      "grad_norm": 2.5956041865577983,
+      "learning_rate": 0.000193123022894092,
+      "loss": 1.6496,
+      "step": 3152
+    },
+    {
+      "epoch": 0.145252683466163,
+      "grad_norm": 2.4132362391824285,
+      "learning_rate": 0.00019311758421762648,
+      "loss": 1.4575,
+      "step": 3153
+    },
+    {
+      "epoch": 0.145298751554798,
+      "grad_norm": 1.9283437908425023,
+      "learning_rate": 0.00019311214346805761,
+      "loss": 1.7317,
+      "step": 3154
+    },
+    {
+      "epoch": 0.145344819643433,
+      "grad_norm": 2.2008835904601365,
+      "learning_rate": 0.00019310670064550653,
+      "loss": 1.6957,
+      "step": 3155
+    },
+    {
+      "epoch": 0.14539088773206799,
+      "grad_norm": 2.130655971922409,
+      "learning_rate": 0.0001931012557500944,
+      "loss": 1.441,
+      "step": 3156
+    },
+    {
+      "epoch": 0.145436955820703,
+      "grad_norm": 3.727504336198602,
+      "learning_rate": 0.00019309580878194247,
+      "loss": 1.4175,
+      "step": 3157
+    },
+    {
+      "epoch": 0.145483023909338,
+      "grad_norm": 3.5645907769516625,
+      "learning_rate": 0.00019309035974117195,
+      "loss": 1.5221,
+      "step": 3158
+    },
+    {
+      "epoch": 0.145529091997973,
+      "grad_norm": 3.1785499972098807,
+      "learning_rate": 0.0001930849086279042,
+      "loss": 1.3673,
+      "step": 3159
+    },
+    {
+      "epoch": 0.145575160086608,
+      "grad_norm": 1.905300579635911,
+      "learning_rate": 0.0001930794554422606,
+      "loss": 1.4387,
+      "step": 3160
+    },
+    {
+      "epoch": 0.145621228175243,
+      "grad_norm": 1.3115132992761036,
+      "learning_rate": 0.00019307400018436246,
+      "loss": 1.6833,
+      "step": 3161
+    },
+    {
+      "epoch": 0.145667296263878,
+      "grad_norm": 2.0338599444484795,
+      "learning_rate": 0.00019306854285433133,
+      "loss": 1.3884,
+      "step": 3162
+    },
+    {
+      "epoch": 0.145713364352513,
+      "grad_norm": 6.243656832552135,
+      "learning_rate": 0.00019306308345228866,
+      "loss": 1.6154,
+      "step": 3163
+    },
+    {
+      "epoch": 0.14575943244114803,
+      "grad_norm": 3.59615119743648,
+      "learning_rate": 0.000193057621978356,
+      "loss": 1.2926,
+      "step": 3164
+    },
+    {
+      "epoch": 0.14580550052978303,
+      "grad_norm": 1.7142737732217073,
+      "learning_rate": 0.00019305215843265497,
+      "loss": 1.5389,
+      "step": 3165
+    },
+    {
+      "epoch": 0.14585156861841803,
+      "grad_norm": 4.351064118177928,
+      "learning_rate": 0.00019304669281530715,
+      "loss": 1.5097,
+      "step": 3166
+    },
+    {
+      "epoch": 0.14589763670705302,
+      "grad_norm": 3.381108262464172,
+      "learning_rate": 0.00019304122512643428,
+      "loss": 1.3703,
+      "step": 3167
+    },
+    {
+      "epoch": 0.14594370479568802,
+      "grad_norm": 5.437209570875597,
+      "learning_rate": 0.00019303575536615806,
+      "loss": 1.6991,
+      "step": 3168
+    },
+    {
+      "epoch": 0.14598977288432302,
+      "grad_norm": 2.3468024547820847,
+      "learning_rate": 0.00019303028353460024,
+      "loss": 1.4937,
+      "step": 3169
+    },
+    {
+      "epoch": 0.14603584097295802,
+      "grad_norm": 3.175325849466768,
+      "learning_rate": 0.0001930248096318827,
+      "loss": 1.4516,
+      "step": 3170
+    },
+    {
+      "epoch": 0.14608190906159305,
+      "grad_norm": 3.079745754221913,
+      "learning_rate": 0.00019301933365812728,
+      "loss": 1.7213,
+      "step": 3171
+    },
+    {
+      "epoch": 0.14612797715022804,
+      "grad_norm": 2.907725426738984,
+      "learning_rate": 0.0001930138556134559,
+      "loss": 1.4335,
+      "step": 3172
+    },
+    {
+      "epoch": 0.14617404523886304,
+      "grad_norm": 3.163070257538416,
+      "learning_rate": 0.00019300837549799047,
+      "loss": 1.371,
+      "step": 3173
+    },
+    {
+      "epoch": 0.14622011332749804,
+      "grad_norm": 3.3328112563286947,
+      "learning_rate": 0.00019300289331185307,
+      "loss": 1.7183,
+      "step": 3174
+    },
+    {
+      "epoch": 0.14626618141613304,
+      "grad_norm": 4.306985739346174,
+      "learning_rate": 0.0001929974090551657,
+      "loss": 1.5582,
+      "step": 3175
+    },
+    {
+      "epoch": 0.14631224950476804,
+      "grad_norm": 2.6298237163710287,
+      "learning_rate": 0.00019299192272805046,
+      "loss": 1.6348,
+      "step": 3176
+    },
+    {
+      "epoch": 0.14635831759340304,
+      "grad_norm": 5.470149461159147,
+      "learning_rate": 0.00019298643433062954,
+      "loss": 1.6662,
+      "step": 3177
+    },
+    {
+      "epoch": 0.14640438568203806,
+      "grad_norm": 3.330678669156452,
+      "learning_rate": 0.00019298094386302505,
+      "loss": 1.3804,
+      "step": 3178
+    },
+    {
+      "epoch": 0.14645045377067306,
+      "grad_norm": 6.909224681364918,
+      "learning_rate": 0.00019297545132535932,
+      "loss": 1.8473,
+      "step": 3179
+    },
+    {
+      "epoch": 0.14649652185930806,
+      "grad_norm": 7.476040177978619,
+      "learning_rate": 0.00019296995671775458,
+      "loss": 1.3146,
+      "step": 3180
+    },
+    {
+      "epoch": 0.14654258994794306,
+      "grad_norm": 8.042951921040144,
+      "learning_rate": 0.00019296446004033312,
+      "loss": 1.6434,
+      "step": 3181
+    },
+    {
+      "epoch": 0.14658865803657806,
+      "grad_norm": 5.39052072346152,
+      "learning_rate": 0.0001929589612932174,
+      "loss": 1.8314,
+      "step": 3182
+    },
+    {
+      "epoch": 0.14663472612521306,
+      "grad_norm": 2.839083768032184,
+      "learning_rate": 0.0001929534604765298,
+      "loss": 1.3797,
+      "step": 3183
+    },
+    {
+      "epoch": 0.14668079421384805,
+      "grad_norm": 4.1924040667514335,
+      "learning_rate": 0.00019294795759039276,
+      "loss": 2.0212,
+      "step": 3184
+    },
+    {
+      "epoch": 0.14672686230248308,
+      "grad_norm": 3.4027809890362906,
+      "learning_rate": 0.00019294245263492882,
+      "loss": 1.9281,
+      "step": 3185
+    },
+    {
+      "epoch": 0.14677293039111808,
+      "grad_norm": 2.8674285645343467,
+      "learning_rate": 0.00019293694561026056,
+      "loss": 1.6851,
+      "step": 3186
+    },
+    {
+      "epoch": 0.14681899847975308,
+      "grad_norm": 20.17403389026359,
+      "learning_rate": 0.00019293143651651055,
+      "loss": 1.1372,
+      "step": 3187
+    },
+    {
+      "epoch": 0.14686506656838808,
+      "grad_norm": 9.764156854327267,
+      "learning_rate": 0.00019292592535380144,
+      "loss": 1.542,
+      "step": 3188
+    },
+    {
+      "epoch": 0.14691113465702307,
+      "grad_norm": 3.9334559989832845,
+      "learning_rate": 0.0001929204121222559,
+      "loss": 1.7015,
+      "step": 3189
+    },
+    {
+      "epoch": 0.14695720274565807,
+      "grad_norm": 3.2279259330232657,
+      "learning_rate": 0.00019291489682199674,
+      "loss": 2.1903,
+      "step": 3190
+    },
+    {
+      "epoch": 0.1470032708342931,
+      "grad_norm": 2.59764028783326,
+      "learning_rate": 0.00019290937945314673,
+      "loss": 1.7479,
+      "step": 3191
+    },
+    {
+      "epoch": 0.1470493389229281,
+      "grad_norm": 7.315428489121801,
+      "learning_rate": 0.00019290386001582867,
+      "loss": 1.9314,
+      "step": 3192
+    },
+    {
+      "epoch": 0.1470954070115631,
+      "grad_norm": 2.460637132969145,
+      "learning_rate": 0.00019289833851016548,
+      "loss": 1.8275,
+      "step": 3193
+    },
+    {
+      "epoch": 0.1471414751001981,
+      "grad_norm": 2.7466741317017647,
+      "learning_rate": 0.00019289281493628006,
+      "loss": 1.9602,
+      "step": 3194
+    },
+    {
+      "epoch": 0.1471875431888331,
+      "grad_norm": 3.8782027882403027,
+      "learning_rate": 0.0001928872892942954,
+      "loss": 1.7111,
+      "step": 3195
+    },
+    {
+      "epoch": 0.1472336112774681,
+      "grad_norm": 3.832063005059847,
+      "learning_rate": 0.0001928817615843345,
+      "loss": 1.7238,
+      "step": 3196
+    },
+    {
+      "epoch": 0.1472796793661031,
+      "grad_norm": 2.408246213342245,
+      "learning_rate": 0.00019287623180652043,
+      "loss": 1.7661,
+      "step": 3197
+    },
+    {
+      "epoch": 0.14732574745473812,
+      "grad_norm": 2.333973372616471,
+      "learning_rate": 0.00019287069996097635,
+      "loss": 1.5728,
+      "step": 3198
+    },
+    {
+      "epoch": 0.1473718155433731,
+      "grad_norm": 1.888497245990771,
+      "learning_rate": 0.00019286516604782533,
+      "loss": 1.7873,
+      "step": 3199
+    },
+    {
+      "epoch": 0.1474178836320081,
+      "grad_norm": 3.1118596019930997,
+      "learning_rate": 0.00019285963006719065,
+      "loss": 1.5799,
+      "step": 3200
+    },
+    {
+      "epoch": 0.1474639517206431,
+      "grad_norm": 2.626984634591532,
+      "learning_rate": 0.00019285409201919552,
+      "loss": 1.8115,
+      "step": 3201
+    },
+    {
+      "epoch": 0.1475100198092781,
+      "grad_norm": 3.893354395841899,
+      "learning_rate": 0.0001928485519039632,
+      "loss": 1.554,
+      "step": 3202
+    },
+    {
+      "epoch": 0.1475560878979131,
+      "grad_norm": 4.125952461145819,
+      "learning_rate": 0.00019284300972161712,
+      "loss": 1.5119,
+      "step": 3203
+    },
+    {
+      "epoch": 0.1476021559865481,
+      "grad_norm": 1.4560682346883125,
+      "learning_rate": 0.0001928374654722806,
+      "loss": 1.6793,
+      "step": 3204
+    },
+    {
+      "epoch": 0.14764822407518313,
+      "grad_norm": 3.644387551246586,
+      "learning_rate": 0.0001928319191560771,
+      "loss": 1.8897,
+      "step": 3205
+    },
+    {
+      "epoch": 0.14769429216381813,
+      "grad_norm": 2.381874675943646,
+      "learning_rate": 0.00019282637077313008,
+      "loss": 1.7807,
+      "step": 3206
+    },
+    {
+      "epoch": 0.14774036025245313,
+      "grad_norm": 1.7411798988974403,
+      "learning_rate": 0.00019282082032356308,
+      "loss": 1.562,
+      "step": 3207
+    },
+    {
+      "epoch": 0.14778642834108813,
+      "grad_norm": 2.3991113647817714,
+      "learning_rate": 0.0001928152678074997,
+      "loss": 1.4765,
+      "step": 3208
+    },
+    {
+      "epoch": 0.14783249642972313,
+      "grad_norm": 1.7309427816073895,
+      "learning_rate": 0.00019280971322506348,
+      "loss": 1.6548,
+      "step": 3209
+    },
+    {
+      "epoch": 0.14787856451835812,
+      "grad_norm": 1.8417653468978086,
+      "learning_rate": 0.00019280415657637817,
+      "loss": 1.4818,
+      "step": 3210
+    },
+    {
+      "epoch": 0.14792463260699312,
+      "grad_norm": 2.7196936263607054,
+      "learning_rate": 0.00019279859786156743,
+      "loss": 1.3444,
+      "step": 3211
+    },
+    {
+      "epoch": 0.14797070069562815,
+      "grad_norm": 1.452988818801096,
+      "learning_rate": 0.000192793037080755,
+      "loss": 1.7091,
+      "step": 3212
+    },
+    {
+      "epoch": 0.14801676878426315,
+      "grad_norm": 2.381350640939192,
+      "learning_rate": 0.00019278747423406473,
+      "loss": 1.6341,
+      "step": 3213
+    },
+    {
+      "epoch": 0.14806283687289815,
+      "grad_norm": 1.3702332456265776,
+      "learning_rate": 0.00019278190932162044,
+      "loss": 1.4664,
+      "step": 3214
+    },
+    {
+      "epoch": 0.14810890496153314,
+      "grad_norm": 3.2545327310087564,
+      "learning_rate": 0.00019277634234354603,
+      "loss": 1.6721,
+      "step": 3215
+    },
+    {
+      "epoch": 0.14815497305016814,
+      "grad_norm": 1.057910739387227,
+      "learning_rate": 0.0001927707732999654,
+      "loss": 1.5912,
+      "step": 3216
+    },
+    {
+      "epoch": 0.14820104113880314,
+      "grad_norm": 1.6163184599252438,
+      "learning_rate": 0.0001927652021910026,
+      "loss": 1.4773,
+      "step": 3217
+    },
+    {
+      "epoch": 0.14824710922743817,
+      "grad_norm": 1.9670363732005383,
+      "learning_rate": 0.00019275962901678164,
+      "loss": 1.5536,
+      "step": 3218
+    },
+    {
+      "epoch": 0.14829317731607317,
+      "grad_norm": 2.4446069053375896,
+      "learning_rate": 0.00019275405377742658,
+      "loss": 1.6669,
+      "step": 3219
+    },
+    {
+      "epoch": 0.14833924540470816,
+      "grad_norm": 1.8180734530646168,
+      "learning_rate": 0.00019274847647306152,
+      "loss": 1.3905,
+      "step": 3220
+    },
+    {
+      "epoch": 0.14838531349334316,
+      "grad_norm": 1.7805314437985216,
+      "learning_rate": 0.00019274289710381068,
+      "loss": 1.664,
+      "step": 3221
+    },
+    {
+      "epoch": 0.14843138158197816,
+      "grad_norm": 3.801881930741384,
+      "learning_rate": 0.00019273731566979828,
+      "loss": 1.5525,
+      "step": 3222
+    },
+    {
+      "epoch": 0.14847744967061316,
+      "grad_norm": 4.0804427919522395,
+      "learning_rate": 0.00019273173217114855,
+      "loss": 1.4263,
+      "step": 3223
+    },
+    {
+      "epoch": 0.14852351775924816,
+      "grad_norm": 2.532660517900373,
+      "learning_rate": 0.00019272614660798579,
+      "loss": 1.6537,
+      "step": 3224
+    },
+    {
+      "epoch": 0.14856958584788318,
+      "grad_norm": 1.3767268190909525,
+      "learning_rate": 0.00019272055898043435,
+      "loss": 1.7121,
+      "step": 3225
+    },
+    {
+      "epoch": 0.14861565393651818,
+      "grad_norm": 2.4067493790694474,
+      "learning_rate": 0.0001927149692886187,
+      "loss": 1.6578,
+      "step": 3226
+    },
+    {
+      "epoch": 0.14866172202515318,
+      "grad_norm": 2.527444547999843,
+      "learning_rate": 0.00019270937753266317,
+      "loss": 1.6285,
+      "step": 3227
+    },
+    {
+      "epoch": 0.14870779011378818,
+      "grad_norm": 4.432645359328785,
+      "learning_rate": 0.00019270378371269236,
+      "loss": 1.4167,
+      "step": 3228
+    },
+    {
+      "epoch": 0.14875385820242318,
+      "grad_norm": 2.3668575796946496,
+      "learning_rate": 0.00019269818782883074,
+      "loss": 1.7037,
+      "step": 3229
+    },
+    {
+      "epoch": 0.14879992629105818,
+      "grad_norm": 2.5972689104447353,
+      "learning_rate": 0.00019269258988120292,
+      "loss": 1.5448,
+      "step": 3230
+    },
+    {
+      "epoch": 0.14884599437969318,
+      "grad_norm": 3.8899763462492998,
+      "learning_rate": 0.0001926869898699335,
+      "loss": 1.5006,
+      "step": 3231
+    },
+    {
+      "epoch": 0.1488920624683282,
+      "grad_norm": 6.184971634708749,
+      "learning_rate": 0.00019268138779514717,
+      "loss": 1.4519,
+      "step": 3232
+    },
+    {
+      "epoch": 0.1489381305569632,
+      "grad_norm": 1.7391249128016297,
+      "learning_rate": 0.00019267578365696867,
+      "loss": 1.893,
+      "step": 3233
+    },
+    {
+      "epoch": 0.1489841986455982,
+      "grad_norm": 1.8408927907086001,
+      "learning_rate": 0.00019267017745552275,
+      "loss": 1.5544,
+      "step": 3234
+    },
+    {
+      "epoch": 0.1490302667342332,
+      "grad_norm": 2.2528978174974688,
+      "learning_rate": 0.00019266456919093424,
+      "loss": 1.5122,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1490763348228682,
+      "grad_norm": 3.425764684805732,
+      "learning_rate": 0.00019265895886332797,
+      "loss": 1.4221,
+      "step": 3236
+    },
+    {
+      "epoch": 0.1491224029115032,
+      "grad_norm": 1.4365215254725898,
+      "learning_rate": 0.00019265334647282885,
+      "loss": 1.5764,
+      "step": 3237
+    },
+    {
+      "epoch": 0.1491684710001382,
+      "grad_norm": 3.051979670413973,
+      "learning_rate": 0.00019264773201956183,
+      "loss": 1.6782,
+      "step": 3238
+    },
+    {
+      "epoch": 0.14921453908877322,
+      "grad_norm": 2.3934593811663056,
+      "learning_rate": 0.00019264211550365192,
+      "loss": 1.6351,
+      "step": 3239
+    },
+    {
+      "epoch": 0.14926060717740822,
+      "grad_norm": 1.7253722192330287,
+      "learning_rate": 0.00019263649692522416,
+      "loss": 1.3987,
+      "step": 3240
+    },
+    {
+      "epoch": 0.14930667526604322,
+      "grad_norm": 2.1395888332660915,
+      "learning_rate": 0.00019263087628440366,
+      "loss": 1.2913,
+      "step": 3241
+    },
+    {
+      "epoch": 0.1493527433546782,
+      "grad_norm": 1.2404811409284344,
+      "learning_rate": 0.00019262525358131548,
+      "loss": 1.5512,
+      "step": 3242
+    },
+    {
+      "epoch": 0.1493988114433132,
+      "grad_norm": 2.163343226998092,
+      "learning_rate": 0.0001926196288160849,
+      "loss": 1.7631,
+      "step": 3243
+    },
+    {
+      "epoch": 0.1494448795319482,
+      "grad_norm": 1.7335527903080474,
+      "learning_rate": 0.00019261400198883704,
+      "loss": 1.567,
+      "step": 3244
+    },
+    {
+      "epoch": 0.1494909476205832,
+      "grad_norm": 5.028555254084781,
+      "learning_rate": 0.00019260837309969725,
+      "loss": 1.351,
+      "step": 3245
+    },
+    {
+      "epoch": 0.14953701570921824,
+      "grad_norm": 3.136559337216204,
+      "learning_rate": 0.00019260274214879082,
+      "loss": 1.6315,
+      "step": 3246
+    },
+    {
+      "epoch": 0.14958308379785323,
+      "grad_norm": 2.756531132687648,
+      "learning_rate": 0.00019259710913624312,
+      "loss": 1.967,
+      "step": 3247
+    },
+    {
+      "epoch": 0.14962915188648823,
+      "grad_norm": 9.840097492414056,
+      "learning_rate": 0.00019259147406217955,
+      "loss": 1.2724,
+      "step": 3248
+    },
+    {
+      "epoch": 0.14967521997512323,
+      "grad_norm": 2.81960751950768,
+      "learning_rate": 0.0001925858369267256,
+      "loss": 1.6706,
+      "step": 3249
+    },
+    {
+      "epoch": 0.14972128806375823,
+      "grad_norm": 13.374717966959349,
+      "learning_rate": 0.0001925801977300067,
+      "loss": 1.4287,
+      "step": 3250
+    },
+    {
+      "epoch": 0.14976735615239323,
+      "grad_norm": 1.6448991720115962,
+      "learning_rate": 0.00019257455647214843,
+      "loss": 1.7427,
+      "step": 3251
+    },
+    {
+      "epoch": 0.14981342424102825,
+      "grad_norm": 1.7531467286537472,
+      "learning_rate": 0.0001925689131532764,
+      "loss": 1.5289,
+      "step": 3252
+    },
+    {
+      "epoch": 0.14985949232966325,
+      "grad_norm": 5.414986407082177,
+      "learning_rate": 0.00019256326777351626,
+      "loss": 1.6204,
+      "step": 3253
+    },
+    {
+      "epoch": 0.14990556041829825,
+      "grad_norm": 2.890340652333637,
+      "learning_rate": 0.0001925576203329937,
+      "loss": 1.7108,
+      "step": 3254
+    },
+    {
+      "epoch": 0.14995162850693325,
+      "grad_norm": 3.9551613279461026,
+      "learning_rate": 0.00019255197083183436,
+      "loss": 1.6626,
+      "step": 3255
+    },
+    {
+      "epoch": 0.14999769659556825,
+      "grad_norm": 3.2547021022155374,
+      "learning_rate": 0.0001925463192701641,
+      "loss": 1.7467,
+      "step": 3256
+    },
+    {
+      "epoch": 0.15004376468420325,
+      "grad_norm": 4.972313513147654,
+      "learning_rate": 0.00019254066564810876,
+      "loss": 1.5087,
+      "step": 3257
+    },
+    {
+      "epoch": 0.15008983277283824,
+      "grad_norm": 2.3301776004429264,
+      "learning_rate": 0.00019253500996579414,
+      "loss": 1.396,
+      "step": 3258
+    },
+    {
+      "epoch": 0.15013590086147327,
+      "grad_norm": 2.715692465485371,
+      "learning_rate": 0.00019252935222334623,
+      "loss": 1.4403,
+      "step": 3259
+    },
+    {
+      "epoch": 0.15018196895010827,
+      "grad_norm": 2.177307520847773,
+      "learning_rate": 0.0001925236924208909,
+      "loss": 1.8041,
+      "step": 3260
+    },
+    {
+      "epoch": 0.15022803703874327,
+      "grad_norm": 2.773408653218445,
+      "learning_rate": 0.00019251803055855422,
+      "loss": 1.8005,
+      "step": 3261
+    },
+    {
+      "epoch": 0.15027410512737827,
+      "grad_norm": 1.995258914391861,
+      "learning_rate": 0.00019251236663646223,
+      "loss": 1.6259,
+      "step": 3262
+    },
+    {
+      "epoch": 0.15032017321601326,
+      "grad_norm": 4.061256580448886,
+      "learning_rate": 0.00019250670065474102,
+      "loss": 1.7772,
+      "step": 3263
+    },
+    {
+      "epoch": 0.15036624130464826,
+      "grad_norm": 3.549969924430775,
+      "learning_rate": 0.00019250103261351674,
+      "loss": 1.7048,
+      "step": 3264
+    },
+    {
+      "epoch": 0.15041230939328326,
+      "grad_norm": 2.5987848153077526,
+      "learning_rate": 0.00019249536251291554,
+      "loss": 1.261,
+      "step": 3265
+    },
+    {
+      "epoch": 0.1504583774819183,
+      "grad_norm": 2.4798164332484665,
+      "learning_rate": 0.0001924896903530637,
+      "loss": 1.5916,
+      "step": 3266
+    },
+    {
+      "epoch": 0.15050444557055329,
+      "grad_norm": 1.7057289354618326,
+      "learning_rate": 0.0001924840161340875,
+      "loss": 1.5725,
+      "step": 3267
+    },
+    {
+      "epoch": 0.15055051365918828,
+      "grad_norm": 3.0027260804777254,
+      "learning_rate": 0.00019247833985611327,
+      "loss": 1.5434,
+      "step": 3268
+    },
+    {
+      "epoch": 0.15059658174782328,
+      "grad_norm": 2.1530605697838547,
+      "learning_rate": 0.00019247266151926735,
+      "loss": 1.3282,
+      "step": 3269
+    },
+    {
+      "epoch": 0.15064264983645828,
+      "grad_norm": 2.392493359381271,
+      "learning_rate": 0.0001924669811236762,
+      "loss": 1.487,
+      "step": 3270
+    },
+    {
+      "epoch": 0.15068871792509328,
+      "grad_norm": 1.8089066457561889,
+      "learning_rate": 0.00019246129866946622,
+      "loss": 1.3988,
+      "step": 3271
+    },
+    {
+      "epoch": 0.15073478601372828,
+      "grad_norm": 1.8821899707756362,
+      "learning_rate": 0.000192455614156764,
+      "loss": 1.5423,
+      "step": 3272
+    },
+    {
+      "epoch": 0.1507808541023633,
+      "grad_norm": 2.873227619109946,
+      "learning_rate": 0.00019244992758569605,
+      "loss": 1.2789,
+      "step": 3273
+    },
+    {
+      "epoch": 0.1508269221909983,
+      "grad_norm": 2.068201449148195,
+      "learning_rate": 0.000192444238956389,
+      "loss": 1.3325,
+      "step": 3274
+    },
+    {
+      "epoch": 0.1508729902796333,
+      "grad_norm": 1.742628759093769,
+      "learning_rate": 0.00019243854826896947,
+      "loss": 1.4262,
+      "step": 3275
+    },
+    {
+      "epoch": 0.1509190583682683,
+      "grad_norm": 2.6397608693507477,
+      "learning_rate": 0.0001924328555235641,
+      "loss": 1.5157,
+      "step": 3276
+    },
+    {
+      "epoch": 0.1509651264569033,
+      "grad_norm": 2.628188118513209,
+      "learning_rate": 0.00019242716072029976,
+      "loss": 1.578,
+      "step": 3277
+    },
+    {
+      "epoch": 0.1510111945455383,
+      "grad_norm": 1.513007247603839,
+      "learning_rate": 0.00019242146385930315,
+      "loss": 1.662,
+      "step": 3278
+    },
+    {
+      "epoch": 0.1510572626341733,
+      "grad_norm": 2.198086258181576,
+      "learning_rate": 0.0001924157649407011,
+      "loss": 1.294,
+      "step": 3279
+    },
+    {
+      "epoch": 0.15110333072280832,
+      "grad_norm": 1.679132430168179,
+      "learning_rate": 0.0001924100639646205,
+      "loss": 1.3293,
+      "step": 3280
+    },
+    {
+      "epoch": 0.15114939881144332,
+      "grad_norm": 1.3431852426890776,
+      "learning_rate": 0.0001924043609311883,
+      "loss": 1.6801,
+      "step": 3281
+    },
+    {
+      "epoch": 0.15119546690007832,
+      "grad_norm": 3.1975835462991022,
+      "learning_rate": 0.00019239865584053142,
+      "loss": 1.4443,
+      "step": 3282
+    },
+    {
+      "epoch": 0.15124153498871332,
+      "grad_norm": 2.128762701772459,
+      "learning_rate": 0.00019239294869277692,
+      "loss": 1.1517,
+      "step": 3283
+    },
+    {
+      "epoch": 0.15128760307734831,
+      "grad_norm": 1.7779321375880472,
+      "learning_rate": 0.0001923872394880518,
+      "loss": 1.4354,
+      "step": 3284
+    },
+    {
+      "epoch": 0.1513336711659833,
+      "grad_norm": 2.436496167098978,
+      "learning_rate": 0.00019238152822648324,
+      "loss": 1.5682,
+      "step": 3285
+    },
+    {
+      "epoch": 0.15137973925461834,
+      "grad_norm": 2.3027679494033957,
+      "learning_rate": 0.00019237581490819835,
+      "loss": 1.4387,
+      "step": 3286
+    },
+    {
+      "epoch": 0.15142580734325334,
+      "grad_norm": 2.2203445421745287,
+      "learning_rate": 0.00019237009953332432,
+      "loss": 1.3951,
+      "step": 3287
+    },
+    {
+      "epoch": 0.15147187543188834,
+      "grad_norm": 2.711953179229129,
+      "learning_rate": 0.00019236438210198839,
+      "loss": 1.6097,
+      "step": 3288
+    },
+    {
+      "epoch": 0.15151794352052333,
+      "grad_norm": 2.591869995145536,
+      "learning_rate": 0.0001923586626143179,
+      "loss": 1.4704,
+      "step": 3289
+    },
+    {
+      "epoch": 0.15156401160915833,
+      "grad_norm": 4.002102503239456,
+      "learning_rate": 0.0001923529410704401,
+      "loss": 1.3853,
+      "step": 3290
+    },
+    {
+      "epoch": 0.15161007969779333,
+      "grad_norm": 4.450806085588317,
+      "learning_rate": 0.00019234721747048248,
+      "loss": 1.6511,
+      "step": 3291
+    },
+    {
+      "epoch": 0.15165614778642833,
+      "grad_norm": 1.8890013372225232,
+      "learning_rate": 0.00019234149181457236,
+      "loss": 1.334,
+      "step": 3292
+    },
+    {
+      "epoch": 0.15170221587506336,
+      "grad_norm": 1.6505913842167879,
+      "learning_rate": 0.00019233576410283726,
+      "loss": 1.4717,
+      "step": 3293
+    },
+    {
+      "epoch": 0.15174828396369835,
+      "grad_norm": 2.4243011842703575,
+      "learning_rate": 0.00019233003433540471,
+      "loss": 1.6977,
+      "step": 3294
+    },
+    {
+      "epoch": 0.15179435205233335,
+      "grad_norm": 1.8427377466046098,
+      "learning_rate": 0.00019232430251240226,
+      "loss": 1.4348,
+      "step": 3295
+    },
+    {
+      "epoch": 0.15184042014096835,
+      "grad_norm": 1.3948083739306258,
+      "learning_rate": 0.0001923185686339575,
+      "loss": 1.5299,
+      "step": 3296
+    },
+    {
+      "epoch": 0.15188648822960335,
+      "grad_norm": 2.1776611607823417,
+      "learning_rate": 0.0001923128327001981,
+      "loss": 1.6114,
+      "step": 3297
+    },
+    {
+      "epoch": 0.15193255631823835,
+      "grad_norm": 1.7258824237462094,
+      "learning_rate": 0.0001923070947112518,
+      "loss": 1.5225,
+      "step": 3298
+    },
+    {
+      "epoch": 0.15197862440687335,
+      "grad_norm": 1.6346078488983151,
+      "learning_rate": 0.00019230135466724626,
+      "loss": 1.5489,
+      "step": 3299
+    },
+    {
+      "epoch": 0.15202469249550837,
+      "grad_norm": 1.4245927213368106,
+      "learning_rate": 0.00019229561256830935,
+      "loss": 1.7607,
+      "step": 3300
+    },
+    {
+      "epoch": 0.15207076058414337,
+      "grad_norm": 1.6069759252316544,
+      "learning_rate": 0.00019228986841456885,
+      "loss": 1.6261,
+      "step": 3301
+    },
+    {
+      "epoch": 0.15211682867277837,
+      "grad_norm": 1.9189188889318707,
+      "learning_rate": 0.0001922841222061527,
+      "loss": 1.384,
+      "step": 3302
+    },
+    {
+      "epoch": 0.15216289676141337,
+      "grad_norm": 1.8494241805029945,
+      "learning_rate": 0.00019227837394318882,
+      "loss": 1.4869,
+      "step": 3303
+    },
+    {
+      "epoch": 0.15220896485004837,
+      "grad_norm": 2.6163029760943606,
+      "learning_rate": 0.00019227262362580514,
+      "loss": 1.308,
+      "step": 3304
+    },
+    {
+      "epoch": 0.15225503293868337,
+      "grad_norm": 1.3995626902604432,
+      "learning_rate": 0.0001922668712541297,
+      "loss": 1.3628,
+      "step": 3305
+    },
+    {
+      "epoch": 0.15230110102731836,
+      "grad_norm": 2.7989426725114437,
+      "learning_rate": 0.00019226111682829058,
+      "loss": 1.3798,
+      "step": 3306
+    },
+    {
+      "epoch": 0.1523471691159534,
+      "grad_norm": 1.2501900007454236,
+      "learning_rate": 0.00019225536034841588,
+      "loss": 1.4436,
+      "step": 3307
+    },
+    {
+      "epoch": 0.1523932372045884,
+      "grad_norm": 1.9974924025092868,
+      "learning_rate": 0.0001922496018146338,
+      "loss": 1.2376,
+      "step": 3308
+    },
+    {
+      "epoch": 0.1524393052932234,
+      "grad_norm": 1.5076813940966498,
+      "learning_rate": 0.0001922438412270725,
+      "loss": 1.5471,
+      "step": 3309
+    },
+    {
+      "epoch": 0.15248537338185839,
+      "grad_norm": 1.9412713498153924,
+      "learning_rate": 0.00019223807858586022,
+      "loss": 1.3984,
+      "step": 3310
+    },
+    {
+      "epoch": 0.15253144147049338,
+      "grad_norm": 1.3107655309906348,
+      "learning_rate": 0.00019223231389112528,
+      "loss": 1.6512,
+      "step": 3311
+    },
+    {
+      "epoch": 0.15257750955912838,
+      "grad_norm": 1.6044140942878835,
+      "learning_rate": 0.00019222654714299604,
+      "loss": 1.3806,
+      "step": 3312
+    },
+    {
+      "epoch": 0.15262357764776338,
+      "grad_norm": 1.620291713370008,
+      "learning_rate": 0.00019222077834160083,
+      "loss": 1.3351,
+      "step": 3313
+    },
+    {
+      "epoch": 0.1526696457363984,
+      "grad_norm": 1.1270526543861863,
+      "learning_rate": 0.00019221500748706812,
+      "loss": 1.0114,
+      "step": 3314
+    },
+    {
+      "epoch": 0.1527157138250334,
+      "grad_norm": 1.5215224622616081,
+      "learning_rate": 0.0001922092345795264,
+      "loss": 1.5488,
+      "step": 3315
+    },
+    {
+      "epoch": 0.1527617819136684,
+      "grad_norm": 1.6336045172396252,
+      "learning_rate": 0.00019220345961910415,
+      "loss": 1.4576,
+      "step": 3316
+    },
+    {
+      "epoch": 0.1528078500023034,
+      "grad_norm": 1.7614385945323516,
+      "learning_rate": 0.00019219768260593,
+      "loss": 1.3879,
+      "step": 3317
+    },
+    {
+      "epoch": 0.1528539180909384,
+      "grad_norm": 1.7706869954174562,
+      "learning_rate": 0.00019219190354013253,
+      "loss": 1.4914,
+      "step": 3318
+    },
+    {
+      "epoch": 0.1528999861795734,
+      "grad_norm": 1.456253776486168,
+      "learning_rate": 0.00019218612242184038,
+      "loss": 1.4588,
+      "step": 3319
+    },
+    {
+      "epoch": 0.15294605426820843,
+      "grad_norm": 2.903602266859782,
+      "learning_rate": 0.0001921803392511823,
+      "loss": 1.2882,
+      "step": 3320
+    },
+    {
+      "epoch": 0.15299212235684342,
+      "grad_norm": 1.7733106442299438,
+      "learning_rate": 0.000192174554028287,
+      "loss": 1.3293,
+      "step": 3321
+    },
+    {
+      "epoch": 0.15303819044547842,
+      "grad_norm": 1.1401087854166263,
+      "learning_rate": 0.0001921687667532833,
+      "loss": 1.544,
+      "step": 3322
+    },
+    {
+      "epoch": 0.15308425853411342,
+      "grad_norm": 1.3388774279154625,
+      "learning_rate": 0.00019216297742630006,
+      "loss": 1.4493,
+      "step": 3323
+    },
+    {
+      "epoch": 0.15313032662274842,
+      "grad_norm": 1.0807830151899671,
+      "learning_rate": 0.00019215718604746615,
+      "loss": 1.6726,
+      "step": 3324
+    },
+    {
+      "epoch": 0.15317639471138342,
+      "grad_norm": 1.4851363803246058,
+      "learning_rate": 0.0001921513926169105,
+      "loss": 1.2829,
+      "step": 3325
+    },
+    {
+      "epoch": 0.15322246280001842,
+      "grad_norm": 2.116425306527824,
+      "learning_rate": 0.0001921455971347621,
+      "loss": 1.3613,
+      "step": 3326
+    },
+    {
+      "epoch": 0.15326853088865344,
+      "grad_norm": 2.410454791336599,
+      "learning_rate": 0.00019213979960115,
+      "loss": 1.6347,
+      "step": 3327
+    },
+    {
+      "epoch": 0.15331459897728844,
+      "grad_norm": 1.3955220235904209,
+      "learning_rate": 0.00019213400001620324,
+      "loss": 1.2809,
+      "step": 3328
+    },
+    {
+      "epoch": 0.15336066706592344,
+      "grad_norm": 2.2748914864587255,
+      "learning_rate": 0.00019212819838005093,
+      "loss": 1.2432,
+      "step": 3329
+    },
+    {
+      "epoch": 0.15340673515455844,
+      "grad_norm": 1.0661082485810662,
+      "learning_rate": 0.00019212239469282224,
+      "loss": 0.9671,
+      "step": 3330
+    },
+    {
+      "epoch": 0.15345280324319344,
+      "grad_norm": 2.005845254471473,
+      "learning_rate": 0.0001921165889546464,
+      "loss": 1.4292,
+      "step": 3331
+    },
+    {
+      "epoch": 0.15349887133182843,
+      "grad_norm": 2.006470477796999,
+      "learning_rate": 0.00019211078116565267,
+      "loss": 1.4954,
+      "step": 3332
+    },
+    {
+      "epoch": 0.15354493942046343,
+      "grad_norm": 1.312734701005438,
+      "learning_rate": 0.00019210497132597033,
+      "loss": 1.4317,
+      "step": 3333
+    },
+    {
+      "epoch": 0.15359100750909846,
+      "grad_norm": 1.274357380160339,
+      "learning_rate": 0.0001920991594357287,
+      "loss": 1.8662,
+      "step": 3334
+    },
+    {
+      "epoch": 0.15363707559773346,
+      "grad_norm": 1.2126452068597824,
+      "learning_rate": 0.0001920933454950572,
+      "loss": 1.5473,
+      "step": 3335
+    },
+    {
+      "epoch": 0.15368314368636846,
+      "grad_norm": 1.4892789088689309,
+      "learning_rate": 0.00019208752950408532,
+      "loss": 1.3389,
+      "step": 3336
+    },
+    {
+      "epoch": 0.15372921177500345,
+      "grad_norm": 1.8471849704456378,
+      "learning_rate": 0.00019208171146294244,
+      "loss": 1.6896,
+      "step": 3337
+    },
+    {
+      "epoch": 0.15377527986363845,
+      "grad_norm": 1.892963125126257,
+      "learning_rate": 0.00019207589137175816,
+      "loss": 1.4775,
+      "step": 3338
+    },
+    {
+      "epoch": 0.15382134795227345,
+      "grad_norm": 1.4861493162185544,
+      "learning_rate": 0.00019207006923066205,
+      "loss": 1.5574,
+      "step": 3339
+    },
+    {
+      "epoch": 0.15386741604090845,
+      "grad_norm": 2.0153580714982176,
+      "learning_rate": 0.0001920642450397837,
+      "loss": 1.5666,
+      "step": 3340
+    },
+    {
+      "epoch": 0.15391348412954348,
+      "grad_norm": 1.433004915492582,
+      "learning_rate": 0.00019205841879925276,
+      "loss": 1.2362,
+      "step": 3341
+    },
+    {
+      "epoch": 0.15395955221817847,
+      "grad_norm": 1.1299249803284706,
+      "learning_rate": 0.00019205259050919903,
+      "loss": 1.627,
+      "step": 3342
+    },
+    {
+      "epoch": 0.15400562030681347,
+      "grad_norm": 1.2614462711147334,
+      "learning_rate": 0.00019204676016975218,
+      "loss": 1.3248,
+      "step": 3343
+    },
+    {
+      "epoch": 0.15405168839544847,
+      "grad_norm": 2.1701668113833685,
+      "learning_rate": 0.00019204092778104203,
+      "loss": 1.1615,
+      "step": 3344
+    },
+    {
+      "epoch": 0.15409775648408347,
+      "grad_norm": 2.4703011634233834,
+      "learning_rate": 0.00019203509334319847,
+      "loss": 1.5108,
+      "step": 3345
+    },
+    {
+      "epoch": 0.15414382457271847,
+      "grad_norm": 1.3037142111171527,
+      "learning_rate": 0.00019202925685635135,
+      "loss": 1.1195,
+      "step": 3346
+    },
+    {
+      "epoch": 0.15418989266135347,
+      "grad_norm": 1.7236364095234642,
+      "learning_rate": 0.0001920234183206306,
+      "loss": 1.405,
+      "step": 3347
+    },
+    {
+      "epoch": 0.1542359607499885,
+      "grad_norm": 1.388835161369103,
+      "learning_rate": 0.00019201757773616625,
+      "loss": 1.3134,
+      "step": 3348
+    },
+    {
+      "epoch": 0.1542820288386235,
+      "grad_norm": 1.3525009299235993,
+      "learning_rate": 0.0001920117351030883,
+      "loss": 1.2948,
+      "step": 3349
+    },
+    {
+      "epoch": 0.1543280969272585,
+      "grad_norm": 1.2748278885015034,
+      "learning_rate": 0.00019200589042152685,
+      "loss": 1.4147,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1543741650158935,
+      "grad_norm": 1.5541882601148722,
+      "learning_rate": 0.00019200004369161197,
+      "loss": 1.3955,
+      "step": 3351
+    },
+    {
+      "epoch": 0.1544202331045285,
+      "grad_norm": 2.849288378853641,
+      "learning_rate": 0.0001919941949134739,
+      "loss": 1.2163,
+      "step": 3352
+    },
+    {
+      "epoch": 0.15446630119316349,
+      "grad_norm": 1.1460350556438816,
+      "learning_rate": 0.0001919883440872428,
+      "loss": 1.2485,
+      "step": 3353
+    },
+    {
+      "epoch": 0.1545123692817985,
+      "grad_norm": 1.8967673737762012,
+      "learning_rate": 0.00019198249121304896,
+      "loss": 1.6114,
+      "step": 3354
+    },
+    {
+      "epoch": 0.1545584373704335,
+      "grad_norm": 1.233993107828214,
+      "learning_rate": 0.00019197663629102264,
+      "loss": 1.5924,
+      "step": 3355
+    },
+    {
+      "epoch": 0.1546045054590685,
+      "grad_norm": 1.2381820460474826,
+      "learning_rate": 0.00019197077932129423,
+      "loss": 1.1544,
+      "step": 3356
+    },
+    {
+      "epoch": 0.1546505735477035,
+      "grad_norm": 1.47823456402487,
+      "learning_rate": 0.00019196492030399412,
+      "loss": 1.2306,
+      "step": 3357
+    },
+    {
+      "epoch": 0.1546966416363385,
+      "grad_norm": 1.6028692717933442,
+      "learning_rate": 0.00019195905923925275,
+      "loss": 1.1842,
+      "step": 3358
+    },
+    {
+      "epoch": 0.1547427097249735,
+      "grad_norm": 1.6546553877655608,
+      "learning_rate": 0.00019195319612720063,
+      "loss": 1.5367,
+      "step": 3359
+    },
+    {
+      "epoch": 0.1547887778136085,
+      "grad_norm": 2.399570823960321,
+      "learning_rate": 0.00019194733096796818,
+      "loss": 1.1416,
+      "step": 3360
+    },
+    {
+      "epoch": 0.15483484590224353,
+      "grad_norm": 1.2156632536837402,
+      "learning_rate": 0.00019194146376168614,
+      "loss": 1.3075,
+      "step": 3361
+    },
+    {
+      "epoch": 0.15488091399087853,
+      "grad_norm": 1.8430196666211376,
+      "learning_rate": 0.00019193559450848504,
+      "loss": 1.3611,
+      "step": 3362
+    },
+    {
+      "epoch": 0.15492698207951353,
+      "grad_norm": 1.8010833562948878,
+      "learning_rate": 0.00019192972320849556,
+      "loss": 1.522,
+      "step": 3363
+    },
+    {
+      "epoch": 0.15497305016814852,
+      "grad_norm": 1.7268574937179408,
+      "learning_rate": 0.0001919238498618484,
+      "loss": 1.3467,
+      "step": 3364
+    },
+    {
+      "epoch": 0.15501911825678352,
+      "grad_norm": 1.5643970192516596,
+      "learning_rate": 0.00019191797446867437,
+      "loss": 1.2889,
+      "step": 3365
+    },
+    {
+      "epoch": 0.15506518634541852,
+      "grad_norm": 1.2773842045992967,
+      "learning_rate": 0.0001919120970291042,
+      "loss": 1.3087,
+      "step": 3366
+    },
+    {
+      "epoch": 0.15511125443405352,
+      "grad_norm": 2.1811397220431603,
+      "learning_rate": 0.00019190621754326884,
+      "loss": 1.5545,
+      "step": 3367
+    },
+    {
+      "epoch": 0.15515732252268855,
+      "grad_norm": 1.6911437539682679,
+      "learning_rate": 0.00019190033601129906,
+      "loss": 1.6449,
+      "step": 3368
+    },
+    {
+      "epoch": 0.15520339061132354,
+      "grad_norm": 1.2407293275616207,
+      "learning_rate": 0.00019189445243332593,
+      "loss": 1.2107,
+      "step": 3369
+    },
+    {
+      "epoch": 0.15524945869995854,
+      "grad_norm": 1.5105197846401948,
+      "learning_rate": 0.00019188856680948035,
+      "loss": 1.2076,
+      "step": 3370
+    },
+    {
+      "epoch": 0.15529552678859354,
+      "grad_norm": 2.5564500720535035,
+      "learning_rate": 0.00019188267913989343,
+      "loss": 1.3225,
+      "step": 3371
+    },
+    {
+      "epoch": 0.15534159487722854,
+      "grad_norm": 1.5990928723401197,
+      "learning_rate": 0.0001918767894246962,
+      "loss": 1.402,
+      "step": 3372
+    },
+    {
+      "epoch": 0.15538766296586354,
+      "grad_norm": 1.8432263958011852,
+      "learning_rate": 0.00019187089766401974,
+      "loss": 1.5431,
+      "step": 3373
+    },
+    {
+      "epoch": 0.15543373105449854,
+      "grad_norm": 2.6709081363260228,
+      "learning_rate": 0.0001918650038579953,
+      "loss": 1.2089,
+      "step": 3374
+    },
+    {
+      "epoch": 0.15547979914313356,
+      "grad_norm": 4.686853348264449,
+      "learning_rate": 0.00019185910800675405,
+      "loss": 1.6125,
+      "step": 3375
+    },
+    {
+      "epoch": 0.15552586723176856,
+      "grad_norm": 1.813112719822547,
+      "learning_rate": 0.00019185321011042726,
+      "loss": 0.9371,
+      "step": 3376
+    },
+    {
+      "epoch": 0.15557193532040356,
+      "grad_norm": 1.2134607733841625,
+      "learning_rate": 0.00019184731016914626,
+      "loss": 1.2086,
+      "step": 3377
+    },
+    {
+      "epoch": 0.15561800340903856,
+      "grad_norm": 1.589140679672068,
+      "learning_rate": 0.00019184140818304237,
+      "loss": 1.3863,
+      "step": 3378
+    },
+    {
+      "epoch": 0.15566407149767356,
+      "grad_norm": 3.6767913805462085,
+      "learning_rate": 0.000191835504152247,
+      "loss": 1.4016,
+      "step": 3379
+    },
+    {
+      "epoch": 0.15571013958630855,
+      "grad_norm": 1.5580549508101047,
+      "learning_rate": 0.0001918295980768916,
+      "loss": 1.422,
+      "step": 3380
+    },
+    {
+      "epoch": 0.15575620767494355,
+      "grad_norm": 2.2048978374994994,
+      "learning_rate": 0.00019182368995710768,
+      "loss": 1.3672,
+      "step": 3381
+    },
+    {
+      "epoch": 0.15580227576357858,
+      "grad_norm": 2.319564059664891,
+      "learning_rate": 0.0001918177797930267,
+      "loss": 1.368,
+      "step": 3382
+    },
+    {
+      "epoch": 0.15584834385221358,
+      "grad_norm": 2.3306712620147065,
+      "learning_rate": 0.00019181186758478031,
+      "loss": 1.3337,
+      "step": 3383
+    },
+    {
+      "epoch": 0.15589441194084858,
+      "grad_norm": 1.4876203032070292,
+      "learning_rate": 0.0001918059533325001,
+      "loss": 1.5389,
+      "step": 3384
+    },
+    {
+      "epoch": 0.15594048002948357,
+      "grad_norm": 2.3807464133905585,
+      "learning_rate": 0.00019180003703631778,
+      "loss": 1.6003,
+      "step": 3385
+    },
+    {
+      "epoch": 0.15598654811811857,
+      "grad_norm": 1.6782334609199705,
+      "learning_rate": 0.00019179411869636503,
+      "loss": 1.402,
+      "step": 3386
+    },
+    {
+      "epoch": 0.15603261620675357,
+      "grad_norm": 1.4140506572932923,
+      "learning_rate": 0.00019178819831277362,
+      "loss": 1.2916,
+      "step": 3387
+    },
+    {
+      "epoch": 0.1560786842953886,
+      "grad_norm": 1.7068180705980232,
+      "learning_rate": 0.00019178227588567532,
+      "loss": 1.4545,
+      "step": 3388
+    },
+    {
+      "epoch": 0.1561247523840236,
+      "grad_norm": 1.704762098763529,
+      "learning_rate": 0.00019177635141520206,
+      "loss": 1.3588,
+      "step": 3389
+    },
+    {
+      "epoch": 0.1561708204726586,
+      "grad_norm": 1.0618565126528112,
+      "learning_rate": 0.0001917704249014857,
+      "loss": 1.626,
+      "step": 3390
+    },
+    {
+      "epoch": 0.1562168885612936,
+      "grad_norm": 2.238643400468843,
+      "learning_rate": 0.00019176449634465814,
+      "loss": 1.6665,
+      "step": 3391
+    },
+    {
+      "epoch": 0.1562629566499286,
+      "grad_norm": 1.4054268053543426,
+      "learning_rate": 0.00019175856574485148,
+      "loss": 1.6826,
+      "step": 3392
+    },
+    {
+      "epoch": 0.1563090247385636,
+      "grad_norm": 1.3517008484044115,
+      "learning_rate": 0.00019175263310219762,
+      "loss": 1.5857,
+      "step": 3393
+    },
+    {
+      "epoch": 0.1563550928271986,
+      "grad_norm": 1.4632752468368417,
+      "learning_rate": 0.00019174669841682873,
+      "loss": 1.5432,
+      "step": 3394
+    },
+    {
+      "epoch": 0.15640116091583361,
+      "grad_norm": 1.1745484999573215,
+      "learning_rate": 0.00019174076168887693,
+      "loss": 1.4033,
+      "step": 3395
+    },
+    {
+      "epoch": 0.1564472290044686,
+      "grad_norm": 3.687012412167252,
+      "learning_rate": 0.00019173482291847432,
+      "loss": 1.4921,
+      "step": 3396
+    },
+    {
+      "epoch": 0.1564932970931036,
+      "grad_norm": 4.0420790186420845,
+      "learning_rate": 0.00019172888210575325,
+      "loss": 1.6306,
+      "step": 3397
+    },
+    {
+      "epoch": 0.1565393651817386,
+      "grad_norm": 2.410697808588139,
+      "learning_rate": 0.00019172293925084585,
+      "loss": 1.2258,
+      "step": 3398
+    },
+    {
+      "epoch": 0.1565854332703736,
+      "grad_norm": 1.6267188038097629,
+      "learning_rate": 0.00019171699435388447,
+      "loss": 1.613,
+      "step": 3399
+    },
+    {
+      "epoch": 0.1566315013590086,
+      "grad_norm": 1.2744451538193038,
+      "learning_rate": 0.0001917110474150015,
+      "loss": 1.6488,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1566775694476436,
+      "grad_norm": 1.2946478048212746,
+      "learning_rate": 0.00019170509843432929,
+      "loss": 1.475,
+      "step": 3401
+    },
+    {
+      "epoch": 0.15672363753627863,
+      "grad_norm": 1.65336789540812,
+      "learning_rate": 0.0001916991474120003,
+      "loss": 1.239,
+      "step": 3402
+    },
+    {
+      "epoch": 0.15676970562491363,
+      "grad_norm": 2.6632731316431197,
+      "learning_rate": 0.00019169319434814703,
+      "loss": 1.0553,
+      "step": 3403
+    },
+    {
+      "epoch": 0.15681577371354863,
+      "grad_norm": 1.4389131206465586,
+      "learning_rate": 0.00019168723924290203,
+      "loss": 1.5235,
+      "step": 3404
+    },
+    {
+      "epoch": 0.15686184180218363,
+      "grad_norm": 1.3543032142734575,
+      "learning_rate": 0.00019168128209639783,
+      "loss": 1.42,
+      "step": 3405
+    },
+    {
+      "epoch": 0.15690790989081863,
+      "grad_norm": 3.119981798979316,
+      "learning_rate": 0.00019167532290876712,
+      "loss": 1.1519,
+      "step": 3406
+    },
+    {
+      "epoch": 0.15695397797945362,
+      "grad_norm": 1.629921903659617,
+      "learning_rate": 0.0001916693616801425,
+      "loss": 1.1924,
+      "step": 3407
+    },
+    {
+      "epoch": 0.15700004606808862,
+      "grad_norm": 1.684907242062247,
+      "learning_rate": 0.00019166339841065672,
+      "loss": 1.4993,
+      "step": 3408
+    },
+    {
+      "epoch": 0.15704611415672365,
+      "grad_norm": 5.082897278994203,
+      "learning_rate": 0.00019165743310044255,
+      "loss": 1.4927,
+      "step": 3409
+    },
+    {
+      "epoch": 0.15709218224535865,
+      "grad_norm": 2.320625392043012,
+      "learning_rate": 0.0001916514657496328,
+      "loss": 1.2762,
+      "step": 3410
+    },
+    {
+      "epoch": 0.15713825033399365,
+      "grad_norm": 1.7429458310723334,
+      "learning_rate": 0.0001916454963583603,
+      "loss": 1.023,
+      "step": 3411
+    },
+    {
+      "epoch": 0.15718431842262864,
+      "grad_norm": 2.0224787790726553,
+      "learning_rate": 0.000191639524926758,
+      "loss": 1.2236,
+      "step": 3412
+    },
+    {
+      "epoch": 0.15723038651126364,
+      "grad_norm": 1.6871582689367253,
+      "learning_rate": 0.00019163355145495876,
+      "loss": 1.6404,
+      "step": 3413
+    },
+    {
+      "epoch": 0.15727645459989864,
+      "grad_norm": 2.4015222184116536,
+      "learning_rate": 0.00019162757594309564,
+      "loss": 1.2532,
+      "step": 3414
+    },
+    {
+      "epoch": 0.15732252268853364,
+      "grad_norm": 1.5948190047119308,
+      "learning_rate": 0.00019162159839130162,
+      "loss": 1.4275,
+      "step": 3415
+    },
+    {
+      "epoch": 0.15736859077716867,
+      "grad_norm": 1.5488040783054766,
+      "learning_rate": 0.00019161561879970984,
+      "loss": 1.5446,
+      "step": 3416
+    },
+    {
+      "epoch": 0.15741465886580366,
+      "grad_norm": 1.7981926198268325,
+      "learning_rate": 0.0001916096371684534,
+      "loss": 1.3066,
+      "step": 3417
+    },
+    {
+      "epoch": 0.15746072695443866,
+      "grad_norm": 1.7432240001340567,
+      "learning_rate": 0.00019160365349766545,
+      "loss": 1.0641,
+      "step": 3418
+    },
+    {
+      "epoch": 0.15750679504307366,
+      "grad_norm": 1.1751375489081362,
+      "learning_rate": 0.0001915976677874792,
+      "loss": 1.2451,
+      "step": 3419
+    },
+    {
+      "epoch": 0.15755286313170866,
+      "grad_norm": 1.5257543185201776,
+      "learning_rate": 0.00019159168003802797,
+      "loss": 1.7188,
+      "step": 3420
+    },
+    {
+      "epoch": 0.15759893122034366,
+      "grad_norm": 1.044445227768071,
+      "learning_rate": 0.000191585690249445,
+      "loss": 1.673,
+      "step": 3421
+    },
+    {
+      "epoch": 0.15764499930897868,
+      "grad_norm": 3.216509229984196,
+      "learning_rate": 0.0001915796984218637,
+      "loss": 1.473,
+      "step": 3422
+    },
+    {
+      "epoch": 0.15769106739761368,
+      "grad_norm": 2.3167780951498194,
+      "learning_rate": 0.00019157370455541743,
+      "loss": 1.3889,
+      "step": 3423
+    },
+    {
+      "epoch": 0.15773713548624868,
+      "grad_norm": 1.239863513120091,
+      "learning_rate": 0.00019156770865023964,
+      "loss": 1.3343,
+      "step": 3424
+    },
+    {
+      "epoch": 0.15778320357488368,
+      "grad_norm": 1.5790534813087396,
+      "learning_rate": 0.00019156171070646382,
+      "loss": 1.7267,
+      "step": 3425
+    },
+    {
+      "epoch": 0.15782927166351868,
+      "grad_norm": 1.7801508061192597,
+      "learning_rate": 0.0001915557107242235,
+      "loss": 1.3956,
+      "step": 3426
+    },
+    {
+      "epoch": 0.15787533975215368,
+      "grad_norm": 1.567849776540534,
+      "learning_rate": 0.00019154970870365227,
+      "loss": 1.335,
+      "step": 3427
+    },
+    {
+      "epoch": 0.15792140784078867,
+      "grad_norm": 3.951893602078586,
+      "learning_rate": 0.00019154370464488375,
+      "loss": 1.2439,
+      "step": 3428
+    },
+    {
+      "epoch": 0.1579674759294237,
+      "grad_norm": 2.317765492117221,
+      "learning_rate": 0.0001915376985480516,
+      "loss": 1.104,
+      "step": 3429
+    },
+    {
+      "epoch": 0.1580135440180587,
+      "grad_norm": 4.105947468529396,
+      "learning_rate": 0.00019153169041328957,
+      "loss": 1.4831,
+      "step": 3430
+    },
+    {
+      "epoch": 0.1580596121066937,
+      "grad_norm": 3.3420134994433974,
+      "learning_rate": 0.0001915256802407314,
+      "loss": 1.3371,
+      "step": 3431
+    },
+    {
+      "epoch": 0.1581056801953287,
+      "grad_norm": 5.001237567002368,
+      "learning_rate": 0.00019151966803051087,
+      "loss": 1.3916,
+      "step": 3432
+    },
+    {
+      "epoch": 0.1581517482839637,
+      "grad_norm": 2.3386336001544534,
+      "learning_rate": 0.00019151365378276185,
+      "loss": 1.5675,
+      "step": 3433
+    },
+    {
+      "epoch": 0.1581978163725987,
+      "grad_norm": 1.5460642438992154,
+      "learning_rate": 0.00019150763749761827,
+      "loss": 1.1733,
+      "step": 3434
+    },
+    {
+      "epoch": 0.1582438844612337,
+      "grad_norm": 1.8694854407643233,
+      "learning_rate": 0.000191501619175214,
+      "loss": 1.3362,
+      "step": 3435
+    },
+    {
+      "epoch": 0.15828995254986872,
+      "grad_norm": 3.3539226910809457,
+      "learning_rate": 0.0001914955988156831,
+      "loss": 1.3157,
+      "step": 3436
+    },
+    {
+      "epoch": 0.15833602063850372,
+      "grad_norm": 2.5335649263253575,
+      "learning_rate": 0.0001914895764191596,
+      "loss": 1.4488,
+      "step": 3437
+    },
+    {
+      "epoch": 0.15838208872713871,
+      "grad_norm": 3.4045185315767963,
+      "learning_rate": 0.0001914835519857775,
+      "loss": 1.3767,
+      "step": 3438
+    },
+    {
+      "epoch": 0.1584281568157737,
+      "grad_norm": 1.279448289220382,
+      "learning_rate": 0.000191477525515671,
+      "loss": 1.4883,
+      "step": 3439
+    },
+    {
+      "epoch": 0.1584742249044087,
+      "grad_norm": 5.55304769449519,
+      "learning_rate": 0.00019147149700897426,
+      "loss": 1.7012,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1585202929930437,
+      "grad_norm": 1.8684817081977725,
+      "learning_rate": 0.00019146546646582144,
+      "loss": 1.5513,
+      "step": 3441
+    },
+    {
+      "epoch": 0.1585663610816787,
+      "grad_norm": 3.4673738464978268,
+      "learning_rate": 0.00019145943388634685,
+      "loss": 1.3116,
+      "step": 3442
+    },
+    {
+      "epoch": 0.15861242917031373,
+      "grad_norm": 2.120219289049499,
+      "learning_rate": 0.0001914533992706848,
+      "loss": 1.512,
+      "step": 3443
+    },
+    {
+      "epoch": 0.15865849725894873,
+      "grad_norm": 1.1427070658505816,
+      "learning_rate": 0.0001914473626189696,
+      "loss": 1.4825,
+      "step": 3444
+    },
+    {
+      "epoch": 0.15870456534758373,
+      "grad_norm": 1.5707726178353254,
+      "learning_rate": 0.00019144132393133568,
+      "loss": 1.5836,
+      "step": 3445
+    },
+    {
+      "epoch": 0.15875063343621873,
+      "grad_norm": 1.3458571682240648,
+      "learning_rate": 0.0001914352832079175,
+      "loss": 1.7936,
+      "step": 3446
+    },
+    {
+      "epoch": 0.15879670152485373,
+      "grad_norm": 1.314797516343799,
+      "learning_rate": 0.00019142924044884948,
+      "loss": 1.6445,
+      "step": 3447
+    },
+    {
+      "epoch": 0.15884276961348873,
+      "grad_norm": 1.6963346634456005,
+      "learning_rate": 0.0001914231956542662,
+      "loss": 1.6006,
+      "step": 3448
+    },
+    {
+      "epoch": 0.15888883770212373,
+      "grad_norm": 1.3783952892511526,
+      "learning_rate": 0.00019141714882430223,
+      "loss": 1.5274,
+      "step": 3449
+    },
+    {
+      "epoch": 0.15893490579075875,
+      "grad_norm": 0.9777245355979717,
+      "learning_rate": 0.00019141109995909222,
+      "loss": 1.4164,
+      "step": 3450
+    },
+    {
+      "epoch": 0.15898097387939375,
+      "grad_norm": 1.5655167541037345,
+      "learning_rate": 0.00019140504905877076,
+      "loss": 1.2536,
+      "step": 3451
+    },
+    {
+      "epoch": 0.15902704196802875,
+      "grad_norm": 1.327935031342264,
+      "learning_rate": 0.00019139899612347262,
+      "loss": 1.1947,
+      "step": 3452
+    },
+    {
+      "epoch": 0.15907311005666375,
+      "grad_norm": 1.2057453849641095,
+      "learning_rate": 0.00019139294115333254,
+      "loss": 1.3989,
+      "step": 3453
+    },
+    {
+      "epoch": 0.15911917814529875,
+      "grad_norm": 2.2443632151533994,
+      "learning_rate": 0.00019138688414848534,
+      "loss": 1.3532,
+      "step": 3454
+    },
+    {
+      "epoch": 0.15916524623393374,
+      "grad_norm": 2.0890054607375212,
+      "learning_rate": 0.00019138082510906584,
+      "loss": 1.08,
+      "step": 3455
+    },
+    {
+      "epoch": 0.15921131432256877,
+      "grad_norm": 1.2035273167804847,
+      "learning_rate": 0.000191374764035209,
+      "loss": 1.4623,
+      "step": 3456
+    },
+    {
+      "epoch": 0.15925738241120377,
+      "grad_norm": 1.16283908635631,
+      "learning_rate": 0.0001913687009270497,
+      "loss": 1.4055,
+      "step": 3457
+    },
+    {
+      "epoch": 0.15930345049983877,
+      "grad_norm": 1.5586667529224292,
+      "learning_rate": 0.0001913626357847229,
+      "loss": 1.4693,
+      "step": 3458
+    },
+    {
+      "epoch": 0.15934951858847377,
+      "grad_norm": 0.9832208386785619,
+      "learning_rate": 0.00019135656860836368,
+      "loss": 1.3871,
+      "step": 3459
+    },
+    {
+      "epoch": 0.15939558667710876,
+      "grad_norm": 1.2171638386949268,
+      "learning_rate": 0.00019135049939810714,
+      "loss": 1.3223,
+      "step": 3460
+    },
+    {
+      "epoch": 0.15944165476574376,
+      "grad_norm": 1.2980955197497601,
+      "learning_rate": 0.00019134442815408833,
+      "loss": 1.6901,
+      "step": 3461
+    },
+    {
+      "epoch": 0.15948772285437876,
+      "grad_norm": 2.6115243326581554,
+      "learning_rate": 0.00019133835487644246,
+      "loss": 1.3193,
+      "step": 3462
+    },
+    {
+      "epoch": 0.1595337909430138,
+      "grad_norm": 2.034345089708598,
+      "learning_rate": 0.0001913322795653047,
+      "loss": 1.7009,
+      "step": 3463
+    },
+    {
+      "epoch": 0.15957985903164879,
+      "grad_norm": 2.576638462405724,
+      "learning_rate": 0.0001913262022208104,
+      "loss": 1.6572,
+      "step": 3464
+    },
+    {
+      "epoch": 0.15962592712028378,
+      "grad_norm": 1.2712714080337386,
+      "learning_rate": 0.00019132012284309475,
+      "loss": 1.4174,
+      "step": 3465
+    },
+    {
+      "epoch": 0.15967199520891878,
+      "grad_norm": 1.8544455219582885,
+      "learning_rate": 0.00019131404143229313,
+      "loss": 1.6727,
+      "step": 3466
+    },
+    {
+      "epoch": 0.15971806329755378,
+      "grad_norm": 1.5788752684884781,
+      "learning_rate": 0.000191307957988541,
+      "loss": 1.3595,
+      "step": 3467
+    },
+    {
+      "epoch": 0.15976413138618878,
+      "grad_norm": 1.5801995379671478,
+      "learning_rate": 0.0001913018725119737,
+      "loss": 1.244,
+      "step": 3468
+    },
+    {
+      "epoch": 0.15981019947482378,
+      "grad_norm": 1.728323987428084,
+      "learning_rate": 0.00019129578500272677,
+      "loss": 1.715,
+      "step": 3469
+    },
+    {
+      "epoch": 0.1598562675634588,
+      "grad_norm": 1.719847880104492,
+      "learning_rate": 0.00019128969546093575,
+      "loss": 1.4711,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1599023356520938,
+      "grad_norm": 1.4316532360156768,
+      "learning_rate": 0.00019128360388673618,
+      "loss": 1.5496,
+      "step": 3471
+    },
+    {
+      "epoch": 0.1599484037407288,
+      "grad_norm": 3.6598387934170025,
+      "learning_rate": 0.0001912775102802637,
+      "loss": 1.4795,
+      "step": 3472
+    },
+    {
+      "epoch": 0.1599944718293638,
+      "grad_norm": 1.340019431198466,
+      "learning_rate": 0.00019127141464165393,
+      "loss": 1.5593,
+      "step": 3473
+    },
+    {
+      "epoch": 0.1600405399179988,
+      "grad_norm": 1.4198478334026627,
+      "learning_rate": 0.0001912653169710426,
+      "loss": 1.3692,
+      "step": 3474
+    },
+    {
+      "epoch": 0.1600866080066338,
+      "grad_norm": 1.9687010590699414,
+      "learning_rate": 0.00019125921726856552,
+      "loss": 1.2024,
+      "step": 3475
+    },
+    {
+      "epoch": 0.1601326760952688,
+      "grad_norm": 1.9016160895146175,
+      "learning_rate": 0.00019125311553435837,
+      "loss": 1.5989,
+      "step": 3476
+    },
+    {
+      "epoch": 0.16017874418390382,
+      "grad_norm": 1.3893753589959714,
+      "learning_rate": 0.00019124701176855715,
+      "loss": 1.0764,
+      "step": 3477
+    },
+    {
+      "epoch": 0.16022481227253882,
+      "grad_norm": 1.4972404531662962,
+      "learning_rate": 0.00019124090597129762,
+      "loss": 1.3996,
+      "step": 3478
+    },
+    {
+      "epoch": 0.16027088036117382,
+      "grad_norm": 2.2060935956569563,
+      "learning_rate": 0.0001912347981427158,
+      "loss": 0.8319,
+      "step": 3479
+    },
+    {
+      "epoch": 0.16031694844980882,
+      "grad_norm": 1.4130309799657366,
+      "learning_rate": 0.0001912286882829476,
+      "loss": 1.0044,
+      "step": 3480
+    },
+    {
+      "epoch": 0.16036301653844381,
+      "grad_norm": 5.143398014187552,
+      "learning_rate": 0.00019122257639212907,
+      "loss": 1.1401,
+      "step": 3481
+    },
+    {
+      "epoch": 0.1604090846270788,
+      "grad_norm": 1.1676784130867117,
+      "learning_rate": 0.00019121646247039634,
+      "loss": 1.464,
+      "step": 3482
+    },
+    {
+      "epoch": 0.1604551527157138,
+      "grad_norm": 1.4282771195954287,
+      "learning_rate": 0.00019121034651788544,
+      "loss": 1.4013,
+      "step": 3483
+    },
+    {
+      "epoch": 0.16050122080434884,
+      "grad_norm": 0.9788465420212445,
+      "learning_rate": 0.00019120422853473255,
+      "loss": 1.8276,
+      "step": 3484
+    },
+    {
+      "epoch": 0.16054728889298384,
+      "grad_norm": 1.0908492580449167,
+      "learning_rate": 0.00019119810852107394,
+      "loss": 1.2835,
+      "step": 3485
+    },
+    {
+      "epoch": 0.16059335698161883,
+      "grad_norm": 1.4578024960389755,
+      "learning_rate": 0.00019119198647704578,
+      "loss": 1.4151,
+      "step": 3486
+    },
+    {
+      "epoch": 0.16063942507025383,
+      "grad_norm": 1.4291413659017973,
+      "learning_rate": 0.00019118586240278443,
+      "loss": 1.3273,
+      "step": 3487
+    },
+    {
+      "epoch": 0.16068549315888883,
+      "grad_norm": 2.049866534385862,
+      "learning_rate": 0.00019117973629842618,
+      "loss": 1.5078,
+      "step": 3488
+    },
+    {
+      "epoch": 0.16073156124752383,
+      "grad_norm": 1.4020712078406454,
+      "learning_rate": 0.00019117360816410742,
+      "loss": 1.3988,
+      "step": 3489
+    },
+    {
+      "epoch": 0.16077762933615886,
+      "grad_norm": 1.4112266189530707,
+      "learning_rate": 0.00019116747799996465,
+      "loss": 1.4096,
+      "step": 3490
+    },
+    {
+      "epoch": 0.16082369742479385,
+      "grad_norm": 3.2032643429521666,
+      "learning_rate": 0.00019116134580613426,
+      "loss": 1.3644,
+      "step": 3491
+    },
+    {
+      "epoch": 0.16086976551342885,
+      "grad_norm": 1.1733852564068643,
+      "learning_rate": 0.00019115521158275285,
+      "loss": 1.3855,
+      "step": 3492
+    },
+    {
+      "epoch": 0.16091583360206385,
+      "grad_norm": 1.432908343693461,
+      "learning_rate": 0.00019114907532995695,
+      "loss": 1.467,
+      "step": 3493
+    },
+    {
+      "epoch": 0.16096190169069885,
+      "grad_norm": 1.6116449639533605,
+      "learning_rate": 0.00019114293704788316,
+      "loss": 1.298,
+      "step": 3494
+    },
+    {
+      "epoch": 0.16100796977933385,
+      "grad_norm": 1.2440052916140385,
+      "learning_rate": 0.00019113679673666813,
+      "loss": 1.7294,
+      "step": 3495
+    },
+    {
+      "epoch": 0.16105403786796885,
+      "grad_norm": 1.1770102430889762,
+      "learning_rate": 0.00019113065439644862,
+      "loss": 1.6192,
+      "step": 3496
+    },
+    {
+      "epoch": 0.16110010595660387,
+      "grad_norm": 1.3940284439097381,
+      "learning_rate": 0.00019112451002736136,
+      "loss": 1.5456,
+      "step": 3497
+    },
+    {
+      "epoch": 0.16114617404523887,
+      "grad_norm": 1.3594033069697242,
+      "learning_rate": 0.00019111836362954308,
+      "loss": 1.3106,
+      "step": 3498
+    },
+    {
+      "epoch": 0.16119224213387387,
+      "grad_norm": 1.2583467938134576,
+      "learning_rate": 0.0001911122152031307,
+      "loss": 1.3569,
+      "step": 3499
+    },
+    {
+      "epoch": 0.16123831022250887,
+      "grad_norm": 1.8965105787170635,
+      "learning_rate": 0.00019110606474826108,
+      "loss": 1.262,
+      "step": 3500
+    },
+    {
+      "epoch": 0.16128437831114387,
+      "grad_norm": 1.1989125945149095,
+      "learning_rate": 0.00019109991226507112,
+      "loss": 1.14,
+      "step": 3501
+    },
+    {
+      "epoch": 0.16133044639977887,
+      "grad_norm": 1.290583350044107,
+      "learning_rate": 0.00019109375775369785,
+      "loss": 1.2552,
+      "step": 3502
+    },
+    {
+      "epoch": 0.16137651448841386,
+      "grad_norm": 1.480816863914924,
+      "learning_rate": 0.0001910876012142782,
+      "loss": 1.7366,
+      "step": 3503
+    },
+    {
+      "epoch": 0.1614225825770489,
+      "grad_norm": 1.4935355783817734,
+      "learning_rate": 0.00019108144264694935,
+      "loss": 1.3884,
+      "step": 3504
+    },
+    {
+      "epoch": 0.1614686506656839,
+      "grad_norm": 1.9153491853867963,
+      "learning_rate": 0.0001910752820518483,
+      "loss": 1.5193,
+      "step": 3505
+    },
+    {
+      "epoch": 0.1615147187543189,
+      "grad_norm": 0.9485882327628752,
+      "learning_rate": 0.00019106911942911228,
+      "loss": 1.3963,
+      "step": 3506
+    },
+    {
+      "epoch": 0.16156078684295389,
+      "grad_norm": 1.4839634617244157,
+      "learning_rate": 0.00019106295477887847,
+      "loss": 1.4976,
+      "step": 3507
+    },
+    {
+      "epoch": 0.16160685493158888,
+      "grad_norm": 1.3125037321453337,
+      "learning_rate": 0.0001910567881012841,
+      "loss": 1.5868,
+      "step": 3508
+    },
+    {
+      "epoch": 0.16165292302022388,
+      "grad_norm": 2.1572259534994167,
+      "learning_rate": 0.00019105061939646647,
+      "loss": 1.3685,
+      "step": 3509
+    },
+    {
+      "epoch": 0.16169899110885888,
+      "grad_norm": 1.1098021199852908,
+      "learning_rate": 0.00019104444866456294,
+      "loss": 1.6742,
+      "step": 3510
+    },
+    {
+      "epoch": 0.1617450591974939,
+      "grad_norm": 1.5827233222639443,
+      "learning_rate": 0.00019103827590571086,
+      "loss": 1.2522,
+      "step": 3511
+    },
+    {
+      "epoch": 0.1617911272861289,
+      "grad_norm": 1.3214880572305017,
+      "learning_rate": 0.00019103210112004766,
+      "loss": 1.5432,
+      "step": 3512
+    },
+    {
+      "epoch": 0.1618371953747639,
+      "grad_norm": 2.27891914834201,
+      "learning_rate": 0.0001910259243077108,
+      "loss": 1.3175,
+      "step": 3513
+    },
+    {
+      "epoch": 0.1618832634633989,
+      "grad_norm": 1.19206310607214,
+      "learning_rate": 0.00019101974546883782,
+      "loss": 1.7605,
+      "step": 3514
+    },
+    {
+      "epoch": 0.1619293315520339,
+      "grad_norm": 7.541675283363203,
+      "learning_rate": 0.0001910135646035663,
+      "loss": 1.5101,
+      "step": 3515
+    },
+    {
+      "epoch": 0.1619753996406689,
+      "grad_norm": 2.1822090568929258,
+      "learning_rate": 0.0001910073817120338,
+      "loss": 1.4257,
+      "step": 3516
+    },
+    {
+      "epoch": 0.1620214677293039,
+      "grad_norm": 1.9719800017731022,
+      "learning_rate": 0.000191001196794378,
+      "loss": 1.1085,
+      "step": 3517
+    },
+    {
+      "epoch": 0.16206753581793892,
+      "grad_norm": 1.6672444864598945,
+      "learning_rate": 0.00019099500985073653,
+      "loss": 1.4963,
+      "step": 3518
+    },
+    {
+      "epoch": 0.16211360390657392,
+      "grad_norm": 1.6135471373203776,
+      "learning_rate": 0.00019098882088124724,
+      "loss": 1.2381,
+      "step": 3519
+    },
+    {
+      "epoch": 0.16215967199520892,
+      "grad_norm": 2.913655092425904,
+      "learning_rate": 0.00019098262988604786,
+      "loss": 1.5529,
+      "step": 3520
+    },
+    {
+      "epoch": 0.16220574008384392,
+      "grad_norm": 3.8609704098498603,
+      "learning_rate": 0.00019097643686527622,
+      "loss": 1.4022,
+      "step": 3521
+    },
+    {
+      "epoch": 0.16225180817247892,
+      "grad_norm": 1.3807419475846898,
+      "learning_rate": 0.0001909702418190702,
+      "loss": 1.5476,
+      "step": 3522
+    },
+    {
+      "epoch": 0.16229787626111392,
+      "grad_norm": 1.4823298834242429,
+      "learning_rate": 0.00019096404474756775,
+      "loss": 1.4936,
+      "step": 3523
+    },
+    {
+      "epoch": 0.16234394434974894,
+      "grad_norm": 1.190246635759897,
+      "learning_rate": 0.00019095784565090678,
+      "loss": 1.4888,
+      "step": 3524
+    },
+    {
+      "epoch": 0.16239001243838394,
+      "grad_norm": 2.0059608424777737,
+      "learning_rate": 0.00019095164452922537,
+      "loss": 1.1539,
+      "step": 3525
+    },
+    {
+      "epoch": 0.16243608052701894,
+      "grad_norm": 2.2396303544306417,
+      "learning_rate": 0.00019094544138266152,
+      "loss": 1.5314,
+      "step": 3526
+    },
+    {
+      "epoch": 0.16248214861565394,
+      "grad_norm": 1.7364900597596045,
+      "learning_rate": 0.00019093923621135337,
+      "loss": 1.5645,
+      "step": 3527
+    },
+    {
+      "epoch": 0.16252821670428894,
+      "grad_norm": 1.1691060762254415,
+      "learning_rate": 0.00019093302901543907,
+      "loss": 1.6187,
+      "step": 3528
+    },
+    {
+      "epoch": 0.16257428479292393,
+      "grad_norm": 1.421458387773932,
+      "learning_rate": 0.00019092681979505674,
+      "loss": 1.5215,
+      "step": 3529
+    },
+    {
+      "epoch": 0.16262035288155893,
+      "grad_norm": 1.2360935121430967,
+      "learning_rate": 0.00019092060855034471,
+      "loss": 1.3555,
+      "step": 3530
+    },
+    {
+      "epoch": 0.16266642097019396,
+      "grad_norm": 1.3567631095220816,
+      "learning_rate": 0.00019091439528144121,
+      "loss": 1.5683,
+      "step": 3531
+    },
+    {
+      "epoch": 0.16271248905882896,
+      "grad_norm": 1.3019548137151047,
+      "learning_rate": 0.00019090817998848462,
+      "loss": 1.3729,
+      "step": 3532
+    },
+    {
+      "epoch": 0.16275855714746396,
+      "grad_norm": 1.36149550456797,
+      "learning_rate": 0.00019090196267161324,
+      "loss": 1.1679,
+      "step": 3533
+    },
+    {
+      "epoch": 0.16280462523609895,
+      "grad_norm": 2.4806937622036003,
+      "learning_rate": 0.00019089574333096553,
+      "loss": 1.3031,
+      "step": 3534
+    },
+    {
+      "epoch": 0.16285069332473395,
+      "grad_norm": 1.395281479270793,
+      "learning_rate": 0.00019088952196667997,
+      "loss": 1.3819,
+      "step": 3535
+    },
+    {
+      "epoch": 0.16289676141336895,
+      "grad_norm": 1.323537620967235,
+      "learning_rate": 0.00019088329857889502,
+      "loss": 1.4153,
+      "step": 3536
+    },
+    {
+      "epoch": 0.16294282950200395,
+      "grad_norm": 2.8315398221934758,
+      "learning_rate": 0.00019087707316774928,
+      "loss": 1.647,
+      "step": 3537
+    },
+    {
+      "epoch": 0.16298889759063898,
+      "grad_norm": 1.4919787191289393,
+      "learning_rate": 0.0001908708457333813,
+      "loss": 1.1298,
+      "step": 3538
+    },
+    {
+      "epoch": 0.16303496567927397,
+      "grad_norm": 2.1273989648308964,
+      "learning_rate": 0.00019086461627592976,
+      "loss": 1.4933,
+      "step": 3539
+    },
+    {
+      "epoch": 0.16308103376790897,
+      "grad_norm": 2.0426523264069942,
+      "learning_rate": 0.00019085838479553335,
+      "loss": 1.5238,
+      "step": 3540
+    },
+    {
+      "epoch": 0.16312710185654397,
+      "grad_norm": 1.9585447519348649,
+      "learning_rate": 0.00019085215129233078,
+      "loss": 1.4368,
+      "step": 3541
+    },
+    {
+      "epoch": 0.16317316994517897,
+      "grad_norm": 1.9420275117340435,
+      "learning_rate": 0.00019084591576646085,
+      "loss": 1.0945,
+      "step": 3542
+    },
+    {
+      "epoch": 0.16321923803381397,
+      "grad_norm": 4.664317975827059,
+      "learning_rate": 0.00019083967821806238,
+      "loss": 1.1221,
+      "step": 3543
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 7.440726642305388,
+      "learning_rate": 0.00019083343864727422,
+      "loss": 1.5175,
+      "step": 3544
+    },
+    {
+      "epoch": 0.163311374211084,
+      "grad_norm": 2.253265813213899,
+      "learning_rate": 0.00019082719705423528,
+      "loss": 1.6834,
+      "step": 3545
+    },
+    {
+      "epoch": 0.163357442299719,
+      "grad_norm": 1.326157698698033,
+      "learning_rate": 0.00019082095343908454,
+      "loss": 1.244,
+      "step": 3546
+    },
+    {
+      "epoch": 0.163403510388354,
+      "grad_norm": 1.6739582445750412,
+      "learning_rate": 0.000190814707801961,
+      "loss": 1.1689,
+      "step": 3547
+    },
+    {
+      "epoch": 0.163449578476989,
+      "grad_norm": 2.082192741995367,
+      "learning_rate": 0.00019080846014300372,
+      "loss": 1.589,
+      "step": 3548
+    },
+    {
+      "epoch": 0.163495646565624,
+      "grad_norm": 1.976888940361307,
+      "learning_rate": 0.00019080221046235174,
+      "loss": 1.3129,
+      "step": 3549
+    },
+    {
+      "epoch": 0.16354171465425899,
+      "grad_norm": 2.878827495323535,
+      "learning_rate": 0.00019079595876014428,
+      "loss": 1.6043,
+      "step": 3550
+    },
+    {
+      "epoch": 0.16358778274289398,
+      "grad_norm": 1.7590763498354813,
+      "learning_rate": 0.00019078970503652045,
+      "loss": 1.397,
+      "step": 3551
+    },
+    {
+      "epoch": 0.163633850831529,
+      "grad_norm": 7.272774832128972,
+      "learning_rate": 0.00019078344929161954,
+      "loss": 1.2006,
+      "step": 3552
+    },
+    {
+      "epoch": 0.163679918920164,
+      "grad_norm": 2.7019123383594223,
+      "learning_rate": 0.00019077719152558076,
+      "loss": 1.3299,
+      "step": 3553
+    },
+    {
+      "epoch": 0.163725987008799,
+      "grad_norm": 4.889001323682043,
+      "learning_rate": 0.00019077093173854346,
+      "loss": 1.6444,
+      "step": 3554
+    },
+    {
+      "epoch": 0.163772055097434,
+      "grad_norm": 5.225035523941143,
+      "learning_rate": 0.00019076466993064704,
+      "loss": 1.6099,
+      "step": 3555
+    },
+    {
+      "epoch": 0.163818123186069,
+      "grad_norm": 1.6505251471996036,
+      "learning_rate": 0.00019075840610203082,
+      "loss": 1.7248,
+      "step": 3556
+    },
+    {
+      "epoch": 0.163864191274704,
+      "grad_norm": 2.5798820317813096,
+      "learning_rate": 0.0001907521402528343,
+      "loss": 1.3173,
+      "step": 3557
+    },
+    {
+      "epoch": 0.16391025936333903,
+      "grad_norm": 1.7021096543783838,
+      "learning_rate": 0.00019074587238319704,
+      "loss": 1.617,
+      "step": 3558
+    },
+    {
+      "epoch": 0.16395632745197403,
+      "grad_norm": 6.621380558402467,
+      "learning_rate": 0.0001907396024932585,
+      "loss": 1.4778,
+      "step": 3559
+    },
+    {
+      "epoch": 0.16400239554060903,
+      "grad_norm": 2.3468100707906348,
+      "learning_rate": 0.00019073333058315832,
+      "loss": 1.4485,
+      "step": 3560
+    },
+    {
+      "epoch": 0.16404846362924402,
+      "grad_norm": 2.78518609981551,
+      "learning_rate": 0.00019072705665303607,
+      "loss": 1.5377,
+      "step": 3561
+    },
+    {
+      "epoch": 0.16409453171787902,
+      "grad_norm": 1.889672769889482,
+      "learning_rate": 0.00019072078070303148,
+      "loss": 1.4659,
+      "step": 3562
+    },
+    {
+      "epoch": 0.16414059980651402,
+      "grad_norm": 1.7360872409703774,
+      "learning_rate": 0.00019071450273328426,
+      "loss": 1.5073,
+      "step": 3563
+    },
+    {
+      "epoch": 0.16418666789514902,
+      "grad_norm": 1.5326029851021432,
+      "learning_rate": 0.0001907082227439342,
+      "loss": 1.2743,
+      "step": 3564
+    },
+    {
+      "epoch": 0.16423273598378405,
+      "grad_norm": 2.450696467725253,
+      "learning_rate": 0.00019070194073512106,
+      "loss": 1.4508,
+      "step": 3565
+    },
+    {
+      "epoch": 0.16427880407241904,
+      "grad_norm": 1.9612641821289343,
+      "learning_rate": 0.00019069565670698477,
+      "loss": 1.3854,
+      "step": 3566
+    },
+    {
+      "epoch": 0.16432487216105404,
+      "grad_norm": 1.7252295045768307,
+      "learning_rate": 0.00019068937065966517,
+      "loss": 1.3829,
+      "step": 3567
+    },
+    {
+      "epoch": 0.16437094024968904,
+      "grad_norm": 1.377442891444874,
+      "learning_rate": 0.0001906830825933022,
+      "loss": 1.5635,
+      "step": 3568
+    },
+    {
+      "epoch": 0.16441700833832404,
+      "grad_norm": 1.8276172456881756,
+      "learning_rate": 0.00019067679250803595,
+      "loss": 1.5107,
+      "step": 3569
+    },
+    {
+      "epoch": 0.16446307642695904,
+      "grad_norm": 1.4382466995341394,
+      "learning_rate": 0.00019067050040400633,
+      "loss": 1.0616,
+      "step": 3570
+    },
+    {
+      "epoch": 0.16450914451559404,
+      "grad_norm": 2.0579427945620057,
+      "learning_rate": 0.00019066420628135352,
+      "loss": 1.6043,
+      "step": 3571
+    },
+    {
+      "epoch": 0.16455521260422906,
+      "grad_norm": 1.2867954912396675,
+      "learning_rate": 0.0001906579101402176,
+      "loss": 1.4339,
+      "step": 3572
+    },
+    {
+      "epoch": 0.16460128069286406,
+      "grad_norm": 1.9386944349997919,
+      "learning_rate": 0.00019065161198073875,
+      "loss": 1.3442,
+      "step": 3573
+    },
+    {
+      "epoch": 0.16464734878149906,
+      "grad_norm": 1.5853585379009645,
+      "learning_rate": 0.0001906453118030572,
+      "loss": 1.2127,
+      "step": 3574
+    },
+    {
+      "epoch": 0.16469341687013406,
+      "grad_norm": 2.940427156849088,
+      "learning_rate": 0.00019063900960731322,
+      "loss": 1.4235,
+      "step": 3575
+    },
+    {
+      "epoch": 0.16473948495876906,
+      "grad_norm": 1.8583602833936288,
+      "learning_rate": 0.00019063270539364707,
+      "loss": 0.9734,
+      "step": 3576
+    },
+    {
+      "epoch": 0.16478555304740405,
+      "grad_norm": 1.7353255512910322,
+      "learning_rate": 0.00019062639916219917,
+      "loss": 1.4498,
+      "step": 3577
+    },
+    {
+      "epoch": 0.16483162113603905,
+      "grad_norm": 1.4526111820270675,
+      "learning_rate": 0.00019062009091310987,
+      "loss": 1.5624,
+      "step": 3578
+    },
+    {
+      "epoch": 0.16487768922467408,
+      "grad_norm": 1.3477422147650646,
+      "learning_rate": 0.00019061378064651963,
+      "loss": 1.3229,
+      "step": 3579
+    },
+    {
+      "epoch": 0.16492375731330908,
+      "grad_norm": 1.8770293765555595,
+      "learning_rate": 0.00019060746836256894,
+      "loss": 1.3043,
+      "step": 3580
+    },
+    {
+      "epoch": 0.16496982540194408,
+      "grad_norm": 0.9881630551938196,
+      "learning_rate": 0.0001906011540613983,
+      "loss": 1.637,
+      "step": 3581
+    },
+    {
+      "epoch": 0.16501589349057907,
+      "grad_norm": 1.3229637913539862,
+      "learning_rate": 0.00019059483774314836,
+      "loss": 1.2461,
+      "step": 3582
+    },
+    {
+      "epoch": 0.16506196157921407,
+      "grad_norm": 1.3331949718781508,
+      "learning_rate": 0.00019058851940795965,
+      "loss": 1.1302,
+      "step": 3583
+    },
+    {
+      "epoch": 0.16510802966784907,
+      "grad_norm": 1.4154847691185601,
+      "learning_rate": 0.0001905821990559729,
+      "loss": 1.5668,
+      "step": 3584
+    },
+    {
+      "epoch": 0.1651540977564841,
+      "grad_norm": 1.7657964091860567,
+      "learning_rate": 0.00019057587668732878,
+      "loss": 1.2303,
+      "step": 3585
+    },
+    {
+      "epoch": 0.1652001658451191,
+      "grad_norm": 0.8658480865065935,
+      "learning_rate": 0.00019056955230216808,
+      "loss": 1.8347,
+      "step": 3586
+    },
+    {
+      "epoch": 0.1652462339337541,
+      "grad_norm": 2.7130550635612907,
+      "learning_rate": 0.00019056322590063163,
+      "loss": 1.4961,
+      "step": 3587
+    },
+    {
+      "epoch": 0.1652923020223891,
+      "grad_norm": 1.3063399927177062,
+      "learning_rate": 0.0001905568974828602,
+      "loss": 1.7497,
+      "step": 3588
+    },
+    {
+      "epoch": 0.1653383701110241,
+      "grad_norm": 1.3896379370093273,
+      "learning_rate": 0.00019055056704899472,
+      "loss": 1.2777,
+      "step": 3589
+    },
+    {
+      "epoch": 0.1653844381996591,
+      "grad_norm": 3.175530020865243,
+      "learning_rate": 0.00019054423459917614,
+      "loss": 1.5057,
+      "step": 3590
+    },
+    {
+      "epoch": 0.1654305062882941,
+      "grad_norm": 4.21183860923458,
+      "learning_rate": 0.0001905379001335454,
+      "loss": 1.2265,
+      "step": 3591
+    },
+    {
+      "epoch": 0.16547657437692911,
+      "grad_norm": 1.5516155040579487,
+      "learning_rate": 0.0001905315636522436,
+      "loss": 1.3758,
+      "step": 3592
+    },
+    {
+      "epoch": 0.1655226424655641,
+      "grad_norm": 2.309060050622076,
+      "learning_rate": 0.0001905252251554117,
+      "loss": 1.4682,
+      "step": 3593
+    },
+    {
+      "epoch": 0.1655687105541991,
+      "grad_norm": 1.637118318223231,
+      "learning_rate": 0.00019051888464319093,
+      "loss": 1.7023,
+      "step": 3594
+    },
+    {
+      "epoch": 0.1656147786428341,
+      "grad_norm": 1.2863034392313277,
+      "learning_rate": 0.00019051254211572237,
+      "loss": 1.6213,
+      "step": 3595
+    },
+    {
+      "epoch": 0.1656608467314691,
+      "grad_norm": 2.041700027724201,
+      "learning_rate": 0.0001905061975731473,
+      "loss": 1.394,
+      "step": 3596
+    },
+    {
+      "epoch": 0.1657069148201041,
+      "grad_norm": 1.6676757236350208,
+      "learning_rate": 0.0001904998510156069,
+      "loss": 1.2552,
+      "step": 3597
+    },
+    {
+      "epoch": 0.1657529829087391,
+      "grad_norm": 2.2719834038357583,
+      "learning_rate": 0.00019049350244324247,
+      "loss": 1.3761,
+      "step": 3598
+    },
+    {
+      "epoch": 0.16579905099737413,
+      "grad_norm": 1.7231564841305227,
+      "learning_rate": 0.00019048715185619538,
+      "loss": 1.3468,
+      "step": 3599
+    },
+    {
+      "epoch": 0.16584511908600913,
+      "grad_norm": 2.0814163246910917,
+      "learning_rate": 0.00019048079925460703,
+      "loss": 1.0223,
+      "step": 3600
+    },
+    {
+      "epoch": 0.16589118717464413,
+      "grad_norm": 1.1046413719058845,
+      "learning_rate": 0.0001904744446386188,
+      "loss": 1.7705,
+      "step": 3601
+    },
+    {
+      "epoch": 0.16593725526327913,
+      "grad_norm": 9.222936399871397,
+      "learning_rate": 0.00019046808800837222,
+      "loss": 1.4346,
+      "step": 3602
+    },
+    {
+      "epoch": 0.16598332335191412,
+      "grad_norm": 1.9345779637291745,
+      "learning_rate": 0.00019046172936400876,
+      "loss": 1.2948,
+      "step": 3603
+    },
+    {
+      "epoch": 0.16602939144054912,
+      "grad_norm": 1.455878535800414,
+      "learning_rate": 0.00019045536870567,
+      "loss": 1.3625,
+      "step": 3604
+    },
+    {
+      "epoch": 0.16607545952918412,
+      "grad_norm": 1.2069376395542946,
+      "learning_rate": 0.00019044900603349755,
+      "loss": 1.5644,
+      "step": 3605
+    },
+    {
+      "epoch": 0.16612152761781915,
+      "grad_norm": 2.925076667033953,
+      "learning_rate": 0.0001904426413476331,
+      "loss": 1.2408,
+      "step": 3606
+    },
+    {
+      "epoch": 0.16616759570645415,
+      "grad_norm": 1.570336805610306,
+      "learning_rate": 0.00019043627464821828,
+      "loss": 1.7742,
+      "step": 3607
+    },
+    {
+      "epoch": 0.16621366379508914,
+      "grad_norm": 7.2424794035293445,
+      "learning_rate": 0.0001904299059353949,
+      "loss": 1.1486,
+      "step": 3608
+    },
+    {
+      "epoch": 0.16625973188372414,
+      "grad_norm": 4.622037320157197,
+      "learning_rate": 0.00019042353520930467,
+      "loss": 1.1494,
+      "step": 3609
+    },
+    {
+      "epoch": 0.16630579997235914,
+      "grad_norm": 1.4663819618780738,
+      "learning_rate": 0.0001904171624700895,
+      "loss": 1.6054,
+      "step": 3610
+    },
+    {
+      "epoch": 0.16635186806099414,
+      "grad_norm": 2.5037124442835808,
+      "learning_rate": 0.00019041078771789126,
+      "loss": 1.3845,
+      "step": 3611
+    },
+    {
+      "epoch": 0.16639793614962914,
+      "grad_norm": 2.7345575616361275,
+      "learning_rate": 0.00019040441095285183,
+      "loss": 1.3422,
+      "step": 3612
+    },
+    {
+      "epoch": 0.16644400423826416,
+      "grad_norm": 1.8765999277372851,
+      "learning_rate": 0.00019039803217511323,
+      "loss": 1.1361,
+      "step": 3613
+    },
+    {
+      "epoch": 0.16649007232689916,
+      "grad_norm": 1.5891155108282877,
+      "learning_rate": 0.0001903916513848174,
+      "loss": 1.3316,
+      "step": 3614
+    },
+    {
+      "epoch": 0.16653614041553416,
+      "grad_norm": 1.3259245991591133,
+      "learning_rate": 0.00019038526858210645,
+      "loss": 1.475,
+      "step": 3615
+    },
+    {
+      "epoch": 0.16658220850416916,
+      "grad_norm": 1.4231719259586908,
+      "learning_rate": 0.0001903788837671225,
+      "loss": 1.5467,
+      "step": 3616
+    },
+    {
+      "epoch": 0.16662827659280416,
+      "grad_norm": 2.317419199889156,
+      "learning_rate": 0.00019037249694000763,
+      "loss": 1.1039,
+      "step": 3617
+    },
+    {
+      "epoch": 0.16667434468143916,
+      "grad_norm": 1.6426153629725766,
+      "learning_rate": 0.0001903661081009041,
+      "loss": 1.3809,
+      "step": 3618
+    },
+    {
+      "epoch": 0.16672041277007418,
+      "grad_norm": 2.0008475742669027,
+      "learning_rate": 0.00019035971724995412,
+      "loss": 1.2739,
+      "step": 3619
+    },
+    {
+      "epoch": 0.16676648085870918,
+      "grad_norm": 1.3496374868002916,
+      "learning_rate": 0.00019035332438729997,
+      "loss": 1.268,
+      "step": 3620
+    },
+    {
+      "epoch": 0.16681254894734418,
+      "grad_norm": 1.1090381482667213,
+      "learning_rate": 0.00019034692951308394,
+      "loss": 1.4396,
+      "step": 3621
+    },
+    {
+      "epoch": 0.16685861703597918,
+      "grad_norm": 1.7103824976931203,
+      "learning_rate": 0.00019034053262744846,
+      "loss": 1.4555,
+      "step": 3622
+    },
+    {
+      "epoch": 0.16690468512461418,
+      "grad_norm": 1.4691327378478427,
+      "learning_rate": 0.00019033413373053593,
+      "loss": 1.433,
+      "step": 3623
+    },
+    {
+      "epoch": 0.16695075321324918,
+      "grad_norm": 1.9147879803547037,
+      "learning_rate": 0.0001903277328224888,
+      "loss": 1.4623,
+      "step": 3624
+    },
+    {
+      "epoch": 0.16699682130188417,
+      "grad_norm": 2.0794051053531164,
+      "learning_rate": 0.00019032132990344958,
+      "loss": 1.6013,
+      "step": 3625
+    },
+    {
+      "epoch": 0.1670428893905192,
+      "grad_norm": 1.3634864731842236,
+      "learning_rate": 0.0001903149249735608,
+      "loss": 1.5425,
+      "step": 3626
+    },
+    {
+      "epoch": 0.1670889574791542,
+      "grad_norm": 2.4605950184386414,
+      "learning_rate": 0.00019030851803296508,
+      "loss": 1.3874,
+      "step": 3627
+    },
+    {
+      "epoch": 0.1671350255677892,
+      "grad_norm": 2.050963477457186,
+      "learning_rate": 0.00019030210908180505,
+      "loss": 0.9918,
+      "step": 3628
+    },
+    {
+      "epoch": 0.1671810936564242,
+      "grad_norm": 2.438116369298337,
+      "learning_rate": 0.0001902956981202234,
+      "loss": 1.474,
+      "step": 3629
+    },
+    {
+      "epoch": 0.1672271617450592,
+      "grad_norm": 2.9461074162701855,
+      "learning_rate": 0.00019028928514836286,
+      "loss": 1.1653,
+      "step": 3630
+    },
+    {
+      "epoch": 0.1672732298336942,
+      "grad_norm": 1.4662251199110463,
+      "learning_rate": 0.0001902828701663662,
+      "loss": 1.4993,
+      "step": 3631
+    },
+    {
+      "epoch": 0.1673192979223292,
+      "grad_norm": 1.828504002633899,
+      "learning_rate": 0.00019027645317437625,
+      "loss": 1.7286,
+      "step": 3632
+    },
+    {
+      "epoch": 0.16736536601096422,
+      "grad_norm": 1.5891206821234516,
+      "learning_rate": 0.00019027003417253582,
+      "loss": 1.6074,
+      "step": 3633
+    },
+    {
+      "epoch": 0.16741143409959922,
+      "grad_norm": 1.58225753954628,
+      "learning_rate": 0.0001902636131609879,
+      "loss": 1.4324,
+      "step": 3634
+    },
+    {
+      "epoch": 0.16745750218823421,
+      "grad_norm": 1.9490418660326352,
+      "learning_rate": 0.00019025719013987538,
+      "loss": 1.5771,
+      "step": 3635
+    },
+    {
+      "epoch": 0.1675035702768692,
+      "grad_norm": 2.296574659500645,
+      "learning_rate": 0.00019025076510934127,
+      "loss": 1.4653,
+      "step": 3636
+    },
+    {
+      "epoch": 0.1675496383655042,
+      "grad_norm": 2.4233215741522027,
+      "learning_rate": 0.00019024433806952865,
+      "loss": 1.427,
+      "step": 3637
+    },
+    {
+      "epoch": 0.1675957064541392,
+      "grad_norm": 2.387227565765917,
+      "learning_rate": 0.00019023790902058054,
+      "loss": 1.3124,
+      "step": 3638
+    },
+    {
+      "epoch": 0.1676417745427742,
+      "grad_norm": 2.1438254615595413,
+      "learning_rate": 0.00019023147796264013,
+      "loss": 1.3403,
+      "step": 3639
+    },
+    {
+      "epoch": 0.16768784263140923,
+      "grad_norm": 1.4417802138901359,
+      "learning_rate": 0.0001902250448958506,
+      "loss": 1.5049,
+      "step": 3640
+    },
+    {
+      "epoch": 0.16773391072004423,
+      "grad_norm": 1.559191147665353,
+      "learning_rate": 0.00019021860982035513,
+      "loss": 1.485,
+      "step": 3641
+    },
+    {
+      "epoch": 0.16777997880867923,
+      "grad_norm": 1.4394258737399306,
+      "learning_rate": 0.00019021217273629701,
+      "loss": 1.6294,
+      "step": 3642
+    },
+    {
+      "epoch": 0.16782604689731423,
+      "grad_norm": 1.6248293988814864,
+      "learning_rate": 0.00019020573364381954,
+      "loss": 1.3397,
+      "step": 3643
+    },
+    {
+      "epoch": 0.16787211498594923,
+      "grad_norm": 1.8739557468475787,
+      "learning_rate": 0.00019019929254306608,
+      "loss": 1.2459,
+      "step": 3644
+    },
+    {
+      "epoch": 0.16791818307458423,
+      "grad_norm": 1.793271507940532,
+      "learning_rate": 0.00019019284943418003,
+      "loss": 1.649,
+      "step": 3645
+    },
+    {
+      "epoch": 0.16796425116321922,
+      "grad_norm": 5.538314921780239,
+      "learning_rate": 0.0001901864043173048,
+      "loss": 1.2181,
+      "step": 3646
+    },
+    {
+      "epoch": 0.16801031925185425,
+      "grad_norm": 1.5996463017363014,
+      "learning_rate": 0.00019017995719258397,
+      "loss": 1.5298,
+      "step": 3647
+    },
+    {
+      "epoch": 0.16805638734048925,
+      "grad_norm": 1.1334281101707078,
+      "learning_rate": 0.000190173508060161,
+      "loss": 1.7792,
+      "step": 3648
+    },
+    {
+      "epoch": 0.16810245542912425,
+      "grad_norm": 1.3427605594345853,
+      "learning_rate": 0.0001901670569201795,
+      "loss": 1.601,
+      "step": 3649
+    },
+    {
+      "epoch": 0.16814852351775925,
+      "grad_norm": 5.033721009010165,
+      "learning_rate": 0.00019016060377278305,
+      "loss": 1.5012,
+      "step": 3650
+    },
+    {
+      "epoch": 0.16819459160639424,
+      "grad_norm": 1.765576480861213,
+      "learning_rate": 0.00019015414861811537,
+      "loss": 1.2674,
+      "step": 3651
+    },
+    {
+      "epoch": 0.16824065969502924,
+      "grad_norm": 2.5561282759001216,
+      "learning_rate": 0.00019014769145632015,
+      "loss": 1.2595,
+      "step": 3652
+    },
+    {
+      "epoch": 0.16828672778366427,
+      "grad_norm": 3.1524967428853397,
+      "learning_rate": 0.00019014123228754115,
+      "loss": 1.2334,
+      "step": 3653
+    },
+    {
+      "epoch": 0.16833279587229927,
+      "grad_norm": 2.8741948102416823,
+      "learning_rate": 0.00019013477111192218,
+      "loss": 1.2234,
+      "step": 3654
+    },
+    {
+      "epoch": 0.16837886396093427,
+      "grad_norm": 1.9804776505268977,
+      "learning_rate": 0.00019012830792960704,
+      "loss": 1.6435,
+      "step": 3655
+    },
+    {
+      "epoch": 0.16842493204956926,
+      "grad_norm": 9.755657032337941,
+      "learning_rate": 0.0001901218427407397,
+      "loss": 1.2574,
+      "step": 3656
+    },
+    {
+      "epoch": 0.16847100013820426,
+      "grad_norm": 3.2728913996226896,
+      "learning_rate": 0.00019011537554546406,
+      "loss": 1.342,
+      "step": 3657
+    },
+    {
+      "epoch": 0.16851706822683926,
+      "grad_norm": 1.6830915084260496,
+      "learning_rate": 0.00019010890634392407,
+      "loss": 1.4956,
+      "step": 3658
+    },
+    {
+      "epoch": 0.16856313631547426,
+      "grad_norm": 2.1404776904270895,
+      "learning_rate": 0.0001901024351362638,
+      "loss": 1.4209,
+      "step": 3659
+    },
+    {
+      "epoch": 0.1686092044041093,
+      "grad_norm": 1.737791640030624,
+      "learning_rate": 0.0001900959619226273,
+      "loss": 1.6482,
+      "step": 3660
+    },
+    {
+      "epoch": 0.16865527249274428,
+      "grad_norm": 3.177368576504768,
+      "learning_rate": 0.00019008948670315868,
+      "loss": 1.2763,
+      "step": 3661
+    },
+    {
+      "epoch": 0.16870134058137928,
+      "grad_norm": 4.226033081852363,
+      "learning_rate": 0.00019008300947800214,
+      "loss": 1.2406,
+      "step": 3662
+    },
+    {
+      "epoch": 0.16874740867001428,
+      "grad_norm": 6.377129007616977,
+      "learning_rate": 0.0001900765302473018,
+      "loss": 1.5693,
+      "step": 3663
+    },
+    {
+      "epoch": 0.16879347675864928,
+      "grad_norm": 2.204836690555917,
+      "learning_rate": 0.00019007004901120203,
+      "loss": 1.4207,
+      "step": 3664
+    },
+    {
+      "epoch": 0.16883954484728428,
+      "grad_norm": 1.4607062221430622,
+      "learning_rate": 0.00019006356576984704,
+      "loss": 1.6905,
+      "step": 3665
+    },
+    {
+      "epoch": 0.16888561293591928,
+      "grad_norm": 1.717729141420124,
+      "learning_rate": 0.00019005708052338114,
+      "loss": 1.4172,
+      "step": 3666
+    },
+    {
+      "epoch": 0.1689316810245543,
+      "grad_norm": 1.7639353472581185,
+      "learning_rate": 0.00019005059327194878,
+      "loss": 1.5161,
+      "step": 3667
+    },
+    {
+      "epoch": 0.1689777491131893,
+      "grad_norm": 1.6934630453446455,
+      "learning_rate": 0.00019004410401569436,
+      "loss": 1.4079,
+      "step": 3668
+    },
+    {
+      "epoch": 0.1690238172018243,
+      "grad_norm": 2.5085816370081098,
+      "learning_rate": 0.00019003761275476237,
+      "loss": 1.759,
+      "step": 3669
+    },
+    {
+      "epoch": 0.1690698852904593,
+      "grad_norm": 1.4873737079943632,
+      "learning_rate": 0.00019003111948929732,
+      "loss": 1.4472,
+      "step": 3670
+    },
+    {
+      "epoch": 0.1691159533790943,
+      "grad_norm": 2.065561649502632,
+      "learning_rate": 0.00019002462421944375,
+      "loss": 1.362,
+      "step": 3671
+    },
+    {
+      "epoch": 0.1691620214677293,
+      "grad_norm": 2.3732216359474365,
+      "learning_rate": 0.00019001812694534627,
+      "loss": 1.3238,
+      "step": 3672
+    },
+    {
+      "epoch": 0.1692080895563643,
+      "grad_norm": 3.247069966837735,
+      "learning_rate": 0.00019001162766714956,
+      "loss": 1.5477,
+      "step": 3673
+    },
+    {
+      "epoch": 0.16925415764499932,
+      "grad_norm": 1.388484019233073,
+      "learning_rate": 0.00019000512638499829,
+      "loss": 1.5543,
+      "step": 3674
+    },
+    {
+      "epoch": 0.16930022573363432,
+      "grad_norm": 2.444622693799421,
+      "learning_rate": 0.0001899986230990372,
+      "loss": 1.3289,
+      "step": 3675
+    },
+    {
+      "epoch": 0.16934629382226932,
+      "grad_norm": 2.3682887322500017,
+      "learning_rate": 0.00018999211780941112,
+      "loss": 1.6597,
+      "step": 3676
+    },
+    {
+      "epoch": 0.16939236191090432,
+      "grad_norm": 2.509756233744145,
+      "learning_rate": 0.00018998561051626479,
+      "loss": 1.178,
+      "step": 3677
+    },
+    {
+      "epoch": 0.16943842999953931,
+      "grad_norm": 1.7370560401911521,
+      "learning_rate": 0.00018997910121974315,
+      "loss": 1.3159,
+      "step": 3678
+    },
+    {
+      "epoch": 0.1694844980881743,
+      "grad_norm": 3.459563180142823,
+      "learning_rate": 0.00018997258991999111,
+      "loss": 1.3041,
+      "step": 3679
+    },
+    {
+      "epoch": 0.1695305661768093,
+      "grad_norm": 2.2161413321265675,
+      "learning_rate": 0.0001899660766171536,
+      "loss": 1.3838,
+      "step": 3680
+    },
+    {
+      "epoch": 0.16957663426544434,
+      "grad_norm": 1.4572987247950024,
+      "learning_rate": 0.00018995956131137568,
+      "loss": 1.3634,
+      "step": 3681
+    },
+    {
+      "epoch": 0.16962270235407934,
+      "grad_norm": 8.815586868365083,
+      "learning_rate": 0.00018995304400280238,
+      "loss": 1.3249,
+      "step": 3682
+    },
+    {
+      "epoch": 0.16966877044271433,
+      "grad_norm": 2.6406877807669953,
+      "learning_rate": 0.0001899465246915788,
+      "loss": 1.5153,
+      "step": 3683
+    },
+    {
+      "epoch": 0.16971483853134933,
+      "grad_norm": 2.175116553798199,
+      "learning_rate": 0.00018994000337785004,
+      "loss": 1.4203,
+      "step": 3684
+    },
+    {
+      "epoch": 0.16976090661998433,
+      "grad_norm": 2.08589040432425,
+      "learning_rate": 0.00018993348006176132,
+      "loss": 1.6129,
+      "step": 3685
+    },
+    {
+      "epoch": 0.16980697470861933,
+      "grad_norm": 3.9948110787627362,
+      "learning_rate": 0.0001899269547434579,
+      "loss": 1.4699,
+      "step": 3686
+    },
+    {
+      "epoch": 0.16985304279725436,
+      "grad_norm": 3.247632568990002,
+      "learning_rate": 0.00018992042742308501,
+      "loss": 1.4157,
+      "step": 3687
+    },
+    {
+      "epoch": 0.16989911088588935,
+      "grad_norm": 2.2127420092685783,
+      "learning_rate": 0.00018991389810078796,
+      "loss": 1.5811,
+      "step": 3688
+    },
+    {
+      "epoch": 0.16994517897452435,
+      "grad_norm": 2.3585253060570444,
+      "learning_rate": 0.00018990736677671217,
+      "loss": 1.4968,
+      "step": 3689
+    },
+    {
+      "epoch": 0.16999124706315935,
+      "grad_norm": 5.042229978449478,
+      "learning_rate": 0.00018990083345100302,
+      "loss": 1.4337,
+      "step": 3690
+    },
+    {
+      "epoch": 0.17003731515179435,
+      "grad_norm": 2.7577784806708086,
+      "learning_rate": 0.00018989429812380594,
+      "loss": 1.4733,
+      "step": 3691
+    },
+    {
+      "epoch": 0.17008338324042935,
+      "grad_norm": 2.6241393658454615,
+      "learning_rate": 0.00018988776079526643,
+      "loss": 1.423,
+      "step": 3692
+    },
+    {
+      "epoch": 0.17012945132906435,
+      "grad_norm": 2.7935927984081954,
+      "learning_rate": 0.00018988122146553009,
+      "loss": 1.5223,
+      "step": 3693
+    },
+    {
+      "epoch": 0.17017551941769937,
+      "grad_norm": 3.7835714074361144,
+      "learning_rate": 0.00018987468013474245,
+      "loss": 1.5098,
+      "step": 3694
+    },
+    {
+      "epoch": 0.17022158750633437,
+      "grad_norm": 2.277641131088796,
+      "learning_rate": 0.0001898681368030492,
+      "loss": 1.283,
+      "step": 3695
+    },
+    {
+      "epoch": 0.17026765559496937,
+      "grad_norm": 6.096509853378525,
+      "learning_rate": 0.0001898615914705959,
+      "loss": 1.6162,
+      "step": 3696
+    },
+    {
+      "epoch": 0.17031372368360437,
+      "grad_norm": 2.1194701024680467,
+      "learning_rate": 0.00018985504413752838,
+      "loss": 1.57,
+      "step": 3697
+    },
+    {
+      "epoch": 0.17035979177223937,
+      "grad_norm": 2.400148833846878,
+      "learning_rate": 0.00018984849480399238,
+      "loss": 1.3676,
+      "step": 3698
+    },
+    {
+      "epoch": 0.17040585986087436,
+      "grad_norm": 2.438472113185096,
+      "learning_rate": 0.0001898419434701337,
+      "loss": 1.638,
+      "step": 3699
+    },
+    {
+      "epoch": 0.17045192794950936,
+      "grad_norm": 2.4005356114576055,
+      "learning_rate": 0.00018983539013609817,
+      "loss": 1.6296,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1704979960381444,
+      "grad_norm": 1.90053003256291,
+      "learning_rate": 0.00018982883480203174,
+      "loss": 1.2747,
+      "step": 3701
+    },
+    {
+      "epoch": 0.1705440641267794,
+      "grad_norm": 4.239240055279759,
+      "learning_rate": 0.0001898222774680803,
+      "loss": 1.3654,
+      "step": 3702
+    },
+    {
+      "epoch": 0.1705901322154144,
+      "grad_norm": 2.1437178403030055,
+      "learning_rate": 0.0001898157181343899,
+      "loss": 1.5615,
+      "step": 3703
+    },
+    {
+      "epoch": 0.17063620030404938,
+      "grad_norm": 2.1882743533130053,
+      "learning_rate": 0.00018980915680110648,
+      "loss": 1.154,
+      "step": 3704
+    },
+    {
+      "epoch": 0.17068226839268438,
+      "grad_norm": 3.5873947514276896,
+      "learning_rate": 0.00018980259346837623,
+      "loss": 1.2746,
+      "step": 3705
+    },
+    {
+      "epoch": 0.17072833648131938,
+      "grad_norm": 2.392189272381574,
+      "learning_rate": 0.0001897960281363452,
+      "loss": 1.4692,
+      "step": 3706
+    },
+    {
+      "epoch": 0.17077440456995438,
+      "grad_norm": 2.145354401758873,
+      "learning_rate": 0.00018978946080515957,
+      "loss": 1.5202,
+      "step": 3707
+    },
+    {
+      "epoch": 0.1708204726585894,
+      "grad_norm": 1.6111177054312567,
+      "learning_rate": 0.00018978289147496554,
+      "loss": 1.236,
+      "step": 3708
+    },
+    {
+      "epoch": 0.1708665407472244,
+      "grad_norm": 3.627751297110911,
+      "learning_rate": 0.00018977632014590938,
+      "loss": 1.501,
+      "step": 3709
+    },
+    {
+      "epoch": 0.1709126088358594,
+      "grad_norm": 5.384987825328456,
+      "learning_rate": 0.00018976974681813737,
+      "loss": 1.4157,
+      "step": 3710
+    },
+    {
+      "epoch": 0.1709586769244944,
+      "grad_norm": 2.035155309791946,
+      "learning_rate": 0.0001897631714917959,
+      "loss": 1.5177,
+      "step": 3711
+    },
+    {
+      "epoch": 0.1710047450131294,
+      "grad_norm": 2.871278713946103,
+      "learning_rate": 0.00018975659416703132,
+      "loss": 1.5327,
+      "step": 3712
+    },
+    {
+      "epoch": 0.1710508131017644,
+      "grad_norm": 2.628542315506275,
+      "learning_rate": 0.00018975001484399006,
+      "loss": 1.2183,
+      "step": 3713
+    },
+    {
+      "epoch": 0.1710968811903994,
+      "grad_norm": 8.091126249028477,
+      "learning_rate": 0.00018974343352281864,
+      "loss": 1.572,
+      "step": 3714
+    },
+    {
+      "epoch": 0.17114294927903442,
+      "grad_norm": 9.78636992456099,
+      "learning_rate": 0.0001897368502036635,
+      "loss": 1.505,
+      "step": 3715
+    },
+    {
+      "epoch": 0.17118901736766942,
+      "grad_norm": 2.740242444377654,
+      "learning_rate": 0.00018973026488667126,
+      "loss": 1.446,
+      "step": 3716
+    },
+    {
+      "epoch": 0.17123508545630442,
+      "grad_norm": 3.0380769581676463,
+      "learning_rate": 0.00018972367757198857,
+      "loss": 1.4626,
+      "step": 3717
+    },
+    {
+      "epoch": 0.17128115354493942,
+      "grad_norm": 1.9206605345054573,
+      "learning_rate": 0.000189717088259762,
+      "loss": 1.5976,
+      "step": 3718
+    },
+    {
+      "epoch": 0.17132722163357442,
+      "grad_norm": 9.488721689055192,
+      "learning_rate": 0.00018971049695013832,
+      "loss": 1.6703,
+      "step": 3719
+    },
+    {
+      "epoch": 0.17137328972220942,
+      "grad_norm": 1.8219927833463923,
+      "learning_rate": 0.00018970390364326422,
+      "loss": 1.5393,
+      "step": 3720
+    },
+    {
+      "epoch": 0.17141935781084444,
+      "grad_norm": 5.674841226683073,
+      "learning_rate": 0.00018969730833928657,
+      "loss": 1.665,
+      "step": 3721
+    },
+    {
+      "epoch": 0.17146542589947944,
+      "grad_norm": 2.8038258711745927,
+      "learning_rate": 0.00018969071103835207,
+      "loss": 1.3172,
+      "step": 3722
+    },
+    {
+      "epoch": 0.17151149398811444,
+      "grad_norm": 3.637232921271436,
+      "learning_rate": 0.00018968411174060773,
+      "loss": 1.6615,
+      "step": 3723
+    },
+    {
+      "epoch": 0.17155756207674944,
+      "grad_norm": 2.1918802668730826,
+      "learning_rate": 0.00018967751044620038,
+      "loss": 1.7103,
+      "step": 3724
+    },
+    {
+      "epoch": 0.17160363016538444,
+      "grad_norm": 5.492703949095959,
+      "learning_rate": 0.00018967090715527708,
+      "loss": 1.6679,
+      "step": 3725
+    },
+    {
+      "epoch": 0.17164969825401943,
+      "grad_norm": 1.7598512620295998,
+      "learning_rate": 0.00018966430186798472,
+      "loss": 1.4678,
+      "step": 3726
+    },
+    {
+      "epoch": 0.17169576634265443,
+      "grad_norm": 3.5977089110591653,
+      "learning_rate": 0.00018965769458447046,
+      "loss": 1.695,
+      "step": 3727
+    },
+    {
+      "epoch": 0.17174183443128946,
+      "grad_norm": 4.736200593525998,
+      "learning_rate": 0.00018965108530488137,
+      "loss": 1.5818,
+      "step": 3728
+    },
+    {
+      "epoch": 0.17178790251992446,
+      "grad_norm": 2.813371698827411,
+      "learning_rate": 0.00018964447402936455,
+      "loss": 1.5966,
+      "step": 3729
+    },
+    {
+      "epoch": 0.17183397060855946,
+      "grad_norm": 2.3682935923176625,
+      "learning_rate": 0.00018963786075806724,
+      "loss": 1.5194,
+      "step": 3730
+    },
+    {
+      "epoch": 0.17188003869719445,
+      "grad_norm": 2.4378750997005714,
+      "learning_rate": 0.00018963124549113664,
+      "loss": 1.5412,
+      "step": 3731
+    },
+    {
+      "epoch": 0.17192610678582945,
+      "grad_norm": 5.3396574413230224,
+      "learning_rate": 0.00018962462822872005,
+      "loss": 1.5514,
+      "step": 3732
+    },
+    {
+      "epoch": 0.17197217487446445,
+      "grad_norm": 2.3759282925223943,
+      "learning_rate": 0.0001896180089709648,
+      "loss": 1.3899,
+      "step": 3733
+    },
+    {
+      "epoch": 0.17201824296309945,
+      "grad_norm": 6.959363772358684,
+      "learning_rate": 0.00018961138771801824,
+      "loss": 1.9714,
+      "step": 3734
+    },
+    {
+      "epoch": 0.17206431105173448,
+      "grad_norm": 4.559740164361293,
+      "learning_rate": 0.00018960476447002776,
+      "loss": 1.4268,
+      "step": 3735
+    },
+    {
+      "epoch": 0.17211037914036947,
+      "grad_norm": 5.242616508043689,
+      "learning_rate": 0.00018959813922714088,
+      "loss": 1.6717,
+      "step": 3736
+    },
+    {
+      "epoch": 0.17215644722900447,
+      "grad_norm": 4.642172122032716,
+      "learning_rate": 0.000189591511989505,
+      "loss": 1.4829,
+      "step": 3737
+    },
+    {
+      "epoch": 0.17220251531763947,
+      "grad_norm": 2.085154754164407,
+      "learning_rate": 0.00018958488275726778,
+      "loss": 1.2877,
+      "step": 3738
+    },
+    {
+      "epoch": 0.17224858340627447,
+      "grad_norm": 3.141358315197958,
+      "learning_rate": 0.0001895782515305767,
+      "loss": 1.3929,
+      "step": 3739
+    },
+    {
+      "epoch": 0.17229465149490947,
+      "grad_norm": 4.422483191930276,
+      "learning_rate": 0.00018957161830957949,
+      "loss": 1.2249,
+      "step": 3740
+    },
+    {
+      "epoch": 0.17234071958354447,
+      "grad_norm": 2.457199469863745,
+      "learning_rate": 0.00018956498309442377,
+      "loss": 1.8353,
+      "step": 3741
+    },
+    {
+      "epoch": 0.1723867876721795,
+      "grad_norm": 3.4377887875127726,
+      "learning_rate": 0.00018955834588525725,
+      "loss": 1.4432,
+      "step": 3742
+    },
+    {
+      "epoch": 0.1724328557608145,
+      "grad_norm": 4.029490600752828,
+      "learning_rate": 0.00018955170668222775,
+      "loss": 1.4915,
+      "step": 3743
+    },
+    {
+      "epoch": 0.1724789238494495,
+      "grad_norm": 9.2359670065704,
+      "learning_rate": 0.00018954506548548303,
+      "loss": 1.5585,
+      "step": 3744
+    },
+    {
+      "epoch": 0.1725249919380845,
+      "grad_norm": 2.4892328406634148,
+      "learning_rate": 0.00018953842229517095,
+      "loss": 1.2579,
+      "step": 3745
+    },
+    {
+      "epoch": 0.17257106002671949,
+      "grad_norm": 5.707611770040545,
+      "learning_rate": 0.00018953177711143943,
+      "loss": 1.1442,
+      "step": 3746
+    },
+    {
+      "epoch": 0.17261712811535448,
+      "grad_norm": 3.555374298426179,
+      "learning_rate": 0.0001895251299344364,
+      "loss": 1.8075,
+      "step": 3747
+    },
+    {
+      "epoch": 0.17266319620398948,
+      "grad_norm": 1.636198891112745,
+      "learning_rate": 0.00018951848076430988,
+      "loss": 1.5034,
+      "step": 3748
+    },
+    {
+      "epoch": 0.1727092642926245,
+      "grad_norm": 2.553814002509398,
+      "learning_rate": 0.0001895118296012079,
+      "loss": 1.5016,
+      "step": 3749
+    },
+    {
+      "epoch": 0.1727553323812595,
+      "grad_norm": 4.1644285959516845,
+      "learning_rate": 0.00018950517644527844,
+      "loss": 1.5925,
+      "step": 3750
+    },
+    {
+      "epoch": 0.1728014004698945,
+      "grad_norm": 3.8549508990484136,
+      "learning_rate": 0.00018949852129666975,
+      "loss": 1.4583,
+      "step": 3751
+    },
+    {
+      "epoch": 0.1728474685585295,
+      "grad_norm": 3.323857553348553,
+      "learning_rate": 0.00018949186415552992,
+      "loss": 1.6286,
+      "step": 3752
+    },
+    {
+      "epoch": 0.1728935366471645,
+      "grad_norm": 2.3045382600381292,
+      "learning_rate": 0.0001894852050220072,
+      "loss": 1.7007,
+      "step": 3753
+    },
+    {
+      "epoch": 0.1729396047357995,
+      "grad_norm": 2.228106326236709,
+      "learning_rate": 0.0001894785438962498,
+      "loss": 1.4604,
+      "step": 3754
+    },
+    {
+      "epoch": 0.17298567282443453,
+      "grad_norm": 4.6822030761234315,
+      "learning_rate": 0.00018947188077840605,
+      "loss": 1.4395,
+      "step": 3755
+    },
+    {
+      "epoch": 0.17303174091306953,
+      "grad_norm": 2.629236431124643,
+      "learning_rate": 0.0001894652156686243,
+      "loss": 1.9469,
+      "step": 3756
+    },
+    {
+      "epoch": 0.17307780900170452,
+      "grad_norm": 6.9046193865334216,
+      "learning_rate": 0.0001894585485670529,
+      "loss": 1.4961,
+      "step": 3757
+    },
+    {
+      "epoch": 0.17312387709033952,
+      "grad_norm": 5.221374763054016,
+      "learning_rate": 0.00018945187947384034,
+      "loss": 1.4883,
+      "step": 3758
+    },
+    {
+      "epoch": 0.17316994517897452,
+      "grad_norm": 6.693823414865886,
+      "learning_rate": 0.00018944520838913502,
+      "loss": 1.4642,
+      "step": 3759
+    },
+    {
+      "epoch": 0.17321601326760952,
+      "grad_norm": 2.0221363825858174,
+      "learning_rate": 0.0001894385353130855,
+      "loss": 1.6313,
+      "step": 3760
+    },
+    {
+      "epoch": 0.17326208135624452,
+      "grad_norm": 2.5792448929918446,
+      "learning_rate": 0.00018943186024584038,
+      "loss": 1.3881,
+      "step": 3761
+    },
+    {
+      "epoch": 0.17330814944487954,
+      "grad_norm": 4.67767707823469,
+      "learning_rate": 0.0001894251831875482,
+      "loss": 1.6225,
+      "step": 3762
+    },
+    {
+      "epoch": 0.17335421753351454,
+      "grad_norm": 4.3982299968463,
+      "learning_rate": 0.0001894185041383577,
+      "loss": 1.5082,
+      "step": 3763
+    },
+    {
+      "epoch": 0.17340028562214954,
+      "grad_norm": 2.3662395382377683,
+      "learning_rate": 0.00018941182309841747,
+      "loss": 1.6146,
+      "step": 3764
+    },
+    {
+      "epoch": 0.17344635371078454,
+      "grad_norm": 2.2995527138356544,
+      "learning_rate": 0.00018940514006787632,
+      "loss": 1.9485,
+      "step": 3765
+    },
+    {
+      "epoch": 0.17349242179941954,
+      "grad_norm": 3.3770339023620273,
+      "learning_rate": 0.00018939845504688303,
+      "loss": 1.4083,
+      "step": 3766
+    },
+    {
+      "epoch": 0.17353848988805454,
+      "grad_norm": 4.288546078157293,
+      "learning_rate": 0.00018939176803558643,
+      "loss": 1.1594,
+      "step": 3767
+    },
+    {
+      "epoch": 0.17358455797668954,
+      "grad_norm": 4.230791758057504,
+      "learning_rate": 0.0001893850790341354,
+      "loss": 1.5541,
+      "step": 3768
+    },
+    {
+      "epoch": 0.17363062606532456,
+      "grad_norm": 5.868017465545887,
+      "learning_rate": 0.0001893783880426788,
+      "loss": 1.2734,
+      "step": 3769
+    },
+    {
+      "epoch": 0.17367669415395956,
+      "grad_norm": 5.134143953287588,
+      "learning_rate": 0.00018937169506136567,
+      "loss": 1.7783,
+      "step": 3770
+    },
+    {
+      "epoch": 0.17372276224259456,
+      "grad_norm": 2.793860782669875,
+      "learning_rate": 0.000189365000090345,
+      "loss": 1.9226,
+      "step": 3771
+    },
+    {
+      "epoch": 0.17376883033122956,
+      "grad_norm": 2.8072809507794005,
+      "learning_rate": 0.00018935830312976585,
+      "loss": 1.4548,
+      "step": 3772
+    },
+    {
+      "epoch": 0.17381489841986456,
+      "grad_norm": 1.9445234448289934,
+      "learning_rate": 0.00018935160417977724,
+      "loss": 1.768,
+      "step": 3773
+    },
+    {
+      "epoch": 0.17386096650849955,
+      "grad_norm": 22.772188595939177,
+      "learning_rate": 0.00018934490324052841,
+      "loss": 1.4884,
+      "step": 3774
+    },
+    {
+      "epoch": 0.17390703459713455,
+      "grad_norm": 2.4999448383167695,
+      "learning_rate": 0.00018933820031216849,
+      "loss": 1.5485,
+      "step": 3775
+    },
+    {
+      "epoch": 0.17395310268576958,
+      "grad_norm": 2.843177584923673,
+      "learning_rate": 0.00018933149539484672,
+      "loss": 1.8162,
+      "step": 3776
+    },
+    {
+      "epoch": 0.17399917077440458,
+      "grad_norm": 2.516599668643897,
+      "learning_rate": 0.00018932478848871238,
+      "loss": 1.6548,
+      "step": 3777
+    },
+    {
+      "epoch": 0.17404523886303958,
+      "grad_norm": 3.2304145141342415,
+      "learning_rate": 0.0001893180795939148,
+      "loss": 1.5873,
+      "step": 3778
+    },
+    {
+      "epoch": 0.17409130695167457,
+      "grad_norm": 5.478311253859563,
+      "learning_rate": 0.00018931136871060329,
+      "loss": 1.2681,
+      "step": 3779
+    },
+    {
+      "epoch": 0.17413737504030957,
+      "grad_norm": 2.1213940920799486,
+      "learning_rate": 0.0001893046558389273,
+      "loss": 1.5266,
+      "step": 3780
+    },
+    {
+      "epoch": 0.17418344312894457,
+      "grad_norm": 12.005184944458023,
+      "learning_rate": 0.0001892979409790363,
+      "loss": 1.4613,
+      "step": 3781
+    },
+    {
+      "epoch": 0.17422951121757957,
+      "grad_norm": 2.8506886975969423,
+      "learning_rate": 0.00018929122413107973,
+      "loss": 1.6601,
+      "step": 3782
+    },
+    {
+      "epoch": 0.1742755793062146,
+      "grad_norm": 2.2030163617385328,
+      "learning_rate": 0.00018928450529520716,
+      "loss": 1.6352,
+      "step": 3783
+    },
+    {
+      "epoch": 0.1743216473948496,
+      "grad_norm": 2.9303133340937793,
+      "learning_rate": 0.00018927778447156817,
+      "loss": 1.7934,
+      "step": 3784
+    },
+    {
+      "epoch": 0.1743677154834846,
+      "grad_norm": 5.141179366363891,
+      "learning_rate": 0.0001892710616603124,
+      "loss": 1.316,
+      "step": 3785
+    },
+    {
+      "epoch": 0.1744137835721196,
+      "grad_norm": 8.261849683449725,
+      "learning_rate": 0.0001892643368615895,
+      "loss": 1.6783,
+      "step": 3786
+    },
+    {
+      "epoch": 0.1744598516607546,
+      "grad_norm": 4.992301806678925,
+      "learning_rate": 0.0001892576100755492,
+      "loss": 1.4917,
+      "step": 3787
+    },
+    {
+      "epoch": 0.1745059197493896,
+      "grad_norm": 5.90380022673505,
+      "learning_rate": 0.00018925088130234126,
+      "loss": 1.6548,
+      "step": 3788
+    },
+    {
+      "epoch": 0.1745519878380246,
+      "grad_norm": 17.99815152712215,
+      "learning_rate": 0.00018924415054211546,
+      "loss": 1.3698,
+      "step": 3789
+    },
+    {
+      "epoch": 0.1745980559266596,
+      "grad_norm": 3.4612527714458254,
+      "learning_rate": 0.0001892374177950217,
+      "loss": 1.6938,
+      "step": 3790
+    },
+    {
+      "epoch": 0.1746441240152946,
+      "grad_norm": 6.654208186649339,
+      "learning_rate": 0.0001892306830612098,
+      "loss": 1.5591,
+      "step": 3791
+    },
+    {
+      "epoch": 0.1746901921039296,
+      "grad_norm": 3.180971295556422,
+      "learning_rate": 0.00018922394634082978,
+      "loss": 1.5307,
+      "step": 3792
+    },
+    {
+      "epoch": 0.1747362601925646,
+      "grad_norm": 4.054283778946922,
+      "learning_rate": 0.00018921720763403156,
+      "loss": 1.6556,
+      "step": 3793
+    },
+    {
+      "epoch": 0.1747823282811996,
+      "grad_norm": 17.87128866920148,
+      "learning_rate": 0.0001892104669409652,
+      "loss": 1.3158,
+      "step": 3794
+    },
+    {
+      "epoch": 0.1748283963698346,
+      "grad_norm": 3.8418110990344427,
+      "learning_rate": 0.00018920372426178074,
+      "loss": 1.7754,
+      "step": 3795
+    },
+    {
+      "epoch": 0.17487446445846963,
+      "grad_norm": 5.809603136562946,
+      "learning_rate": 0.00018919697959662832,
+      "loss": 1.3744,
+      "step": 3796
+    },
+    {
+      "epoch": 0.17492053254710463,
+      "grad_norm": 3.2817638800113564,
+      "learning_rate": 0.00018919023294565807,
+      "loss": 1.5458,
+      "step": 3797
+    },
+    {
+      "epoch": 0.17496660063573963,
+      "grad_norm": 30.80201749709619,
+      "learning_rate": 0.00018918348430902024,
+      "loss": 1.6213,
+      "step": 3798
+    },
+    {
+      "epoch": 0.17501266872437463,
+      "grad_norm": 6.778752082909534,
+      "learning_rate": 0.00018917673368686505,
+      "loss": 1.4948,
+      "step": 3799
+    },
+    {
+      "epoch": 0.17505873681300962,
+      "grad_norm": 11.604667114826844,
+      "learning_rate": 0.00018916998107934278,
+      "loss": 1.7616,
+      "step": 3800
+    },
+    {
+      "epoch": 0.17510480490164462,
+      "grad_norm": 37.00739935899308,
+      "learning_rate": 0.00018916322648660375,
+      "loss": 1.8709,
+      "step": 3801
+    },
+    {
+      "epoch": 0.17515087299027962,
+      "grad_norm": 7.943820012310054,
+      "learning_rate": 0.00018915646990879843,
+      "loss": 1.9778,
+      "step": 3802
+    },
+    {
+      "epoch": 0.17519694107891465,
+      "grad_norm": 14.08200048872388,
+      "learning_rate": 0.00018914971134607712,
+      "loss": 1.3467,
+      "step": 3803
+    },
+    {
+      "epoch": 0.17524300916754965,
+      "grad_norm": 2.6167907781849835,
+      "learning_rate": 0.00018914295079859036,
+      "loss": 1.6593,
+      "step": 3804
+    },
+    {
+      "epoch": 0.17528907725618464,
+      "grad_norm": 35.12083973509924,
+      "learning_rate": 0.00018913618826648865,
+      "loss": 1.7033,
+      "step": 3805
+    },
+    {
+      "epoch": 0.17533514534481964,
+      "grad_norm": 6.611994375313864,
+      "learning_rate": 0.00018912942374992256,
+      "loss": 1.3937,
+      "step": 3806
+    },
+    {
+      "epoch": 0.17538121343345464,
+      "grad_norm": 7.248515547125241,
+      "learning_rate": 0.00018912265724904264,
+      "loss": 1.4958,
+      "step": 3807
+    },
+    {
+      "epoch": 0.17542728152208964,
+      "grad_norm": 17.048378159802855,
+      "learning_rate": 0.0001891158887639996,
+      "loss": 1.6021,
+      "step": 3808
+    },
+    {
+      "epoch": 0.17547334961072464,
+      "grad_norm": 4.118656370341137,
+      "learning_rate": 0.0001891091182949441,
+      "loss": 1.6558,
+      "step": 3809
+    },
+    {
+      "epoch": 0.17551941769935966,
+      "grad_norm": 3.777019179907758,
+      "learning_rate": 0.00018910234584202683,
+      "loss": 1.5628,
+      "step": 3810
+    },
+    {
+      "epoch": 0.17556548578799466,
+      "grad_norm": 4.564338745043047,
+      "learning_rate": 0.00018909557140539863,
+      "loss": 1.6421,
+      "step": 3811
+    },
+    {
+      "epoch": 0.17561155387662966,
+      "grad_norm": 3.3603932960768996,
+      "learning_rate": 0.0001890887949852103,
+      "loss": 1.8282,
+      "step": 3812
+    },
+    {
+      "epoch": 0.17565762196526466,
+      "grad_norm": 3.886977087013116,
+      "learning_rate": 0.00018908201658161274,
+      "loss": 1.4951,
+      "step": 3813
+    },
+    {
+      "epoch": 0.17570369005389966,
+      "grad_norm": 3.1752353053256432,
+      "learning_rate": 0.00018907523619475676,
+      "loss": 1.6563,
+      "step": 3814
+    },
+    {
+      "epoch": 0.17574975814253466,
+      "grad_norm": 2.867997420834869,
+      "learning_rate": 0.00018906845382479342,
+      "loss": 1.7773,
+      "step": 3815
+    },
+    {
+      "epoch": 0.17579582623116966,
+      "grad_norm": 5.325998895607667,
+      "learning_rate": 0.00018906166947187366,
+      "loss": 1.8767,
+      "step": 3816
+    },
+    {
+      "epoch": 0.17584189431980468,
+      "grad_norm": 4.0020708470428366,
+      "learning_rate": 0.00018905488313614854,
+      "loss": 1.4889,
+      "step": 3817
+    },
+    {
+      "epoch": 0.17588796240843968,
+      "grad_norm": 4.575121877172815,
+      "learning_rate": 0.00018904809481776912,
+      "loss": 1.4537,
+      "step": 3818
+    },
+    {
+      "epoch": 0.17593403049707468,
+      "grad_norm": 2.664863279205828,
+      "learning_rate": 0.0001890413045168866,
+      "loss": 1.6125,
+      "step": 3819
+    },
+    {
+      "epoch": 0.17598009858570968,
+      "grad_norm": 4.2999102849764474,
+      "learning_rate": 0.00018903451223365207,
+      "loss": 1.6734,
+      "step": 3820
+    },
+    {
+      "epoch": 0.17602616667434468,
+      "grad_norm": 5.167006746822973,
+      "learning_rate": 0.0001890277179682168,
+      "loss": 1.4556,
+      "step": 3821
+    },
+    {
+      "epoch": 0.17607223476297967,
+      "grad_norm": 3.311248990185076,
+      "learning_rate": 0.00018902092172073208,
+      "loss": 1.6149,
+      "step": 3822
+    },
+    {
+      "epoch": 0.1761183028516147,
+      "grad_norm": 3.018084675831601,
+      "learning_rate": 0.0001890141234913491,
+      "loss": 1.7307,
+      "step": 3823
+    },
+    {
+      "epoch": 0.1761643709402497,
+      "grad_norm": 3.1535512352544353,
+      "learning_rate": 0.00018900732328021934,
+      "loss": 1.5856,
+      "step": 3824
+    },
+    {
+      "epoch": 0.1762104390288847,
+      "grad_norm": 6.612398738653752,
+      "learning_rate": 0.0001890005210874941,
+      "loss": 1.5184,
+      "step": 3825
+    },
+    {
+      "epoch": 0.1762565071175197,
+      "grad_norm": 2.82757771444697,
+      "learning_rate": 0.0001889937169133249,
+      "loss": 1.3224,
+      "step": 3826
+    },
+    {
+      "epoch": 0.1763025752061547,
+      "grad_norm": 3.1900700919801643,
+      "learning_rate": 0.00018898691075786318,
+      "loss": 1.7234,
+      "step": 3827
+    },
+    {
+      "epoch": 0.1763486432947897,
+      "grad_norm": 4.0997160705994995,
+      "learning_rate": 0.00018898010262126044,
+      "loss": 1.5391,
+      "step": 3828
+    },
+    {
+      "epoch": 0.1763947113834247,
+      "grad_norm": 3.655067193853615,
+      "learning_rate": 0.00018897329250366834,
+      "loss": 1.6697,
+      "step": 3829
+    },
+    {
+      "epoch": 0.17644077947205972,
+      "grad_norm": 4.356455976754294,
+      "learning_rate": 0.0001889664804052384,
+      "loss": 1.6675,
+      "step": 3830
+    },
+    {
+      "epoch": 0.17648684756069472,
+      "grad_norm": 3.828919732583753,
+      "learning_rate": 0.0001889596663261223,
+      "loss": 1.5063,
+      "step": 3831
+    },
+    {
+      "epoch": 0.1765329156493297,
+      "grad_norm": 2.134418679654642,
+      "learning_rate": 0.0001889528502664718,
+      "loss": 1.4885,
+      "step": 3832
+    },
+    {
+      "epoch": 0.1765789837379647,
+      "grad_norm": 4.326906214913746,
+      "learning_rate": 0.00018894603222643862,
+      "loss": 1.4542,
+      "step": 3833
+    },
+    {
+      "epoch": 0.1766250518265997,
+      "grad_norm": 3.999911872682243,
+      "learning_rate": 0.0001889392122061745,
+      "loss": 1.4281,
+      "step": 3834
+    },
+    {
+      "epoch": 0.1766711199152347,
+      "grad_norm": 4.54568843163096,
+      "learning_rate": 0.00018893239020583132,
+      "loss": 1.6103,
+      "step": 3835
+    },
+    {
+      "epoch": 0.1767171880038697,
+      "grad_norm": 2.904519591458787,
+      "learning_rate": 0.00018892556622556098,
+      "loss": 1.5873,
+      "step": 3836
+    },
+    {
+      "epoch": 0.17676325609250473,
+      "grad_norm": 3.6543746526766308,
+      "learning_rate": 0.0001889187402655154,
+      "loss": 1.5237,
+      "step": 3837
+    },
+    {
+      "epoch": 0.17680932418113973,
+      "grad_norm": 2.25767974365507,
+      "learning_rate": 0.00018891191232584648,
+      "loss": 1.5293,
+      "step": 3838
+    },
+    {
+      "epoch": 0.17685539226977473,
+      "grad_norm": 2.8244579152607976,
+      "learning_rate": 0.00018890508240670633,
+      "loss": 1.4452,
+      "step": 3839
+    },
+    {
+      "epoch": 0.17690146035840973,
+      "grad_norm": 3.0936806730575634,
+      "learning_rate": 0.00018889825050824694,
+      "loss": 1.4396,
+      "step": 3840
+    },
+    {
+      "epoch": 0.17694752844704473,
+      "grad_norm": 2.9224993946830673,
+      "learning_rate": 0.00018889141663062042,
+      "loss": 1.7436,
+      "step": 3841
+    },
+    {
+      "epoch": 0.17699359653567973,
+      "grad_norm": 2.142836834875738,
+      "learning_rate": 0.00018888458077397896,
+      "loss": 1.6497,
+      "step": 3842
+    },
+    {
+      "epoch": 0.17703966462431472,
+      "grad_norm": 2.3980688687861598,
+      "learning_rate": 0.00018887774293847466,
+      "loss": 1.2161,
+      "step": 3843
+    },
+    {
+      "epoch": 0.17708573271294975,
+      "grad_norm": 4.6114154860797685,
+      "learning_rate": 0.00018887090312425986,
+      "loss": 1.5155,
+      "step": 3844
+    },
+    {
+      "epoch": 0.17713180080158475,
+      "grad_norm": 1.5349536453744814,
+      "learning_rate": 0.00018886406133148675,
+      "loss": 1.4677,
+      "step": 3845
+    },
+    {
+      "epoch": 0.17717786889021975,
+      "grad_norm": 2.20842594294518,
+      "learning_rate": 0.0001888572175603077,
+      "loss": 1.4477,
+      "step": 3846
+    },
+    {
+      "epoch": 0.17722393697885475,
+      "grad_norm": 1.6521918992720759,
+      "learning_rate": 0.00018885037181087504,
+      "loss": 1.4645,
+      "step": 3847
+    },
+    {
+      "epoch": 0.17727000506748974,
+      "grad_norm": 3.410321357702148,
+      "learning_rate": 0.00018884352408334124,
+      "loss": 1.2944,
+      "step": 3848
+    },
+    {
+      "epoch": 0.17731607315612474,
+      "grad_norm": 2.890762137945975,
+      "learning_rate": 0.00018883667437785866,
+      "loss": 1.4311,
+      "step": 3849
+    },
+    {
+      "epoch": 0.17736214124475974,
+      "grad_norm": 3.001712758571395,
+      "learning_rate": 0.00018882982269457984,
+      "loss": 1.1831,
+      "step": 3850
+    },
+    {
+      "epoch": 0.17740820933339477,
+      "grad_norm": 2.5565978814830776,
+      "learning_rate": 0.00018882296903365738,
+      "loss": 1.6987,
+      "step": 3851
+    },
+    {
+      "epoch": 0.17745427742202977,
+      "grad_norm": 7.188195779817521,
+      "learning_rate": 0.00018881611339524379,
+      "loss": 1.3579,
+      "step": 3852
+    },
+    {
+      "epoch": 0.17750034551066476,
+      "grad_norm": 4.685454997784705,
+      "learning_rate": 0.00018880925577949173,
+      "loss": 1.3977,
+      "step": 3853
+    },
+    {
+      "epoch": 0.17754641359929976,
+      "grad_norm": 2.7904173030843813,
+      "learning_rate": 0.00018880239618655385,
+      "loss": 1.3183,
+      "step": 3854
+    },
+    {
+      "epoch": 0.17759248168793476,
+      "grad_norm": 4.223842098807439,
+      "learning_rate": 0.0001887955346165829,
+      "loss": 1.406,
+      "step": 3855
+    },
+    {
+      "epoch": 0.17763854977656976,
+      "grad_norm": 2.0942163660799964,
+      "learning_rate": 0.00018878867106973159,
+      "loss": 1.7462,
+      "step": 3856
+    },
+    {
+      "epoch": 0.17768461786520479,
+      "grad_norm": 3.847691617643813,
+      "learning_rate": 0.00018878180554615277,
+      "loss": 1.3969,
+      "step": 3857
+    },
+    {
+      "epoch": 0.17773068595383978,
+      "grad_norm": 2.344172070297375,
+      "learning_rate": 0.00018877493804599928,
+      "loss": 1.4623,
+      "step": 3858
+    },
+    {
+      "epoch": 0.17777675404247478,
+      "grad_norm": 1.9522390864513899,
+      "learning_rate": 0.00018876806856942403,
+      "loss": 1.3156,
+      "step": 3859
+    },
+    {
+      "epoch": 0.17782282213110978,
+      "grad_norm": 1.9472194306802069,
+      "learning_rate": 0.00018876119711657994,
+      "loss": 1.5674,
+      "step": 3860
+    },
+    {
+      "epoch": 0.17786889021974478,
+      "grad_norm": 2.6469268520306555,
+      "learning_rate": 0.00018875432368761996,
+      "loss": 1.2504,
+      "step": 3861
+    },
+    {
+      "epoch": 0.17791495830837978,
+      "grad_norm": 2.1049837340544397,
+      "learning_rate": 0.00018874744828269716,
+      "loss": 1.5758,
+      "step": 3862
+    },
+    {
+      "epoch": 0.17796102639701478,
+      "grad_norm": 3.1220611138679173,
+      "learning_rate": 0.0001887405709019646,
+      "loss": 1.1531,
+      "step": 3863
+    },
+    {
+      "epoch": 0.1780070944856498,
+      "grad_norm": 2.1597608248472677,
+      "learning_rate": 0.0001887336915455754,
+      "loss": 1.2568,
+      "step": 3864
+    },
+    {
+      "epoch": 0.1780531625742848,
+      "grad_norm": 2.903968096444386,
+      "learning_rate": 0.00018872681021368267,
+      "loss": 1.6229,
+      "step": 3865
+    },
+    {
+      "epoch": 0.1780992306629198,
+      "grad_norm": 4.397994892078951,
+      "learning_rate": 0.00018871992690643966,
+      "loss": 1.2886,
+      "step": 3866
+    },
+    {
+      "epoch": 0.1781452987515548,
+      "grad_norm": 2.172958866675456,
+      "learning_rate": 0.0001887130416239996,
+      "loss": 1.4644,
+      "step": 3867
+    },
+    {
+      "epoch": 0.1781913668401898,
+      "grad_norm": 1.4416252609767937,
+      "learning_rate": 0.00018870615436651576,
+      "loss": 1.712,
+      "step": 3868
+    },
+    {
+      "epoch": 0.1782374349288248,
+      "grad_norm": 2.4004049524429023,
+      "learning_rate": 0.0001886992651341415,
+      "loss": 1.6585,
+      "step": 3869
+    },
+    {
+      "epoch": 0.1782835030174598,
+      "grad_norm": 4.114785067350773,
+      "learning_rate": 0.0001886923739270302,
+      "loss": 1.4152,
+      "step": 3870
+    },
+    {
+      "epoch": 0.17832957110609482,
+      "grad_norm": 2.1797235585127455,
+      "learning_rate": 0.00018868548074533526,
+      "loss": 1.5087,
+      "step": 3871
+    },
+    {
+      "epoch": 0.17837563919472982,
+      "grad_norm": 2.622074273450739,
+      "learning_rate": 0.00018867858558921016,
+      "loss": 1.4535,
+      "step": 3872
+    },
+    {
+      "epoch": 0.17842170728336482,
+      "grad_norm": 1.904814293102743,
+      "learning_rate": 0.0001886716884588084,
+      "loss": 1.4357,
+      "step": 3873
+    },
+    {
+      "epoch": 0.17846777537199982,
+      "grad_norm": 1.7344033055736077,
+      "learning_rate": 0.00018866478935428352,
+      "loss": 1.4187,
+      "step": 3874
+    },
+    {
+      "epoch": 0.1785138434606348,
+      "grad_norm": 2.2689266297905455,
+      "learning_rate": 0.00018865788827578917,
+      "loss": 1.5432,
+      "step": 3875
+    },
+    {
+      "epoch": 0.1785599115492698,
+      "grad_norm": 4.304465822744723,
+      "learning_rate": 0.00018865098522347892,
+      "loss": 1.2021,
+      "step": 3876
+    },
+    {
+      "epoch": 0.1786059796379048,
+      "grad_norm": 3.43199275775582,
+      "learning_rate": 0.0001886440801975065,
+      "loss": 1.3997,
+      "step": 3877
+    },
+    {
+      "epoch": 0.17865204772653984,
+      "grad_norm": 2.1850068938414338,
+      "learning_rate": 0.00018863717319802565,
+      "loss": 1.2024,
+      "step": 3878
+    },
+    {
+      "epoch": 0.17869811581517484,
+      "grad_norm": 1.7849229757496141,
+      "learning_rate": 0.00018863026422519008,
+      "loss": 1.7481,
+      "step": 3879
+    },
+    {
+      "epoch": 0.17874418390380983,
+      "grad_norm": 2.434808228818981,
+      "learning_rate": 0.00018862335327915367,
+      "loss": 1.2689,
+      "step": 3880
+    },
+    {
+      "epoch": 0.17879025199244483,
+      "grad_norm": 2.1614804503528746,
+      "learning_rate": 0.00018861644036007025,
+      "loss": 1.4177,
+      "step": 3881
+    },
+    {
+      "epoch": 0.17883632008107983,
+      "grad_norm": 3.67194725271085,
+      "learning_rate": 0.00018860952546809372,
+      "loss": 1.538,
+      "step": 3882
+    },
+    {
+      "epoch": 0.17888238816971483,
+      "grad_norm": 3.2138990697969647,
+      "learning_rate": 0.00018860260860337804,
+      "loss": 1.6117,
+      "step": 3883
+    },
+    {
+      "epoch": 0.17892845625834983,
+      "grad_norm": 2.309400364884018,
+      "learning_rate": 0.00018859568976607722,
+      "loss": 1.3155,
+      "step": 3884
+    },
+    {
+      "epoch": 0.17897452434698485,
+      "grad_norm": 2.3085501871894736,
+      "learning_rate": 0.00018858876895634524,
+      "loss": 1.291,
+      "step": 3885
+    },
+    {
+      "epoch": 0.17902059243561985,
+      "grad_norm": 2.2905752245074376,
+      "learning_rate": 0.00018858184617433625,
+      "loss": 1.507,
+      "step": 3886
+    },
+    {
+      "epoch": 0.17906666052425485,
+      "grad_norm": 2.880865658296318,
+      "learning_rate": 0.00018857492142020435,
+      "loss": 1.3753,
+      "step": 3887
+    },
+    {
+      "epoch": 0.17911272861288985,
+      "grad_norm": 2.2256317909784293,
+      "learning_rate": 0.00018856799469410364,
+      "loss": 1.2248,
+      "step": 3888
+    },
+    {
+      "epoch": 0.17915879670152485,
+      "grad_norm": 1.1070300442797625,
+      "learning_rate": 0.00018856106599618845,
+      "loss": 1.4702,
+      "step": 3889
+    },
+    {
+      "epoch": 0.17920486479015985,
+      "grad_norm": 1.5183921719128226,
+      "learning_rate": 0.00018855413532661294,
+      "loss": 1.4832,
+      "step": 3890
+    },
+    {
+      "epoch": 0.17925093287879487,
+      "grad_norm": 1.4487188014178083,
+      "learning_rate": 0.00018854720268553145,
+      "loss": 1.6174,
+      "step": 3891
+    },
+    {
+      "epoch": 0.17929700096742987,
+      "grad_norm": 1.5049630529715772,
+      "learning_rate": 0.00018854026807309831,
+      "loss": 1.3646,
+      "step": 3892
+    },
+    {
+      "epoch": 0.17934306905606487,
+      "grad_norm": 4.026903806243556,
+      "learning_rate": 0.00018853333148946793,
+      "loss": 1.9038,
+      "step": 3893
+    },
+    {
+      "epoch": 0.17938913714469987,
+      "grad_norm": 2.1475051631802478,
+      "learning_rate": 0.00018852639293479476,
+      "loss": 1.4858,
+      "step": 3894
+    },
+    {
+      "epoch": 0.17943520523333487,
+      "grad_norm": 1.8442460188826242,
+      "learning_rate": 0.0001885194524092332,
+      "loss": 1.6643,
+      "step": 3895
+    },
+    {
+      "epoch": 0.17948127332196986,
+      "grad_norm": 3.282827484651445,
+      "learning_rate": 0.0001885125099129378,
+      "loss": 1.374,
+      "step": 3896
+    },
+    {
+      "epoch": 0.17952734141060486,
+      "grad_norm": 4.8418679222452745,
+      "learning_rate": 0.00018850556544606315,
+      "loss": 1.2133,
+      "step": 3897
+    },
+    {
+      "epoch": 0.1795734094992399,
+      "grad_norm": 2.339675374983888,
+      "learning_rate": 0.00018849861900876387,
+      "loss": 1.4451,
+      "step": 3898
+    },
+    {
+      "epoch": 0.1796194775878749,
+      "grad_norm": 1.6808364245018277,
+      "learning_rate": 0.00018849167060119455,
+      "loss": 1.1963,
+      "step": 3899
+    },
+    {
+      "epoch": 0.17966554567650989,
+      "grad_norm": 6.4565328731586575,
+      "learning_rate": 0.00018848472022350992,
+      "loss": 1.5563,
+      "step": 3900
+    },
+    {
+      "epoch": 0.17971161376514488,
+      "grad_norm": 2.9331302458792976,
+      "learning_rate": 0.0001884777678758647,
+      "loss": 1.3728,
+      "step": 3901
+    },
+    {
+      "epoch": 0.17975768185377988,
+      "grad_norm": 2.1567283992337996,
+      "learning_rate": 0.00018847081355841366,
+      "loss": 1.668,
+      "step": 3902
+    },
+    {
+      "epoch": 0.17980374994241488,
+      "grad_norm": 3.219458219850426,
+      "learning_rate": 0.00018846385727131173,
+      "loss": 1.5782,
+      "step": 3903
+    },
+    {
+      "epoch": 0.17984981803104988,
+      "grad_norm": 2.0330426229012337,
+      "learning_rate": 0.00018845689901471365,
+      "loss": 1.3732,
+      "step": 3904
+    },
+    {
+      "epoch": 0.1798958861196849,
+      "grad_norm": 2.146806615327944,
+      "learning_rate": 0.00018844993878877438,
+      "loss": 1.8599,
+      "step": 3905
+    },
+    {
+      "epoch": 0.1799419542083199,
+      "grad_norm": 4.306651727452975,
+      "learning_rate": 0.0001884429765936489,
+      "loss": 1.5169,
+      "step": 3906
+    },
+    {
+      "epoch": 0.1799880222969549,
+      "grad_norm": 6.423215608605416,
+      "learning_rate": 0.00018843601242949216,
+      "loss": 1.7007,
+      "step": 3907
+    },
+    {
+      "epoch": 0.1800340903855899,
+      "grad_norm": 2.468478590064037,
+      "learning_rate": 0.00018842904629645928,
+      "loss": 1.4345,
+      "step": 3908
+    },
+    {
+      "epoch": 0.1800801584742249,
+      "grad_norm": 4.7134621256584985,
+      "learning_rate": 0.0001884220781947053,
+      "loss": 1.224,
+      "step": 3909
+    },
+    {
+      "epoch": 0.1801262265628599,
+      "grad_norm": 2.7049041292486575,
+      "learning_rate": 0.00018841510812438533,
+      "loss": 1.6017,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1801722946514949,
+      "grad_norm": 1.9588028975986496,
+      "learning_rate": 0.00018840813608565462,
+      "loss": 1.7764,
+      "step": 3911
+    },
+    {
+      "epoch": 0.18021836274012992,
+      "grad_norm": 2.8146298073554328,
+      "learning_rate": 0.00018840116207866832,
+      "loss": 1.5066,
+      "step": 3912
+    },
+    {
+      "epoch": 0.18026443082876492,
+      "grad_norm": 1.8585468818389936,
+      "learning_rate": 0.00018839418610358172,
+      "loss": 1.7039,
+      "step": 3913
+    },
+    {
+      "epoch": 0.18031049891739992,
+      "grad_norm": 1.4534271155510523,
+      "learning_rate": 0.00018838720816055014,
+      "loss": 1.8034,
+      "step": 3914
+    },
+    {
+      "epoch": 0.18035656700603492,
+      "grad_norm": 1.9705511206632549,
+      "learning_rate": 0.00018838022824972894,
+      "loss": 1.6098,
+      "step": 3915
+    },
+    {
+      "epoch": 0.18040263509466992,
+      "grad_norm": 1.5503882466051242,
+      "learning_rate": 0.00018837324637127346,
+      "loss": 1.5094,
+      "step": 3916
+    },
+    {
+      "epoch": 0.18044870318330491,
+      "grad_norm": 2.9274643879578406,
+      "learning_rate": 0.0001883662625253392,
+      "loss": 1.6033,
+      "step": 3917
+    },
+    {
+      "epoch": 0.18049477127193994,
+      "grad_norm": 1.8708087010473187,
+      "learning_rate": 0.0001883592767120816,
+      "loss": 1.2257,
+      "step": 3918
+    },
+    {
+      "epoch": 0.18054083936057494,
+      "grad_norm": 2.191700536442718,
+      "learning_rate": 0.00018835228893165624,
+      "loss": 1.3863,
+      "step": 3919
+    },
+    {
+      "epoch": 0.18058690744920994,
+      "grad_norm": 1.1758445242473,
+      "learning_rate": 0.00018834529918421863,
+      "loss": 1.5116,
+      "step": 3920
+    },
+    {
+      "epoch": 0.18063297553784494,
+      "grad_norm": 3.202841490744232,
+      "learning_rate": 0.00018833830746992444,
+      "loss": 1.4636,
+      "step": 3921
+    },
+    {
+      "epoch": 0.18067904362647993,
+      "grad_norm": 2.6291777262796154,
+      "learning_rate": 0.00018833131378892927,
+      "loss": 2.0225,
+      "step": 3922
+    },
+    {
+      "epoch": 0.18072511171511493,
+      "grad_norm": 2.756583522602746,
+      "learning_rate": 0.00018832431814138888,
+      "loss": 1.5222,
+      "step": 3923
+    },
+    {
+      "epoch": 0.18077117980374993,
+      "grad_norm": 1.601721840309813,
+      "learning_rate": 0.00018831732052745898,
+      "loss": 1.6047,
+      "step": 3924
+    },
+    {
+      "epoch": 0.18081724789238496,
+      "grad_norm": 2.615030552694145,
+      "learning_rate": 0.00018831032094729537,
+      "loss": 1.2766,
+      "step": 3925
+    },
+    {
+      "epoch": 0.18086331598101996,
+      "grad_norm": 1.1666338420524376,
+      "learning_rate": 0.0001883033194010539,
+      "loss": 1.7157,
+      "step": 3926
+    },
+    {
+      "epoch": 0.18090938406965495,
+      "grad_norm": 1.8394394237339236,
+      "learning_rate": 0.00018829631588889042,
+      "loss": 1.3825,
+      "step": 3927
+    },
+    {
+      "epoch": 0.18095545215828995,
+      "grad_norm": 2.3024732101897287,
+      "learning_rate": 0.00018828931041096086,
+      "loss": 1.4995,
+      "step": 3928
+    },
+    {
+      "epoch": 0.18100152024692495,
+      "grad_norm": 2.64605998571444,
+      "learning_rate": 0.00018828230296742118,
+      "loss": 1.2305,
+      "step": 3929
+    },
+    {
+      "epoch": 0.18104758833555995,
+      "grad_norm": 1.9898268223529247,
+      "learning_rate": 0.0001882752935584274,
+      "loss": 1.3799,
+      "step": 3930
+    },
+    {
+      "epoch": 0.18109365642419495,
+      "grad_norm": 5.940294153280154,
+      "learning_rate": 0.00018826828218413558,
+      "loss": 1.4699,
+      "step": 3931
+    },
+    {
+      "epoch": 0.18113972451282997,
+      "grad_norm": 1.6089721716047636,
+      "learning_rate": 0.0001882612688447018,
+      "loss": 1.8055,
+      "step": 3932
+    },
+    {
+      "epoch": 0.18118579260146497,
+      "grad_norm": 2.415781005580156,
+      "learning_rate": 0.0001882542535402822,
+      "loss": 1.0667,
+      "step": 3933
+    },
+    {
+      "epoch": 0.18123186069009997,
+      "grad_norm": 2.3678337347360663,
+      "learning_rate": 0.00018824723627103296,
+      "loss": 1.3314,
+      "step": 3934
+    },
+    {
+      "epoch": 0.18127792877873497,
+      "grad_norm": 2.287361725487971,
+      "learning_rate": 0.00018824021703711032,
+      "loss": 1.2722,
+      "step": 3935
+    },
+    {
+      "epoch": 0.18132399686736997,
+      "grad_norm": 2.3402446947131152,
+      "learning_rate": 0.00018823319583867056,
+      "loss": 1.1934,
+      "step": 3936
+    },
+    {
+      "epoch": 0.18137006495600497,
+      "grad_norm": 2.581767461336684,
+      "learning_rate": 0.00018822617267586996,
+      "loss": 1.1232,
+      "step": 3937
+    },
+    {
+      "epoch": 0.18141613304463997,
+      "grad_norm": 1.4607276976159393,
+      "learning_rate": 0.0001882191475488649,
+      "loss": 1.567,
+      "step": 3938
+    },
+    {
+      "epoch": 0.181462201133275,
+      "grad_norm": 2.5905446300376944,
+      "learning_rate": 0.00018821212045781182,
+      "loss": 1.491,
+      "step": 3939
+    },
+    {
+      "epoch": 0.18150826922191,
+      "grad_norm": 1.496418393905333,
+      "learning_rate": 0.00018820509140286712,
+      "loss": 1.2478,
+      "step": 3940
+    },
+    {
+      "epoch": 0.181554337310545,
+      "grad_norm": 1.4861928326122844,
+      "learning_rate": 0.00018819806038418728,
+      "loss": 1.5984,
+      "step": 3941
+    },
+    {
+      "epoch": 0.18160040539918,
+      "grad_norm": 2.096792597468499,
+      "learning_rate": 0.00018819102740192886,
+      "loss": 1.5273,
+      "step": 3942
+    },
+    {
+      "epoch": 0.18164647348781499,
+      "grad_norm": 4.495345240976964,
+      "learning_rate": 0.00018818399245624842,
+      "loss": 1.5946,
+      "step": 3943
+    },
+    {
+      "epoch": 0.18169254157644998,
+      "grad_norm": 3.094191152234101,
+      "learning_rate": 0.0001881769555473026,
+      "loss": 1.5975,
+      "step": 3944
+    },
+    {
+      "epoch": 0.18173860966508498,
+      "grad_norm": 1.1950982751910115,
+      "learning_rate": 0.00018816991667524807,
+      "loss": 1.4418,
+      "step": 3945
+    },
+    {
+      "epoch": 0.18178467775372,
+      "grad_norm": 1.3716764832790105,
+      "learning_rate": 0.00018816287584024152,
+      "loss": 1.5897,
+      "step": 3946
+    },
+    {
+      "epoch": 0.181830745842355,
+      "grad_norm": 1.01377512591959,
+      "learning_rate": 0.00018815583304243968,
+      "loss": 1.722,
+      "step": 3947
+    },
+    {
+      "epoch": 0.18187681393099,
+      "grad_norm": 1.4090501041842922,
+      "learning_rate": 0.0001881487882819994,
+      "loss": 1.4698,
+      "step": 3948
+    },
+    {
+      "epoch": 0.181922882019625,
+      "grad_norm": 1.589336238807379,
+      "learning_rate": 0.00018814174155907748,
+      "loss": 1.5252,
+      "step": 3949
+    },
+    {
+      "epoch": 0.18196895010826,
+      "grad_norm": 1.5156496465263936,
+      "learning_rate": 0.00018813469287383083,
+      "loss": 1.3132,
+      "step": 3950
+    },
+    {
+      "epoch": 0.182015018196895,
+      "grad_norm": 0.8982567409578451,
+      "learning_rate": 0.00018812764222641634,
+      "loss": 1.3219,
+      "step": 3951
+    },
+    {
+      "epoch": 0.18206108628553003,
+      "grad_norm": 1.3614620015292287,
+      "learning_rate": 0.000188120589616991,
+      "loss": 1.3754,
+      "step": 3952
+    },
+    {
+      "epoch": 0.18210715437416503,
+      "grad_norm": 3.882619671595325,
+      "learning_rate": 0.00018811353504571184,
+      "loss": 1.3062,
+      "step": 3953
+    },
+    {
+      "epoch": 0.18215322246280002,
+      "grad_norm": 1.3544100686561358,
+      "learning_rate": 0.0001881064785127359,
+      "loss": 1.7041,
+      "step": 3954
+    },
+    {
+      "epoch": 0.18219929055143502,
+      "grad_norm": 1.521236953604276,
+      "learning_rate": 0.00018809942001822032,
+      "loss": 1.5724,
+      "step": 3955
+    },
+    {
+      "epoch": 0.18224535864007002,
+      "grad_norm": 1.9590297612251628,
+      "learning_rate": 0.00018809235956232216,
+      "loss": 1.4649,
+      "step": 3956
+    },
+    {
+      "epoch": 0.18229142672870502,
+      "grad_norm": 1.2222464900777066,
+      "learning_rate": 0.0001880852971451987,
+      "loss": 1.6991,
+      "step": 3957
+    },
+    {
+      "epoch": 0.18233749481734002,
+      "grad_norm": 1.5221846801291208,
+      "learning_rate": 0.00018807823276700708,
+      "loss": 1.5906,
+      "step": 3958
+    },
+    {
+      "epoch": 0.18238356290597504,
+      "grad_norm": 1.1362355758070288,
+      "learning_rate": 0.00018807116642790468,
+      "loss": 1.6681,
+      "step": 3959
+    },
+    {
+      "epoch": 0.18242963099461004,
+      "grad_norm": 4.474374086890713,
+      "learning_rate": 0.00018806409812804876,
+      "loss": 1.5757,
+      "step": 3960
+    },
+    {
+      "epoch": 0.18247569908324504,
+      "grad_norm": 1.7669054231833374,
+      "learning_rate": 0.00018805702786759668,
+      "loss": 1.3557,
+      "step": 3961
+    },
+    {
+      "epoch": 0.18252176717188004,
+      "grad_norm": 1.8436168299421454,
+      "learning_rate": 0.00018804995564670587,
+      "loss": 1.4089,
+      "step": 3962
+    },
+    {
+      "epoch": 0.18256783526051504,
+      "grad_norm": 1.251869095841085,
+      "learning_rate": 0.00018804288146553376,
+      "loss": 1.3536,
+      "step": 3963
+    },
+    {
+      "epoch": 0.18261390334915004,
+      "grad_norm": 1.9150159423825275,
+      "learning_rate": 0.00018803580532423782,
+      "loss": 1.4002,
+      "step": 3964
+    },
+    {
+      "epoch": 0.18265997143778503,
+      "grad_norm": 1.7993226218199645,
+      "learning_rate": 0.00018802872722297568,
+      "loss": 1.6053,
+      "step": 3965
+    },
+    {
+      "epoch": 0.18270603952642006,
+      "grad_norm": 23.440766160577798,
+      "learning_rate": 0.00018802164716190487,
+      "loss": 1.2224,
+      "step": 3966
+    },
+    {
+      "epoch": 0.18275210761505506,
+      "grad_norm": 2.0692800009265566,
+      "learning_rate": 0.00018801456514118298,
+      "loss": 1.4704,
+      "step": 3967
+    },
+    {
+      "epoch": 0.18279817570369006,
+      "grad_norm": 8.20750053945833,
+      "learning_rate": 0.00018800748116096772,
+      "loss": 1.9504,
+      "step": 3968
+    },
+    {
+      "epoch": 0.18284424379232506,
+      "grad_norm": 1.5502759679853324,
+      "learning_rate": 0.00018800039522141678,
+      "loss": 1.5525,
+      "step": 3969
+    },
+    {
+      "epoch": 0.18289031188096005,
+      "grad_norm": 1.2583679770709741,
+      "learning_rate": 0.00018799330732268796,
+      "loss": 1.2175,
+      "step": 3970
+    },
+    {
+      "epoch": 0.18293637996959505,
+      "grad_norm": 1.017950875041328,
+      "learning_rate": 0.00018798621746493897,
+      "loss": 1.412,
+      "step": 3971
+    },
+    {
+      "epoch": 0.18298244805823005,
+      "grad_norm": 1.5927254083141102,
+      "learning_rate": 0.00018797912564832777,
+      "loss": 1.3597,
+      "step": 3972
+    },
+    {
+      "epoch": 0.18302851614686508,
+      "grad_norm": 1.6753233312413944,
+      "learning_rate": 0.0001879720318730122,
+      "loss": 1.6534,
+      "step": 3973
+    },
+    {
+      "epoch": 0.18307458423550008,
+      "grad_norm": 1.4631473441594924,
+      "learning_rate": 0.00018796493613915013,
+      "loss": 1.8916,
+      "step": 3974
+    },
+    {
+      "epoch": 0.18312065232413507,
+      "grad_norm": 1.7459974697440728,
+      "learning_rate": 0.0001879578384468996,
+      "loss": 2.0084,
+      "step": 3975
+    },
+    {
+      "epoch": 0.18316672041277007,
+      "grad_norm": 9.180189858582308,
+      "learning_rate": 0.00018795073879641862,
+      "loss": 1.9237,
+      "step": 3976
+    },
+    {
+      "epoch": 0.18321278850140507,
+      "grad_norm": 1.2577620182325937,
+      "learning_rate": 0.00018794363718786526,
+      "loss": 1.7341,
+      "step": 3977
+    },
+    {
+      "epoch": 0.18325885659004007,
+      "grad_norm": 1.3347147619477684,
+      "learning_rate": 0.00018793653362139757,
+      "loss": 1.4953,
+      "step": 3978
+    },
+    {
+      "epoch": 0.18330492467867507,
+      "grad_norm": 1.0442819809536434,
+      "learning_rate": 0.00018792942809717377,
+      "loss": 1.6534,
+      "step": 3979
+    },
+    {
+      "epoch": 0.1833509927673101,
+      "grad_norm": 0.8767024859886807,
+      "learning_rate": 0.000187922320615352,
+      "loss": 1.3286,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1833970608559451,
+      "grad_norm": 1.204605764171159,
+      "learning_rate": 0.0001879152111760905,
+      "loss": 1.2892,
+      "step": 3981
+    },
+    {
+      "epoch": 0.1834431289445801,
+      "grad_norm": 0.9114096960686865,
+      "learning_rate": 0.0001879080997795476,
+      "loss": 1.4944,
+      "step": 3982
+    },
+    {
+      "epoch": 0.1834891970332151,
+      "grad_norm": 1.1529435685298262,
+      "learning_rate": 0.0001879009864258816,
+      "loss": 1.2521,
+      "step": 3983
+    },
+    {
+      "epoch": 0.1835352651218501,
+      "grad_norm": 1.229817181317135,
+      "learning_rate": 0.00018789387111525083,
+      "loss": 1.3825,
+      "step": 3984
+    },
+    {
+      "epoch": 0.1835813332104851,
+      "grad_norm": 1.374130211637771,
+      "learning_rate": 0.00018788675384781375,
+      "loss": 1.396,
+      "step": 3985
+    },
+    {
+      "epoch": 0.1836274012991201,
+      "grad_norm": 1.7179198919078766,
+      "learning_rate": 0.00018787963462372874,
+      "loss": 1.3703,
+      "step": 3986
+    },
+    {
+      "epoch": 0.1836734693877551,
+      "grad_norm": 1.0715404508579047,
+      "learning_rate": 0.00018787251344315437,
+      "loss": 1.552,
+      "step": 3987
+    },
+    {
+      "epoch": 0.1837195374763901,
+      "grad_norm": 1.541651162200383,
+      "learning_rate": 0.00018786539030624916,
+      "loss": 1.6997,
+      "step": 3988
+    },
+    {
+      "epoch": 0.1837656055650251,
+      "grad_norm": 1.0215779282401867,
+      "learning_rate": 0.00018785826521317169,
+      "loss": 1.3787,
+      "step": 3989
+    },
+    {
+      "epoch": 0.1838116736536601,
+      "grad_norm": 1.2426764967222796,
+      "learning_rate": 0.0001878511381640806,
+      "loss": 1.3389,
+      "step": 3990
+    },
+    {
+      "epoch": 0.1838577417422951,
+      "grad_norm": 1.6487440590155602,
+      "learning_rate": 0.00018784400915913455,
+      "loss": 1.3229,
+      "step": 3991
+    },
+    {
+      "epoch": 0.1839038098309301,
+      "grad_norm": 1.2207380368536416,
+      "learning_rate": 0.00018783687819849223,
+      "loss": 1.4354,
+      "step": 3992
+    },
+    {
+      "epoch": 0.18394987791956513,
+      "grad_norm": 0.9431300888630361,
+      "learning_rate": 0.00018782974528231246,
+      "loss": 1.2823,
+      "step": 3993
+    },
+    {
+      "epoch": 0.18399594600820013,
+      "grad_norm": 3.114419818549334,
+      "learning_rate": 0.00018782261041075398,
+      "loss": 1.4894,
+      "step": 3994
+    },
+    {
+      "epoch": 0.18404201409683513,
+      "grad_norm": 10.39015829289029,
+      "learning_rate": 0.0001878154735839757,
+      "loss": 1.3145,
+      "step": 3995
+    },
+    {
+      "epoch": 0.18408808218547013,
+      "grad_norm": 0.8740967009768503,
+      "learning_rate": 0.00018780833480213646,
+      "loss": 1.3064,
+      "step": 3996
+    },
+    {
+      "epoch": 0.18413415027410512,
+      "grad_norm": 2.2161931614812875,
+      "learning_rate": 0.00018780119406539516,
+      "loss": 1.6302,
+      "step": 3997
+    },
+    {
+      "epoch": 0.18418021836274012,
+      "grad_norm": 1.5106623934247694,
+      "learning_rate": 0.00018779405137391085,
+      "loss": 0.9858,
+      "step": 3998
+    },
+    {
+      "epoch": 0.18422628645137512,
+      "grad_norm": 4.7168172130220585,
+      "learning_rate": 0.0001877869067278425,
+      "loss": 1.3495,
+      "step": 3999
+    },
+    {
+      "epoch": 0.18427235454001015,
+      "grad_norm": 1.8277999834369505,
+      "learning_rate": 0.00018777976012734925,
+      "loss": 1.1491,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18431842262864515,
+      "grad_norm": 1.570241998841316,
+      "learning_rate": 0.0001877726115725901,
+      "loss": 1.4053,
+      "step": 4001
+    },
+    {
+      "epoch": 0.18436449071728014,
+      "grad_norm": 4.631968767462027,
+      "learning_rate": 0.00018776546106372428,
+      "loss": 1.7102,
+      "step": 4002
+    },
+    {
+      "epoch": 0.18441055880591514,
+      "grad_norm": 2.1851207670069854,
+      "learning_rate": 0.00018775830860091094,
+      "loss": 1.6502,
+      "step": 4003
+    },
+    {
+      "epoch": 0.18445662689455014,
+      "grad_norm": 2.282392485459576,
+      "learning_rate": 0.00018775115418430933,
+      "loss": 1.6196,
+      "step": 4004
+    },
+    {
+      "epoch": 0.18450269498318514,
+      "grad_norm": 1.3894212896073523,
+      "learning_rate": 0.00018774399781407874,
+      "loss": 1.5292,
+      "step": 4005
+    },
+    {
+      "epoch": 0.18454876307182014,
+      "grad_norm": 7.42321980015007,
+      "learning_rate": 0.0001877368394903785,
+      "loss": 1.1513,
+      "step": 4006
+    },
+    {
+      "epoch": 0.18459483116045516,
+      "grad_norm": 1.9175199094477497,
+      "learning_rate": 0.00018772967921336795,
+      "loss": 1.6566,
+      "step": 4007
+    },
+    {
+      "epoch": 0.18464089924909016,
+      "grad_norm": 2.530295975717321,
+      "learning_rate": 0.00018772251698320652,
+      "loss": 1.4539,
+      "step": 4008
+    },
+    {
+      "epoch": 0.18468696733772516,
+      "grad_norm": 2.0755343060748017,
+      "learning_rate": 0.00018771535280005366,
+      "loss": 1.3981,
+      "step": 4009
+    },
+    {
+      "epoch": 0.18473303542636016,
+      "grad_norm": 1.5813298674379968,
+      "learning_rate": 0.00018770818666406886,
+      "loss": 1.4961,
+      "step": 4010
+    },
+    {
+      "epoch": 0.18477910351499516,
+      "grad_norm": 10.941272510601213,
+      "learning_rate": 0.00018770101857541166,
+      "loss": 1.2922,
+      "step": 4011
+    },
+    {
+      "epoch": 0.18482517160363016,
+      "grad_norm": 1.7998885436295242,
+      "learning_rate": 0.00018769384853424168,
+      "loss": 1.7398,
+      "step": 4012
+    },
+    {
+      "epoch": 0.18487123969226515,
+      "grad_norm": 6.589175494772908,
+      "learning_rate": 0.00018768667654071855,
+      "loss": 1.4173,
+      "step": 4013
+    },
+    {
+      "epoch": 0.18491730778090018,
+      "grad_norm": 1.0413361220751487,
+      "learning_rate": 0.00018767950259500186,
+      "loss": 1.7924,
+      "step": 4014
+    },
+    {
+      "epoch": 0.18496337586953518,
+      "grad_norm": 1.6945387839739934,
+      "learning_rate": 0.00018767232669725142,
+      "loss": 1.6019,
+      "step": 4015
+    },
+    {
+      "epoch": 0.18500944395817018,
+      "grad_norm": 4.988972367817158,
+      "learning_rate": 0.00018766514884762693,
+      "loss": 1.5052,
+      "step": 4016
+    },
+    {
+      "epoch": 0.18505551204680518,
+      "grad_norm": 3.305636256909509,
+      "learning_rate": 0.00018765796904628826,
+      "loss": 1.5987,
+      "step": 4017
+    },
+    {
+      "epoch": 0.18510158013544017,
+      "grad_norm": 13.657663721791861,
+      "learning_rate": 0.00018765078729339517,
+      "loss": 1.5404,
+      "step": 4018
+    },
+    {
+      "epoch": 0.18514764822407517,
+      "grad_norm": 5.8683294514298,
+      "learning_rate": 0.0001876436035891076,
+      "loss": 1.8198,
+      "step": 4019
+    },
+    {
+      "epoch": 0.1851937163127102,
+      "grad_norm": 9.177286887330714,
+      "learning_rate": 0.00018763641793358545,
+      "loss": 1.4785,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1852397844013452,
+      "grad_norm": 3.8643814884600327,
+      "learning_rate": 0.00018762923032698874,
+      "loss": 1.3836,
+      "step": 4021
+    },
+    {
+      "epoch": 0.1852858524899802,
+      "grad_norm": 6.427625018254589,
+      "learning_rate": 0.00018762204076947748,
+      "loss": 1.8171,
+      "step": 4022
+    },
+    {
+      "epoch": 0.1853319205786152,
+      "grad_norm": 26.763358509460694,
+      "learning_rate": 0.0001876148492612117,
+      "loss": 1.2969,
+      "step": 4023
+    },
+    {
+      "epoch": 0.1853779886672502,
+      "grad_norm": 10.916501239678961,
+      "learning_rate": 0.00018760765580235153,
+      "loss": 1.4151,
+      "step": 4024
+    },
+    {
+      "epoch": 0.1854240567558852,
+      "grad_norm": 4.576206825904063,
+      "learning_rate": 0.00018760046039305715,
+      "loss": 1.2323,
+      "step": 4025
+    },
+    {
+      "epoch": 0.1854701248445202,
+      "grad_norm": 1.9350602206301728,
+      "learning_rate": 0.00018759326303348868,
+      "loss": 1.5249,
+      "step": 4026
+    },
+    {
+      "epoch": 0.18551619293315522,
+      "grad_norm": 1.575548196206712,
+      "learning_rate": 0.0001875860637238064,
+      "loss": 1.5844,
+      "step": 4027
+    },
+    {
+      "epoch": 0.18556226102179021,
+      "grad_norm": 1.950825072858649,
+      "learning_rate": 0.00018757886246417062,
+      "loss": 1.6756,
+      "step": 4028
+    },
+    {
+      "epoch": 0.1856083291104252,
+      "grad_norm": 2.5004822884155455,
+      "learning_rate": 0.00018757165925474163,
+      "loss": 1.3006,
+      "step": 4029
+    },
+    {
+      "epoch": 0.1856543971990602,
+      "grad_norm": 2.8295238718707867,
+      "learning_rate": 0.00018756445409567977,
+      "loss": 1.587,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1857004652876952,
+      "grad_norm": 5.255241012727881,
+      "learning_rate": 0.0001875572469871455,
+      "loss": 1.1582,
+      "step": 4031
+    },
+    {
+      "epoch": 0.1857465333763302,
+      "grad_norm": 3.506007281382299,
+      "learning_rate": 0.00018755003792929927,
+      "loss": 1.3486,
+      "step": 4032
+    },
+    {
+      "epoch": 0.1857926014649652,
+      "grad_norm": 10.074400267373102,
+      "learning_rate": 0.00018754282692230154,
+      "loss": 1.226,
+      "step": 4033
+    },
+    {
+      "epoch": 0.18583866955360023,
+      "grad_norm": 1.4667600102641092,
+      "learning_rate": 0.00018753561396631286,
+      "loss": 1.5497,
+      "step": 4034
+    },
+    {
+      "epoch": 0.18588473764223523,
+      "grad_norm": 1.6028522548931723,
+      "learning_rate": 0.00018752839906149388,
+      "loss": 1.3717,
+      "step": 4035
+    },
+    {
+      "epoch": 0.18593080573087023,
+      "grad_norm": 17.317322844624574,
+      "learning_rate": 0.00018752118220800512,
+      "loss": 1.4611,
+      "step": 4036
+    },
+    {
+      "epoch": 0.18597687381950523,
+      "grad_norm": 1.5091189362436521,
+      "learning_rate": 0.00018751396340600733,
+      "loss": 1.4631,
+      "step": 4037
+    },
+    {
+      "epoch": 0.18602294190814023,
+      "grad_norm": 1.896393487326917,
+      "learning_rate": 0.0001875067426556612,
+      "loss": 1.4898,
+      "step": 4038
+    },
+    {
+      "epoch": 0.18606900999677523,
+      "grad_norm": 2.7180123564116325,
+      "learning_rate": 0.00018749951995712745,
+      "loss": 1.6702,
+      "step": 4039
+    },
+    {
+      "epoch": 0.18611507808541022,
+      "grad_norm": 1.8391554474416227,
+      "learning_rate": 0.00018749229531056695,
+      "loss": 1.6127,
+      "step": 4040
+    },
+    {
+      "epoch": 0.18616114617404525,
+      "grad_norm": 3.337059696127694,
+      "learning_rate": 0.0001874850687161405,
+      "loss": 1.3747,
+      "step": 4041
+    },
+    {
+      "epoch": 0.18620721426268025,
+      "grad_norm": 1.668343444012805,
+      "learning_rate": 0.00018747784017400903,
+      "loss": 1.5073,
+      "step": 4042
+    },
+    {
+      "epoch": 0.18625328235131525,
+      "grad_norm": 4.243497253041649,
+      "learning_rate": 0.00018747060968433342,
+      "loss": 1.7095,
+      "step": 4043
+    },
+    {
+      "epoch": 0.18629935043995025,
+      "grad_norm": 2.080736177774694,
+      "learning_rate": 0.00018746337724727464,
+      "loss": 1.8049,
+      "step": 4044
+    },
+    {
+      "epoch": 0.18634541852858524,
+      "grad_norm": 1.6175941682081643,
+      "learning_rate": 0.00018745614286299378,
+      "loss": 1.6559,
+      "step": 4045
+    },
+    {
+      "epoch": 0.18639148661722024,
+      "grad_norm": 1.4963700680526957,
+      "learning_rate": 0.0001874489065316518,
+      "loss": 1.7382,
+      "step": 4046
+    },
+    {
+      "epoch": 0.18643755470585524,
+      "grad_norm": 4.609049482521492,
+      "learning_rate": 0.0001874416682534099,
+      "loss": 1.7457,
+      "step": 4047
+    },
+    {
+      "epoch": 0.18648362279449027,
+      "grad_norm": 5.337827035096778,
+      "learning_rate": 0.00018743442802842918,
+      "loss": 1.48,
+      "step": 4048
+    },
+    {
+      "epoch": 0.18652969088312527,
+      "grad_norm": 1.5834255015183865,
+      "learning_rate": 0.00018742718585687085,
+      "loss": 1.3165,
+      "step": 4049
+    },
+    {
+      "epoch": 0.18657575897176026,
+      "grad_norm": 1.7103667508320646,
+      "learning_rate": 0.00018741994173889612,
+      "loss": 1.3845,
+      "step": 4050
+    },
+    {
+      "epoch": 0.18662182706039526,
+      "grad_norm": 1.8416012858640183,
+      "learning_rate": 0.00018741269567466628,
+      "loss": 1.7344,
+      "step": 4051
+    },
+    {
+      "epoch": 0.18666789514903026,
+      "grad_norm": 2.3579102532585705,
+      "learning_rate": 0.00018740544766434263,
+      "loss": 1.5947,
+      "step": 4052
+    },
+    {
+      "epoch": 0.18671396323766526,
+      "grad_norm": 3.088654505235596,
+      "learning_rate": 0.00018739819770808658,
+      "loss": 1.3443,
+      "step": 4053
+    },
+    {
+      "epoch": 0.18676003132630029,
+      "grad_norm": 1.400769435234432,
+      "learning_rate": 0.0001873909458060595,
+      "loss": 1.7026,
+      "step": 4054
+    },
+    {
+      "epoch": 0.18680609941493528,
+      "grad_norm": 1.5066957389807867,
+      "learning_rate": 0.00018738369195842286,
+      "loss": 1.5408,
+      "step": 4055
+    },
+    {
+      "epoch": 0.18685216750357028,
+      "grad_norm": 6.040638068339019,
+      "learning_rate": 0.00018737643616533814,
+      "loss": 1.4252,
+      "step": 4056
+    },
+    {
+      "epoch": 0.18689823559220528,
+      "grad_norm": 2.193179113557414,
+      "learning_rate": 0.00018736917842696694,
+      "loss": 1.4897,
+      "step": 4057
+    },
+    {
+      "epoch": 0.18694430368084028,
+      "grad_norm": 2.0844713372744237,
+      "learning_rate": 0.00018736191874347074,
+      "loss": 1.5015,
+      "step": 4058
+    },
+    {
+      "epoch": 0.18699037176947528,
+      "grad_norm": 1.5340848092866837,
+      "learning_rate": 0.00018735465711501123,
+      "loss": 1.5193,
+      "step": 4059
+    },
+    {
+      "epoch": 0.18703643985811028,
+      "grad_norm": 1.8509499295518193,
+      "learning_rate": 0.00018734739354175003,
+      "loss": 1.5131,
+      "step": 4060
+    },
+    {
+      "epoch": 0.1870825079467453,
+      "grad_norm": 1.9767891803202537,
+      "learning_rate": 0.00018734012802384893,
+      "loss": 1.646,
+      "step": 4061
+    },
+    {
+      "epoch": 0.1871285760353803,
+      "grad_norm": 2.5927578556466218,
+      "learning_rate": 0.0001873328605614696,
+      "loss": 1.5934,
+      "step": 4062
+    },
+    {
+      "epoch": 0.1871746441240153,
+      "grad_norm": 2.0150846468846804,
+      "learning_rate": 0.00018732559115477388,
+      "loss": 1.5735,
+      "step": 4063
+    },
+    {
+      "epoch": 0.1872207122126503,
+      "grad_norm": 2.4531969549855246,
+      "learning_rate": 0.00018731831980392362,
+      "loss": 1.5506,
+      "step": 4064
+    },
+    {
+      "epoch": 0.1872667803012853,
+      "grad_norm": 3.45372538110111,
+      "learning_rate": 0.00018731104650908064,
+      "loss": 1.5001,
+      "step": 4065
+    },
+    {
+      "epoch": 0.1873128483899203,
+      "grad_norm": 1.805283795519339,
+      "learning_rate": 0.00018730377127040696,
+      "loss": 1.2716,
+      "step": 4066
+    },
+    {
+      "epoch": 0.1873589164785553,
+      "grad_norm": 1.407037769258917,
+      "learning_rate": 0.0001872964940880645,
+      "loss": 1.3286,
+      "step": 4067
+    },
+    {
+      "epoch": 0.18740498456719032,
+      "grad_norm": 2.3290503446013213,
+      "learning_rate": 0.00018728921496221528,
+      "loss": 1.589,
+      "step": 4068
+    },
+    {
+      "epoch": 0.18745105265582532,
+      "grad_norm": 1.443650325265951,
+      "learning_rate": 0.00018728193389302134,
+      "loss": 1.3454,
+      "step": 4069
+    },
+    {
+      "epoch": 0.18749712074446032,
+      "grad_norm": 3.159573757292214,
+      "learning_rate": 0.00018727465088064484,
+      "loss": 1.6065,
+      "step": 4070
+    },
+    {
+      "epoch": 0.18754318883309531,
+      "grad_norm": 2.03517960664157,
+      "learning_rate": 0.00018726736592524783,
+      "loss": 1.262,
+      "step": 4071
+    },
+    {
+      "epoch": 0.1875892569217303,
+      "grad_norm": 2.113206712996779,
+      "learning_rate": 0.0001872600790269926,
+      "loss": 1.6585,
+      "step": 4072
+    },
+    {
+      "epoch": 0.1876353250103653,
+      "grad_norm": 2.1506815459402486,
+      "learning_rate": 0.0001872527901860413,
+      "loss": 1.7378,
+      "step": 4073
+    },
+    {
+      "epoch": 0.1876813930990003,
+      "grad_norm": 2.379326063967237,
+      "learning_rate": 0.00018724549940255623,
+      "loss": 1.3585,
+      "step": 4074
+    },
+    {
+      "epoch": 0.18772746118763534,
+      "grad_norm": 2.3428485978692817,
+      "learning_rate": 0.00018723820667669973,
+      "loss": 1.4256,
+      "step": 4075
+    },
+    {
+      "epoch": 0.18777352927627033,
+      "grad_norm": 1.1989702051926865,
+      "learning_rate": 0.00018723091200863412,
+      "loss": 1.5184,
+      "step": 4076
+    },
+    {
+      "epoch": 0.18781959736490533,
+      "grad_norm": 2.116265854345309,
+      "learning_rate": 0.00018722361539852183,
+      "loss": 1.4365,
+      "step": 4077
+    },
+    {
+      "epoch": 0.18786566545354033,
+      "grad_norm": 4.852786148138573,
+      "learning_rate": 0.00018721631684652529,
+      "loss": 1.5835,
+      "step": 4078
+    },
+    {
+      "epoch": 0.18791173354217533,
+      "grad_norm": 3.124839024882297,
+      "learning_rate": 0.000187209016352807,
+      "loss": 1.5529,
+      "step": 4079
+    },
+    {
+      "epoch": 0.18795780163081033,
+      "grad_norm": 2.2143329733186725,
+      "learning_rate": 0.00018720171391752952,
+      "loss": 1.4571,
+      "step": 4080
+    },
+    {
+      "epoch": 0.18800386971944533,
+      "grad_norm": 2.992498579273721,
+      "learning_rate": 0.00018719440954085537,
+      "loss": 1.4172,
+      "step": 4081
+    },
+    {
+      "epoch": 0.18804993780808035,
+      "grad_norm": 1.871531746588849,
+      "learning_rate": 0.0001871871032229472,
+      "loss": 1.5182,
+      "step": 4082
+    },
+    {
+      "epoch": 0.18809600589671535,
+      "grad_norm": 2.7507292950119457,
+      "learning_rate": 0.0001871797949639677,
+      "loss": 1.5207,
+      "step": 4083
+    },
+    {
+      "epoch": 0.18814207398535035,
+      "grad_norm": 1.636287055347011,
+      "learning_rate": 0.0001871724847640795,
+      "loss": 1.3699,
+      "step": 4084
+    },
+    {
+      "epoch": 0.18818814207398535,
+      "grad_norm": 1.489550970747178,
+      "learning_rate": 0.00018716517262344545,
+      "loss": 1.3839,
+      "step": 4085
+    },
+    {
+      "epoch": 0.18823421016262035,
+      "grad_norm": 3.66285111249469,
+      "learning_rate": 0.00018715785854222825,
+      "loss": 1.5087,
+      "step": 4086
+    },
+    {
+      "epoch": 0.18828027825125535,
+      "grad_norm": 2.6534137785470033,
+      "learning_rate": 0.00018715054252059078,
+      "loss": 1.3952,
+      "step": 4087
+    },
+    {
+      "epoch": 0.18832634633989037,
+      "grad_norm": 1.4431904382936065,
+      "learning_rate": 0.0001871432245586959,
+      "loss": 1.4015,
+      "step": 4088
+    },
+    {
+      "epoch": 0.18837241442852537,
+      "grad_norm": 2.1324113125193307,
+      "learning_rate": 0.00018713590465670656,
+      "loss": 1.4507,
+      "step": 4089
+    },
+    {
+      "epoch": 0.18841848251716037,
+      "grad_norm": 1.4317484790451567,
+      "learning_rate": 0.00018712858281478572,
+      "loss": 1.8202,
+      "step": 4090
+    },
+    {
+      "epoch": 0.18846455060579537,
+      "grad_norm": 1.8800296037006805,
+      "learning_rate": 0.00018712125903309635,
+      "loss": 1.1628,
+      "step": 4091
+    },
+    {
+      "epoch": 0.18851061869443037,
+      "grad_norm": 1.4231490394215869,
+      "learning_rate": 0.00018711393331180155,
+      "loss": 1.2667,
+      "step": 4092
+    },
+    {
+      "epoch": 0.18855668678306536,
+      "grad_norm": 1.8619734558414083,
+      "learning_rate": 0.00018710660565106438,
+      "loss": 1.5453,
+      "step": 4093
+    },
+    {
+      "epoch": 0.18860275487170036,
+      "grad_norm": 2.1312211175342455,
+      "learning_rate": 0.000187099276051048,
+      "loss": 1.3905,
+      "step": 4094
+    },
+    {
+      "epoch": 0.1886488229603354,
+      "grad_norm": 1.7113963855060756,
+      "learning_rate": 0.00018709194451191555,
+      "loss": 1.48,
+      "step": 4095
+    },
+    {
+      "epoch": 0.1886948910489704,
+      "grad_norm": 1.5704741891899525,
+      "learning_rate": 0.00018708461103383032,
+      "loss": 1.3265,
+      "step": 4096
+    },
+    {
+      "epoch": 0.18874095913760539,
+      "grad_norm": 1.6380829175262586,
+      "learning_rate": 0.00018707727561695557,
+      "loss": 1.3612,
+      "step": 4097
+    },
+    {
+      "epoch": 0.18878702722624038,
+      "grad_norm": 2.7786184890153516,
+      "learning_rate": 0.00018706993826145453,
+      "loss": 1.188,
+      "step": 4098
+    },
+    {
+      "epoch": 0.18883309531487538,
+      "grad_norm": 1.2602540973399696,
+      "learning_rate": 0.00018706259896749064,
+      "loss": 1.3686,
+      "step": 4099
+    },
+    {
+      "epoch": 0.18887916340351038,
+      "grad_norm": 1.4532901197476356,
+      "learning_rate": 0.00018705525773522725,
+      "loss": 1.3828,
+      "step": 4100
+    },
+    {
+      "epoch": 0.18892523149214538,
+      "grad_norm": 2.144182458143982,
+      "learning_rate": 0.0001870479145648278,
+      "loss": 1.2024,
+      "step": 4101
+    },
+    {
+      "epoch": 0.1889712995807804,
+      "grad_norm": 2.7370050780993256,
+      "learning_rate": 0.00018704056945645584,
+      "loss": 1.4133,
+      "step": 4102
+    },
+    {
+      "epoch": 0.1890173676694154,
+      "grad_norm": 2.632729684309105,
+      "learning_rate": 0.0001870332224102748,
+      "loss": 1.3985,
+      "step": 4103
+    },
+    {
+      "epoch": 0.1890634357580504,
+      "grad_norm": 1.6304220495325064,
+      "learning_rate": 0.00018702587342644828,
+      "loss": 1.4407,
+      "step": 4104
+    },
+    {
+      "epoch": 0.1891095038466854,
+      "grad_norm": 1.9506099804432881,
+      "learning_rate": 0.00018701852250513993,
+      "loss": 1.5186,
+      "step": 4105
+    },
+    {
+      "epoch": 0.1891555719353204,
+      "grad_norm": 21.77866122364967,
+      "learning_rate": 0.00018701116964651337,
+      "loss": 1.2278,
+      "step": 4106
+    },
+    {
+      "epoch": 0.1892016400239554,
+      "grad_norm": 1.5727266327702405,
+      "learning_rate": 0.00018700381485073234,
+      "loss": 1.6128,
+      "step": 4107
+    },
+    {
+      "epoch": 0.1892477081125904,
+      "grad_norm": 1.319955367145242,
+      "learning_rate": 0.00018699645811796051,
+      "loss": 1.1886,
+      "step": 4108
+    },
+    {
+      "epoch": 0.18929377620122542,
+      "grad_norm": 1.9951889455065457,
+      "learning_rate": 0.00018698909944836174,
+      "loss": 1.539,
+      "step": 4109
+    },
+    {
+      "epoch": 0.18933984428986042,
+      "grad_norm": 3.614712631500923,
+      "learning_rate": 0.0001869817388420998,
+      "loss": 1.3974,
+      "step": 4110
+    },
+    {
+      "epoch": 0.18938591237849542,
+      "grad_norm": 1.3259135891966598,
+      "learning_rate": 0.0001869743762993386,
+      "loss": 1.4701,
+      "step": 4111
+    },
+    {
+      "epoch": 0.18943198046713042,
+      "grad_norm": 1.9333070521442877,
+      "learning_rate": 0.00018696701182024207,
+      "loss": 1.4665,
+      "step": 4112
+    },
+    {
+      "epoch": 0.18947804855576542,
+      "grad_norm": 2.215051885881088,
+      "learning_rate": 0.0001869596454049741,
+      "loss": 1.775,
+      "step": 4113
+    },
+    {
+      "epoch": 0.18952411664440041,
+      "grad_norm": 2.8719054052113777,
+      "learning_rate": 0.00018695227705369872,
+      "loss": 1.5579,
+      "step": 4114
+    },
+    {
+      "epoch": 0.1895701847330354,
+      "grad_norm": 1.3107709967318146,
+      "learning_rate": 0.00018694490676658,
+      "loss": 1.5422,
+      "step": 4115
+    },
+    {
+      "epoch": 0.18961625282167044,
+      "grad_norm": 6.157950144828371,
+      "learning_rate": 0.000186937534543782,
+      "loss": 1.1264,
+      "step": 4116
+    },
+    {
+      "epoch": 0.18966232091030544,
+      "grad_norm": 3.46746065298582,
+      "learning_rate": 0.00018693016038546888,
+      "loss": 1.0973,
+      "step": 4117
+    },
+    {
+      "epoch": 0.18970838899894044,
+      "grad_norm": 12.769019530852619,
+      "learning_rate": 0.0001869227842918048,
+      "loss": 1.2139,
+      "step": 4118
+    },
+    {
+      "epoch": 0.18975445708757543,
+      "grad_norm": 4.281078947345235,
+      "learning_rate": 0.00018691540626295392,
+      "loss": 1.5222,
+      "step": 4119
+    },
+    {
+      "epoch": 0.18980052517621043,
+      "grad_norm": 4.752088122510855,
+      "learning_rate": 0.00018690802629908057,
+      "loss": 1.4938,
+      "step": 4120
+    },
+    {
+      "epoch": 0.18984659326484543,
+      "grad_norm": 2.239529075935936,
+      "learning_rate": 0.00018690064440034902,
+      "loss": 1.3462,
+      "step": 4121
+    },
+    {
+      "epoch": 0.18989266135348046,
+      "grad_norm": 2.32031487984469,
+      "learning_rate": 0.00018689326056692366,
+      "loss": 1.473,
+      "step": 4122
+    },
+    {
+      "epoch": 0.18993872944211546,
+      "grad_norm": 2.9496160492446157,
+      "learning_rate": 0.00018688587479896883,
+      "loss": 1.3088,
+      "step": 4123
+    },
+    {
+      "epoch": 0.18998479753075045,
+      "grad_norm": 3.0414633789633427,
+      "learning_rate": 0.00018687848709664898,
+      "loss": 1.3124,
+      "step": 4124
+    },
+    {
+      "epoch": 0.19003086561938545,
+      "grad_norm": 1.976694676705152,
+      "learning_rate": 0.00018687109746012856,
+      "loss": 1.3258,
+      "step": 4125
+    },
+    {
+      "epoch": 0.19007693370802045,
+      "grad_norm": 2.3513456714386076,
+      "learning_rate": 0.00018686370588957209,
+      "loss": 1.3999,
+      "step": 4126
+    },
+    {
+      "epoch": 0.19012300179665545,
+      "grad_norm": 5.973091328065671,
+      "learning_rate": 0.0001868563123851442,
+      "loss": 1.4308,
+      "step": 4127
+    },
+    {
+      "epoch": 0.19016906988529045,
+      "grad_norm": 7.485142487583017,
+      "learning_rate": 0.0001868489169470094,
+      "loss": 1.3462,
+      "step": 4128
+    },
+    {
+      "epoch": 0.19021513797392547,
+      "grad_norm": 2.748269726268668,
+      "learning_rate": 0.0001868415195753324,
+      "loss": 1.44,
+      "step": 4129
+    },
+    {
+      "epoch": 0.19026120606256047,
+      "grad_norm": 4.314315986816833,
+      "learning_rate": 0.00018683412027027787,
+      "loss": 1.2647,
+      "step": 4130
+    },
+    {
+      "epoch": 0.19030727415119547,
+      "grad_norm": 2.633063808711455,
+      "learning_rate": 0.00018682671903201052,
+      "loss": 1.7852,
+      "step": 4131
+    },
+    {
+      "epoch": 0.19035334223983047,
+      "grad_norm": 10.140705068256354,
+      "learning_rate": 0.00018681931586069514,
+      "loss": 1.6976,
+      "step": 4132
+    },
+    {
+      "epoch": 0.19039941032846547,
+      "grad_norm": 2.235732774602849,
+      "learning_rate": 0.0001868119107564966,
+      "loss": 1.7238,
+      "step": 4133
+    },
+    {
+      "epoch": 0.19044547841710047,
+      "grad_norm": 2.7474942033020398,
+      "learning_rate": 0.00018680450371957966,
+      "loss": 1.8357,
+      "step": 4134
+    },
+    {
+      "epoch": 0.19049154650573547,
+      "grad_norm": 3.666087563056715,
+      "learning_rate": 0.00018679709475010934,
+      "loss": 1.8189,
+      "step": 4135
+    },
+    {
+      "epoch": 0.1905376145943705,
+      "grad_norm": 2.6819380562123536,
+      "learning_rate": 0.0001867896838482505,
+      "loss": 1.6418,
+      "step": 4136
+    },
+    {
+      "epoch": 0.1905836826830055,
+      "grad_norm": 2.3747064120878534,
+      "learning_rate": 0.00018678227101416817,
+      "loss": 1.9795,
+      "step": 4137
+    },
+    {
+      "epoch": 0.1906297507716405,
+      "grad_norm": 2.6663687985635223,
+      "learning_rate": 0.00018677485624802738,
+      "loss": 1.3933,
+      "step": 4138
+    },
+    {
+      "epoch": 0.1906758188602755,
+      "grad_norm": 2.110669432435171,
+      "learning_rate": 0.00018676743954999317,
+      "loss": 1.5844,
+      "step": 4139
+    },
+    {
+      "epoch": 0.19072188694891049,
+      "grad_norm": 2.515521434946675,
+      "learning_rate": 0.00018676002092023072,
+      "loss": 1.6505,
+      "step": 4140
+    },
+    {
+      "epoch": 0.19076795503754548,
+      "grad_norm": 1.5652712298636868,
+      "learning_rate": 0.00018675260035890516,
+      "loss": 1.2718,
+      "step": 4141
+    },
+    {
+      "epoch": 0.19081402312618048,
+      "grad_norm": 3.066628995137191,
+      "learning_rate": 0.0001867451778661817,
+      "loss": 1.6116,
+      "step": 4142
+    },
+    {
+      "epoch": 0.1908600912148155,
+      "grad_norm": 2.0303170410053504,
+      "learning_rate": 0.00018673775344222563,
+      "loss": 1.4758,
+      "step": 4143
+    },
+    {
+      "epoch": 0.1909061593034505,
+      "grad_norm": 1.131987855806461,
+      "learning_rate": 0.00018673032708720215,
+      "loss": 1.6518,
+      "step": 4144
+    },
+    {
+      "epoch": 0.1909522273920855,
+      "grad_norm": 0.9404074638119211,
+      "learning_rate": 0.00018672289880127666,
+      "loss": 1.752,
+      "step": 4145
+    },
+    {
+      "epoch": 0.1909982954807205,
+      "grad_norm": 5.130615154724814,
+      "learning_rate": 0.00018671546858461453,
+      "loss": 1.4974,
+      "step": 4146
+    },
+    {
+      "epoch": 0.1910443635693555,
+      "grad_norm": 4.473480503897341,
+      "learning_rate": 0.00018670803643738116,
+      "loss": 1.6582,
+      "step": 4147
+    },
+    {
+      "epoch": 0.1910904316579905,
+      "grad_norm": 2.6344773483317865,
+      "learning_rate": 0.00018670060235974207,
+      "loss": 1.4565,
+      "step": 4148
+    },
+    {
+      "epoch": 0.1911364997466255,
+      "grad_norm": 1.2553405582420092,
+      "learning_rate": 0.0001866931663518627,
+      "loss": 1.571,
+      "step": 4149
+    },
+    {
+      "epoch": 0.19118256783526053,
+      "grad_norm": 1.7626176716894757,
+      "learning_rate": 0.00018668572841390866,
+      "loss": 1.733,
+      "step": 4150
+    },
+    {
+      "epoch": 0.19122863592389552,
+      "grad_norm": 1.7121310530492706,
+      "learning_rate": 0.00018667828854604547,
+      "loss": 1.6071,
+      "step": 4151
+    },
+    {
+      "epoch": 0.19127470401253052,
+      "grad_norm": 1.786617578568972,
+      "learning_rate": 0.0001866708467484388,
+      "loss": 1.7558,
+      "step": 4152
+    },
+    {
+      "epoch": 0.19132077210116552,
+      "grad_norm": 3.8607966939261775,
+      "learning_rate": 0.0001866634030212544,
+      "loss": 1.3998,
+      "step": 4153
+    },
+    {
+      "epoch": 0.19136684018980052,
+      "grad_norm": 2.0613017401118765,
+      "learning_rate": 0.00018665595736465788,
+      "loss": 1.4389,
+      "step": 4154
+    },
+    {
+      "epoch": 0.19141290827843552,
+      "grad_norm": 1.3415767951343305,
+      "learning_rate": 0.00018664850977881506,
+      "loss": 1.6816,
+      "step": 4155
+    },
+    {
+      "epoch": 0.19145897636707054,
+      "grad_norm": 1.6593060311692436,
+      "learning_rate": 0.00018664106026389173,
+      "loss": 1.2462,
+      "step": 4156
+    },
+    {
+      "epoch": 0.19150504445570554,
+      "grad_norm": 2.491631929587008,
+      "learning_rate": 0.00018663360882005377,
+      "loss": 1.4676,
+      "step": 4157
+    },
+    {
+      "epoch": 0.19155111254434054,
+      "grad_norm": 1.7386647665452968,
+      "learning_rate": 0.00018662615544746703,
+      "loss": 1.4984,
+      "step": 4158
+    },
+    {
+      "epoch": 0.19159718063297554,
+      "grad_norm": 2.1108523874647207,
+      "learning_rate": 0.0001866187001462975,
+      "loss": 1.7768,
+      "step": 4159
+    },
+    {
+      "epoch": 0.19164324872161054,
+      "grad_norm": 1.4385426442771214,
+      "learning_rate": 0.0001866112429167111,
+      "loss": 1.1777,
+      "step": 4160
+    },
+    {
+      "epoch": 0.19168931681024554,
+      "grad_norm": 1.3221880978727838,
+      "learning_rate": 0.0001866037837588739,
+      "loss": 1.5755,
+      "step": 4161
+    },
+    {
+      "epoch": 0.19173538489888053,
+      "grad_norm": 1.4201712379431506,
+      "learning_rate": 0.00018659632267295194,
+      "loss": 1.3535,
+      "step": 4162
+    },
+    {
+      "epoch": 0.19178145298751556,
+      "grad_norm": 4.983640643749719,
+      "learning_rate": 0.00018658885965911135,
+      "loss": 1.8074,
+      "step": 4163
+    },
+    {
+      "epoch": 0.19182752107615056,
+      "grad_norm": 1.5397812672430793,
+      "learning_rate": 0.00018658139471751828,
+      "loss": 1.6479,
+      "step": 4164
+    },
+    {
+      "epoch": 0.19187358916478556,
+      "grad_norm": 1.059170575300199,
+      "learning_rate": 0.0001865739278483389,
+      "loss": 1.284,
+      "step": 4165
+    },
+    {
+      "epoch": 0.19191965725342056,
+      "grad_norm": 14.872250854958128,
+      "learning_rate": 0.00018656645905173943,
+      "loss": 1.6572,
+      "step": 4166
+    },
+    {
+      "epoch": 0.19196572534205555,
+      "grad_norm": 14.597585630084794,
+      "learning_rate": 0.00018655898832788622,
+      "loss": 1.4839,
+      "step": 4167
+    },
+    {
+      "epoch": 0.19201179343069055,
+      "grad_norm": 8.501141401154925,
+      "learning_rate": 0.0001865515156769455,
+      "loss": 1.4077,
+      "step": 4168
+    },
+    {
+      "epoch": 0.19205786151932555,
+      "grad_norm": 3.335089496978012,
+      "learning_rate": 0.00018654404109908374,
+      "loss": 1.345,
+      "step": 4169
+    },
+    {
+      "epoch": 0.19210392960796058,
+      "grad_norm": 7.17493090999926,
+      "learning_rate": 0.0001865365645944673,
+      "loss": 1.6108,
+      "step": 4170
+    },
+    {
+      "epoch": 0.19214999769659558,
+      "grad_norm": 2.747321534876721,
+      "learning_rate": 0.0001865290861632626,
+      "loss": 1.7242,
+      "step": 4171
+    },
+    {
+      "epoch": 0.19219606578523057,
+      "grad_norm": 2.784600333261108,
+      "learning_rate": 0.00018652160580563616,
+      "loss": 1.7196,
+      "step": 4172
+    },
+    {
+      "epoch": 0.19224213387386557,
+      "grad_norm": 3.1901382792538806,
+      "learning_rate": 0.00018651412352175452,
+      "loss": 1.8381,
+      "step": 4173
+    },
+    {
+      "epoch": 0.19228820196250057,
+      "grad_norm": 4.494526866242237,
+      "learning_rate": 0.00018650663931178428,
+      "loss": 1.5902,
+      "step": 4174
+    },
+    {
+      "epoch": 0.19233427005113557,
+      "grad_norm": 3.0012070763464567,
+      "learning_rate": 0.00018649915317589202,
+      "loss": 1.5813,
+      "step": 4175
+    },
+    {
+      "epoch": 0.19238033813977057,
+      "grad_norm": 4.781938881223064,
+      "learning_rate": 0.00018649166511424444,
+      "loss": 1.6242,
+      "step": 4176
+    },
+    {
+      "epoch": 0.1924264062284056,
+      "grad_norm": 2.3721281118759054,
+      "learning_rate": 0.00018648417512700824,
+      "loss": 1.7118,
+      "step": 4177
+    },
+    {
+      "epoch": 0.1924724743170406,
+      "grad_norm": 2.656477138071557,
+      "learning_rate": 0.00018647668321435017,
+      "loss": 1.6339,
+      "step": 4178
+    },
+    {
+      "epoch": 0.1925185424056756,
+      "grad_norm": 3.891865294838672,
+      "learning_rate": 0.000186469189376437,
+      "loss": 1.6502,
+      "step": 4179
+    },
+    {
+      "epoch": 0.1925646104943106,
+      "grad_norm": 3.707899241792333,
+      "learning_rate": 0.0001864616936134356,
+      "loss": 1.6262,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1926106785829456,
+      "grad_norm": 11.544917435160457,
+      "learning_rate": 0.0001864541959255129,
+      "loss": 1.3346,
+      "step": 4181
+    },
+    {
+      "epoch": 0.1926567466715806,
+      "grad_norm": 17.761764653708298,
+      "learning_rate": 0.0001864466963128357,
+      "loss": 1.6933,
+      "step": 4182
+    },
+    {
+      "epoch": 0.19270281476021558,
+      "grad_norm": 5.586682068986965,
+      "learning_rate": 0.00018643919477557105,
+      "loss": 1.436,
+      "step": 4183
+    },
+    {
+      "epoch": 0.1927488828488506,
+      "grad_norm": 3.111300324635457,
+      "learning_rate": 0.00018643169131388592,
+      "loss": 1.6415,
+      "step": 4184
+    },
+    {
+      "epoch": 0.1927949509374856,
+      "grad_norm": 4.080922187698729,
+      "learning_rate": 0.0001864241859279474,
+      "loss": 1.4602,
+      "step": 4185
+    },
+    {
+      "epoch": 0.1928410190261206,
+      "grad_norm": 4.395074093506386,
+      "learning_rate": 0.00018641667861792256,
+      "loss": 1.5689,
+      "step": 4186
+    },
+    {
+      "epoch": 0.1928870871147556,
+      "grad_norm": 4.0404424265567584,
+      "learning_rate": 0.00018640916938397853,
+      "loss": 1.4745,
+      "step": 4187
+    },
+    {
+      "epoch": 0.1929331552033906,
+      "grad_norm": 3.1506213947690744,
+      "learning_rate": 0.00018640165822628255,
+      "loss": 1.7039,
+      "step": 4188
+    },
+    {
+      "epoch": 0.1929792232920256,
+      "grad_norm": 2.1916822351095058,
+      "learning_rate": 0.00018639414514500175,
+      "loss": 1.6499,
+      "step": 4189
+    },
+    {
+      "epoch": 0.19302529138066063,
+      "grad_norm": 3.859879211499788,
+      "learning_rate": 0.00018638663014030346,
+      "loss": 1.521,
+      "step": 4190
+    },
+    {
+      "epoch": 0.19307135946929563,
+      "grad_norm": 2.3643832032236483,
+      "learning_rate": 0.000186379113212355,
+      "loss": 1.6702,
+      "step": 4191
+    },
+    {
+      "epoch": 0.19311742755793063,
+      "grad_norm": 4.533922789097439,
+      "learning_rate": 0.00018637159436132365,
+      "loss": 1.7506,
+      "step": 4192
+    },
+    {
+      "epoch": 0.19316349564656562,
+      "grad_norm": 3.7010205264634854,
+      "learning_rate": 0.00018636407358737687,
+      "loss": 1.9144,
+      "step": 4193
+    },
+    {
+      "epoch": 0.19320956373520062,
+      "grad_norm": 2.941419276269111,
+      "learning_rate": 0.00018635655089068207,
+      "loss": 1.6925,
+      "step": 4194
+    },
+    {
+      "epoch": 0.19325563182383562,
+      "grad_norm": 4.058919707756113,
+      "learning_rate": 0.00018634902627140672,
+      "loss": 1.7641,
+      "step": 4195
+    },
+    {
+      "epoch": 0.19330169991247062,
+      "grad_norm": 5.034999105681786,
+      "learning_rate": 0.00018634149972971838,
+      "loss": 1.6951,
+      "step": 4196
+    },
+    {
+      "epoch": 0.19334776800110565,
+      "grad_norm": 4.531324741532791,
+      "learning_rate": 0.00018633397126578455,
+      "loss": 1.7677,
+      "step": 4197
+    },
+    {
+      "epoch": 0.19339383608974064,
+      "grad_norm": 7.005612845586174,
+      "learning_rate": 0.0001863264408797729,
+      "loss": 1.5071,
+      "step": 4198
+    },
+    {
+      "epoch": 0.19343990417837564,
+      "grad_norm": 4.178174910399559,
+      "learning_rate": 0.00018631890857185107,
+      "loss": 1.4118,
+      "step": 4199
+    },
+    {
+      "epoch": 0.19348597226701064,
+      "grad_norm": 3.4163311071642997,
+      "learning_rate": 0.00018631137434218676,
+      "loss": 1.5329,
+      "step": 4200
+    },
+    {
+      "epoch": 0.19353204035564564,
+      "grad_norm": 2.369756689878346,
+      "learning_rate": 0.0001863038381909477,
+      "loss": 1.3682,
+      "step": 4201
+    },
+    {
+      "epoch": 0.19357810844428064,
+      "grad_norm": 4.501987308287093,
+      "learning_rate": 0.00018629630011830165,
+      "loss": 1.5563,
+      "step": 4202
+    },
+    {
+      "epoch": 0.19362417653291564,
+      "grad_norm": 2.8878112871561545,
+      "learning_rate": 0.00018628876012441643,
+      "loss": 1.3504,
+      "step": 4203
+    },
+    {
+      "epoch": 0.19367024462155066,
+      "grad_norm": 4.9003035807855575,
+      "learning_rate": 0.00018628121820945992,
+      "loss": 1.6469,
+      "step": 4204
+    },
+    {
+      "epoch": 0.19371631271018566,
+      "grad_norm": 5.529040386498414,
+      "learning_rate": 0.00018627367437360005,
+      "loss": 1.5555,
+      "step": 4205
+    },
+    {
+      "epoch": 0.19376238079882066,
+      "grad_norm": 2.700100173956107,
+      "learning_rate": 0.00018626612861700472,
+      "loss": 1.8949,
+      "step": 4206
+    },
+    {
+      "epoch": 0.19380844888745566,
+      "grad_norm": 5.32652822912339,
+      "learning_rate": 0.00018625858093984196,
+      "loss": 1.8891,
+      "step": 4207
+    },
+    {
+      "epoch": 0.19385451697609066,
+      "grad_norm": 2.6751035850601452,
+      "learning_rate": 0.00018625103134227978,
+      "loss": 1.8132,
+      "step": 4208
+    },
+    {
+      "epoch": 0.19390058506472566,
+      "grad_norm": 2.796207914468844,
+      "learning_rate": 0.00018624347982448632,
+      "loss": 1.6093,
+      "step": 4209
+    },
+    {
+      "epoch": 0.19394665315336065,
+      "grad_norm": 2.6608842773294046,
+      "learning_rate": 0.00018623592638662963,
+      "loss": 1.4808,
+      "step": 4210
+    },
+    {
+      "epoch": 0.19399272124199568,
+      "grad_norm": 3.5822500980351233,
+      "learning_rate": 0.0001862283710288779,
+      "loss": 1.5719,
+      "step": 4211
+    },
+    {
+      "epoch": 0.19403878933063068,
+      "grad_norm": 1.6392409016355636,
+      "learning_rate": 0.00018622081375139934,
+      "loss": 1.4318,
+      "step": 4212
+    },
+    {
+      "epoch": 0.19408485741926568,
+      "grad_norm": 15.077461978734544,
+      "learning_rate": 0.0001862132545543622,
+      "loss": 1.8225,
+      "step": 4213
+    },
+    {
+      "epoch": 0.19413092550790068,
+      "grad_norm": 2.198615919305075,
+      "learning_rate": 0.00018620569343793478,
+      "loss": 1.7761,
+      "step": 4214
+    },
+    {
+      "epoch": 0.19417699359653567,
+      "grad_norm": 3.0727349493742127,
+      "learning_rate": 0.00018619813040228543,
+      "loss": 1.5219,
+      "step": 4215
+    },
+    {
+      "epoch": 0.19422306168517067,
+      "grad_norm": 5.64455936004434,
+      "learning_rate": 0.00018619056544758245,
+      "loss": 1.5499,
+      "step": 4216
+    },
+    {
+      "epoch": 0.19426912977380567,
+      "grad_norm": 4.889470611313948,
+      "learning_rate": 0.00018618299857399434,
+      "loss": 1.6823,
+      "step": 4217
+    },
+    {
+      "epoch": 0.1943151978624407,
+      "grad_norm": 7.9893153280783595,
+      "learning_rate": 0.00018617542978168952,
+      "loss": 1.5213,
+      "step": 4218
+    },
+    {
+      "epoch": 0.1943612659510757,
+      "grad_norm": 5.777219786220017,
+      "learning_rate": 0.00018616785907083657,
+      "loss": 1.8035,
+      "step": 4219
+    },
+    {
+      "epoch": 0.1944073340397107,
+      "grad_norm": 3.680087044346359,
+      "learning_rate": 0.00018616028644160394,
+      "loss": 1.6763,
+      "step": 4220
+    },
+    {
+      "epoch": 0.1944534021283457,
+      "grad_norm": 4.752451586367965,
+      "learning_rate": 0.00018615271189416027,
+      "loss": 1.6324,
+      "step": 4221
+    },
+    {
+      "epoch": 0.1944994702169807,
+      "grad_norm": 3.7419738591010847,
+      "learning_rate": 0.0001861451354286742,
+      "loss": 1.4407,
+      "step": 4222
+    },
+    {
+      "epoch": 0.1945455383056157,
+      "grad_norm": 10.872301016564293,
+      "learning_rate": 0.0001861375570453144,
+      "loss": 1.2653,
+      "step": 4223
+    },
+    {
+      "epoch": 0.19459160639425072,
+      "grad_norm": 19.46596682069967,
+      "learning_rate": 0.00018612997674424957,
+      "loss": 2.0956,
+      "step": 4224
+    },
+    {
+      "epoch": 0.19463767448288571,
+      "grad_norm": 12.157568814751778,
+      "learning_rate": 0.0001861223945256485,
+      "loss": 1.5839,
+      "step": 4225
+    },
+    {
+      "epoch": 0.1946837425715207,
+      "grad_norm": 10.29748387929946,
+      "learning_rate": 0.00018611481038967998,
+      "loss": 1.8423,
+      "step": 4226
+    },
+    {
+      "epoch": 0.1947298106601557,
+      "grad_norm": 33.672068092356206,
+      "learning_rate": 0.00018610722433651285,
+      "loss": 2.0362,
+      "step": 4227
+    },
+    {
+      "epoch": 0.1947758787487907,
+      "grad_norm": 8.082651649896441,
+      "learning_rate": 0.00018609963636631601,
+      "loss": 2.0165,
+      "step": 4228
+    },
+    {
+      "epoch": 0.1948219468374257,
+      "grad_norm": 9.777226887765297,
+      "learning_rate": 0.0001860920464792584,
+      "loss": 1.9897,
+      "step": 4229
+    },
+    {
+      "epoch": 0.1948680149260607,
+      "grad_norm": 4.392510319970188,
+      "learning_rate": 0.00018608445467550904,
+      "loss": 1.949,
+      "step": 4230
+    },
+    {
+      "epoch": 0.19491408301469573,
+      "grad_norm": 5.29538299360704,
+      "learning_rate": 0.00018607686095523684,
+      "loss": 1.8798,
+      "step": 4231
+    },
+    {
+      "epoch": 0.19496015110333073,
+      "grad_norm": 7.488648285449605,
+      "learning_rate": 0.00018606926531861096,
+      "loss": 1.8072,
+      "step": 4232
+    },
+    {
+      "epoch": 0.19500621919196573,
+      "grad_norm": 8.33469039340707,
+      "learning_rate": 0.00018606166776580045,
+      "loss": 2.1411,
+      "step": 4233
+    },
+    {
+      "epoch": 0.19505228728060073,
+      "grad_norm": 6.038269151936536,
+      "learning_rate": 0.00018605406829697447,
+      "loss": 1.8082,
+      "step": 4234
+    },
+    {
+      "epoch": 0.19509835536923573,
+      "grad_norm": 10.084766229612203,
+      "learning_rate": 0.0001860464669123022,
+      "loss": 2.1992,
+      "step": 4235
+    },
+    {
+      "epoch": 0.19514442345787072,
+      "grad_norm": 5.387261063533361,
+      "learning_rate": 0.00018603886361195292,
+      "loss": 1.9202,
+      "step": 4236
+    },
+    {
+      "epoch": 0.19519049154650572,
+      "grad_norm": 3.567131143316026,
+      "learning_rate": 0.00018603125839609583,
+      "loss": 1.8593,
+      "step": 4237
+    },
+    {
+      "epoch": 0.19523655963514075,
+      "grad_norm": 29.956494802139655,
+      "learning_rate": 0.00018602365126490028,
+      "loss": 2.1442,
+      "step": 4238
+    },
+    {
+      "epoch": 0.19528262772377575,
+      "grad_norm": 6.348307498689154,
+      "learning_rate": 0.00018601604221853566,
+      "loss": 1.9926,
+      "step": 4239
+    },
+    {
+      "epoch": 0.19532869581241075,
+      "grad_norm": 10.469561854613078,
+      "learning_rate": 0.00018600843125717132,
+      "loss": 2.3093,
+      "step": 4240
+    },
+    {
+      "epoch": 0.19537476390104574,
+      "grad_norm": 6.561910841798019,
+      "learning_rate": 0.00018600081838097675,
+      "loss": 2.3224,
+      "step": 4241
+    },
+    {
+      "epoch": 0.19542083198968074,
+      "grad_norm": 3.784619907991716,
+      "learning_rate": 0.0001859932035901214,
+      "loss": 1.6567,
+      "step": 4242
+    },
+    {
+      "epoch": 0.19546690007831574,
+      "grad_norm": 6.351172807282012,
+      "learning_rate": 0.00018598558688477483,
+      "loss": 2.1787,
+      "step": 4243
+    },
+    {
+      "epoch": 0.19551296816695074,
+      "grad_norm": 5.274141562092312,
+      "learning_rate": 0.00018597796826510661,
+      "loss": 2.1015,
+      "step": 4244
+    },
+    {
+      "epoch": 0.19555903625558577,
+      "grad_norm": 8.930822249419654,
+      "learning_rate": 0.00018597034773128632,
+      "loss": 1.9696,
+      "step": 4245
+    },
+    {
+      "epoch": 0.19560510434422076,
+      "grad_norm": 4.942809513935499,
+      "learning_rate": 0.00018596272528348368,
+      "loss": 1.6743,
+      "step": 4246
+    },
+    {
+      "epoch": 0.19565117243285576,
+      "grad_norm": 5.651494665568518,
+      "learning_rate": 0.00018595510092186835,
+      "loss": 1.6656,
+      "step": 4247
+    },
+    {
+      "epoch": 0.19569724052149076,
+      "grad_norm": 14.923792018602267,
+      "learning_rate": 0.00018594747464661007,
+      "loss": 2.0014,
+      "step": 4248
+    },
+    {
+      "epoch": 0.19574330861012576,
+      "grad_norm": 15.143433099829018,
+      "learning_rate": 0.00018593984645787863,
+      "loss": 2.0482,
+      "step": 4249
+    },
+    {
+      "epoch": 0.19578937669876076,
+      "grad_norm": 6.602276880204695,
+      "learning_rate": 0.00018593221635584387,
+      "loss": 2.0145,
+      "step": 4250
+    },
+    {
+      "epoch": 0.19583544478739576,
+      "grad_norm": 4.464456820282585,
+      "learning_rate": 0.00018592458434067566,
+      "loss": 1.8011,
+      "step": 4251
+    },
+    {
+      "epoch": 0.19588151287603078,
+      "grad_norm": 4.564912294136775,
+      "learning_rate": 0.00018591695041254389,
+      "loss": 2.1332,
+      "step": 4252
+    },
+    {
+      "epoch": 0.19592758096466578,
+      "grad_norm": 4.269718730608756,
+      "learning_rate": 0.00018590931457161856,
+      "loss": 1.8254,
+      "step": 4253
+    },
+    {
+      "epoch": 0.19597364905330078,
+      "grad_norm": 3.905294922884679,
+      "learning_rate": 0.00018590167681806961,
+      "loss": 1.8858,
+      "step": 4254
+    },
+    {
+      "epoch": 0.19601971714193578,
+      "grad_norm": 10.08766383503339,
+      "learning_rate": 0.00018589403715206715,
+      "loss": 2.0516,
+      "step": 4255
+    },
+    {
+      "epoch": 0.19606578523057078,
+      "grad_norm": 8.752175810782933,
+      "learning_rate": 0.0001858863955737812,
+      "loss": 2.0485,
+      "step": 4256
+    },
+    {
+      "epoch": 0.19611185331920578,
+      "grad_norm": 5.217374532347924,
+      "learning_rate": 0.00018587875208338195,
+      "loss": 1.8783,
+      "step": 4257
+    },
+    {
+      "epoch": 0.1961579214078408,
+      "grad_norm": 4.425730339777566,
+      "learning_rate": 0.00018587110668103951,
+      "loss": 2.1955,
+      "step": 4258
+    },
+    {
+      "epoch": 0.1962039894964758,
+      "grad_norm": 2.9617121652064027,
+      "learning_rate": 0.00018586345936692414,
+      "loss": 1.7138,
+      "step": 4259
+    },
+    {
+      "epoch": 0.1962500575851108,
+      "grad_norm": 3.6107101408426177,
+      "learning_rate": 0.00018585581014120606,
+      "loss": 1.674,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1962961256737458,
+      "grad_norm": 7.2475792227332585,
+      "learning_rate": 0.00018584815900405554,
+      "loss": 1.6812,
+      "step": 4261
+    },
+    {
+      "epoch": 0.1963421937623808,
+      "grad_norm": 5.612486873475808,
+      "learning_rate": 0.000185840505955643,
+      "loss": 2.0087,
+      "step": 4262
+    },
+    {
+      "epoch": 0.1963882618510158,
+      "grad_norm": 6.056552138783922,
+      "learning_rate": 0.0001858328509961388,
+      "loss": 1.9744,
+      "step": 4263
+    },
+    {
+      "epoch": 0.1964343299396508,
+      "grad_norm": 5.1037286842729515,
+      "learning_rate": 0.00018582519412571328,
+      "loss": 1.6197,
+      "step": 4264
+    },
+    {
+      "epoch": 0.19648039802828582,
+      "grad_norm": 2.9590249790877463,
+      "learning_rate": 0.00018581753534453702,
+      "loss": 1.4898,
+      "step": 4265
+    },
+    {
+      "epoch": 0.19652646611692082,
+      "grad_norm": 3.9013730004564393,
+      "learning_rate": 0.00018580987465278048,
+      "loss": 1.8969,
+      "step": 4266
+    },
+    {
+      "epoch": 0.19657253420555582,
+      "grad_norm": 3.0200253459649984,
+      "learning_rate": 0.0001858022120506142,
+      "loss": 1.8545,
+      "step": 4267
+    },
+    {
+      "epoch": 0.19661860229419081,
+      "grad_norm": 7.142393762385729,
+      "learning_rate": 0.0001857945475382088,
+      "loss": 1.9901,
+      "step": 4268
+    },
+    {
+      "epoch": 0.1966646703828258,
+      "grad_norm": 11.485265290760095,
+      "learning_rate": 0.00018578688111573489,
+      "loss": 1.9868,
+      "step": 4269
+    },
+    {
+      "epoch": 0.1967107384714608,
+      "grad_norm": 4.771124966863752,
+      "learning_rate": 0.00018577921278336318,
+      "loss": 1.679,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1967568065600958,
+      "grad_norm": 9.683738889562854,
+      "learning_rate": 0.00018577154254126438,
+      "loss": 1.8089,
+      "step": 4271
+    },
+    {
+      "epoch": 0.19680287464873084,
+      "grad_norm": 13.799524279919844,
+      "learning_rate": 0.00018576387038960924,
+      "loss": 1.8364,
+      "step": 4272
+    },
+    {
+      "epoch": 0.19684894273736583,
+      "grad_norm": 39.249229026294714,
+      "learning_rate": 0.00018575619632856856,
+      "loss": 1.3893,
+      "step": 4273
+    },
+    {
+      "epoch": 0.19689501082600083,
+      "grad_norm": 4.364182348339946,
+      "learning_rate": 0.00018574852035831323,
+      "loss": 1.773,
+      "step": 4274
+    },
+    {
+      "epoch": 0.19694107891463583,
+      "grad_norm": 3.7786087872248655,
+      "learning_rate": 0.00018574084247901414,
+      "loss": 1.7973,
+      "step": 4275
+    },
+    {
+      "epoch": 0.19698714700327083,
+      "grad_norm": 8.502278000713055,
+      "learning_rate": 0.00018573316269084215,
+      "loss": 1.6642,
+      "step": 4276
+    },
+    {
+      "epoch": 0.19703321509190583,
+      "grad_norm": 6.470392535299701,
+      "learning_rate": 0.00018572548099396834,
+      "loss": 1.8399,
+      "step": 4277
+    },
+    {
+      "epoch": 0.19707928318054083,
+      "grad_norm": 69.04482948042536,
+      "learning_rate": 0.00018571779738856366,
+      "loss": 2.381,
+      "step": 4278
+    },
+    {
+      "epoch": 0.19712535126917585,
+      "grad_norm": 16.60021189224688,
+      "learning_rate": 0.0001857101118747992,
+      "loss": 4.0349,
+      "step": 4279
+    },
+    {
+      "epoch": 0.19717141935781085,
+      "grad_norm": 9.412664391045647,
+      "learning_rate": 0.0001857024244528461,
+      "loss": 2.2926,
+      "step": 4280
+    },
+    {
+      "epoch": 0.19721748744644585,
+      "grad_norm": 51.40130677669818,
+      "learning_rate": 0.00018569473512287542,
+      "loss": 2.2196,
+      "step": 4281
+    },
+    {
+      "epoch": 0.19726355553508085,
+      "grad_norm": 16.184094340494383,
+      "learning_rate": 0.00018568704388505843,
+      "loss": 7.0141,
+      "step": 4282
+    },
+    {
+      "epoch": 0.19730962362371585,
+      "grad_norm": 13.735492179202309,
+      "learning_rate": 0.0001856793507395663,
+      "loss": 4.396,
+      "step": 4283
+    },
+    {
+      "epoch": 0.19735569171235084,
+      "grad_norm": 6.1007648962331515,
+      "learning_rate": 0.00018567165568657035,
+      "loss": 3.1826,
+      "step": 4284
+    },
+    {
+      "epoch": 0.19740175980098587,
+      "grad_norm": 6.844232331003923,
+      "learning_rate": 0.00018566395872624187,
+      "loss": 3.0355,
+      "step": 4285
+    },
+    {
+      "epoch": 0.19744782788962087,
+      "grad_norm": 9.437571668527275,
+      "learning_rate": 0.00018565625985875225,
+      "loss": 3.2491,
+      "step": 4286
+    },
+    {
+      "epoch": 0.19749389597825587,
+      "grad_norm": 8.580694032130635,
+      "learning_rate": 0.00018564855908427286,
+      "loss": 2.6722,
+      "step": 4287
+    },
+    {
+      "epoch": 0.19753996406689087,
+      "grad_norm": 5.799353352908637,
+      "learning_rate": 0.00018564085640297514,
+      "loss": 2.2246,
+      "step": 4288
+    },
+    {
+      "epoch": 0.19758603215552586,
+      "grad_norm": 6.214095634358493,
+      "learning_rate": 0.00018563315181503062,
+      "loss": 2.6162,
+      "step": 4289
+    },
+    {
+      "epoch": 0.19763210024416086,
+      "grad_norm": 10.279268682530772,
+      "learning_rate": 0.00018562544532061082,
+      "loss": 2.786,
+      "step": 4290
+    },
+    {
+      "epoch": 0.19767816833279586,
+      "grad_norm": 3.8234107423921837,
+      "learning_rate": 0.00018561773691988726,
+      "loss": 2.3258,
+      "step": 4291
+    },
+    {
+      "epoch": 0.1977242364214309,
+      "grad_norm": 5.312484737854352,
+      "learning_rate": 0.0001856100266130316,
+      "loss": 2.3957,
+      "step": 4292
+    },
+    {
+      "epoch": 0.1977703045100659,
+      "grad_norm": 6.523083229411337,
+      "learning_rate": 0.00018560231440021551,
+      "loss": 2.308,
+      "step": 4293
+    },
+    {
+      "epoch": 0.19781637259870088,
+      "grad_norm": 8.085397425371315,
+      "learning_rate": 0.00018559460028161066,
+      "loss": 2.431,
+      "step": 4294
+    },
+    {
+      "epoch": 0.19786244068733588,
+      "grad_norm": 4.692344672991792,
+      "learning_rate": 0.00018558688425738878,
+      "loss": 2.4273,
+      "step": 4295
+    },
+    {
+      "epoch": 0.19790850877597088,
+      "grad_norm": 5.895912672936421,
+      "learning_rate": 0.00018557916632772168,
+      "loss": 2.3075,
+      "step": 4296
+    },
+    {
+      "epoch": 0.19795457686460588,
+      "grad_norm": 4.921781895870429,
+      "learning_rate": 0.00018557144649278119,
+      "loss": 1.8689,
+      "step": 4297
+    },
+    {
+      "epoch": 0.19800064495324088,
+      "grad_norm": 7.165487856223376,
+      "learning_rate": 0.00018556372475273916,
+      "loss": 2.3599,
+      "step": 4298
+    },
+    {
+      "epoch": 0.1980467130418759,
+      "grad_norm": 9.356838166795225,
+      "learning_rate": 0.0001855560011077675,
+      "loss": 2.286,
+      "step": 4299
+    },
+    {
+      "epoch": 0.1980927811305109,
+      "grad_norm": 11.263600404742533,
+      "learning_rate": 0.00018554827555803817,
+      "loss": 1.718,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1981388492191459,
+      "grad_norm": 3.698775757269605,
+      "learning_rate": 0.00018554054810372318,
+      "loss": 2.3958,
+      "step": 4301
+    },
+    {
+      "epoch": 0.1981849173077809,
+      "grad_norm": 12.379412425241181,
+      "learning_rate": 0.00018553281874499454,
+      "loss": 2.0593,
+      "step": 4302
+    },
+    {
+      "epoch": 0.1982309853964159,
+      "grad_norm": 6.842581046859203,
+      "learning_rate": 0.00018552508748202436,
+      "loss": 2.1999,
+      "step": 4303
+    },
+    {
+      "epoch": 0.1982770534850509,
+      "grad_norm": 5.539659297484667,
+      "learning_rate": 0.00018551735431498475,
+      "loss": 1.6975,
+      "step": 4304
+    },
+    {
+      "epoch": 0.1983231215736859,
+      "grad_norm": 6.1779008353329194,
+      "learning_rate": 0.0001855096192440479,
+      "loss": 2.3076,
+      "step": 4305
+    },
+    {
+      "epoch": 0.19836918966232092,
+      "grad_norm": 3.114666209332088,
+      "learning_rate": 0.000185501882269386,
+      "loss": 2.3695,
+      "step": 4306
+    },
+    {
+      "epoch": 0.19841525775095592,
+      "grad_norm": 7.504482943994321,
+      "learning_rate": 0.00018549414339117125,
+      "loss": 2.0031,
+      "step": 4307
+    },
+    {
+      "epoch": 0.19846132583959092,
+      "grad_norm": 3.609648660242529,
+      "learning_rate": 0.00018548640260957603,
+      "loss": 1.9242,
+      "step": 4308
+    },
+    {
+      "epoch": 0.19850739392822592,
+      "grad_norm": 6.410727482222023,
+      "learning_rate": 0.0001854786599247726,
+      "loss": 1.5645,
+      "step": 4309
+    },
+    {
+      "epoch": 0.19855346201686092,
+      "grad_norm": 2.6095728445294606,
+      "learning_rate": 0.00018547091533693336,
+      "loss": 1.9719,
+      "step": 4310
+    },
+    {
+      "epoch": 0.19859953010549591,
+      "grad_norm": 7.194305925059048,
+      "learning_rate": 0.00018546316884623078,
+      "loss": 2.2171,
+      "step": 4311
+    },
+    {
+      "epoch": 0.1986455981941309,
+      "grad_norm": 4.928537857550111,
+      "learning_rate": 0.0001854554204528373,
+      "loss": 2.0597,
+      "step": 4312
+    },
+    {
+      "epoch": 0.19869166628276594,
+      "grad_norm": 3.511478261677568,
+      "learning_rate": 0.00018544767015692536,
+      "loss": 2.0365,
+      "step": 4313
+    },
+    {
+      "epoch": 0.19873773437140094,
+      "grad_norm": 4.2208632874039,
+      "learning_rate": 0.00018543991795866755,
+      "loss": 2.0335,
+      "step": 4314
+    },
+    {
+      "epoch": 0.19878380246003594,
+      "grad_norm": 4.705033695267873,
+      "learning_rate": 0.0001854321638582365,
+      "loss": 2.0698,
+      "step": 4315
+    },
+    {
+      "epoch": 0.19882987054867093,
+      "grad_norm": 3.0006531771730605,
+      "learning_rate": 0.0001854244078558048,
+      "loss": 1.9103,
+      "step": 4316
+    },
+    {
+      "epoch": 0.19887593863730593,
+      "grad_norm": 4.996153585346472,
+      "learning_rate": 0.0001854166499515451,
+      "loss": 2.0346,
+      "step": 4317
+    },
+    {
+      "epoch": 0.19892200672594093,
+      "grad_norm": 3.0157680868629946,
+      "learning_rate": 0.0001854088901456302,
+      "loss": 2.1027,
+      "step": 4318
+    },
+    {
+      "epoch": 0.19896807481457596,
+      "grad_norm": 6.135104950941696,
+      "learning_rate": 0.00018540112843823276,
+      "loss": 1.9089,
+      "step": 4319
+    },
+    {
+      "epoch": 0.19901414290321096,
+      "grad_norm": 2.5009735997433786,
+      "learning_rate": 0.00018539336482952562,
+      "loss": 1.868,
+      "step": 4320
+    },
+    {
+      "epoch": 0.19906021099184595,
+      "grad_norm": 2.2949897701624304,
+      "learning_rate": 0.00018538559931968168,
+      "loss": 2.0337,
+      "step": 4321
+    },
+    {
+      "epoch": 0.19910627908048095,
+      "grad_norm": 1.879092580203654,
+      "learning_rate": 0.00018537783190887374,
+      "loss": 1.965,
+      "step": 4322
+    },
+    {
+      "epoch": 0.19915234716911595,
+      "grad_norm": 2.784163448957954,
+      "learning_rate": 0.0001853700625972748,
+      "loss": 1.9238,
+      "step": 4323
+    },
+    {
+      "epoch": 0.19919841525775095,
+      "grad_norm": 2.0832599985928217,
+      "learning_rate": 0.00018536229138505777,
+      "loss": 1.7751,
+      "step": 4324
+    },
+    {
+      "epoch": 0.19924448334638595,
+      "grad_norm": 2.226821396711609,
+      "learning_rate": 0.00018535451827239564,
+      "loss": 1.7003,
+      "step": 4325
+    },
+    {
+      "epoch": 0.19929055143502097,
+      "grad_norm": 1.9663537101689017,
+      "learning_rate": 0.0001853467432594616,
+      "loss": 1.5321,
+      "step": 4326
+    },
+    {
+      "epoch": 0.19933661952365597,
+      "grad_norm": 2.1087680370476325,
+      "learning_rate": 0.0001853389663464286,
+      "loss": 1.9695,
+      "step": 4327
+    },
+    {
+      "epoch": 0.19938268761229097,
+      "grad_norm": 6.424939582880378,
+      "learning_rate": 0.00018533118753346985,
+      "loss": 1.7329,
+      "step": 4328
+    },
+    {
+      "epoch": 0.19942875570092597,
+      "grad_norm": 3.259520351180974,
+      "learning_rate": 0.00018532340682075855,
+      "loss": 1.8968,
+      "step": 4329
+    },
+    {
+      "epoch": 0.19947482378956097,
+      "grad_norm": 6.808967350591412,
+      "learning_rate": 0.00018531562420846787,
+      "loss": 1.4614,
+      "step": 4330
+    },
+    {
+      "epoch": 0.19952089187819597,
+      "grad_norm": 7.141294552904444,
+      "learning_rate": 0.00018530783969677108,
+      "loss": 1.8141,
+      "step": 4331
+    },
+    {
+      "epoch": 0.19956695996683096,
+      "grad_norm": 3.6005327809877747,
+      "learning_rate": 0.00018530005328584154,
+      "loss": 1.6132,
+      "step": 4332
+    },
+    {
+      "epoch": 0.199613028055466,
+      "grad_norm": 5.954833900829777,
+      "learning_rate": 0.00018529226497585258,
+      "loss": 1.9515,
+      "step": 4333
+    },
+    {
+      "epoch": 0.199659096144101,
+      "grad_norm": 5.319423132153789,
+      "learning_rate": 0.00018528447476697757,
+      "loss": 2.1221,
+      "step": 4334
+    },
+    {
+      "epoch": 0.199705164232736,
+      "grad_norm": 2.08355746703568,
+      "learning_rate": 0.00018527668265938996,
+      "loss": 1.8697,
+      "step": 4335
+    },
+    {
+      "epoch": 0.199751232321371,
+      "grad_norm": 2.436449709967507,
+      "learning_rate": 0.00018526888865326324,
+      "loss": 1.7561,
+      "step": 4336
+    },
+    {
+      "epoch": 0.19979730041000598,
+      "grad_norm": 6.346416210475136,
+      "learning_rate": 0.00018526109274877092,
+      "loss": 1.8079,
+      "step": 4337
+    },
+    {
+      "epoch": 0.19984336849864098,
+      "grad_norm": 6.5817007228199005,
+      "learning_rate": 0.00018525329494608654,
+      "loss": 1.6095,
+      "step": 4338
+    },
+    {
+      "epoch": 0.19988943658727598,
+      "grad_norm": 8.987548216622477,
+      "learning_rate": 0.00018524549524538374,
+      "loss": 1.7454,
+      "step": 4339
+    },
+    {
+      "epoch": 0.199935504675911,
+      "grad_norm": 4.404951006135343,
+      "learning_rate": 0.00018523769364683617,
+      "loss": 1.6823,
+      "step": 4340
+    },
+    {
+      "epoch": 0.199981572764546,
+      "grad_norm": 2.0476812815921965,
+      "learning_rate": 0.00018522989015061747,
+      "loss": 1.8827,
+      "step": 4341
+    },
+    {
+      "epoch": 0.200027640853181,
+      "grad_norm": 4.7143471903983665,
+      "learning_rate": 0.00018522208475690142,
+      "loss": 2.1035,
+      "step": 4342
+    },
+    {
+      "epoch": 0.200073708941816,
+      "grad_norm": 2.0707432621807347,
+      "learning_rate": 0.00018521427746586176,
+      "loss": 1.4008,
+      "step": 4343
+    },
+    {
+      "epoch": 0.200119777030451,
+      "grad_norm": 8.534614086471505,
+      "learning_rate": 0.00018520646827767235,
+      "loss": 1.6469,
+      "step": 4344
+    },
+    {
+      "epoch": 0.200165845119086,
+      "grad_norm": 2.9587821568681196,
+      "learning_rate": 0.000185198657192507,
+      "loss": 1.6196,
+      "step": 4345
+    },
+    {
+      "epoch": 0.200211913207721,
+      "grad_norm": 4.97354391577364,
+      "learning_rate": 0.00018519084421053964,
+      "loss": 1.6156,
+      "step": 4346
+    },
+    {
+      "epoch": 0.20025798129635602,
+      "grad_norm": 3.011678246086217,
+      "learning_rate": 0.00018518302933194421,
+      "loss": 1.8393,
+      "step": 4347
+    },
+    {
+      "epoch": 0.20030404938499102,
+      "grad_norm": 5.3272237615052465,
+      "learning_rate": 0.0001851752125568947,
+      "loss": 1.8469,
+      "step": 4348
+    },
+    {
+      "epoch": 0.20035011747362602,
+      "grad_norm": 2.3054566639848786,
+      "learning_rate": 0.0001851673938855651,
+      "loss": 1.7275,
+      "step": 4349
+    },
+    {
+      "epoch": 0.20039618556226102,
+      "grad_norm": 2.0315499874698264,
+      "learning_rate": 0.00018515957331812952,
+      "loss": 1.0892,
+      "step": 4350
+    },
+    {
+      "epoch": 0.20044225365089602,
+      "grad_norm": 21.955518177923103,
+      "learning_rate": 0.00018515175085476206,
+      "loss": 1.809,
+      "step": 4351
+    },
+    {
+      "epoch": 0.20048832173953102,
+      "grad_norm": 5.8349425527809515,
+      "learning_rate": 0.00018514392649563688,
+      "loss": 1.9604,
+      "step": 4352
+    },
+    {
+      "epoch": 0.20053438982816604,
+      "grad_norm": 2.092660161108856,
+      "learning_rate": 0.00018513610024092816,
+      "loss": 1.5505,
+      "step": 4353
+    },
+    {
+      "epoch": 0.20058045791680104,
+      "grad_norm": 2.7417660387433997,
+      "learning_rate": 0.00018512827209081012,
+      "loss": 1.8373,
+      "step": 4354
+    },
+    {
+      "epoch": 0.20062652600543604,
+      "grad_norm": 3.1350985307571873,
+      "learning_rate": 0.00018512044204545709,
+      "loss": 1.7878,
+      "step": 4355
+    },
+    {
+      "epoch": 0.20067259409407104,
+      "grad_norm": 2.597115603827202,
+      "learning_rate": 0.00018511261010504336,
+      "loss": 1.8665,
+      "step": 4356
+    },
+    {
+      "epoch": 0.20071866218270604,
+      "grad_norm": 4.148137691257703,
+      "learning_rate": 0.00018510477626974332,
+      "loss": 2.0083,
+      "step": 4357
+    },
+    {
+      "epoch": 0.20076473027134104,
+      "grad_norm": 6.702561663829496,
+      "learning_rate": 0.00018509694053973132,
+      "loss": 1.6465,
+      "step": 4358
+    },
+    {
+      "epoch": 0.20081079835997603,
+      "grad_norm": 3.230898848281337,
+      "learning_rate": 0.0001850891029151819,
+      "loss": 1.9507,
+      "step": 4359
+    },
+    {
+      "epoch": 0.20085686644861106,
+      "grad_norm": 2.2461129738543857,
+      "learning_rate": 0.00018508126339626948,
+      "loss": 1.5112,
+      "step": 4360
+    },
+    {
+      "epoch": 0.20090293453724606,
+      "grad_norm": 12.550703921351735,
+      "learning_rate": 0.0001850734219831686,
+      "loss": 1.8256,
+      "step": 4361
+    },
+    {
+      "epoch": 0.20094900262588106,
+      "grad_norm": 1.5949224798660713,
+      "learning_rate": 0.00018506557867605384,
+      "loss": 1.78,
+      "step": 4362
+    },
+    {
+      "epoch": 0.20099507071451606,
+      "grad_norm": 3.144141098294865,
+      "learning_rate": 0.00018505773347509986,
+      "loss": 1.6793,
+      "step": 4363
+    },
+    {
+      "epoch": 0.20104113880315105,
+      "grad_norm": 2.3609868059831607,
+      "learning_rate": 0.00018504988638048126,
+      "loss": 1.8999,
+      "step": 4364
+    },
+    {
+      "epoch": 0.20108720689178605,
+      "grad_norm": 2.217755311054897,
+      "learning_rate": 0.0001850420373923728,
+      "loss": 1.8039,
+      "step": 4365
+    },
+    {
+      "epoch": 0.20113327498042105,
+      "grad_norm": 2.197364630544301,
+      "learning_rate": 0.00018503418651094915,
+      "loss": 1.7035,
+      "step": 4366
+    },
+    {
+      "epoch": 0.20117934306905608,
+      "grad_norm": 2.086573205484656,
+      "learning_rate": 0.00018502633373638515,
+      "loss": 1.6556,
+      "step": 4367
+    },
+    {
+      "epoch": 0.20122541115769108,
+      "grad_norm": 2.6133288853849437,
+      "learning_rate": 0.00018501847906885567,
+      "loss": 1.6363,
+      "step": 4368
+    },
+    {
+      "epoch": 0.20127147924632607,
+      "grad_norm": 4.935488721670298,
+      "learning_rate": 0.00018501062250853546,
+      "loss": 1.8927,
+      "step": 4369
+    },
+    {
+      "epoch": 0.20131754733496107,
+      "grad_norm": 1.5331385553108074,
+      "learning_rate": 0.00018500276405559955,
+      "loss": 1.8114,
+      "step": 4370
+    },
+    {
+      "epoch": 0.20136361542359607,
+      "grad_norm": 5.855495770117098,
+      "learning_rate": 0.00018499490371022282,
+      "loss": 1.8422,
+      "step": 4371
+    },
+    {
+      "epoch": 0.20140968351223107,
+      "grad_norm": 3.3405493243193227,
+      "learning_rate": 0.00018498704147258032,
+      "loss": 1.6482,
+      "step": 4372
+    },
+    {
+      "epoch": 0.20145575160086607,
+      "grad_norm": 6.474969885787425,
+      "learning_rate": 0.00018497917734284705,
+      "loss": 1.9488,
+      "step": 4373
+    },
+    {
+      "epoch": 0.2015018196895011,
+      "grad_norm": 12.729222628994393,
+      "learning_rate": 0.00018497131132119812,
+      "loss": 1.4762,
+      "step": 4374
+    },
+    {
+      "epoch": 0.2015478877781361,
+      "grad_norm": 3.0917290255985126,
+      "learning_rate": 0.0001849634434078086,
+      "loss": 2.0538,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2015939558667711,
+      "grad_norm": 5.578426068370878,
+      "learning_rate": 0.00018495557360285369,
+      "loss": 1.669,
+      "step": 4376
+    },
+    {
+      "epoch": 0.2016400239554061,
+      "grad_norm": 2.7139493070335994,
+      "learning_rate": 0.00018494770190650864,
+      "loss": 1.9678,
+      "step": 4377
+    },
+    {
+      "epoch": 0.2016860920440411,
+      "grad_norm": 3.2623716004981045,
+      "learning_rate": 0.00018493982831894865,
+      "loss": 1.6964,
+      "step": 4378
+    },
+    {
+      "epoch": 0.20173216013267609,
+      "grad_norm": 2.905302671319306,
+      "learning_rate": 0.00018493195284034902,
+      "loss": 1.7289,
+      "step": 4379
+    },
+    {
+      "epoch": 0.20177822822131108,
+      "grad_norm": 3.773124099050174,
+      "learning_rate": 0.0001849240754708851,
+      "loss": 1.5568,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2018242963099461,
+      "grad_norm": 3.0278950788691614,
+      "learning_rate": 0.0001849161962107322,
+      "loss": 1.9053,
+      "step": 4381
+    },
+    {
+      "epoch": 0.2018703643985811,
+      "grad_norm": 7.207821697520899,
+      "learning_rate": 0.00018490831506006586,
+      "loss": 1.6575,
+      "step": 4382
+    },
+    {
+      "epoch": 0.2019164324872161,
+      "grad_norm": 3.427813656229064,
+      "learning_rate": 0.00018490043201906144,
+      "loss": 1.6367,
+      "step": 4383
+    },
+    {
+      "epoch": 0.2019625005758511,
+      "grad_norm": 3.3263343830084984,
+      "learning_rate": 0.00018489254708789447,
+      "loss": 1.5328,
+      "step": 4384
+    },
+    {
+      "epoch": 0.2020085686644861,
+      "grad_norm": 6.08011769595954,
+      "learning_rate": 0.0001848846602667405,
+      "loss": 1.6348,
+      "step": 4385
+    },
+    {
+      "epoch": 0.2020546367531211,
+      "grad_norm": 4.373051716440591,
+      "learning_rate": 0.00018487677155577513,
+      "loss": 1.6899,
+      "step": 4386
+    },
+    {
+      "epoch": 0.20210070484175613,
+      "grad_norm": 2.748923197338329,
+      "learning_rate": 0.00018486888095517397,
+      "loss": 1.759,
+      "step": 4387
+    },
+    {
+      "epoch": 0.20214677293039113,
+      "grad_norm": 5.001853322787805,
+      "learning_rate": 0.00018486098846511267,
+      "loss": 1.7964,
+      "step": 4388
+    },
+    {
+      "epoch": 0.20219284101902613,
+      "grad_norm": 4.777834314057859,
+      "learning_rate": 0.000184853094085767,
+      "loss": 1.9877,
+      "step": 4389
+    },
+    {
+      "epoch": 0.20223890910766112,
+      "grad_norm": 3.6802536137645516,
+      "learning_rate": 0.00018484519781731263,
+      "loss": 1.831,
+      "step": 4390
+    },
+    {
+      "epoch": 0.20228497719629612,
+      "grad_norm": 3.144331366571087,
+      "learning_rate": 0.00018483729965992546,
+      "loss": 1.9842,
+      "step": 4391
+    },
+    {
+      "epoch": 0.20233104528493112,
+      "grad_norm": 3.356027208000785,
+      "learning_rate": 0.00018482939961378126,
+      "loss": 1.3756,
+      "step": 4392
+    },
+    {
+      "epoch": 0.20237711337356612,
+      "grad_norm": 2.690452087915247,
+      "learning_rate": 0.00018482149767905595,
+      "loss": 1.6581,
+      "step": 4393
+    },
+    {
+      "epoch": 0.20242318146220115,
+      "grad_norm": 3.5890875965038034,
+      "learning_rate": 0.0001848135938559254,
+      "loss": 1.2688,
+      "step": 4394
+    },
+    {
+      "epoch": 0.20246924955083614,
+      "grad_norm": 7.262406173971967,
+      "learning_rate": 0.00018480568814456563,
+      "loss": 1.7957,
+      "step": 4395
+    },
+    {
+      "epoch": 0.20251531763947114,
+      "grad_norm": 5.42769597467853,
+      "learning_rate": 0.00018479778054515263,
+      "loss": 1.7559,
+      "step": 4396
+    },
+    {
+      "epoch": 0.20256138572810614,
+      "grad_norm": 3.2335478356130913,
+      "learning_rate": 0.00018478987105786243,
+      "loss": 1.8933,
+      "step": 4397
+    },
+    {
+      "epoch": 0.20260745381674114,
+      "grad_norm": 3.398667084882407,
+      "learning_rate": 0.00018478195968287114,
+      "loss": 1.6072,
+      "step": 4398
+    },
+    {
+      "epoch": 0.20265352190537614,
+      "grad_norm": 1.799830628227152,
+      "learning_rate": 0.00018477404642035485,
+      "loss": 1.6976,
+      "step": 4399
+    },
+    {
+      "epoch": 0.20269958999401114,
+      "grad_norm": 3.0062097281650906,
+      "learning_rate": 0.0001847661312704898,
+      "loss": 1.5738,
+      "step": 4400
+    },
+    {
+      "epoch": 0.20274565808264616,
+      "grad_norm": 2.1485605879175913,
+      "learning_rate": 0.0001847582142334522,
+      "loss": 1.9561,
+      "step": 4401
+    },
+    {
+      "epoch": 0.20279172617128116,
+      "grad_norm": 4.421816805294009,
+      "learning_rate": 0.00018475029530941827,
+      "loss": 1.5276,
+      "step": 4402
+    },
+    {
+      "epoch": 0.20283779425991616,
+      "grad_norm": 2.11321799072354,
+      "learning_rate": 0.00018474237449856432,
+      "loss": 1.4948,
+      "step": 4403
+    },
+    {
+      "epoch": 0.20288386234855116,
+      "grad_norm": 3.468079564782851,
+      "learning_rate": 0.00018473445180106676,
+      "loss": 1.7232,
+      "step": 4404
+    },
+    {
+      "epoch": 0.20292993043718616,
+      "grad_norm": 6.79620540709792,
+      "learning_rate": 0.00018472652721710184,
+      "loss": 1.3933,
+      "step": 4405
+    },
+    {
+      "epoch": 0.20297599852582116,
+      "grad_norm": 4.37132737393011,
+      "learning_rate": 0.00018471860074684612,
+      "loss": 1.9288,
+      "step": 4406
+    },
+    {
+      "epoch": 0.20302206661445615,
+      "grad_norm": 8.633936706724484,
+      "learning_rate": 0.000184710672390476,
+      "loss": 1.6782,
+      "step": 4407
+    },
+    {
+      "epoch": 0.20306813470309118,
+      "grad_norm": 3.138097246526785,
+      "learning_rate": 0.000184702742148168,
+      "loss": 1.8924,
+      "step": 4408
+    },
+    {
+      "epoch": 0.20311420279172618,
+      "grad_norm": 5.419433849893636,
+      "learning_rate": 0.00018469481002009866,
+      "loss": 1.5614,
+      "step": 4409
+    },
+    {
+      "epoch": 0.20316027088036118,
+      "grad_norm": 3.0033462460261156,
+      "learning_rate": 0.00018468687600644464,
+      "loss": 1.5428,
+      "step": 4410
+    },
+    {
+      "epoch": 0.20320633896899618,
+      "grad_norm": 2.890469685303289,
+      "learning_rate": 0.0001846789401073825,
+      "loss": 1.7774,
+      "step": 4411
+    },
+    {
+      "epoch": 0.20325240705763117,
+      "grad_norm": 5.185257962808281,
+      "learning_rate": 0.00018467100232308895,
+      "loss": 1.5803,
+      "step": 4412
+    },
+    {
+      "epoch": 0.20329847514626617,
+      "grad_norm": 2.336765511819933,
+      "learning_rate": 0.00018466306265374076,
+      "loss": 1.8425,
+      "step": 4413
+    },
+    {
+      "epoch": 0.20334454323490117,
+      "grad_norm": 2.7565042513386673,
+      "learning_rate": 0.00018465512109951463,
+      "loss": 1.5278,
+      "step": 4414
+    },
+    {
+      "epoch": 0.2033906113235362,
+      "grad_norm": 5.665306542207392,
+      "learning_rate": 0.00018464717766058733,
+      "loss": 1.5648,
+      "step": 4415
+    },
+    {
+      "epoch": 0.2034366794121712,
+      "grad_norm": 2.6447715968276424,
+      "learning_rate": 0.0001846392323371358,
+      "loss": 1.732,
+      "step": 4416
+    },
+    {
+      "epoch": 0.2034827475008062,
+      "grad_norm": 4.114569163046179,
+      "learning_rate": 0.0001846312851293369,
+      "loss": 1.7636,
+      "step": 4417
+    },
+    {
+      "epoch": 0.2035288155894412,
+      "grad_norm": 2.61238910429187,
+      "learning_rate": 0.00018462333603736753,
+      "loss": 1.6532,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2035748836780762,
+      "grad_norm": 3.477096967413551,
+      "learning_rate": 0.00018461538506140464,
+      "loss": 1.7976,
+      "step": 4419
+    },
+    {
+      "epoch": 0.2036209517667112,
+      "grad_norm": 2.6543878600653916,
+      "learning_rate": 0.00018460743220162536,
+      "loss": 1.8218,
+      "step": 4420
+    },
+    {
+      "epoch": 0.20366701985534622,
+      "grad_norm": 3.3917456643489916,
+      "learning_rate": 0.00018459947745820662,
+      "loss": 1.915,
+      "step": 4421
+    },
+    {
+      "epoch": 0.2037130879439812,
+      "grad_norm": 3.32540691179705,
+      "learning_rate": 0.00018459152083132559,
+      "loss": 1.6399,
+      "step": 4422
+    },
+    {
+      "epoch": 0.2037591560326162,
+      "grad_norm": 3.883905805164181,
+      "learning_rate": 0.00018458356232115938,
+      "loss": 1.4559,
+      "step": 4423
+    },
+    {
+      "epoch": 0.2038052241212512,
+      "grad_norm": 2.598565493849038,
+      "learning_rate": 0.00018457560192788522,
+      "loss": 1.939,
+      "step": 4424
+    },
+    {
+      "epoch": 0.2038512922098862,
+      "grad_norm": 1.442272291746429,
+      "learning_rate": 0.00018456763965168027,
+      "loss": 1.8552,
+      "step": 4425
+    },
+    {
+      "epoch": 0.2038973602985212,
+      "grad_norm": 2.325221425720077,
+      "learning_rate": 0.0001845596754927218,
+      "loss": 1.5532,
+      "step": 4426
+    },
+    {
+      "epoch": 0.2039434283871562,
+      "grad_norm": 2.4827041034418365,
+      "learning_rate": 0.0001845517094511872,
+      "loss": 1.5631,
+      "step": 4427
+    },
+    {
+      "epoch": 0.20398949647579123,
+      "grad_norm": 6.794014291176859,
+      "learning_rate": 0.00018454374152725373,
+      "loss": 1.2062,
+      "step": 4428
+    },
+    {
+      "epoch": 0.20403556456442623,
+      "grad_norm": 1.6376381555585013,
+      "learning_rate": 0.00018453577172109883,
+      "loss": 1.7999,
+      "step": 4429
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 8.755511354162685,
+      "learning_rate": 0.0001845278000328999,
+      "loss": 1.5579,
+      "step": 4430
+    },
+    {
+      "epoch": 0.20412770074169623,
+      "grad_norm": 1.5707385981897126,
+      "learning_rate": 0.0001845198264628345,
+      "loss": 1.6111,
+      "step": 4431
+    },
+    {
+      "epoch": 0.20417376883033123,
+      "grad_norm": 1.595183458957735,
+      "learning_rate": 0.00018451185101108001,
+      "loss": 1.5844,
+      "step": 4432
+    },
+    {
+      "epoch": 0.20421983691896622,
+      "grad_norm": 2.1559199797152306,
+      "learning_rate": 0.0001845038736778141,
+      "loss": 1.5277,
+      "step": 4433
+    },
+    {
+      "epoch": 0.20426590500760122,
+      "grad_norm": 1.516254302805495,
+      "learning_rate": 0.0001844958944632143,
+      "loss": 1.6735,
+      "step": 4434
+    },
+    {
+      "epoch": 0.20431197309623625,
+      "grad_norm": 1.8532779173485285,
+      "learning_rate": 0.00018448791336745832,
+      "loss": 1.8152,
+      "step": 4435
+    },
+    {
+      "epoch": 0.20435804118487125,
+      "grad_norm": 2.575443852993282,
+      "learning_rate": 0.0001844799303907238,
+      "loss": 1.6937,
+      "step": 4436
+    },
+    {
+      "epoch": 0.20440410927350625,
+      "grad_norm": 2.0673596572924433,
+      "learning_rate": 0.00018447194553318852,
+      "loss": 1.6485,
+      "step": 4437
+    },
+    {
+      "epoch": 0.20445017736214124,
+      "grad_norm": 3.4839126446748296,
+      "learning_rate": 0.00018446395879503019,
+      "loss": 1.7716,
+      "step": 4438
+    },
+    {
+      "epoch": 0.20449624545077624,
+      "grad_norm": 3.248826442661443,
+      "learning_rate": 0.00018445597017642662,
+      "loss": 1.7785,
+      "step": 4439
+    },
+    {
+      "epoch": 0.20454231353941124,
+      "grad_norm": 2.1704689721615265,
+      "learning_rate": 0.0001844479796775557,
+      "loss": 1.6657,
+      "step": 4440
+    },
+    {
+      "epoch": 0.20458838162804624,
+      "grad_norm": 1.806010461405363,
+      "learning_rate": 0.0001844399872985953,
+      "loss": 1.7019,
+      "step": 4441
+    },
+    {
+      "epoch": 0.20463444971668127,
+      "grad_norm": 2.9175500723817933,
+      "learning_rate": 0.00018443199303972338,
+      "loss": 1.717,
+      "step": 4442
+    },
+    {
+      "epoch": 0.20468051780531626,
+      "grad_norm": 2.215476334214772,
+      "learning_rate": 0.00018442399690111788,
+      "loss": 1.3864,
+      "step": 4443
+    },
+    {
+      "epoch": 0.20472658589395126,
+      "grad_norm": 2.1650459519852867,
+      "learning_rate": 0.00018441599888295687,
+      "loss": 1.612,
+      "step": 4444
+    },
+    {
+      "epoch": 0.20477265398258626,
+      "grad_norm": 2.195053740580744,
+      "learning_rate": 0.00018440799898541838,
+      "loss": 1.4637,
+      "step": 4445
+    },
+    {
+      "epoch": 0.20481872207122126,
+      "grad_norm": 3.2205415437291784,
+      "learning_rate": 0.00018439999720868056,
+      "loss": 1.8248,
+      "step": 4446
+    },
+    {
+      "epoch": 0.20486479015985626,
+      "grad_norm": 4.088545049010775,
+      "learning_rate": 0.00018439199355292147,
+      "loss": 1.4735,
+      "step": 4447
+    },
+    {
+      "epoch": 0.20491085824849126,
+      "grad_norm": 5.346779251005505,
+      "learning_rate": 0.00018438398801831935,
+      "loss": 1.3014,
+      "step": 4448
+    },
+    {
+      "epoch": 0.20495692633712628,
+      "grad_norm": 3.8531995370030705,
+      "learning_rate": 0.00018437598060505242,
+      "loss": 1.5031,
+      "step": 4449
+    },
+    {
+      "epoch": 0.20500299442576128,
+      "grad_norm": 1.5295066218154483,
+      "learning_rate": 0.00018436797131329898,
+      "loss": 1.648,
+      "step": 4450
+    },
+    {
+      "epoch": 0.20504906251439628,
+      "grad_norm": 2.6800878641513277,
+      "learning_rate": 0.0001843599601432373,
+      "loss": 1.3606,
+      "step": 4451
+    },
+    {
+      "epoch": 0.20509513060303128,
+      "grad_norm": 1.417849264914954,
+      "learning_rate": 0.00018435194709504575,
+      "loss": 1.7321,
+      "step": 4452
+    },
+    {
+      "epoch": 0.20514119869166628,
+      "grad_norm": 3.198213518300264,
+      "learning_rate": 0.00018434393216890276,
+      "loss": 1.7152,
+      "step": 4453
+    },
+    {
+      "epoch": 0.20518726678030128,
+      "grad_norm": 2.0135559970927925,
+      "learning_rate": 0.0001843359153649867,
+      "loss": 1.3786,
+      "step": 4454
+    },
+    {
+      "epoch": 0.2052333348689363,
+      "grad_norm": 1.748710702903713,
+      "learning_rate": 0.00018432789668347609,
+      "loss": 1.5044,
+      "step": 4455
+    },
+    {
+      "epoch": 0.2052794029575713,
+      "grad_norm": 3.6554439679008603,
+      "learning_rate": 0.0001843198761245495,
+      "loss": 1.4133,
+      "step": 4456
+    },
+    {
+      "epoch": 0.2053254710462063,
+      "grad_norm": 44.540948950416166,
+      "learning_rate": 0.00018431185368838538,
+      "loss": 1.2815,
+      "step": 4457
+    },
+    {
+      "epoch": 0.2053715391348413,
+      "grad_norm": 8.265879494489672,
+      "learning_rate": 0.00018430382937516246,
+      "loss": 1.5885,
+      "step": 4458
+    },
+    {
+      "epoch": 0.2054176072234763,
+      "grad_norm": 4.0770241491044334,
+      "learning_rate": 0.0001842958031850593,
+      "loss": 1.7231,
+      "step": 4459
+    },
+    {
+      "epoch": 0.2054636753121113,
+      "grad_norm": 5.055172213659204,
+      "learning_rate": 0.00018428777511825462,
+      "loss": 1.653,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2055097434007463,
+      "grad_norm": 2.032355419801078,
+      "learning_rate": 0.00018427974517492716,
+      "loss": 1.6827,
+      "step": 4461
+    },
+    {
+      "epoch": 0.20555581148938132,
+      "grad_norm": 10.743649280688746,
+      "learning_rate": 0.00018427171335525568,
+      "loss": 1.9275,
+      "step": 4462
+    },
+    {
+      "epoch": 0.20560187957801632,
+      "grad_norm": 5.5622596671893,
+      "learning_rate": 0.000184263679659419,
+      "loss": 2.0073,
+      "step": 4463
+    },
+    {
+      "epoch": 0.20564794766665132,
+      "grad_norm": 3.3503937983920506,
+      "learning_rate": 0.00018425564408759595,
+      "loss": 1.6544,
+      "step": 4464
+    },
+    {
+      "epoch": 0.2056940157552863,
+      "grad_norm": 2.6540408792890804,
+      "learning_rate": 0.0001842476066399655,
+      "loss": 1.7226,
+      "step": 4465
+    },
+    {
+      "epoch": 0.2057400838439213,
+      "grad_norm": 2.6969682762544678,
+      "learning_rate": 0.0001842395673167065,
+      "loss": 2.0734,
+      "step": 4466
+    },
+    {
+      "epoch": 0.2057861519325563,
+      "grad_norm": 2.604270232062385,
+      "learning_rate": 0.00018423152611799797,
+      "loss": 1.6781,
+      "step": 4467
+    },
+    {
+      "epoch": 0.2058322200211913,
+      "grad_norm": 1.5111102041462239,
+      "learning_rate": 0.00018422348304401894,
+      "loss": 1.7822,
+      "step": 4468
+    },
+    {
+      "epoch": 0.20587828810982634,
+      "grad_norm": 5.973204673520907,
+      "learning_rate": 0.00018421543809494847,
+      "loss": 1.5601,
+      "step": 4469
+    },
+    {
+      "epoch": 0.20592435619846133,
+      "grad_norm": 1.9499753070781767,
+      "learning_rate": 0.0001842073912709657,
+      "loss": 1.6152,
+      "step": 4470
+    },
+    {
+      "epoch": 0.20597042428709633,
+      "grad_norm": 3.0745959650659707,
+      "learning_rate": 0.00018419934257224975,
+      "loss": 1.9089,
+      "step": 4471
+    },
+    {
+      "epoch": 0.20601649237573133,
+      "grad_norm": 3.2251722331769646,
+      "learning_rate": 0.00018419129199897976,
+      "loss": 1.7298,
+      "step": 4472
+    },
+    {
+      "epoch": 0.20606256046436633,
+      "grad_norm": 4.052715009975536,
+      "learning_rate": 0.00018418323955133505,
+      "loss": 1.6353,
+      "step": 4473
+    },
+    {
+      "epoch": 0.20610862855300133,
+      "grad_norm": 2.1100998438992895,
+      "learning_rate": 0.00018417518522949485,
+      "loss": 1.484,
+      "step": 4474
+    },
+    {
+      "epoch": 0.20615469664163633,
+      "grad_norm": 1.5221620321756733,
+      "learning_rate": 0.00018416712903363846,
+      "loss": 1.5572,
+      "step": 4475
+    },
+    {
+      "epoch": 0.20620076473027135,
+      "grad_norm": 6.939104929481654,
+      "learning_rate": 0.00018415907096394526,
+      "loss": 1.4054,
+      "step": 4476
+    },
+    {
+      "epoch": 0.20624683281890635,
+      "grad_norm": 2.8135055053716953,
+      "learning_rate": 0.00018415101102059467,
+      "loss": 1.5396,
+      "step": 4477
+    },
+    {
+      "epoch": 0.20629290090754135,
+      "grad_norm": 1.8252734606243761,
+      "learning_rate": 0.0001841429492037661,
+      "loss": 1.5895,
+      "step": 4478
+    },
+    {
+      "epoch": 0.20633896899617635,
+      "grad_norm": 1.6500626293242187,
+      "learning_rate": 0.00018413488551363905,
+      "loss": 1.8315,
+      "step": 4479
+    },
+    {
+      "epoch": 0.20638503708481135,
+      "grad_norm": 7.098043143289806,
+      "learning_rate": 0.00018412681995039302,
+      "loss": 1.877,
+      "step": 4480
+    },
+    {
+      "epoch": 0.20643110517344634,
+      "grad_norm": 6.588308680577403,
+      "learning_rate": 0.0001841187525142076,
+      "loss": 1.6715,
+      "step": 4481
+    },
+    {
+      "epoch": 0.20647717326208134,
+      "grad_norm": 3.094932851103813,
+      "learning_rate": 0.00018411068320526238,
+      "loss": 1.6417,
+      "step": 4482
+    },
+    {
+      "epoch": 0.20652324135071637,
+      "grad_norm": 2.851862133184747,
+      "learning_rate": 0.000184102612023737,
+      "loss": 1.7372,
+      "step": 4483
+    },
+    {
+      "epoch": 0.20656930943935137,
+      "grad_norm": 2.0142001060804446,
+      "learning_rate": 0.00018409453896981118,
+      "loss": 1.6722,
+      "step": 4484
+    },
+    {
+      "epoch": 0.20661537752798637,
+      "grad_norm": 2.5990471638416537,
+      "learning_rate": 0.00018408646404366469,
+      "loss": 1.6747,
+      "step": 4485
+    },
+    {
+      "epoch": 0.20666144561662136,
+      "grad_norm": 4.3686433457061264,
+      "learning_rate": 0.00018407838724547718,
+      "loss": 1.7447,
+      "step": 4486
+    },
+    {
+      "epoch": 0.20670751370525636,
+      "grad_norm": 5.80148600886325,
+      "learning_rate": 0.00018407030857542857,
+      "loss": 1.6037,
+      "step": 4487
+    },
+    {
+      "epoch": 0.20675358179389136,
+      "grad_norm": 2.742337404029612,
+      "learning_rate": 0.00018406222803369867,
+      "loss": 1.6564,
+      "step": 4488
+    },
+    {
+      "epoch": 0.2067996498825264,
+      "grad_norm": 3.689184414547432,
+      "learning_rate": 0.00018405414562046746,
+      "loss": 1.5313,
+      "step": 4489
+    },
+    {
+      "epoch": 0.20684571797116139,
+      "grad_norm": 2.591475127728515,
+      "learning_rate": 0.00018404606133591477,
+      "loss": 1.6671,
+      "step": 4490
+    },
+    {
+      "epoch": 0.20689178605979638,
+      "grad_norm": 2.2775045284968813,
+      "learning_rate": 0.00018403797518022063,
+      "loss": 1.295,
+      "step": 4491
+    },
+    {
+      "epoch": 0.20693785414843138,
+      "grad_norm": 2.9049772131559797,
+      "learning_rate": 0.0001840298871535651,
+      "loss": 1.6568,
+      "step": 4492
+    },
+    {
+      "epoch": 0.20698392223706638,
+      "grad_norm": 4.749700966575816,
+      "learning_rate": 0.00018402179725612817,
+      "loss": 1.8279,
+      "step": 4493
+    },
+    {
+      "epoch": 0.20702999032570138,
+      "grad_norm": 20.51555059888271,
+      "learning_rate": 0.00018401370548809003,
+      "loss": 1.4189,
+      "step": 4494
+    },
+    {
+      "epoch": 0.20707605841433638,
+      "grad_norm": 3.1104433218548997,
+      "learning_rate": 0.00018400561184963077,
+      "loss": 1.6155,
+      "step": 4495
+    },
+    {
+      "epoch": 0.2071221265029714,
+      "grad_norm": 3.8410882671069455,
+      "learning_rate": 0.0001839975163409306,
+      "loss": 1.869,
+      "step": 4496
+    },
+    {
+      "epoch": 0.2071681945916064,
+      "grad_norm": 5.132444784006931,
+      "learning_rate": 0.00018398941896216977,
+      "loss": 1.7115,
+      "step": 4497
+    },
+    {
+      "epoch": 0.2072142626802414,
+      "grad_norm": 8.902886103908621,
+      "learning_rate": 0.00018398131971352852,
+      "loss": 1.1266,
+      "step": 4498
+    },
+    {
+      "epoch": 0.2072603307688764,
+      "grad_norm": 3.0967382715992113,
+      "learning_rate": 0.00018397321859518718,
+      "loss": 1.749,
+      "step": 4499
+    },
+    {
+      "epoch": 0.2073063988575114,
+      "grad_norm": 6.758695098820963,
+      "learning_rate": 0.00018396511560732614,
+      "loss": 2.0679,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2073524669461464,
+      "grad_norm": 4.465795410537297,
+      "learning_rate": 0.00018395701075012575,
+      "loss": 1.6821,
+      "step": 4501
+    },
+    {
+      "epoch": 0.2073985350347814,
+      "grad_norm": 3.129895027638634,
+      "learning_rate": 0.00018394890402376648,
+      "loss": 1.4746,
+      "step": 4502
+    },
+    {
+      "epoch": 0.20744460312341642,
+      "grad_norm": 2.455643747218541,
+      "learning_rate": 0.00018394079542842878,
+      "loss": 1.5238,
+      "step": 4503
+    },
+    {
+      "epoch": 0.20749067121205142,
+      "grad_norm": 4.100352722000894,
+      "learning_rate": 0.0001839326849642932,
+      "loss": 1.7125,
+      "step": 4504
+    },
+    {
+      "epoch": 0.20753673930068642,
+      "grad_norm": 2.4448334385722545,
+      "learning_rate": 0.00018392457263154033,
+      "loss": 1.8986,
+      "step": 4505
+    },
+    {
+      "epoch": 0.20758280738932142,
+      "grad_norm": 3.1861700994007403,
+      "learning_rate": 0.00018391645843035073,
+      "loss": 1.6157,
+      "step": 4506
+    },
+    {
+      "epoch": 0.20762887547795641,
+      "grad_norm": 2.7501261466121605,
+      "learning_rate": 0.00018390834236090508,
+      "loss": 1.6667,
+      "step": 4507
+    },
+    {
+      "epoch": 0.2076749435665914,
+      "grad_norm": 2.164300512033418,
+      "learning_rate": 0.00018390022442338406,
+      "loss": 1.4722,
+      "step": 4508
+    },
+    {
+      "epoch": 0.2077210116552264,
+      "grad_norm": 2.035170970545087,
+      "learning_rate": 0.0001838921046179684,
+      "loss": 1.5711,
+      "step": 4509
+    },
+    {
+      "epoch": 0.20776707974386144,
+      "grad_norm": 3.8524129760480723,
+      "learning_rate": 0.00018388398294483886,
+      "loss": 1.5612,
+      "step": 4510
+    },
+    {
+      "epoch": 0.20781314783249644,
+      "grad_norm": 7.609935839408824,
+      "learning_rate": 0.00018387585940417628,
+      "loss": 1.8781,
+      "step": 4511
+    },
+    {
+      "epoch": 0.20785921592113143,
+      "grad_norm": 3.7365143350070555,
+      "learning_rate": 0.0001838677339961615,
+      "loss": 1.7713,
+      "step": 4512
+    },
+    {
+      "epoch": 0.20790528400976643,
+      "grad_norm": 5.730278235661422,
+      "learning_rate": 0.00018385960672097544,
+      "loss": 1.813,
+      "step": 4513
+    },
+    {
+      "epoch": 0.20795135209840143,
+      "grad_norm": 3.3708510259766715,
+      "learning_rate": 0.000183851477578799,
+      "loss": 1.6606,
+      "step": 4514
+    },
+    {
+      "epoch": 0.20799742018703643,
+      "grad_norm": 2.4395190605378794,
+      "learning_rate": 0.00018384334656981323,
+      "loss": 1.6043,
+      "step": 4515
+    },
+    {
+      "epoch": 0.20804348827567143,
+      "grad_norm": 1.7487783562902963,
+      "learning_rate": 0.00018383521369419906,
+      "loss": 1.4103,
+      "step": 4516
+    },
+    {
+      "epoch": 0.20808955636430645,
+      "grad_norm": 4.5850354351465725,
+      "learning_rate": 0.00018382707895213765,
+      "loss": 1.8594,
+      "step": 4517
+    },
+    {
+      "epoch": 0.20813562445294145,
+      "grad_norm": 5.053485618324785,
+      "learning_rate": 0.00018381894234381003,
+      "loss": 1.6807,
+      "step": 4518
+    },
+    {
+      "epoch": 0.20818169254157645,
+      "grad_norm": 2.0562011402892924,
+      "learning_rate": 0.00018381080386939736,
+      "loss": 1.5757,
+      "step": 4519
+    },
+    {
+      "epoch": 0.20822776063021145,
+      "grad_norm": 4.535424501400465,
+      "learning_rate": 0.00018380266352908088,
+      "loss": 1.6746,
+      "step": 4520
+    },
+    {
+      "epoch": 0.20827382871884645,
+      "grad_norm": 8.497791366009732,
+      "learning_rate": 0.0001837945213230418,
+      "loss": 1.4194,
+      "step": 4521
+    },
+    {
+      "epoch": 0.20831989680748145,
+      "grad_norm": 8.131390203288348,
+      "learning_rate": 0.00018378637725146132,
+      "loss": 1.6239,
+      "step": 4522
+    },
+    {
+      "epoch": 0.20836596489611647,
+      "grad_norm": 1.7279505260011754,
+      "learning_rate": 0.00018377823131452087,
+      "loss": 1.8491,
+      "step": 4523
+    },
+    {
+      "epoch": 0.20841203298475147,
+      "grad_norm": 3.3383846479821107,
+      "learning_rate": 0.00018377008351240173,
+      "loss": 1.9991,
+      "step": 4524
+    },
+    {
+      "epoch": 0.20845810107338647,
+      "grad_norm": 3.866125945782958,
+      "learning_rate": 0.0001837619338452853,
+      "loss": 1.8516,
+      "step": 4525
+    },
+    {
+      "epoch": 0.20850416916202147,
+      "grad_norm": 2.724087157811203,
+      "learning_rate": 0.00018375378231335306,
+      "loss": 1.9051,
+      "step": 4526
+    },
+    {
+      "epoch": 0.20855023725065647,
+      "grad_norm": 4.591558037002743,
+      "learning_rate": 0.00018374562891678643,
+      "loss": 1.7812,
+      "step": 4527
+    },
+    {
+      "epoch": 0.20859630533929147,
+      "grad_norm": 3.383754502329868,
+      "learning_rate": 0.000183737473655767,
+      "loss": 1.3394,
+      "step": 4528
+    },
+    {
+      "epoch": 0.20864237342792646,
+      "grad_norm": 2.2845245585131866,
+      "learning_rate": 0.0001837293165304763,
+      "loss": 1.5942,
+      "step": 4529
+    },
+    {
+      "epoch": 0.2086884415165615,
+      "grad_norm": 20.05141458756271,
+      "learning_rate": 0.00018372115754109592,
+      "loss": 1.5842,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2087345096051965,
+      "grad_norm": 3.3056686042245658,
+      "learning_rate": 0.0001837129966878075,
+      "loss": 1.7898,
+      "step": 4531
+    },
+    {
+      "epoch": 0.2087805776938315,
+      "grad_norm": 3.765326683733287,
+      "learning_rate": 0.00018370483397079273,
+      "loss": 1.5333,
+      "step": 4532
+    },
+    {
+      "epoch": 0.20882664578246649,
+      "grad_norm": 4.772315929621375,
+      "learning_rate": 0.0001836966693902334,
+      "loss": 1.8909,
+      "step": 4533
+    },
+    {
+      "epoch": 0.20887271387110148,
+      "grad_norm": 3.3363507025844865,
+      "learning_rate": 0.00018368850294631125,
+      "loss": 1.8305,
+      "step": 4534
+    },
+    {
+      "epoch": 0.20891878195973648,
+      "grad_norm": 17.090183802748918,
+      "learning_rate": 0.00018368033463920803,
+      "loss": 1.8969,
+      "step": 4535
+    },
+    {
+      "epoch": 0.20896485004837148,
+      "grad_norm": 2.251994686957423,
+      "learning_rate": 0.00018367216446910567,
+      "loss": 1.3338,
+      "step": 4536
+    },
+    {
+      "epoch": 0.2090109181370065,
+      "grad_norm": 2.198652728510019,
+      "learning_rate": 0.00018366399243618603,
+      "loss": 1.3065,
+      "step": 4537
+    },
+    {
+      "epoch": 0.2090569862256415,
+      "grad_norm": 3.772218869005258,
+      "learning_rate": 0.00018365581854063104,
+      "loss": 1.9753,
+      "step": 4538
+    },
+    {
+      "epoch": 0.2091030543142765,
+      "grad_norm": 2.9598989967131786,
+      "learning_rate": 0.00018364764278262267,
+      "loss": 1.8303,
+      "step": 4539
+    },
+    {
+      "epoch": 0.2091491224029115,
+      "grad_norm": 3.396046083551735,
+      "learning_rate": 0.00018363946516234298,
+      "loss": 1.9933,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2091951904915465,
+      "grad_norm": 2.499257175699716,
+      "learning_rate": 0.00018363128567997403,
+      "loss": 1.9999,
+      "step": 4541
+    },
+    {
+      "epoch": 0.2092412585801815,
+      "grad_norm": 4.380420812518821,
+      "learning_rate": 0.00018362310433569788,
+      "loss": 1.749,
+      "step": 4542
+    },
+    {
+      "epoch": 0.2092873266688165,
+      "grad_norm": 2.6134215803492067,
+      "learning_rate": 0.00018361492112969668,
+      "loss": 1.7842,
+      "step": 4543
+    },
+    {
+      "epoch": 0.20933339475745152,
+      "grad_norm": 1.8048801501409144,
+      "learning_rate": 0.00018360673606215267,
+      "loss": 1.8607,
+      "step": 4544
+    },
+    {
+      "epoch": 0.20937946284608652,
+      "grad_norm": 5.406407192107045,
+      "learning_rate": 0.000183598549133248,
+      "loss": 1.6297,
+      "step": 4545
+    },
+    {
+      "epoch": 0.20942553093472152,
+      "grad_norm": 3.046974073522381,
+      "learning_rate": 0.000183590360343165,
+      "loss": 1.8773,
+      "step": 4546
+    },
+    {
+      "epoch": 0.20947159902335652,
+      "grad_norm": 5.790632371355523,
+      "learning_rate": 0.00018358216969208594,
+      "loss": 1.7076,
+      "step": 4547
+    },
+    {
+      "epoch": 0.20951766711199152,
+      "grad_norm": 2.583702355641612,
+      "learning_rate": 0.00018357397718019318,
+      "loss": 1.4639,
+      "step": 4548
+    },
+    {
+      "epoch": 0.20956373520062652,
+      "grad_norm": 1.8490685937452151,
+      "learning_rate": 0.00018356578280766912,
+      "loss": 1.4728,
+      "step": 4549
+    },
+    {
+      "epoch": 0.20960980328926151,
+      "grad_norm": 3.7158242188249457,
+      "learning_rate": 0.0001835575865746962,
+      "loss": 1.6612,
+      "step": 4550
+    },
+    {
+      "epoch": 0.20965587137789654,
+      "grad_norm": 5.239256721658291,
+      "learning_rate": 0.00018354938848145688,
+      "loss": 1.5804,
+      "step": 4551
+    },
+    {
+      "epoch": 0.20970193946653154,
+      "grad_norm": 2.4800957348782138,
+      "learning_rate": 0.00018354118852813368,
+      "loss": 1.7153,
+      "step": 4552
+    },
+    {
+      "epoch": 0.20974800755516654,
+      "grad_norm": 2.586791300494372,
+      "learning_rate": 0.00018353298671490914,
+      "loss": 1.4292,
+      "step": 4553
+    },
+    {
+      "epoch": 0.20979407564380154,
+      "grad_norm": 9.331506559669139,
+      "learning_rate": 0.00018352478304196592,
+      "loss": 1.594,
+      "step": 4554
+    },
+    {
+      "epoch": 0.20984014373243653,
+      "grad_norm": 2.245247976442041,
+      "learning_rate": 0.0001835165775094866,
+      "loss": 1.2831,
+      "step": 4555
+    },
+    {
+      "epoch": 0.20988621182107153,
+      "grad_norm": 2.345905499366,
+      "learning_rate": 0.00018350837011765387,
+      "loss": 1.6914,
+      "step": 4556
+    },
+    {
+      "epoch": 0.20993227990970656,
+      "grad_norm": 2.681344692067168,
+      "learning_rate": 0.0001835001608666505,
+      "loss": 1.6394,
+      "step": 4557
+    },
+    {
+      "epoch": 0.20997834799834156,
+      "grad_norm": 5.604479815786279,
+      "learning_rate": 0.00018349194975665918,
+      "loss": 1.5179,
+      "step": 4558
+    },
+    {
+      "epoch": 0.21002441608697656,
+      "grad_norm": 21.352217002040938,
+      "learning_rate": 0.00018348373678786278,
+      "loss": 1.4173,
+      "step": 4559
+    },
+    {
+      "epoch": 0.21007048417561155,
+      "grad_norm": 8.760404325252962,
+      "learning_rate": 0.0001834755219604441,
+      "loss": 1.7752,
+      "step": 4560
+    },
+    {
+      "epoch": 0.21011655226424655,
+      "grad_norm": 4.236882438475052,
+      "learning_rate": 0.00018346730527458606,
+      "loss": 1.6765,
+      "step": 4561
+    },
+    {
+      "epoch": 0.21016262035288155,
+      "grad_norm": 5.818309149087024,
+      "learning_rate": 0.00018345908673047162,
+      "loss": 1.4417,
+      "step": 4562
+    },
+    {
+      "epoch": 0.21020868844151655,
+      "grad_norm": 2.413436053758732,
+      "learning_rate": 0.00018345086632828367,
+      "loss": 1.8645,
+      "step": 4563
+    },
+    {
+      "epoch": 0.21025475653015158,
+      "grad_norm": 9.415652615357791,
+      "learning_rate": 0.0001834426440682053,
+      "loss": 1.6813,
+      "step": 4564
+    },
+    {
+      "epoch": 0.21030082461878657,
+      "grad_norm": 4.822839311318031,
+      "learning_rate": 0.00018343441995041952,
+      "loss": 1.8348,
+      "step": 4565
+    },
+    {
+      "epoch": 0.21034689270742157,
+      "grad_norm": 3.4463697254846655,
+      "learning_rate": 0.00018342619397510945,
+      "loss": 1.5927,
+      "step": 4566
+    },
+    {
+      "epoch": 0.21039296079605657,
+      "grad_norm": 4.390360435543998,
+      "learning_rate": 0.0001834179661424582,
+      "loss": 1.758,
+      "step": 4567
+    },
+    {
+      "epoch": 0.21043902888469157,
+      "grad_norm": 11.897747671842549,
+      "learning_rate": 0.000183409736452649,
+      "loss": 1.411,
+      "step": 4568
+    },
+    {
+      "epoch": 0.21048509697332657,
+      "grad_norm": 3.2859311172094054,
+      "learning_rate": 0.000183401504905865,
+      "loss": 1.8516,
+      "step": 4569
+    },
+    {
+      "epoch": 0.21053116506196157,
+      "grad_norm": 5.711217988014021,
+      "learning_rate": 0.00018339327150228952,
+      "loss": 1.9763,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2105772331505966,
+      "grad_norm": 2.1771064111647442,
+      "learning_rate": 0.0001833850362421058,
+      "loss": 1.714,
+      "step": 4571
+    },
+    {
+      "epoch": 0.2106233012392316,
+      "grad_norm": 7.687242332065841,
+      "learning_rate": 0.00018337679912549723,
+      "loss": 1.8874,
+      "step": 4572
+    },
+    {
+      "epoch": 0.2106693693278666,
+      "grad_norm": 2.764964423554615,
+      "learning_rate": 0.00018336856015264722,
+      "loss": 1.6365,
+      "step": 4573
+    },
+    {
+      "epoch": 0.2107154374165016,
+      "grad_norm": 3.2537772893710963,
+      "learning_rate": 0.00018336031932373917,
+      "loss": 1.7017,
+      "step": 4574
+    },
+    {
+      "epoch": 0.2107615055051366,
+      "grad_norm": 2.9112664086575717,
+      "learning_rate": 0.0001833520766389565,
+      "loss": 1.474,
+      "step": 4575
+    },
+    {
+      "epoch": 0.21080757359377159,
+      "grad_norm": 8.246519192443866,
+      "learning_rate": 0.0001833438320984828,
+      "loss": 1.8127,
+      "step": 4576
+    },
+    {
+      "epoch": 0.21085364168240658,
+      "grad_norm": 1.8079767755814,
+      "learning_rate": 0.00018333558570250155,
+      "loss": 1.7238,
+      "step": 4577
+    },
+    {
+      "epoch": 0.2108997097710416,
+      "grad_norm": 5.814928416139414,
+      "learning_rate": 0.00018332733745119638,
+      "loss": 1.4379,
+      "step": 4578
+    },
+    {
+      "epoch": 0.2109457778596766,
+      "grad_norm": 1.799628307248101,
+      "learning_rate": 0.0001833190873447509,
+      "loss": 1.9044,
+      "step": 4579
+    },
+    {
+      "epoch": 0.2109918459483116,
+      "grad_norm": 2.3347069055203797,
+      "learning_rate": 0.00018331083538334885,
+      "loss": 1.6254,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2110379140369466,
+      "grad_norm": 1.8118000759185,
+      "learning_rate": 0.00018330258156717388,
+      "loss": 1.9589,
+      "step": 4581
+    },
+    {
+      "epoch": 0.2110839821255816,
+      "grad_norm": 6.330102247689913,
+      "learning_rate": 0.00018329432589640976,
+      "loss": 1.0586,
+      "step": 4582
+    },
+    {
+      "epoch": 0.2111300502142166,
+      "grad_norm": 6.741261768755454,
+      "learning_rate": 0.0001832860683712403,
+      "loss": 1.8428,
+      "step": 4583
+    },
+    {
+      "epoch": 0.2111761183028516,
+      "grad_norm": 3.1149604075939514,
+      "learning_rate": 0.00018327780899184934,
+      "loss": 1.8829,
+      "step": 4584
+    },
+    {
+      "epoch": 0.21122218639148663,
+      "grad_norm": 4.269349140946456,
+      "learning_rate": 0.00018326954775842074,
+      "loss": 1.4278,
+      "step": 4585
+    },
+    {
+      "epoch": 0.21126825448012163,
+      "grad_norm": 3.183539116699412,
+      "learning_rate": 0.00018326128467113843,
+      "loss": 1.7098,
+      "step": 4586
+    },
+    {
+      "epoch": 0.21131432256875662,
+      "grad_norm": 2.563756748456286,
+      "learning_rate": 0.00018325301973018637,
+      "loss": 1.6453,
+      "step": 4587
+    },
+    {
+      "epoch": 0.21136039065739162,
+      "grad_norm": 3.838821805410065,
+      "learning_rate": 0.0001832447529357486,
+      "loss": 1.5738,
+      "step": 4588
+    },
+    {
+      "epoch": 0.21140645874602662,
+      "grad_norm": 2.845416334291807,
+      "learning_rate": 0.00018323648428800911,
+      "loss": 1.7374,
+      "step": 4589
+    },
+    {
+      "epoch": 0.21145252683466162,
+      "grad_norm": 2.6709850257071537,
+      "learning_rate": 0.00018322821378715206,
+      "loss": 1.886,
+      "step": 4590
+    },
+    {
+      "epoch": 0.21149859492329665,
+      "grad_norm": 3.467919963542419,
+      "learning_rate": 0.0001832199414333615,
+      "loss": 1.6395,
+      "step": 4591
+    },
+    {
+      "epoch": 0.21154466301193164,
+      "grad_norm": 2.7405643267368998,
+      "learning_rate": 0.00018321166722682166,
+      "loss": 1.7546,
+      "step": 4592
+    },
+    {
+      "epoch": 0.21159073110056664,
+      "grad_norm": 4.078039862422518,
+      "learning_rate": 0.0001832033911677167,
+      "loss": 1.6746,
+      "step": 4593
+    },
+    {
+      "epoch": 0.21163679918920164,
+      "grad_norm": 2.2182986335227173,
+      "learning_rate": 0.00018319511325623093,
+      "loss": 1.5494,
+      "step": 4594
+    },
+    {
+      "epoch": 0.21168286727783664,
+      "grad_norm": 3.83259635434467,
+      "learning_rate": 0.0001831868334925486,
+      "loss": 1.5704,
+      "step": 4595
+    },
+    {
+      "epoch": 0.21172893536647164,
+      "grad_norm": 7.605107811935652,
+      "learning_rate": 0.00018317855187685407,
+      "loss": 1.5692,
+      "step": 4596
+    },
+    {
+      "epoch": 0.21177500345510664,
+      "grad_norm": 2.8983649936674154,
+      "learning_rate": 0.00018317026840933168,
+      "loss": 1.6272,
+      "step": 4597
+    },
+    {
+      "epoch": 0.21182107154374166,
+      "grad_norm": 3.998139692781574,
+      "learning_rate": 0.00018316198309016585,
+      "loss": 1.6925,
+      "step": 4598
+    },
+    {
+      "epoch": 0.21186713963237666,
+      "grad_norm": 3.436417092827452,
+      "learning_rate": 0.0001831536959195411,
+      "loss": 1.5623,
+      "step": 4599
+    },
+    {
+      "epoch": 0.21191320772101166,
+      "grad_norm": 3.5115768023404375,
+      "learning_rate": 0.0001831454068976419,
+      "loss": 1.9569,
+      "step": 4600
+    },
+    {
+      "epoch": 0.21195927580964666,
+      "grad_norm": 3.717948255898704,
+      "learning_rate": 0.0001831371160246527,
+      "loss": 1.4036,
+      "step": 4601
+    },
+    {
+      "epoch": 0.21200534389828166,
+      "grad_norm": 3.492140061283508,
+      "learning_rate": 0.00018312882330075826,
+      "loss": 1.7713,
+      "step": 4602
+    },
+    {
+      "epoch": 0.21205141198691665,
+      "grad_norm": 3.856682664409372,
+      "learning_rate": 0.00018312052872614305,
+      "loss": 1.6931,
+      "step": 4603
+    },
+    {
+      "epoch": 0.21209748007555165,
+      "grad_norm": 11.527569226285278,
+      "learning_rate": 0.0001831122323009918,
+      "loss": 1.6709,
+      "step": 4604
+    },
+    {
+      "epoch": 0.21214354816418668,
+      "grad_norm": 6.295227628837269,
+      "learning_rate": 0.0001831039340254892,
+      "loss": 1.88,
+      "step": 4605
+    },
+    {
+      "epoch": 0.21218961625282168,
+      "grad_norm": 4.456825340999828,
+      "learning_rate": 0.00018309563389981999,
+      "loss": 1.5772,
+      "step": 4606
+    },
+    {
+      "epoch": 0.21223568434145668,
+      "grad_norm": 4.149612351553757,
+      "learning_rate": 0.000183087331924169,
+      "loss": 1.8157,
+      "step": 4607
+    },
+    {
+      "epoch": 0.21228175243009167,
+      "grad_norm": 2.4884381949523458,
+      "learning_rate": 0.000183079028098721,
+      "loss": 1.7651,
+      "step": 4608
+    },
+    {
+      "epoch": 0.21232782051872667,
+      "grad_norm": 3.3541042647570123,
+      "learning_rate": 0.0001830707224236609,
+      "loss": 1.5485,
+      "step": 4609
+    },
+    {
+      "epoch": 0.21237388860736167,
+      "grad_norm": 7.149113508505121,
+      "learning_rate": 0.00018306241489917361,
+      "loss": 1.5837,
+      "step": 4610
+    },
+    {
+      "epoch": 0.21241995669599667,
+      "grad_norm": 10.937974380544667,
+      "learning_rate": 0.00018305410552544405,
+      "loss": 1.4427,
+      "step": 4611
+    },
+    {
+      "epoch": 0.2124660247846317,
+      "grad_norm": 7.437583439647635,
+      "learning_rate": 0.00018304579430265726,
+      "loss": 1.725,
+      "step": 4612
+    },
+    {
+      "epoch": 0.2125120928732667,
+      "grad_norm": 7.984889232748964,
+      "learning_rate": 0.00018303748123099826,
+      "loss": 1.7798,
+      "step": 4613
+    },
+    {
+      "epoch": 0.2125581609619017,
+      "grad_norm": 4.906117386628533,
+      "learning_rate": 0.0001830291663106521,
+      "loss": 1.8765,
+      "step": 4614
+    },
+    {
+      "epoch": 0.2126042290505367,
+      "grad_norm": 7.136344514344814,
+      "learning_rate": 0.00018302084954180395,
+      "loss": 1.5931,
+      "step": 4615
+    },
+    {
+      "epoch": 0.2126502971391717,
+      "grad_norm": 5.368697320634226,
+      "learning_rate": 0.0001830125309246389,
+      "loss": 1.2591,
+      "step": 4616
+    },
+    {
+      "epoch": 0.2126963652278067,
+      "grad_norm": 4.058021412926005,
+      "learning_rate": 0.00018300421045934219,
+      "loss": 1.722,
+      "step": 4617
+    },
+    {
+      "epoch": 0.2127424333164417,
+      "grad_norm": 2.4821987696691084,
+      "learning_rate": 0.00018299588814609904,
+      "loss": 1.849,
+      "step": 4618
+    },
+    {
+      "epoch": 0.2127885014050767,
+      "grad_norm": 5.483763901923326,
+      "learning_rate": 0.00018298756398509478,
+      "loss": 1.9555,
+      "step": 4619
+    },
+    {
+      "epoch": 0.2128345694937117,
+      "grad_norm": 2.213330319850431,
+      "learning_rate": 0.00018297923797651469,
+      "loss": 1.7902,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2128806375823467,
+      "grad_norm": 7.51592244651267,
+      "learning_rate": 0.00018297091012054413,
+      "loss": 1.5126,
+      "step": 4621
+    },
+    {
+      "epoch": 0.2129267056709817,
+      "grad_norm": 7.704058262184992,
+      "learning_rate": 0.00018296258041736855,
+      "loss": 1.2991,
+      "step": 4622
+    },
+    {
+      "epoch": 0.2129727737596167,
+      "grad_norm": 9.184912452711822,
+      "learning_rate": 0.00018295424886717333,
+      "loss": 1.7336,
+      "step": 4623
+    },
+    {
+      "epoch": 0.2130188418482517,
+      "grad_norm": 4.014584341340152,
+      "learning_rate": 0.00018294591547014403,
+      "loss": 1.6106,
+      "step": 4624
+    },
+    {
+      "epoch": 0.21306490993688673,
+      "grad_norm": 3.373750505066092,
+      "learning_rate": 0.00018293758022646612,
+      "loss": 1.8107,
+      "step": 4625
+    },
+    {
+      "epoch": 0.21311097802552173,
+      "grad_norm": 4.046742030422276,
+      "learning_rate": 0.00018292924313632522,
+      "loss": 1.7104,
+      "step": 4626
+    },
+    {
+      "epoch": 0.21315704611415673,
+      "grad_norm": 3.662118846188741,
+      "learning_rate": 0.0001829209041999069,
+      "loss": 1.6807,
+      "step": 4627
+    },
+    {
+      "epoch": 0.21320311420279173,
+      "grad_norm": 2.3330662371134823,
+      "learning_rate": 0.00018291256341739683,
+      "loss": 1.6805,
+      "step": 4628
+    },
+    {
+      "epoch": 0.21324918229142673,
+      "grad_norm": 2.7641846343729313,
+      "learning_rate": 0.0001829042207889807,
+      "loss": 1.5975,
+      "step": 4629
+    },
+    {
+      "epoch": 0.21329525038006172,
+      "grad_norm": 4.782639734368663,
+      "learning_rate": 0.00018289587631484423,
+      "loss": 1.8851,
+      "step": 4630
+    },
+    {
+      "epoch": 0.21334131846869672,
+      "grad_norm": 7.17952330357705,
+      "learning_rate": 0.0001828875299951732,
+      "loss": 1.7738,
+      "step": 4631
+    },
+    {
+      "epoch": 0.21338738655733175,
+      "grad_norm": 7.537037443579318,
+      "learning_rate": 0.0001828791818301535,
+      "loss": 1.465,
+      "step": 4632
+    },
+    {
+      "epoch": 0.21343345464596675,
+      "grad_norm": 2.1951344264400725,
+      "learning_rate": 0.00018287083181997086,
+      "loss": 1.8385,
+      "step": 4633
+    },
+    {
+      "epoch": 0.21347952273460175,
+      "grad_norm": 2.5094583885056037,
+      "learning_rate": 0.00018286247996481125,
+      "loss": 1.6837,
+      "step": 4634
+    },
+    {
+      "epoch": 0.21352559082323674,
+      "grad_norm": 5.505654361799716,
+      "learning_rate": 0.00018285412626486064,
+      "loss": 1.5923,
+      "step": 4635
+    },
+    {
+      "epoch": 0.21357165891187174,
+      "grad_norm": 2.182871920680815,
+      "learning_rate": 0.00018284577072030493,
+      "loss": 1.56,
+      "step": 4636
+    },
+    {
+      "epoch": 0.21361772700050674,
+      "grad_norm": 2.577923074737655,
+      "learning_rate": 0.0001828374133313302,
+      "loss": 1.7156,
+      "step": 4637
+    },
+    {
+      "epoch": 0.21366379508914174,
+      "grad_norm": 2.989075191466587,
+      "learning_rate": 0.00018282905409812251,
+      "loss": 1.5438,
+      "step": 4638
+    },
+    {
+      "epoch": 0.21370986317777677,
+      "grad_norm": 2.272479355802785,
+      "learning_rate": 0.00018282069302086792,
+      "loss": 1.5955,
+      "step": 4639
+    },
+    {
+      "epoch": 0.21375593126641176,
+      "grad_norm": 2.9649376745776044,
+      "learning_rate": 0.00018281233009975265,
+      "loss": 1.7527,
+      "step": 4640
+    },
+    {
+      "epoch": 0.21380199935504676,
+      "grad_norm": 6.159152685380261,
+      "learning_rate": 0.00018280396533496284,
+      "loss": 1.6325,
+      "step": 4641
+    },
+    {
+      "epoch": 0.21384806744368176,
+      "grad_norm": 4.265912950456718,
+      "learning_rate": 0.0001827955987266847,
+      "loss": 1.6808,
+      "step": 4642
+    },
+    {
+      "epoch": 0.21389413553231676,
+      "grad_norm": 3.0735721401376024,
+      "learning_rate": 0.0001827872302751045,
+      "loss": 1.6505,
+      "step": 4643
+    },
+    {
+      "epoch": 0.21394020362095176,
+      "grad_norm": 2.472424657831009,
+      "learning_rate": 0.0001827788599804086,
+      "loss": 1.6095,
+      "step": 4644
+    },
+    {
+      "epoch": 0.21398627170958676,
+      "grad_norm": 4.069084793889558,
+      "learning_rate": 0.0001827704878427833,
+      "loss": 1.7356,
+      "step": 4645
+    },
+    {
+      "epoch": 0.21403233979822178,
+      "grad_norm": 3.615115520025387,
+      "learning_rate": 0.00018276211386241502,
+      "loss": 1.6777,
+      "step": 4646
+    },
+    {
+      "epoch": 0.21407840788685678,
+      "grad_norm": 5.703780448488673,
+      "learning_rate": 0.0001827537380394902,
+      "loss": 1.5805,
+      "step": 4647
+    },
+    {
+      "epoch": 0.21412447597549178,
+      "grad_norm": 2.8886579448159675,
+      "learning_rate": 0.00018274536037419522,
+      "loss": 1.6313,
+      "step": 4648
+    },
+    {
+      "epoch": 0.21417054406412678,
+      "grad_norm": 2.1900909352146494,
+      "learning_rate": 0.00018273698086671675,
+      "loss": 1.9506,
+      "step": 4649
+    },
+    {
+      "epoch": 0.21421661215276178,
+      "grad_norm": 4.270144595957271,
+      "learning_rate": 0.00018272859951724125,
+      "loss": 1.1766,
+      "step": 4650
+    },
+    {
+      "epoch": 0.21426268024139677,
+      "grad_norm": 2.4548369181657534,
+      "learning_rate": 0.00018272021632595532,
+      "loss": 1.7859,
+      "step": 4651
+    },
+    {
+      "epoch": 0.2143087483300318,
+      "grad_norm": 2.5782363127420407,
+      "learning_rate": 0.00018271183129304558,
+      "loss": 1.7969,
+      "step": 4652
+    },
+    {
+      "epoch": 0.2143548164186668,
+      "grad_norm": 1.8680640120126404,
+      "learning_rate": 0.00018270344441869874,
+      "loss": 1.6019,
+      "step": 4653
+    },
+    {
+      "epoch": 0.2144008845073018,
+      "grad_norm": 2.7830367887950285,
+      "learning_rate": 0.00018269505570310158,
+      "loss": 1.5581,
+      "step": 4654
+    },
+    {
+      "epoch": 0.2144469525959368,
+      "grad_norm": 4.985023357622415,
+      "learning_rate": 0.00018268666514644074,
+      "loss": 1.6417,
+      "step": 4655
+    },
+    {
+      "epoch": 0.2144930206845718,
+      "grad_norm": 2.4514918690823375,
+      "learning_rate": 0.0001826782727489031,
+      "loss": 1.7089,
+      "step": 4656
+    },
+    {
+      "epoch": 0.2145390887732068,
+      "grad_norm": 2.8073399488596738,
+      "learning_rate": 0.00018266987851067547,
+      "loss": 1.7098,
+      "step": 4657
+    },
+    {
+      "epoch": 0.2145851568618418,
+      "grad_norm": 1.9648329372784243,
+      "learning_rate": 0.00018266148243194474,
+      "loss": 1.6588,
+      "step": 4658
+    },
+    {
+      "epoch": 0.21463122495047682,
+      "grad_norm": 1.628075132624749,
+      "learning_rate": 0.00018265308451289783,
+      "loss": 1.8244,
+      "step": 4659
+    },
+    {
+      "epoch": 0.21467729303911182,
+      "grad_norm": 1.7860608502515734,
+      "learning_rate": 0.00018264468475372175,
+      "loss": 1.513,
+      "step": 4660
+    },
+    {
+      "epoch": 0.21472336112774681,
+      "grad_norm": 1.9865861794844042,
+      "learning_rate": 0.00018263628315460343,
+      "loss": 1.3657,
+      "step": 4661
+    },
+    {
+      "epoch": 0.2147694292163818,
+      "grad_norm": 2.08166472780451,
+      "learning_rate": 0.00018262787971572996,
+      "loss": 1.5764,
+      "step": 4662
+    },
+    {
+      "epoch": 0.2148154973050168,
+      "grad_norm": 1.8025346115463365,
+      "learning_rate": 0.00018261947443728844,
+      "loss": 1.529,
+      "step": 4663
+    },
+    {
+      "epoch": 0.2148615653936518,
+      "grad_norm": 3.2482159867211804,
+      "learning_rate": 0.000182611067319466,
+      "loss": 1.436,
+      "step": 4664
+    },
+    {
+      "epoch": 0.2149076334822868,
+      "grad_norm": 2.975589962722503,
+      "learning_rate": 0.00018260265836244976,
+      "loss": 1.4393,
+      "step": 4665
+    },
+    {
+      "epoch": 0.21495370157092183,
+      "grad_norm": 2.1076686417207413,
+      "learning_rate": 0.00018259424756642697,
+      "loss": 1.5129,
+      "step": 4666
+    },
+    {
+      "epoch": 0.21499976965955683,
+      "grad_norm": 1.455612413233475,
+      "learning_rate": 0.0001825858349315849,
+      "loss": 1.4813,
+      "step": 4667
+    },
+    {
+      "epoch": 0.21504583774819183,
+      "grad_norm": 3.911485298808748,
+      "learning_rate": 0.0001825774204581108,
+      "loss": 1.8018,
+      "step": 4668
+    },
+    {
+      "epoch": 0.21509190583682683,
+      "grad_norm": 1.5615979649443685,
+      "learning_rate": 0.00018256900414619204,
+      "loss": 1.2888,
+      "step": 4669
+    },
+    {
+      "epoch": 0.21513797392546183,
+      "grad_norm": 1.4591180632459317,
+      "learning_rate": 0.00018256058599601597,
+      "loss": 1.721,
+      "step": 4670
+    },
+    {
+      "epoch": 0.21518404201409683,
+      "grad_norm": 1.8488868937265939,
+      "learning_rate": 0.00018255216600777,
+      "loss": 1.6845,
+      "step": 4671
+    },
+    {
+      "epoch": 0.21523011010273183,
+      "grad_norm": 1.6248128701065505,
+      "learning_rate": 0.0001825437441816416,
+      "loss": 1.7958,
+      "step": 4672
+    },
+    {
+      "epoch": 0.21527617819136685,
+      "grad_norm": 3.6030567292184186,
+      "learning_rate": 0.0001825353205178183,
+      "loss": 1.5788,
+      "step": 4673
+    },
+    {
+      "epoch": 0.21532224628000185,
+      "grad_norm": 3.0162736787379574,
+      "learning_rate": 0.00018252689501648757,
+      "loss": 1.4589,
+      "step": 4674
+    },
+    {
+      "epoch": 0.21536831436863685,
+      "grad_norm": 1.7270349927134927,
+      "learning_rate": 0.00018251846767783705,
+      "loss": 1.832,
+      "step": 4675
+    },
+    {
+      "epoch": 0.21541438245727185,
+      "grad_norm": 2.7175229856716245,
+      "learning_rate": 0.00018251003850205431,
+      "loss": 1.6761,
+      "step": 4676
+    },
+    {
+      "epoch": 0.21546045054590685,
+      "grad_norm": 1.36872910331271,
+      "learning_rate": 0.00018250160748932705,
+      "loss": 1.7162,
+      "step": 4677
+    },
+    {
+      "epoch": 0.21550651863454184,
+      "grad_norm": 1.741601818030805,
+      "learning_rate": 0.00018249317463984297,
+      "loss": 1.7878,
+      "step": 4678
+    },
+    {
+      "epoch": 0.21555258672317684,
+      "grad_norm": 1.7526191110428446,
+      "learning_rate": 0.00018248473995378982,
+      "loss": 1.3717,
+      "step": 4679
+    },
+    {
+      "epoch": 0.21559865481181187,
+      "grad_norm": 2.440972887388374,
+      "learning_rate": 0.0001824763034313553,
+      "loss": 1.608,
+      "step": 4680
+    },
+    {
+      "epoch": 0.21564472290044687,
+      "grad_norm": 2.0411849730191154,
+      "learning_rate": 0.00018246786507272736,
+      "loss": 1.4724,
+      "step": 4681
+    },
+    {
+      "epoch": 0.21569079098908187,
+      "grad_norm": 1.882597264401815,
+      "learning_rate": 0.0001824594248780938,
+      "loss": 1.4175,
+      "step": 4682
+    },
+    {
+      "epoch": 0.21573685907771686,
+      "grad_norm": 3.1215473877993043,
+      "learning_rate": 0.0001824509828476425,
+      "loss": 1.5904,
+      "step": 4683
+    },
+    {
+      "epoch": 0.21578292716635186,
+      "grad_norm": 1.6741073023912145,
+      "learning_rate": 0.00018244253898156148,
+      "loss": 1.5761,
+      "step": 4684
+    },
+    {
+      "epoch": 0.21582899525498686,
+      "grad_norm": 2.5635083584999,
+      "learning_rate": 0.00018243409328003868,
+      "loss": 1.3669,
+      "step": 4685
+    },
+    {
+      "epoch": 0.2158750633436219,
+      "grad_norm": 2.9189969716504964,
+      "learning_rate": 0.0001824256457432621,
+      "loss": 1.2612,
+      "step": 4686
+    },
+    {
+      "epoch": 0.21592113143225689,
+      "grad_norm": 1.457020823343036,
+      "learning_rate": 0.00018241719637141987,
+      "loss": 1.5443,
+      "step": 4687
+    },
+    {
+      "epoch": 0.21596719952089188,
+      "grad_norm": 1.9571847030642728,
+      "learning_rate": 0.00018240874516470008,
+      "loss": 1.8078,
+      "step": 4688
+    },
+    {
+      "epoch": 0.21601326760952688,
+      "grad_norm": 8.442320941930435,
+      "learning_rate": 0.00018240029212329088,
+      "loss": 1.5028,
+      "step": 4689
+    },
+    {
+      "epoch": 0.21605933569816188,
+      "grad_norm": 2.7893247319957575,
+      "learning_rate": 0.00018239183724738048,
+      "loss": 1.5624,
+      "step": 4690
+    },
+    {
+      "epoch": 0.21610540378679688,
+      "grad_norm": 2.2070272162405704,
+      "learning_rate": 0.00018238338053715706,
+      "loss": 1.3155,
+      "step": 4691
+    },
+    {
+      "epoch": 0.21615147187543188,
+      "grad_norm": 5.002197249219553,
+      "learning_rate": 0.00018237492199280894,
+      "loss": 1.5181,
+      "step": 4692
+    },
+    {
+      "epoch": 0.2161975399640669,
+      "grad_norm": 3.7586160396112755,
+      "learning_rate": 0.00018236646161452446,
+      "loss": 1.4382,
+      "step": 4693
+    },
+    {
+      "epoch": 0.2162436080527019,
+      "grad_norm": 1.843605704500003,
+      "learning_rate": 0.0001823579994024919,
+      "loss": 1.2822,
+      "step": 4694
+    },
+    {
+      "epoch": 0.2162896761413369,
+      "grad_norm": 1.3750895008664785,
+      "learning_rate": 0.0001823495353568997,
+      "loss": 1.8292,
+      "step": 4695
+    },
+    {
+      "epoch": 0.2163357442299719,
+      "grad_norm": 2.126213450233978,
+      "learning_rate": 0.00018234106947793632,
+      "loss": 1.5692,
+      "step": 4696
+    },
+    {
+      "epoch": 0.2163818123186069,
+      "grad_norm": 2.388595665215315,
+      "learning_rate": 0.0001823326017657902,
+      "loss": 1.2361,
+      "step": 4697
+    },
+    {
+      "epoch": 0.2164278804072419,
+      "grad_norm": 2.7998376595728613,
+      "learning_rate": 0.00018232413222064989,
+      "loss": 1.4973,
+      "step": 4698
+    },
+    {
+      "epoch": 0.2164739484958769,
+      "grad_norm": 1.746080119941745,
+      "learning_rate": 0.0001823156608427039,
+      "loss": 1.5313,
+      "step": 4699
+    },
+    {
+      "epoch": 0.21652001658451192,
+      "grad_norm": 3.0428412888234013,
+      "learning_rate": 0.00018230718763214085,
+      "loss": 1.4371,
+      "step": 4700
+    },
+    {
+      "epoch": 0.21656608467314692,
+      "grad_norm": 4.732621075693956,
+      "learning_rate": 0.0001822987125891494,
+      "loss": 1.8183,
+      "step": 4701
+    },
+    {
+      "epoch": 0.21661215276178192,
+      "grad_norm": 3.2852967779944398,
+      "learning_rate": 0.0001822902357139183,
+      "loss": 1.5868,
+      "step": 4702
+    },
+    {
+      "epoch": 0.21665822085041692,
+      "grad_norm": 3.058788409254842,
+      "learning_rate": 0.0001822817570066361,
+      "loss": 1.6435,
+      "step": 4703
+    },
+    {
+      "epoch": 0.21670428893905191,
+      "grad_norm": 5.464766068274446,
+      "learning_rate": 0.00018227327646749172,
+      "loss": 1.3387,
+      "step": 4704
+    },
+    {
+      "epoch": 0.2167503570276869,
+      "grad_norm": 2.370168185395886,
+      "learning_rate": 0.00018226479409667388,
+      "loss": 1.6181,
+      "step": 4705
+    },
+    {
+      "epoch": 0.2167964251163219,
+      "grad_norm": 2.452069316265759,
+      "learning_rate": 0.00018225630989437147,
+      "loss": 1.3779,
+      "step": 4706
+    },
+    {
+      "epoch": 0.21684249320495694,
+      "grad_norm": 1.7771475031279478,
+      "learning_rate": 0.00018224782386077335,
+      "loss": 1.6845,
+      "step": 4707
+    },
+    {
+      "epoch": 0.21688856129359194,
+      "grad_norm": 2.5976582403224606,
+      "learning_rate": 0.00018223933599606848,
+      "loss": 1.5625,
+      "step": 4708
+    },
+    {
+      "epoch": 0.21693462938222693,
+      "grad_norm": 2.458144631673299,
+      "learning_rate": 0.0001822308463004458,
+      "loss": 1.5456,
+      "step": 4709
+    },
+    {
+      "epoch": 0.21698069747086193,
+      "grad_norm": 2.1972669046699758,
+      "learning_rate": 0.00018222235477409428,
+      "loss": 1.8379,
+      "step": 4710
+    },
+    {
+      "epoch": 0.21702676555949693,
+      "grad_norm": 6.28078535159849,
+      "learning_rate": 0.00018221386141720306,
+      "loss": 1.5012,
+      "step": 4711
+    },
+    {
+      "epoch": 0.21707283364813193,
+      "grad_norm": 3.1170238049118395,
+      "learning_rate": 0.00018220536622996116,
+      "loss": 1.6397,
+      "step": 4712
+    },
+    {
+      "epoch": 0.21711890173676693,
+      "grad_norm": 10.51592203448791,
+      "learning_rate": 0.00018219686921255775,
+      "loss": 1.4519,
+      "step": 4713
+    },
+    {
+      "epoch": 0.21716496982540195,
+      "grad_norm": 1.8427883246050265,
+      "learning_rate": 0.00018218837036518195,
+      "loss": 1.6914,
+      "step": 4714
+    },
+    {
+      "epoch": 0.21721103791403695,
+      "grad_norm": 2.4911713138753746,
+      "learning_rate": 0.00018217986968802302,
+      "loss": 1.6331,
+      "step": 4715
+    },
+    {
+      "epoch": 0.21725710600267195,
+      "grad_norm": 1.9224695476180564,
+      "learning_rate": 0.0001821713671812702,
+      "loss": 1.5711,
+      "step": 4716
+    },
+    {
+      "epoch": 0.21730317409130695,
+      "grad_norm": 2.9270867895586856,
+      "learning_rate": 0.0001821628628451128,
+      "loss": 1.1644,
+      "step": 4717
+    },
+    {
+      "epoch": 0.21734924217994195,
+      "grad_norm": 4.484589077662357,
+      "learning_rate": 0.00018215435667974015,
+      "loss": 1.9051,
+      "step": 4718
+    },
+    {
+      "epoch": 0.21739531026857695,
+      "grad_norm": 1.864084184444098,
+      "learning_rate": 0.00018214584868534158,
+      "loss": 1.6406,
+      "step": 4719
+    },
+    {
+      "epoch": 0.21744137835721197,
+      "grad_norm": 2.9405100930237307,
+      "learning_rate": 0.00018213733886210655,
+      "loss": 1.37,
+      "step": 4720
+    },
+    {
+      "epoch": 0.21748744644584697,
+      "grad_norm": 1.797785736539357,
+      "learning_rate": 0.00018212882721022451,
+      "loss": 1.6517,
+      "step": 4721
+    },
+    {
+      "epoch": 0.21753351453448197,
+      "grad_norm": 2.3322340197850737,
+      "learning_rate": 0.00018212031372988495,
+      "loss": 1.4234,
+      "step": 4722
+    },
+    {
+      "epoch": 0.21757958262311697,
+      "grad_norm": 2.0710913291751107,
+      "learning_rate": 0.0001821117984212774,
+      "loss": 1.5063,
+      "step": 4723
+    },
+    {
+      "epoch": 0.21762565071175197,
+      "grad_norm": 1.4727795714761858,
+      "learning_rate": 0.00018210328128459146,
+      "loss": 1.7029,
+      "step": 4724
+    },
+    {
+      "epoch": 0.21767171880038697,
+      "grad_norm": 4.945116283730176,
+      "learning_rate": 0.00018209476232001673,
+      "loss": 1.6233,
+      "step": 4725
+    },
+    {
+      "epoch": 0.21771778688902196,
+      "grad_norm": 3.2644408101918954,
+      "learning_rate": 0.0001820862415277429,
+      "loss": 1.4758,
+      "step": 4726
+    },
+    {
+      "epoch": 0.217763854977657,
+      "grad_norm": 3.51644557830705,
+      "learning_rate": 0.00018207771890795966,
+      "loss": 1.7859,
+      "step": 4727
+    },
+    {
+      "epoch": 0.217809923066292,
+      "grad_norm": 1.7339381467433723,
+      "learning_rate": 0.0001820691944608567,
+      "loss": 1.3908,
+      "step": 4728
+    },
+    {
+      "epoch": 0.217855991154927,
+      "grad_norm": 3.8491425431505766,
+      "learning_rate": 0.00018206066818662387,
+      "loss": 1.3083,
+      "step": 4729
+    },
+    {
+      "epoch": 0.21790205924356199,
+      "grad_norm": 2.5414394472374893,
+      "learning_rate": 0.00018205214008545095,
+      "loss": 1.2929,
+      "step": 4730
+    },
+    {
+      "epoch": 0.21794812733219698,
+      "grad_norm": 2.9629028445132666,
+      "learning_rate": 0.00018204361015752783,
+      "loss": 1.7164,
+      "step": 4731
+    },
+    {
+      "epoch": 0.21799419542083198,
+      "grad_norm": 3.2445708673611433,
+      "learning_rate": 0.0001820350784030444,
+      "loss": 1.4924,
+      "step": 4732
+    },
+    {
+      "epoch": 0.21804026350946698,
+      "grad_norm": 3.235180124690981,
+      "learning_rate": 0.00018202654482219063,
+      "loss": 1.8514,
+      "step": 4733
+    },
+    {
+      "epoch": 0.218086331598102,
+      "grad_norm": 3.062786743604861,
+      "learning_rate": 0.00018201800941515644,
+      "loss": 1.4494,
+      "step": 4734
+    },
+    {
+      "epoch": 0.218132399686737,
+      "grad_norm": 1.7012469411288413,
+      "learning_rate": 0.00018200947218213195,
+      "loss": 1.8158,
+      "step": 4735
+    },
+    {
+      "epoch": 0.218178467775372,
+      "grad_norm": 2.334018151186614,
+      "learning_rate": 0.00018200093312330717,
+      "loss": 1.7069,
+      "step": 4736
+    },
+    {
+      "epoch": 0.218224535864007,
+      "grad_norm": 2.3761305125037118,
+      "learning_rate": 0.0001819923922388722,
+      "loss": 1.572,
+      "step": 4737
+    },
+    {
+      "epoch": 0.218270603952642,
+      "grad_norm": 3.0871118882786273,
+      "learning_rate": 0.00018198384952901718,
+      "loss": 1.4557,
+      "step": 4738
+    },
+    {
+      "epoch": 0.218316672041277,
+      "grad_norm": 1.641858191025268,
+      "learning_rate": 0.0001819753049939324,
+      "loss": 1.5241,
+      "step": 4739
+    },
+    {
+      "epoch": 0.218362740129912,
+      "grad_norm": 2.052636747489013,
+      "learning_rate": 0.00018196675863380794,
+      "loss": 1.8082,
+      "step": 4740
+    },
+    {
+      "epoch": 0.21840880821854702,
+      "grad_norm": 2.3296036782844447,
+      "learning_rate": 0.00018195821044883416,
+      "loss": 1.6427,
+      "step": 4741
+    },
+    {
+      "epoch": 0.21845487630718202,
+      "grad_norm": 2.643500078410284,
+      "learning_rate": 0.00018194966043920134,
+      "loss": 1.5918,
+      "step": 4742
+    },
+    {
+      "epoch": 0.21850094439581702,
+      "grad_norm": 2.0932660109302486,
+      "learning_rate": 0.00018194110860509987,
+      "loss": 1.6227,
+      "step": 4743
+    },
+    {
+      "epoch": 0.21854701248445202,
+      "grad_norm": 1.1663018996463954,
+      "learning_rate": 0.00018193255494672013,
+      "loss": 1.3666,
+      "step": 4744
+    },
+    {
+      "epoch": 0.21859308057308702,
+      "grad_norm": 4.941428698414219,
+      "learning_rate": 0.0001819239994642525,
+      "loss": 1.137,
+      "step": 4745
+    },
+    {
+      "epoch": 0.21863914866172202,
+      "grad_norm": 1.5287669772351444,
+      "learning_rate": 0.00018191544215788753,
+      "loss": 1.443,
+      "step": 4746
+    },
+    {
+      "epoch": 0.21868521675035701,
+      "grad_norm": 2.239139652556453,
+      "learning_rate": 0.00018190688302781565,
+      "loss": 1.527,
+      "step": 4747
+    },
+    {
+      "epoch": 0.21873128483899204,
+      "grad_norm": 1.7562142136380627,
+      "learning_rate": 0.00018189832207422748,
+      "loss": 1.6897,
+      "step": 4748
+    },
+    {
+      "epoch": 0.21877735292762704,
+      "grad_norm": 1.860935007663604,
+      "learning_rate": 0.00018188975929731357,
+      "loss": 1.2743,
+      "step": 4749
+    },
+    {
+      "epoch": 0.21882342101626204,
+      "grad_norm": 1.8748410665252229,
+      "learning_rate": 0.00018188119469726463,
+      "loss": 1.2917,
+      "step": 4750
+    },
+    {
+      "epoch": 0.21886948910489704,
+      "grad_norm": 2.1322735403011284,
+      "learning_rate": 0.00018187262827427128,
+      "loss": 1.3205,
+      "step": 4751
+    },
+    {
+      "epoch": 0.21891555719353203,
+      "grad_norm": 1.1204650476315492,
+      "learning_rate": 0.0001818640600285242,
+      "loss": 1.8082,
+      "step": 4752
+    },
+    {
+      "epoch": 0.21896162528216703,
+      "grad_norm": 2.382557168419171,
+      "learning_rate": 0.00018185548996021422,
+      "loss": 1.456,
+      "step": 4753
+    },
+    {
+      "epoch": 0.21900769337080206,
+      "grad_norm": 4.995585137602188,
+      "learning_rate": 0.0001818469180695321,
+      "loss": 1.7297,
+      "step": 4754
+    },
+    {
+      "epoch": 0.21905376145943706,
+      "grad_norm": 2.813545096355376,
+      "learning_rate": 0.0001818383443566687,
+      "loss": 1.4496,
+      "step": 4755
+    },
+    {
+      "epoch": 0.21909982954807206,
+      "grad_norm": 1.9987833353155395,
+      "learning_rate": 0.00018182976882181486,
+      "loss": 1.3381,
+      "step": 4756
+    },
+    {
+      "epoch": 0.21914589763670705,
+      "grad_norm": 2.8772065476752813,
+      "learning_rate": 0.00018182119146516155,
+      "loss": 1.2097,
+      "step": 4757
+    },
+    {
+      "epoch": 0.21919196572534205,
+      "grad_norm": 1.4545468697592245,
+      "learning_rate": 0.00018181261228689968,
+      "loss": 1.6609,
+      "step": 4758
+    },
+    {
+      "epoch": 0.21923803381397705,
+      "grad_norm": 2.5285896959750023,
+      "learning_rate": 0.0001818040312872203,
+      "loss": 1.693,
+      "step": 4759
+    },
+    {
+      "epoch": 0.21928410190261205,
+      "grad_norm": 2.7216512406786326,
+      "learning_rate": 0.00018179544846631444,
+      "loss": 1.3295,
+      "step": 4760
+    },
+    {
+      "epoch": 0.21933016999124708,
+      "grad_norm": 2.2995537969783726,
+      "learning_rate": 0.00018178686382437314,
+      "loss": 1.5921,
+      "step": 4761
+    },
+    {
+      "epoch": 0.21937623807988207,
+      "grad_norm": 3.527582607853387,
+      "learning_rate": 0.00018177827736158753,
+      "loss": 1.557,
+      "step": 4762
+    },
+    {
+      "epoch": 0.21942230616851707,
+      "grad_norm": 3.2856384952147843,
+      "learning_rate": 0.00018176968907814884,
+      "loss": 1.4578,
+      "step": 4763
+    },
+    {
+      "epoch": 0.21946837425715207,
+      "grad_norm": 2.0147411085826694,
+      "learning_rate": 0.00018176109897424821,
+      "loss": 1.6275,
+      "step": 4764
+    },
+    {
+      "epoch": 0.21951444234578707,
+      "grad_norm": 1.2990039381369363,
+      "learning_rate": 0.00018175250705007692,
+      "loss": 1.471,
+      "step": 4765
+    },
+    {
+      "epoch": 0.21956051043442207,
+      "grad_norm": 1.9629935935063483,
+      "learning_rate": 0.0001817439133058262,
+      "loss": 1.2803,
+      "step": 4766
+    },
+    {
+      "epoch": 0.21960657852305707,
+      "grad_norm": 1.6692637604011409,
+      "learning_rate": 0.00018173531774168743,
+      "loss": 1.7255,
+      "step": 4767
+    },
+    {
+      "epoch": 0.2196526466116921,
+      "grad_norm": 6.9491107152313685,
+      "learning_rate": 0.00018172672035785196,
+      "loss": 1.5433,
+      "step": 4768
+    },
+    {
+      "epoch": 0.2196987147003271,
+      "grad_norm": 2.1880846411892048,
+      "learning_rate": 0.00018171812115451118,
+      "loss": 1.2662,
+      "step": 4769
+    },
+    {
+      "epoch": 0.2197447827889621,
+      "grad_norm": 3.584454165297971,
+      "learning_rate": 0.00018170952013185656,
+      "loss": 1.4428,
+      "step": 4770
+    },
+    {
+      "epoch": 0.2197908508775971,
+      "grad_norm": 2.45259743641436,
+      "learning_rate": 0.00018170091729007956,
+      "loss": 1.3947,
+      "step": 4771
+    },
+    {
+      "epoch": 0.2198369189662321,
+      "grad_norm": 4.406739979925347,
+      "learning_rate": 0.00018169231262937175,
+      "loss": 1.3546,
+      "step": 4772
+    },
+    {
+      "epoch": 0.21988298705486709,
+      "grad_norm": 2.084452802429144,
+      "learning_rate": 0.00018168370614992467,
+      "loss": 1.5319,
+      "step": 4773
+    },
+    {
+      "epoch": 0.21992905514350208,
+      "grad_norm": 21.345757953499948,
+      "learning_rate": 0.0001816750978519299,
+      "loss": 1.4566,
+      "step": 4774
+    },
+    {
+      "epoch": 0.2199751232321371,
+      "grad_norm": 2.7233270805115324,
+      "learning_rate": 0.00018166648773557914,
+      "loss": 1.3217,
+      "step": 4775
+    },
+    {
+      "epoch": 0.2200211913207721,
+      "grad_norm": 4.468771195312162,
+      "learning_rate": 0.00018165787580106405,
+      "loss": 1.4854,
+      "step": 4776
+    },
+    {
+      "epoch": 0.2200672594094071,
+      "grad_norm": 3.205997657843377,
+      "learning_rate": 0.0001816492620485764,
+      "loss": 1.2735,
+      "step": 4777
+    },
+    {
+      "epoch": 0.2201133274980421,
+      "grad_norm": 4.120713201401263,
+      "learning_rate": 0.0001816406464783079,
+      "loss": 1.5017,
+      "step": 4778
+    },
+    {
+      "epoch": 0.2201593955866771,
+      "grad_norm": 2.617728295575324,
+      "learning_rate": 0.00018163202909045038,
+      "loss": 1.6765,
+      "step": 4779
+    },
+    {
+      "epoch": 0.2202054636753121,
+      "grad_norm": 4.098990606212902,
+      "learning_rate": 0.0001816234098851957,
+      "loss": 1.8365,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2202515317639471,
+      "grad_norm": 2.540503406690186,
+      "learning_rate": 0.0001816147888627358,
+      "loss": 1.6008,
+      "step": 4781
+    },
+    {
+      "epoch": 0.22029759985258213,
+      "grad_norm": 3.170299540521577,
+      "learning_rate": 0.00018160616602326253,
+      "loss": 1.3727,
+      "step": 4782
+    },
+    {
+      "epoch": 0.22034366794121713,
+      "grad_norm": 3.038620798803768,
+      "learning_rate": 0.00018159754136696789,
+      "loss": 1.6777,
+      "step": 4783
+    },
+    {
+      "epoch": 0.22038973602985212,
+      "grad_norm": 4.230668052189593,
+      "learning_rate": 0.0001815889148940439,
+      "loss": 1.6695,
+      "step": 4784
+    },
+    {
+      "epoch": 0.22043580411848712,
+      "grad_norm": 2.928567590917769,
+      "learning_rate": 0.00018158028660468265,
+      "loss": 1.4472,
+      "step": 4785
+    },
+    {
+      "epoch": 0.22048187220712212,
+      "grad_norm": 3.7507517239526678,
+      "learning_rate": 0.00018157165649907616,
+      "loss": 1.3879,
+      "step": 4786
+    },
+    {
+      "epoch": 0.22052794029575712,
+      "grad_norm": 5.73090327659222,
+      "learning_rate": 0.00018156302457741664,
+      "loss": 1.4931,
+      "step": 4787
+    },
+    {
+      "epoch": 0.22057400838439215,
+      "grad_norm": 2.3650014659789984,
+      "learning_rate": 0.00018155439083989618,
+      "loss": 1.4628,
+      "step": 4788
+    },
+    {
+      "epoch": 0.22062007647302714,
+      "grad_norm": 3.3363292032445577,
+      "learning_rate": 0.00018154575528670706,
+      "loss": 1.4945,
+      "step": 4789
+    },
+    {
+      "epoch": 0.22066614456166214,
+      "grad_norm": 2.3659592618971734,
+      "learning_rate": 0.00018153711791804155,
+      "loss": 1.6481,
+      "step": 4790
+    },
+    {
+      "epoch": 0.22071221265029714,
+      "grad_norm": 4.200245297437545,
+      "learning_rate": 0.0001815284787340919,
+      "loss": 1.5709,
+      "step": 4791
+    },
+    {
+      "epoch": 0.22075828073893214,
+      "grad_norm": 5.141068182723186,
+      "learning_rate": 0.0001815198377350505,
+      "loss": 1.3855,
+      "step": 4792
+    },
+    {
+      "epoch": 0.22080434882756714,
+      "grad_norm": 2.5100310468817475,
+      "learning_rate": 0.0001815111949211096,
+      "loss": 1.302,
+      "step": 4793
+    },
+    {
+      "epoch": 0.22085041691620214,
+      "grad_norm": 2.7940910268939123,
+      "learning_rate": 0.00018150255029246177,
+      "loss": 1.5441,
+      "step": 4794
+    },
+    {
+      "epoch": 0.22089648500483716,
+      "grad_norm": 4.658337835176947,
+      "learning_rate": 0.0001814939038492994,
+      "loss": 1.4472,
+      "step": 4795
+    },
+    {
+      "epoch": 0.22094255309347216,
+      "grad_norm": 3.8847023836202452,
+      "learning_rate": 0.00018148525559181497,
+      "loss": 1.7054,
+      "step": 4796
+    },
+    {
+      "epoch": 0.22098862118210716,
+      "grad_norm": 2.4072788248627384,
+      "learning_rate": 0.00018147660552020106,
+      "loss": 1.6677,
+      "step": 4797
+    },
+    {
+      "epoch": 0.22103468927074216,
+      "grad_norm": 3.6042897671373284,
+      "learning_rate": 0.0001814679536346502,
+      "loss": 1.5357,
+      "step": 4798
+    },
+    {
+      "epoch": 0.22108075735937716,
+      "grad_norm": 1.8943522307532945,
+      "learning_rate": 0.0001814592999353551,
+      "loss": 1.5592,
+      "step": 4799
+    },
+    {
+      "epoch": 0.22112682544801215,
+      "grad_norm": 2.5602264404068307,
+      "learning_rate": 0.00018145064442250832,
+      "loss": 1.477,
+      "step": 4800
+    },
+    {
+      "epoch": 0.22117289353664715,
+      "grad_norm": 1.9326931193925077,
+      "learning_rate": 0.00018144198709630258,
+      "loss": 1.354,
+      "step": 4801
+    },
+    {
+      "epoch": 0.22121896162528218,
+      "grad_norm": 2.3917283694961733,
+      "learning_rate": 0.00018143332795693068,
+      "loss": 1.6064,
+      "step": 4802
+    },
+    {
+      "epoch": 0.22126502971391718,
+      "grad_norm": 1.608158059261515,
+      "learning_rate": 0.00018142466700458534,
+      "loss": 1.3462,
+      "step": 4803
+    },
+    {
+      "epoch": 0.22131109780255218,
+      "grad_norm": 3.322465144568839,
+      "learning_rate": 0.00018141600423945942,
+      "loss": 1.1832,
+      "step": 4804
+    },
+    {
+      "epoch": 0.22135716589118717,
+      "grad_norm": 2.3521594484946164,
+      "learning_rate": 0.00018140733966174577,
+      "loss": 1.3862,
+      "step": 4805
+    },
+    {
+      "epoch": 0.22140323397982217,
+      "grad_norm": 1.372222891035379,
+      "learning_rate": 0.00018139867327163724,
+      "loss": 1.7379,
+      "step": 4806
+    },
+    {
+      "epoch": 0.22144930206845717,
+      "grad_norm": 4.3700788242866615,
+      "learning_rate": 0.00018139000506932687,
+      "loss": 1.51,
+      "step": 4807
+    },
+    {
+      "epoch": 0.22149537015709217,
+      "grad_norm": 2.162572180269316,
+      "learning_rate": 0.00018138133505500755,
+      "loss": 1.6224,
+      "step": 4808
+    },
+    {
+      "epoch": 0.2215414382457272,
+      "grad_norm": 5.500661497269863,
+      "learning_rate": 0.00018137266322887236,
+      "loss": 1.2574,
+      "step": 4809
+    },
+    {
+      "epoch": 0.2215875063343622,
+      "grad_norm": 1.9689860490912712,
+      "learning_rate": 0.00018136398959111435,
+      "loss": 1.5058,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2216335744229972,
+      "grad_norm": 3.110292037672002,
+      "learning_rate": 0.00018135531414192662,
+      "loss": 1.4082,
+      "step": 4811
+    },
+    {
+      "epoch": 0.2216796425116322,
+      "grad_norm": 3.4113172690977387,
+      "learning_rate": 0.0001813466368815023,
+      "loss": 1.5763,
+      "step": 4812
+    },
+    {
+      "epoch": 0.2217257106002672,
+      "grad_norm": 4.1957872459386385,
+      "learning_rate": 0.00018133795781003463,
+      "loss": 1.6756,
+      "step": 4813
+    },
+    {
+      "epoch": 0.2217717786889022,
+      "grad_norm": 1.796663124494633,
+      "learning_rate": 0.00018132927692771677,
+      "loss": 1.3581,
+      "step": 4814
+    },
+    {
+      "epoch": 0.2218178467775372,
+      "grad_norm": 3.9538036160714483,
+      "learning_rate": 0.000181320594234742,
+      "loss": 1.8501,
+      "step": 4815
+    },
+    {
+      "epoch": 0.2218639148661722,
+      "grad_norm": 2.1380178670717918,
+      "learning_rate": 0.00018131190973130366,
+      "loss": 1.5404,
+      "step": 4816
+    },
+    {
+      "epoch": 0.2219099829548072,
+      "grad_norm": 3.1230090702162823,
+      "learning_rate": 0.00018130322341759503,
+      "loss": 1.6792,
+      "step": 4817
+    },
+    {
+      "epoch": 0.2219560510434422,
+      "grad_norm": 2.283739456822466,
+      "learning_rate": 0.00018129453529380958,
+      "loss": 1.3864,
+      "step": 4818
+    },
+    {
+      "epoch": 0.2220021191320772,
+      "grad_norm": 4.7224226372505616,
+      "learning_rate": 0.00018128584536014066,
+      "loss": 1.2967,
+      "step": 4819
+    },
+    {
+      "epoch": 0.2220481872207122,
+      "grad_norm": 4.16674991432958,
+      "learning_rate": 0.0001812771536167818,
+      "loss": 1.3889,
+      "step": 4820
+    },
+    {
+      "epoch": 0.2220942553093472,
+      "grad_norm": 2.479361046857653,
+      "learning_rate": 0.0001812684600639264,
+      "loss": 1.6849,
+      "step": 4821
+    },
+    {
+      "epoch": 0.22214032339798223,
+      "grad_norm": 1.8971195722336998,
+      "learning_rate": 0.00018125976470176816,
+      "loss": 1.5023,
+      "step": 4822
+    },
+    {
+      "epoch": 0.22218639148661723,
+      "grad_norm": 2.102170005933197,
+      "learning_rate": 0.0001812510675305006,
+      "loss": 1.6011,
+      "step": 4823
+    },
+    {
+      "epoch": 0.22223245957525223,
+      "grad_norm": 2.481813356115232,
+      "learning_rate": 0.0001812423685503173,
+      "loss": 1.4871,
+      "step": 4824
+    },
+    {
+      "epoch": 0.22227852766388723,
+      "grad_norm": 1.6499044162611263,
+      "learning_rate": 0.00018123366776141194,
+      "loss": 1.5681,
+      "step": 4825
+    },
+    {
+      "epoch": 0.22232459575252222,
+      "grad_norm": 3.80327280959357,
+      "learning_rate": 0.0001812249651639783,
+      "loss": 1.4361,
+      "step": 4826
+    },
+    {
+      "epoch": 0.22237066384115722,
+      "grad_norm": 9.059661382572079,
+      "learning_rate": 0.00018121626075821005,
+      "loss": 1.4399,
+      "step": 4827
+    },
+    {
+      "epoch": 0.22241673192979222,
+      "grad_norm": 2.6891239202369013,
+      "learning_rate": 0.00018120755454430104,
+      "loss": 1.5382,
+      "step": 4828
+    },
+    {
+      "epoch": 0.22246280001842725,
+      "grad_norm": 1.8181010392582986,
+      "learning_rate": 0.00018119884652244503,
+      "loss": 1.2054,
+      "step": 4829
+    },
+    {
+      "epoch": 0.22250886810706225,
+      "grad_norm": 2.726431710136798,
+      "learning_rate": 0.00018119013669283594,
+      "loss": 1.7,
+      "step": 4830
+    },
+    {
+      "epoch": 0.22255493619569724,
+      "grad_norm": 2.056921809585613,
+      "learning_rate": 0.00018118142505566768,
+      "loss": 1.609,
+      "step": 4831
+    },
+    {
+      "epoch": 0.22260100428433224,
+      "grad_norm": 2.493054904095054,
+      "learning_rate": 0.00018117271161113417,
+      "loss": 1.5518,
+      "step": 4832
+    },
+    {
+      "epoch": 0.22264707237296724,
+      "grad_norm": 2.365068702656345,
+      "learning_rate": 0.00018116399635942948,
+      "loss": 1.4763,
+      "step": 4833
+    },
+    {
+      "epoch": 0.22269314046160224,
+      "grad_norm": 3.198549574433039,
+      "learning_rate": 0.00018115527930074752,
+      "loss": 1.6074,
+      "step": 4834
+    },
+    {
+      "epoch": 0.22273920855023724,
+      "grad_norm": 1.9603931849587677,
+      "learning_rate": 0.00018114656043528241,
+      "loss": 1.2375,
+      "step": 4835
+    },
+    {
+      "epoch": 0.22278527663887226,
+      "grad_norm": 1.4283052720081868,
+      "learning_rate": 0.0001811378397632283,
+      "loss": 1.7068,
+      "step": 4836
+    },
+    {
+      "epoch": 0.22283134472750726,
+      "grad_norm": 2.7264037914545867,
+      "learning_rate": 0.00018112911728477928,
+      "loss": 1.6333,
+      "step": 4837
+    },
+    {
+      "epoch": 0.22287741281614226,
+      "grad_norm": 2.1808950971442562,
+      "learning_rate": 0.0001811203930001296,
+      "loss": 1.6411,
+      "step": 4838
+    },
+    {
+      "epoch": 0.22292348090477726,
+      "grad_norm": 1.9491181845287047,
+      "learning_rate": 0.00018111166690947344,
+      "loss": 1.7237,
+      "step": 4839
+    },
+    {
+      "epoch": 0.22296954899341226,
+      "grad_norm": 2.5109342098084113,
+      "learning_rate": 0.0001811029390130051,
+      "loss": 1.6388,
+      "step": 4840
+    },
+    {
+      "epoch": 0.22301561708204726,
+      "grad_norm": 3.4131965956848123,
+      "learning_rate": 0.00018109420931091888,
+      "loss": 1.5746,
+      "step": 4841
+    },
+    {
+      "epoch": 0.22306168517068226,
+      "grad_norm": 13.855297327999432,
+      "learning_rate": 0.00018108547780340913,
+      "loss": 1.6829,
+      "step": 4842
+    },
+    {
+      "epoch": 0.22310775325931728,
+      "grad_norm": 3.202106495526763,
+      "learning_rate": 0.0001810767444906703,
+      "loss": 1.596,
+      "step": 4843
+    },
+    {
+      "epoch": 0.22315382134795228,
+      "grad_norm": 4.809366441049873,
+      "learning_rate": 0.0001810680093728967,
+      "loss": 1.455,
+      "step": 4844
+    },
+    {
+      "epoch": 0.22319988943658728,
+      "grad_norm": 13.068836646417676,
+      "learning_rate": 0.00018105927245028292,
+      "loss": 1.3992,
+      "step": 4845
+    },
+    {
+      "epoch": 0.22324595752522228,
+      "grad_norm": 2.420167201942517,
+      "learning_rate": 0.00018105053372302341,
+      "loss": 1.4218,
+      "step": 4846
+    },
+    {
+      "epoch": 0.22329202561385728,
+      "grad_norm": 3.0834403607873044,
+      "learning_rate": 0.00018104179319131274,
+      "loss": 1.6249,
+      "step": 4847
+    },
+    {
+      "epoch": 0.22333809370249227,
+      "grad_norm": 2.1022074404305657,
+      "learning_rate": 0.0001810330508553455,
+      "loss": 1.4038,
+      "step": 4848
+    },
+    {
+      "epoch": 0.22338416179112727,
+      "grad_norm": 2.912404824155079,
+      "learning_rate": 0.00018102430671531632,
+      "loss": 1.4222,
+      "step": 4849
+    },
+    {
+      "epoch": 0.2234302298797623,
+      "grad_norm": 4.31027818116193,
+      "learning_rate": 0.0001810155607714199,
+      "loss": 1.6091,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2234762979683973,
+      "grad_norm": 1.93027712907654,
+      "learning_rate": 0.0001810068130238509,
+      "loss": 1.5493,
+      "step": 4851
+    },
+    {
+      "epoch": 0.2235223660570323,
+      "grad_norm": 2.199953377837314,
+      "learning_rate": 0.00018099806347280414,
+      "loss": 1.7293,
+      "step": 4852
+    },
+    {
+      "epoch": 0.2235684341456673,
+      "grad_norm": 3.266255758284734,
+      "learning_rate": 0.00018098931211847435,
+      "loss": 1.4235,
+      "step": 4853
+    },
+    {
+      "epoch": 0.2236145022343023,
+      "grad_norm": 3.7308333381178245,
+      "learning_rate": 0.0001809805589610564,
+      "loss": 1.6448,
+      "step": 4854
+    },
+    {
+      "epoch": 0.2236605703229373,
+      "grad_norm": 2.0937783066468874,
+      "learning_rate": 0.00018097180400074514,
+      "loss": 1.5031,
+      "step": 4855
+    },
+    {
+      "epoch": 0.22370663841157232,
+      "grad_norm": 2.96542391859009,
+      "learning_rate": 0.00018096304723773553,
+      "loss": 1.8246,
+      "step": 4856
+    },
+    {
+      "epoch": 0.22375270650020732,
+      "grad_norm": 2.8664237801989554,
+      "learning_rate": 0.00018095428867222245,
+      "loss": 1.082,
+      "step": 4857
+    },
+    {
+      "epoch": 0.22379877458884231,
+      "grad_norm": 2.2454728505402466,
+      "learning_rate": 0.000180945528304401,
+      "loss": 1.5776,
+      "step": 4858
+    },
+    {
+      "epoch": 0.2238448426774773,
+      "grad_norm": 5.209414573572071,
+      "learning_rate": 0.00018093676613446612,
+      "loss": 1.7759,
+      "step": 4859
+    },
+    {
+      "epoch": 0.2238909107661123,
+      "grad_norm": 5.17297348955424,
+      "learning_rate": 0.00018092800216261293,
+      "loss": 1.4194,
+      "step": 4860
+    },
+    {
+      "epoch": 0.2239369788547473,
+      "grad_norm": 2.358587309034008,
+      "learning_rate": 0.00018091923638903653,
+      "loss": 1.2031,
+      "step": 4861
+    },
+    {
+      "epoch": 0.2239830469433823,
+      "grad_norm": 1.9252557084611588,
+      "learning_rate": 0.00018091046881393207,
+      "loss": 1.7559,
+      "step": 4862
+    },
+    {
+      "epoch": 0.22402911503201733,
+      "grad_norm": 2.255404222979017,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 1.4412,
+      "step": 4863
+    },
+    {
+      "epoch": 0.22407518312065233,
+      "grad_norm": 2.101998867709625,
+      "learning_rate": 0.00018089292825991985,
+      "loss": 1.4466,
+      "step": 4864
+    },
+    {
+      "epoch": 0.22412125120928733,
+      "grad_norm": 1.6966644614979869,
+      "learning_rate": 0.00018088415528140253,
+      "loss": 1.8185,
+      "step": 4865
+    },
+    {
+      "epoch": 0.22416731929792233,
+      "grad_norm": 2.7497934759577887,
+      "learning_rate": 0.00018087538050213823,
+      "loss": 1.5282,
+      "step": 4866
+    },
+    {
+      "epoch": 0.22421338738655733,
+      "grad_norm": 2.6790236331641006,
+      "learning_rate": 0.00018086660392232227,
+      "loss": 1.5767,
+      "step": 4867
+    },
+    {
+      "epoch": 0.22425945547519233,
+      "grad_norm": 3.153875421560093,
+      "learning_rate": 0.00018085782554214997,
+      "loss": 1.5673,
+      "step": 4868
+    },
+    {
+      "epoch": 0.22430552356382732,
+      "grad_norm": 3.2392472814911977,
+      "learning_rate": 0.00018084904536181689,
+      "loss": 1.2338,
+      "step": 4869
+    },
+    {
+      "epoch": 0.22435159165246235,
+      "grad_norm": 7.4107674484604225,
+      "learning_rate": 0.0001808402633815184,
+      "loss": 1.4347,
+      "step": 4870
+    },
+    {
+      "epoch": 0.22439765974109735,
+      "grad_norm": 1.7959428337678045,
+      "learning_rate": 0.00018083147960145008,
+      "loss": 1.568,
+      "step": 4871
+    },
+    {
+      "epoch": 0.22444372782973235,
+      "grad_norm": 2.3605385157568417,
+      "learning_rate": 0.00018082269402180744,
+      "loss": 1.56,
+      "step": 4872
+    },
+    {
+      "epoch": 0.22448979591836735,
+      "grad_norm": 4.069852377505225,
+      "learning_rate": 0.00018081390664278612,
+      "loss": 2.0828,
+      "step": 4873
+    },
+    {
+      "epoch": 0.22453586400700234,
+      "grad_norm": 2.5118443786818916,
+      "learning_rate": 0.00018080511746458174,
+      "loss": 1.4631,
+      "step": 4874
+    },
+    {
+      "epoch": 0.22458193209563734,
+      "grad_norm": 1.620690765006355,
+      "learning_rate": 0.00018079632648738996,
+      "loss": 1.1122,
+      "step": 4875
+    },
+    {
+      "epoch": 0.22462800018427234,
+      "grad_norm": 1.5196297822697022,
+      "learning_rate": 0.0001807875337114065,
+      "loss": 1.521,
+      "step": 4876
+    },
+    {
+      "epoch": 0.22467406827290737,
+      "grad_norm": 1.6035106023952685,
+      "learning_rate": 0.0001807787391368271,
+      "loss": 1.7858,
+      "step": 4877
+    },
+    {
+      "epoch": 0.22472013636154237,
+      "grad_norm": 3.603645200497409,
+      "learning_rate": 0.00018076994276384762,
+      "loss": 1.6845,
+      "step": 4878
+    },
+    {
+      "epoch": 0.22476620445017736,
+      "grad_norm": 2.092688953927915,
+      "learning_rate": 0.00018076114459266384,
+      "loss": 1.3854,
+      "step": 4879
+    },
+    {
+      "epoch": 0.22481227253881236,
+      "grad_norm": 2.3198686906928794,
+      "learning_rate": 0.00018075234462347167,
+      "loss": 1.486,
+      "step": 4880
+    },
+    {
+      "epoch": 0.22485834062744736,
+      "grad_norm": 2.3458563557150396,
+      "learning_rate": 0.000180743542856467,
+      "loss": 1.4672,
+      "step": 4881
+    },
+    {
+      "epoch": 0.22490440871608236,
+      "grad_norm": 1.7735094204961628,
+      "learning_rate": 0.00018073473929184578,
+      "loss": 1.9349,
+      "step": 4882
+    },
+    {
+      "epoch": 0.22495047680471736,
+      "grad_norm": 3.5636053807986707,
+      "learning_rate": 0.000180725933929804,
+      "loss": 1.338,
+      "step": 4883
+    },
+    {
+      "epoch": 0.22499654489335238,
+      "grad_norm": 1.9377765074431552,
+      "learning_rate": 0.00018071712677053775,
+      "loss": 1.3421,
+      "step": 4884
+    },
+    {
+      "epoch": 0.22504261298198738,
+      "grad_norm": 7.233453814811935,
+      "learning_rate": 0.00018070831781424307,
+      "loss": 1.4806,
+      "step": 4885
+    },
+    {
+      "epoch": 0.22508868107062238,
+      "grad_norm": 2.017127934593473,
+      "learning_rate": 0.00018069950706111606,
+      "loss": 1.5603,
+      "step": 4886
+    },
+    {
+      "epoch": 0.22513474915925738,
+      "grad_norm": 6.276690476014607,
+      "learning_rate": 0.00018069069451135292,
+      "loss": 1.7119,
+      "step": 4887
+    },
+    {
+      "epoch": 0.22518081724789238,
+      "grad_norm": 2.1567022585776483,
+      "learning_rate": 0.00018068188016514983,
+      "loss": 1.6119,
+      "step": 4888
+    },
+    {
+      "epoch": 0.22522688533652738,
+      "grad_norm": 4.344603993046121,
+      "learning_rate": 0.00018067306402270297,
+      "loss": 1.638,
+      "step": 4889
+    },
+    {
+      "epoch": 0.2252729534251624,
+      "grad_norm": 2.742932648146329,
+      "learning_rate": 0.00018066424608420868,
+      "loss": 1.4457,
+      "step": 4890
+    },
+    {
+      "epoch": 0.2253190215137974,
+      "grad_norm": 3.0655326146396993,
+      "learning_rate": 0.0001806554263498633,
+      "loss": 1.6588,
+      "step": 4891
+    },
+    {
+      "epoch": 0.2253650896024324,
+      "grad_norm": 5.118618264427824,
+      "learning_rate": 0.00018064660481986313,
+      "loss": 1.417,
+      "step": 4892
+    },
+    {
+      "epoch": 0.2254111576910674,
+      "grad_norm": 4.292614417190956,
+      "learning_rate": 0.00018063778149440455,
+      "loss": 1.3586,
+      "step": 4893
+    },
+    {
+      "epoch": 0.2254572257797024,
+      "grad_norm": 1.644860567439699,
+      "learning_rate": 0.00018062895637368406,
+      "loss": 1.6323,
+      "step": 4894
+    },
+    {
+      "epoch": 0.2255032938683374,
+      "grad_norm": 2.6985795634277925,
+      "learning_rate": 0.0001806201294578981,
+      "loss": 1.394,
+      "step": 4895
+    },
+    {
+      "epoch": 0.2255493619569724,
+      "grad_norm": 3.3120788229555655,
+      "learning_rate": 0.0001806113007472432,
+      "loss": 1.3294,
+      "step": 4896
+    },
+    {
+      "epoch": 0.22559543004560742,
+      "grad_norm": 1.8313075462095032,
+      "learning_rate": 0.00018060247024191588,
+      "loss": 1.6006,
+      "step": 4897
+    },
+    {
+      "epoch": 0.22564149813424242,
+      "grad_norm": 23.883830252692185,
+      "learning_rate": 0.00018059363794211276,
+      "loss": 1.5419,
+      "step": 4898
+    },
+    {
+      "epoch": 0.22568756622287742,
+      "grad_norm": 10.916810217518771,
+      "learning_rate": 0.0001805848038480305,
+      "loss": 1.7726,
+      "step": 4899
+    },
+    {
+      "epoch": 0.22573363431151242,
+      "grad_norm": 2.701700096574407,
+      "learning_rate": 0.00018057596795986576,
+      "loss": 1.7393,
+      "step": 4900
+    },
+    {
+      "epoch": 0.22577970240014741,
+      "grad_norm": 2.18334184144237,
+      "learning_rate": 0.00018056713027781528,
+      "loss": 1.718,
+      "step": 4901
+    },
+    {
+      "epoch": 0.2258257704887824,
+      "grad_norm": 7.442468473022997,
+      "learning_rate": 0.00018055829080207574,
+      "loss": 1.7155,
+      "step": 4902
+    },
+    {
+      "epoch": 0.2258718385774174,
+      "grad_norm": 6.614294163988126,
+      "learning_rate": 0.00018054944953284398,
+      "loss": 1.5985,
+      "step": 4903
+    },
+    {
+      "epoch": 0.22591790666605244,
+      "grad_norm": 9.912280706018874,
+      "learning_rate": 0.00018054060647031687,
+      "loss": 1.3308,
+      "step": 4904
+    },
+    {
+      "epoch": 0.22596397475468744,
+      "grad_norm": 3.8608263414998545,
+      "learning_rate": 0.00018053176161469123,
+      "loss": 1.3826,
+      "step": 4905
+    },
+    {
+      "epoch": 0.22601004284332243,
+      "grad_norm": 2.6114630342400242,
+      "learning_rate": 0.00018052291496616402,
+      "loss": 1.7755,
+      "step": 4906
+    },
+    {
+      "epoch": 0.22605611093195743,
+      "grad_norm": 2.8209263303894767,
+      "learning_rate": 0.00018051406652493216,
+      "loss": 1.6404,
+      "step": 4907
+    },
+    {
+      "epoch": 0.22610217902059243,
+      "grad_norm": 2.582487727596951,
+      "learning_rate": 0.00018050521629119266,
+      "loss": 1.8865,
+      "step": 4908
+    },
+    {
+      "epoch": 0.22614824710922743,
+      "grad_norm": 2.2991103057361073,
+      "learning_rate": 0.0001804963642651426,
+      "loss": 1.8439,
+      "step": 4909
+    },
+    {
+      "epoch": 0.22619431519786243,
+      "grad_norm": 3.1091495846595816,
+      "learning_rate": 0.00018048751044697894,
+      "loss": 1.7108,
+      "step": 4910
+    },
+    {
+      "epoch": 0.22624038328649745,
+      "grad_norm": 4.301720122685178,
+      "learning_rate": 0.0001804786548368989,
+      "loss": 1.589,
+      "step": 4911
+    },
+    {
+      "epoch": 0.22628645137513245,
+      "grad_norm": 6.229586330173048,
+      "learning_rate": 0.00018046979743509958,
+      "loss": 1.4059,
+      "step": 4912
+    },
+    {
+      "epoch": 0.22633251946376745,
+      "grad_norm": 2.035371693937343,
+      "learning_rate": 0.00018046093824177823,
+      "loss": 1.7028,
+      "step": 4913
+    },
+    {
+      "epoch": 0.22637858755240245,
+      "grad_norm": 5.0964222811136635,
+      "learning_rate": 0.00018045207725713203,
+      "loss": 1.3924,
+      "step": 4914
+    },
+    {
+      "epoch": 0.22642465564103745,
+      "grad_norm": 2.545744946535896,
+      "learning_rate": 0.0001804432144813583,
+      "loss": 1.6191,
+      "step": 4915
+    },
+    {
+      "epoch": 0.22647072372967245,
+      "grad_norm": 2.0678578291846703,
+      "learning_rate": 0.00018043434991465431,
+      "loss": 1.6299,
+      "step": 4916
+    },
+    {
+      "epoch": 0.22651679181830744,
+      "grad_norm": 3.404673814152892,
+      "learning_rate": 0.00018042548355721744,
+      "loss": 1.5872,
+      "step": 4917
+    },
+    {
+      "epoch": 0.22656285990694247,
+      "grad_norm": 1.8910096473876095,
+      "learning_rate": 0.00018041661540924508,
+      "loss": 1.4373,
+      "step": 4918
+    },
+    {
+      "epoch": 0.22660892799557747,
+      "grad_norm": 4.834802559159588,
+      "learning_rate": 0.00018040774547093465,
+      "loss": 1.4024,
+      "step": 4919
+    },
+    {
+      "epoch": 0.22665499608421247,
+      "grad_norm": 1.7159134763489086,
+      "learning_rate": 0.00018039887374248365,
+      "loss": 1.6129,
+      "step": 4920
+    },
+    {
+      "epoch": 0.22670106417284747,
+      "grad_norm": 1.818397604682267,
+      "learning_rate": 0.00018039000022408957,
+      "loss": 1.6807,
+      "step": 4921
+    },
+    {
+      "epoch": 0.22674713226148246,
+      "grad_norm": 3.751739936933711,
+      "learning_rate": 0.00018038112491594997,
+      "loss": 1.7985,
+      "step": 4922
+    },
+    {
+      "epoch": 0.22679320035011746,
+      "grad_norm": 1.466195359485693,
+      "learning_rate": 0.0001803722478182625,
+      "loss": 1.3456,
+      "step": 4923
+    },
+    {
+      "epoch": 0.2268392684387525,
+      "grad_norm": 1.5886669724184392,
+      "learning_rate": 0.0001803633689312247,
+      "loss": 1.4037,
+      "step": 4924
+    },
+    {
+      "epoch": 0.2268853365273875,
+      "grad_norm": 5.049182964243304,
+      "learning_rate": 0.00018035448825503428,
+      "loss": 1.0408,
+      "step": 4925
+    },
+    {
+      "epoch": 0.2269314046160225,
+      "grad_norm": 1.9310899835493156,
+      "learning_rate": 0.00018034560578988895,
+      "loss": 1.6366,
+      "step": 4926
+    },
+    {
+      "epoch": 0.22697747270465748,
+      "grad_norm": 2.5833653071711287,
+      "learning_rate": 0.0001803367215359865,
+      "loss": 1.3192,
+      "step": 4927
+    },
+    {
+      "epoch": 0.22702354079329248,
+      "grad_norm": 2.713730754664407,
+      "learning_rate": 0.0001803278354935247,
+      "loss": 1.8183,
+      "step": 4928
+    },
+    {
+      "epoch": 0.22706960888192748,
+      "grad_norm": 2.082589060327675,
+      "learning_rate": 0.00018031894766270133,
+      "loss": 1.6771,
+      "step": 4929
+    },
+    {
+      "epoch": 0.22711567697056248,
+      "grad_norm": 2.005853998462955,
+      "learning_rate": 0.00018031005804371433,
+      "loss": 1.9287,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2271617450591975,
+      "grad_norm": 2.8604734719060514,
+      "learning_rate": 0.00018030116663676158,
+      "loss": 1.7417,
+      "step": 4931
+    },
+    {
+      "epoch": 0.2272078131478325,
+      "grad_norm": 1.8259784486100104,
+      "learning_rate": 0.00018029227344204104,
+      "loss": 1.4729,
+      "step": 4932
+    },
+    {
+      "epoch": 0.2272538812364675,
+      "grad_norm": 17.84442862695898,
+      "learning_rate": 0.00018028337845975073,
+      "loss": 1.4099,
+      "step": 4933
+    },
+    {
+      "epoch": 0.2272999493251025,
+      "grad_norm": 6.26104497312099,
+      "learning_rate": 0.00018027448169008863,
+      "loss": 1.4852,
+      "step": 4934
+    },
+    {
+      "epoch": 0.2273460174137375,
+      "grad_norm": 2.7837862383657854,
+      "learning_rate": 0.00018026558313325282,
+      "loss": 1.6523,
+      "step": 4935
+    },
+    {
+      "epoch": 0.2273920855023725,
+      "grad_norm": 2.598433826175222,
+      "learning_rate": 0.00018025668278944147,
+      "loss": 1.6614,
+      "step": 4936
+    },
+    {
+      "epoch": 0.2274381535910075,
+      "grad_norm": 2.038006395108329,
+      "learning_rate": 0.0001802477806588527,
+      "loss": 1.5265,
+      "step": 4937
+    },
+    {
+      "epoch": 0.22748422167964252,
+      "grad_norm": 33.9769538676196,
+      "learning_rate": 0.00018023887674168464,
+      "loss": 1.6261,
+      "step": 4938
+    },
+    {
+      "epoch": 0.22753028976827752,
+      "grad_norm": 4.923330305308953,
+      "learning_rate": 0.00018022997103813563,
+      "loss": 1.735,
+      "step": 4939
+    },
+    {
+      "epoch": 0.22757635785691252,
+      "grad_norm": 10.13439608183118,
+      "learning_rate": 0.0001802210635484038,
+      "loss": 1.7693,
+      "step": 4940
+    },
+    {
+      "epoch": 0.22762242594554752,
+      "grad_norm": 11.82366555568307,
+      "learning_rate": 0.00018021215427268758,
+      "loss": 1.7294,
+      "step": 4941
+    },
+    {
+      "epoch": 0.22766849403418252,
+      "grad_norm": 3.3165671533086583,
+      "learning_rate": 0.0001802032432111853,
+      "loss": 1.3476,
+      "step": 4942
+    },
+    {
+      "epoch": 0.22771456212281752,
+      "grad_norm": 2.5065618103157963,
+      "learning_rate": 0.0001801943303640953,
+      "loss": 1.9098,
+      "step": 4943
+    },
+    {
+      "epoch": 0.2277606302114525,
+      "grad_norm": 6.827289667708476,
+      "learning_rate": 0.000180185415731616,
+      "loss": 1.6237,
+      "step": 4944
+    },
+    {
+      "epoch": 0.22780669830008754,
+      "grad_norm": 3.383924949305938,
+      "learning_rate": 0.000180176499313946,
+      "loss": 1.7873,
+      "step": 4945
+    },
+    {
+      "epoch": 0.22785276638872254,
+      "grad_norm": 11.663541340160563,
+      "learning_rate": 0.00018016758111128367,
+      "loss": 1.6754,
+      "step": 4946
+    },
+    {
+      "epoch": 0.22789883447735754,
+      "grad_norm": 5.597257628300199,
+      "learning_rate": 0.0001801586611238276,
+      "loss": 1.6232,
+      "step": 4947
+    },
+    {
+      "epoch": 0.22794490256599254,
+      "grad_norm": 3.503869175404472,
+      "learning_rate": 0.0001801497393517764,
+      "loss": 1.7853,
+      "step": 4948
+    },
+    {
+      "epoch": 0.22799097065462753,
+      "grad_norm": 19.01896049441371,
+      "learning_rate": 0.00018014081579532864,
+      "loss": 1.911,
+      "step": 4949
+    },
+    {
+      "epoch": 0.22803703874326253,
+      "grad_norm": 6.690242101163941,
+      "learning_rate": 0.00018013189045468307,
+      "loss": 1.762,
+      "step": 4950
+    },
+    {
+      "epoch": 0.22808310683189753,
+      "grad_norm": 3.068527059886826,
+      "learning_rate": 0.00018012296333003836,
+      "loss": 1.7697,
+      "step": 4951
+    },
+    {
+      "epoch": 0.22812917492053256,
+      "grad_norm": 3.1987405776017086,
+      "learning_rate": 0.00018011403442159322,
+      "loss": 1.6337,
+      "step": 4952
+    },
+    {
+      "epoch": 0.22817524300916756,
+      "grad_norm": 3.980831763840996,
+      "learning_rate": 0.00018010510372954647,
+      "loss": 1.808,
+      "step": 4953
+    },
+    {
+      "epoch": 0.22822131109780255,
+      "grad_norm": 3.0608755259233065,
+      "learning_rate": 0.00018009617125409695,
+      "loss": 1.6566,
+      "step": 4954
+    },
+    {
+      "epoch": 0.22826737918643755,
+      "grad_norm": 3.79828320653693,
+      "learning_rate": 0.00018008723699544353,
+      "loss": 1.6499,
+      "step": 4955
+    },
+    {
+      "epoch": 0.22831344727507255,
+      "grad_norm": 4.364199111409528,
+      "learning_rate": 0.00018007830095378505,
+      "loss": 1.6538,
+      "step": 4956
+    },
+    {
+      "epoch": 0.22835951536370755,
+      "grad_norm": 2.166272426127184,
+      "learning_rate": 0.00018006936312932057,
+      "loss": 1.6032,
+      "step": 4957
+    },
+    {
+      "epoch": 0.22840558345234258,
+      "grad_norm": 5.276951677635791,
+      "learning_rate": 0.00018006042352224896,
+      "loss": 2.1008,
+      "step": 4958
+    },
+    {
+      "epoch": 0.22845165154097757,
+      "grad_norm": 2.730258871988346,
+      "learning_rate": 0.0001800514821327693,
+      "loss": 1.6769,
+      "step": 4959
+    },
+    {
+      "epoch": 0.22849771962961257,
+      "grad_norm": 4.213174781924843,
+      "learning_rate": 0.00018004253896108068,
+      "loss": 1.9753,
+      "step": 4960
+    },
+    {
+      "epoch": 0.22854378771824757,
+      "grad_norm": 2.633172640727933,
+      "learning_rate": 0.00018003359400738214,
+      "loss": 1.4411,
+      "step": 4961
+    },
+    {
+      "epoch": 0.22858985580688257,
+      "grad_norm": 4.9166868159404915,
+      "learning_rate": 0.00018002464727187288,
+      "loss": 1.8502,
+      "step": 4962
+    },
+    {
+      "epoch": 0.22863592389551757,
+      "grad_norm": 2.86255991018753,
+      "learning_rate": 0.00018001569875475206,
+      "loss": 1.4148,
+      "step": 4963
+    },
+    {
+      "epoch": 0.22868199198415257,
+      "grad_norm": 2.355554169314356,
+      "learning_rate": 0.0001800067484562189,
+      "loss": 1.7942,
+      "step": 4964
+    },
+    {
+      "epoch": 0.2287280600727876,
+      "grad_norm": 5.922743235768164,
+      "learning_rate": 0.00017999779637647265,
+      "loss": 1.6396,
+      "step": 4965
+    },
+    {
+      "epoch": 0.2287741281614226,
+      "grad_norm": 2.3957905061123443,
+      "learning_rate": 0.00017998884251571264,
+      "loss": 1.809,
+      "step": 4966
+    },
+    {
+      "epoch": 0.2288201962500576,
+      "grad_norm": 5.399916207344086,
+      "learning_rate": 0.0001799798868741382,
+      "loss": 1.4839,
+      "step": 4967
+    },
+    {
+      "epoch": 0.2288662643386926,
+      "grad_norm": 5.395018160259882,
+      "learning_rate": 0.00017997092945194873,
+      "loss": 1.5592,
+      "step": 4968
+    },
+    {
+      "epoch": 0.2289123324273276,
+      "grad_norm": 2.4850767286385556,
+      "learning_rate": 0.00017996197024934362,
+      "loss": 1.5289,
+      "step": 4969
+    },
+    {
+      "epoch": 0.22895840051596258,
+      "grad_norm": 2.4486220623958643,
+      "learning_rate": 0.00017995300926652237,
+      "loss": 1.8418,
+      "step": 4970
+    },
+    {
+      "epoch": 0.22900446860459758,
+      "grad_norm": 5.12025203901543,
+      "learning_rate": 0.00017994404650368445,
+      "loss": 1.6149,
+      "step": 4971
+    },
+    {
+      "epoch": 0.2290505366932326,
+      "grad_norm": 2.929781102164631,
+      "learning_rate": 0.0001799350819610294,
+      "loss": 1.565,
+      "step": 4972
+    },
+    {
+      "epoch": 0.2290966047818676,
+      "grad_norm": 4.324977582729226,
+      "learning_rate": 0.0001799261156387568,
+      "loss": 1.8653,
+      "step": 4973
+    },
+    {
+      "epoch": 0.2291426728705026,
+      "grad_norm": 3.889369402720255,
+      "learning_rate": 0.0001799171475370663,
+      "loss": 1.6633,
+      "step": 4974
+    },
+    {
+      "epoch": 0.2291887409591376,
+      "grad_norm": 2.412218051064014,
+      "learning_rate": 0.00017990817765615753,
+      "loss": 1.8669,
+      "step": 4975
+    },
+    {
+      "epoch": 0.2292348090477726,
+      "grad_norm": 2.9626985499745193,
+      "learning_rate": 0.0001798992059962302,
+      "loss": 1.6493,
+      "step": 4976
+    },
+    {
+      "epoch": 0.2292808771364076,
+      "grad_norm": 4.01796921114513,
+      "learning_rate": 0.00017989023255748405,
+      "loss": 1.5157,
+      "step": 4977
+    },
+    {
+      "epoch": 0.2293269452250426,
+      "grad_norm": 12.041025067451773,
+      "learning_rate": 0.00017988125734011884,
+      "loss": 1.7746,
+      "step": 4978
+    },
+    {
+      "epoch": 0.22937301331367763,
+      "grad_norm": 2.0184159718868258,
+      "learning_rate": 0.00017987228034433442,
+      "loss": 1.6104,
+      "step": 4979
+    },
+    {
+      "epoch": 0.22941908140231262,
+      "grad_norm": 4.6701397135191645,
+      "learning_rate": 0.00017986330157033063,
+      "loss": 1.6442,
+      "step": 4980
+    },
+    {
+      "epoch": 0.22946514949094762,
+      "grad_norm": 3.0619860349186974,
+      "learning_rate": 0.00017985432101830736,
+      "loss": 1.7011,
+      "step": 4981
+    },
+    {
+      "epoch": 0.22951121757958262,
+      "grad_norm": 3.709260966588875,
+      "learning_rate": 0.00017984533868846457,
+      "loss": 1.7011,
+      "step": 4982
+    },
+    {
+      "epoch": 0.22955728566821762,
+      "grad_norm": 2.7383272584313096,
+      "learning_rate": 0.00017983635458100218,
+      "loss": 1.57,
+      "step": 4983
+    },
+    {
+      "epoch": 0.22960335375685262,
+      "grad_norm": 2.3248629927763913,
+      "learning_rate": 0.00017982736869612026,
+      "loss": 1.6287,
+      "step": 4984
+    },
+    {
+      "epoch": 0.22964942184548764,
+      "grad_norm": 4.2141502728179745,
+      "learning_rate": 0.00017981838103401886,
+      "loss": 1.3781,
+      "step": 4985
+    },
+    {
+      "epoch": 0.22969548993412264,
+      "grad_norm": 4.314957207635838,
+      "learning_rate": 0.00017980939159489807,
+      "loss": 1.7507,
+      "step": 4986
+    },
+    {
+      "epoch": 0.22974155802275764,
+      "grad_norm": 4.81363589268383,
+      "learning_rate": 0.000179800400378958,
+      "loss": 1.504,
+      "step": 4987
+    },
+    {
+      "epoch": 0.22978762611139264,
+      "grad_norm": 2.995062587970764,
+      "learning_rate": 0.00017979140738639887,
+      "loss": 1.4061,
+      "step": 4988
+    },
+    {
+      "epoch": 0.22983369420002764,
+      "grad_norm": 2.5048386525583335,
+      "learning_rate": 0.00017978241261742083,
+      "loss": 1.8222,
+      "step": 4989
+    },
+    {
+      "epoch": 0.22987976228866264,
+      "grad_norm": 8.153729998095354,
+      "learning_rate": 0.0001797734160722242,
+      "loss": 1.6205,
+      "step": 4990
+    },
+    {
+      "epoch": 0.22992583037729764,
+      "grad_norm": 3.2912990151978647,
+      "learning_rate": 0.00017976441775100925,
+      "loss": 1.5858,
+      "step": 4991
+    },
+    {
+      "epoch": 0.22997189846593266,
+      "grad_norm": 3.10417636120083,
+      "learning_rate": 0.0001797554176539763,
+      "loss": 1.7253,
+      "step": 4992
+    },
+    {
+      "epoch": 0.23001796655456766,
+      "grad_norm": 2.4725039987766166,
+      "learning_rate": 0.0001797464157813257,
+      "loss": 1.4444,
+      "step": 4993
+    },
+    {
+      "epoch": 0.23006403464320266,
+      "grad_norm": 2.924409925013682,
+      "learning_rate": 0.00017973741213325793,
+      "loss": 1.4039,
+      "step": 4994
+    },
+    {
+      "epoch": 0.23011010273183766,
+      "grad_norm": 9.550653344960535,
+      "learning_rate": 0.00017972840670997338,
+      "loss": 1.937,
+      "step": 4995
+    },
+    {
+      "epoch": 0.23015617082047266,
+      "grad_norm": 3.0875927050947927,
+      "learning_rate": 0.00017971939951167256,
+      "loss": 1.6116,
+      "step": 4996
+    },
+    {
+      "epoch": 0.23020223890910765,
+      "grad_norm": 1.6320131052828097,
+      "learning_rate": 0.00017971039053855597,
+      "loss": 1.7099,
+      "step": 4997
+    },
+    {
+      "epoch": 0.23024830699774265,
+      "grad_norm": 4.268838276559864,
+      "learning_rate": 0.00017970137979082424,
+      "loss": 1.9023,
+      "step": 4998
+    },
+    {
+      "epoch": 0.23029437508637768,
+      "grad_norm": 7.782110396597422,
+      "learning_rate": 0.00017969236726867794,
+      "loss": 1.6064,
+      "step": 4999
+    },
+    {
+      "epoch": 0.23034044317501268,
+      "grad_norm": 2.5123652650442025,
+      "learning_rate": 0.00017968335297231772,
+      "loss": 1.636,
+      "step": 5000
+    },
+    {
+      "epoch": 0.23038651126364768,
+      "grad_norm": 7.785622891427669,
+      "learning_rate": 0.0001796743369019443,
+      "loss": 1.5924,
+      "step": 5001
+    },
+    {
+      "epoch": 0.23043257935228267,
+      "grad_norm": 1.9986811127304207,
+      "learning_rate": 0.00017966531905775837,
+      "loss": 1.557,
+      "step": 5002
+    },
+    {
+      "epoch": 0.23047864744091767,
+      "grad_norm": 10.968527112084685,
+      "learning_rate": 0.0001796562994399607,
+      "loss": 1.7919,
+      "step": 5003
+    },
+    {
+      "epoch": 0.23052471552955267,
+      "grad_norm": 18.032564855232756,
+      "learning_rate": 0.0001796472780487521,
+      "loss": 1.5551,
+      "step": 5004
+    },
+    {
+      "epoch": 0.23057078361818767,
+      "grad_norm": 1.367656649278642,
+      "learning_rate": 0.00017963825488433342,
+      "loss": 1.4405,
+      "step": 5005
+    },
+    {
+      "epoch": 0.2306168517068227,
+      "grad_norm": 1.8915798992847779,
+      "learning_rate": 0.00017962922994690556,
+      "loss": 1.5748,
+      "step": 5006
+    },
+    {
+      "epoch": 0.2306629197954577,
+      "grad_norm": 2.203498015145907,
+      "learning_rate": 0.00017962020323666943,
+      "loss": 1.8843,
+      "step": 5007
+    },
+    {
+      "epoch": 0.2307089878840927,
+      "grad_norm": 2.1393367916432986,
+      "learning_rate": 0.00017961117475382598,
+      "loss": 1.4602,
+      "step": 5008
+    },
+    {
+      "epoch": 0.2307550559727277,
+      "grad_norm": 2.574632881459042,
+      "learning_rate": 0.00017960214449857622,
+      "loss": 1.6622,
+      "step": 5009
+    },
+    {
+      "epoch": 0.2308011240613627,
+      "grad_norm": 1.8859132420657676,
+      "learning_rate": 0.00017959311247112123,
+      "loss": 1.3211,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2308471921499977,
+      "grad_norm": 3.688595003367873,
+      "learning_rate": 0.00017958407867166206,
+      "loss": 1.6108,
+      "step": 5011
+    },
+    {
+      "epoch": 0.23089326023863269,
+      "grad_norm": 1.9730968140230456,
+      "learning_rate": 0.0001795750431003998,
+      "loss": 1.6626,
+      "step": 5012
+    },
+    {
+      "epoch": 0.2309393283272677,
+      "grad_norm": 2.55407753418264,
+      "learning_rate": 0.00017956600575753568,
+      "loss": 1.3538,
+      "step": 5013
+    },
+    {
+      "epoch": 0.2309853964159027,
+      "grad_norm": 8.240581840526902,
+      "learning_rate": 0.00017955696664327087,
+      "loss": 1.2655,
+      "step": 5014
+    },
+    {
+      "epoch": 0.2310314645045377,
+      "grad_norm": 3.2267653536317744,
+      "learning_rate": 0.0001795479257578066,
+      "loss": 1.4936,
+      "step": 5015
+    },
+    {
+      "epoch": 0.2310775325931727,
+      "grad_norm": 3.4573755322000927,
+      "learning_rate": 0.00017953888310134414,
+      "loss": 1.371,
+      "step": 5016
+    },
+    {
+      "epoch": 0.2311236006818077,
+      "grad_norm": 2.325064600104719,
+      "learning_rate": 0.00017952983867408487,
+      "loss": 1.4852,
+      "step": 5017
+    },
+    {
+      "epoch": 0.2311696687704427,
+      "grad_norm": 2.813526124444756,
+      "learning_rate": 0.00017952079247623005,
+      "loss": 1.5659,
+      "step": 5018
+    },
+    {
+      "epoch": 0.23121573685907773,
+      "grad_norm": 2.9249242943829805,
+      "learning_rate": 0.00017951174450798118,
+      "loss": 1.3882,
+      "step": 5019
+    },
+    {
+      "epoch": 0.23126180494771273,
+      "grad_norm": 2.449863263076148,
+      "learning_rate": 0.00017950269476953965,
+      "loss": 1.5957,
+      "step": 5020
+    },
+    {
+      "epoch": 0.23130787303634773,
+      "grad_norm": 2.1326338355303918,
+      "learning_rate": 0.00017949364326110692,
+      "loss": 1.3334,
+      "step": 5021
+    },
+    {
+      "epoch": 0.23135394112498273,
+      "grad_norm": 2.3247913542954066,
+      "learning_rate": 0.00017948458998288453,
+      "loss": 1.8083,
+      "step": 5022
+    },
+    {
+      "epoch": 0.23140000921361772,
+      "grad_norm": 1.5079633154287597,
+      "learning_rate": 0.00017947553493507403,
+      "loss": 1.5021,
+      "step": 5023
+    },
+    {
+      "epoch": 0.23144607730225272,
+      "grad_norm": 1.991671316938538,
+      "learning_rate": 0.00017946647811787703,
+      "loss": 1.5962,
+      "step": 5024
+    },
+    {
+      "epoch": 0.23149214539088772,
+      "grad_norm": 3.3350032611815172,
+      "learning_rate": 0.00017945741953149513,
+      "loss": 1.705,
+      "step": 5025
+    },
+    {
+      "epoch": 0.23153821347952275,
+      "grad_norm": 1.5604034961705964,
+      "learning_rate": 0.00017944835917613002,
+      "loss": 1.6609,
+      "step": 5026
+    },
+    {
+      "epoch": 0.23158428156815775,
+      "grad_norm": 7.3255536363408575,
+      "learning_rate": 0.00017943929705198342,
+      "loss": 1.4107,
+      "step": 5027
+    },
+    {
+      "epoch": 0.23163034965679274,
+      "grad_norm": 2.7903368502028614,
+      "learning_rate": 0.0001794302331592571,
+      "loss": 1.4811,
+      "step": 5028
+    },
+    {
+      "epoch": 0.23167641774542774,
+      "grad_norm": 1.6302570199767323,
+      "learning_rate": 0.00017942116749815283,
+      "loss": 1.6784,
+      "step": 5029
+    },
+    {
+      "epoch": 0.23172248583406274,
+      "grad_norm": 5.125482053386532,
+      "learning_rate": 0.00017941210006887243,
+      "loss": 1.2506,
+      "step": 5030
+    },
+    {
+      "epoch": 0.23176855392269774,
+      "grad_norm": 4.882194920539343,
+      "learning_rate": 0.00017940303087161778,
+      "loss": 1.5881,
+      "step": 5031
+    },
+    {
+      "epoch": 0.23181462201133274,
+      "grad_norm": 1.4425819087849021,
+      "learning_rate": 0.00017939395990659081,
+      "loss": 1.5626,
+      "step": 5032
+    },
+    {
+      "epoch": 0.23186069009996776,
+      "grad_norm": 2.9827263145554115,
+      "learning_rate": 0.00017938488717399346,
+      "loss": 1.381,
+      "step": 5033
+    },
+    {
+      "epoch": 0.23190675818860276,
+      "grad_norm": 1.6426928041635611,
+      "learning_rate": 0.0001793758126740277,
+      "loss": 1.4465,
+      "step": 5034
+    },
+    {
+      "epoch": 0.23195282627723776,
+      "grad_norm": 1.6758726802410695,
+      "learning_rate": 0.0001793667364068956,
+      "loss": 1.4183,
+      "step": 5035
+    },
+    {
+      "epoch": 0.23199889436587276,
+      "grad_norm": 1.4414569380854318,
+      "learning_rate": 0.00017935765837279917,
+      "loss": 1.5792,
+      "step": 5036
+    },
+    {
+      "epoch": 0.23204496245450776,
+      "grad_norm": 1.551897545520471,
+      "learning_rate": 0.00017934857857194054,
+      "loss": 1.8841,
+      "step": 5037
+    },
+    {
+      "epoch": 0.23209103054314276,
+      "grad_norm": 1.4066180492602953,
+      "learning_rate": 0.00017933949700452186,
+      "loss": 1.4177,
+      "step": 5038
+    },
+    {
+      "epoch": 0.23213709863177776,
+      "grad_norm": 2.15813482156287,
+      "learning_rate": 0.00017933041367074534,
+      "loss": 1.707,
+      "step": 5039
+    },
+    {
+      "epoch": 0.23218316672041278,
+      "grad_norm": 1.2891984933920788,
+      "learning_rate": 0.00017932132857081316,
+      "loss": 1.2566,
+      "step": 5040
+    },
+    {
+      "epoch": 0.23222923480904778,
+      "grad_norm": 1.6431570751409206,
+      "learning_rate": 0.00017931224170492763,
+      "loss": 1.392,
+      "step": 5041
+    },
+    {
+      "epoch": 0.23227530289768278,
+      "grad_norm": 1.58725106088266,
+      "learning_rate": 0.000179303153073291,
+      "loss": 1.5473,
+      "step": 5042
+    },
+    {
+      "epoch": 0.23232137098631778,
+      "grad_norm": 2.5466133719347934,
+      "learning_rate": 0.00017929406267610566,
+      "loss": 1.4781,
+      "step": 5043
+    },
+    {
+      "epoch": 0.23236743907495278,
+      "grad_norm": 11.184807865903482,
+      "learning_rate": 0.00017928497051357398,
+      "loss": 1.5151,
+      "step": 5044
+    },
+    {
+      "epoch": 0.23241350716358777,
+      "grad_norm": 4.0938172371402395,
+      "learning_rate": 0.00017927587658589839,
+      "loss": 1.5213,
+      "step": 5045
+    },
+    {
+      "epoch": 0.23245957525222277,
+      "grad_norm": 1.2435889381031142,
+      "learning_rate": 0.0001792667808932813,
+      "loss": 1.6281,
+      "step": 5046
+    },
+    {
+      "epoch": 0.2325056433408578,
+      "grad_norm": 2.525515406816452,
+      "learning_rate": 0.00017925768343592525,
+      "loss": 1.443,
+      "step": 5047
+    },
+    {
+      "epoch": 0.2325517114294928,
+      "grad_norm": 1.3902523298201324,
+      "learning_rate": 0.0001792485842140328,
+      "loss": 1.5286,
+      "step": 5048
+    },
+    {
+      "epoch": 0.2325977795181278,
+      "grad_norm": 1.6836197164203086,
+      "learning_rate": 0.0001792394832278065,
+      "loss": 1.5253,
+      "step": 5049
+    },
+    {
+      "epoch": 0.2326438476067628,
+      "grad_norm": 1.304241625684314,
+      "learning_rate": 0.00017923038047744895,
+      "loss": 1.6224,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2326899156953978,
+      "grad_norm": 1.5092242238645148,
+      "learning_rate": 0.00017922127596316285,
+      "loss": 1.5038,
+      "step": 5051
+    },
+    {
+      "epoch": 0.2327359837840328,
+      "grad_norm": 1.5889916348724926,
+      "learning_rate": 0.00017921216968515088,
+      "loss": 1.5971,
+      "step": 5052
+    },
+    {
+      "epoch": 0.23278205187266782,
+      "grad_norm": 2.338180405452097,
+      "learning_rate": 0.00017920306164361576,
+      "loss": 1.6455,
+      "step": 5053
+    },
+    {
+      "epoch": 0.23282811996130282,
+      "grad_norm": 2.1259981532791956,
+      "learning_rate": 0.00017919395183876028,
+      "loss": 1.758,
+      "step": 5054
+    },
+    {
+      "epoch": 0.2328741880499378,
+      "grad_norm": 2.0280144833851845,
+      "learning_rate": 0.00017918484027078723,
+      "loss": 1.1542,
+      "step": 5055
+    },
+    {
+      "epoch": 0.2329202561385728,
+      "grad_norm": 1.2718292272514533,
+      "learning_rate": 0.00017917572693989953,
+      "loss": 1.4955,
+      "step": 5056
+    },
+    {
+      "epoch": 0.2329663242272078,
+      "grad_norm": 1.35349196791481,
+      "learning_rate": 0.00017916661184630002,
+      "loss": 1.5165,
+      "step": 5057
+    },
+    {
+      "epoch": 0.2330123923158428,
+      "grad_norm": 2.908453394772926,
+      "learning_rate": 0.00017915749499019162,
+      "loss": 1.46,
+      "step": 5058
+    },
+    {
+      "epoch": 0.2330584604044778,
+      "grad_norm": 2.8515712306824668,
+      "learning_rate": 0.00017914837637177732,
+      "loss": 1.1848,
+      "step": 5059
+    },
+    {
+      "epoch": 0.23310452849311283,
+      "grad_norm": 2.123523589039672,
+      "learning_rate": 0.00017913925599126013,
+      "loss": 1.6006,
+      "step": 5060
+    },
+    {
+      "epoch": 0.23315059658174783,
+      "grad_norm": 1.5190503195935727,
+      "learning_rate": 0.0001791301338488431,
+      "loss": 1.6931,
+      "step": 5061
+    },
+    {
+      "epoch": 0.23319666467038283,
+      "grad_norm": 1.1129055952715428,
+      "learning_rate": 0.0001791210099447293,
+      "loss": 1.6618,
+      "step": 5062
+    },
+    {
+      "epoch": 0.23324273275901783,
+      "grad_norm": 2.295015327628432,
+      "learning_rate": 0.0001791118842791219,
+      "loss": 1.4871,
+      "step": 5063
+    },
+    {
+      "epoch": 0.23328880084765283,
+      "grad_norm": 0.997566729458358,
+      "learning_rate": 0.00017910275685222406,
+      "loss": 1.4151,
+      "step": 5064
+    },
+    {
+      "epoch": 0.23333486893628783,
+      "grad_norm": 1.3783397927553194,
+      "learning_rate": 0.00017909362766423893,
+      "loss": 1.472,
+      "step": 5065
+    },
+    {
+      "epoch": 0.23338093702492282,
+      "grad_norm": 1.0547283797993825,
+      "learning_rate": 0.00017908449671536985,
+      "loss": 1.4007,
+      "step": 5066
+    },
+    {
+      "epoch": 0.23342700511355785,
+      "grad_norm": 1.72428547122667,
+      "learning_rate": 0.00017907536400581998,
+      "loss": 1.3057,
+      "step": 5067
+    },
+    {
+      "epoch": 0.23347307320219285,
+      "grad_norm": 2.9113381224298984,
+      "learning_rate": 0.00017906622953579277,
+      "loss": 1.6637,
+      "step": 5068
+    },
+    {
+      "epoch": 0.23351914129082785,
+      "grad_norm": 1.7821327319336855,
+      "learning_rate": 0.00017905709330549152,
+      "loss": 1.1271,
+      "step": 5069
+    },
+    {
+      "epoch": 0.23356520937946285,
+      "grad_norm": 1.2549671026463534,
+      "learning_rate": 0.00017904795531511965,
+      "loss": 1.4539,
+      "step": 5070
+    },
+    {
+      "epoch": 0.23361127746809784,
+      "grad_norm": 1.4734043857604768,
+      "learning_rate": 0.00017903881556488057,
+      "loss": 1.6442,
+      "step": 5071
+    },
+    {
+      "epoch": 0.23365734555673284,
+      "grad_norm": 2.414120409728032,
+      "learning_rate": 0.0001790296740549778,
+      "loss": 1.5819,
+      "step": 5072
+    },
+    {
+      "epoch": 0.23370341364536784,
+      "grad_norm": 1.7415978722573537,
+      "learning_rate": 0.00017902053078561482,
+      "loss": 1.6714,
+      "step": 5073
+    },
+    {
+      "epoch": 0.23374948173400287,
+      "grad_norm": 1.6451276604214873,
+      "learning_rate": 0.00017901138575699522,
+      "loss": 1.65,
+      "step": 5074
+    },
+    {
+      "epoch": 0.23379554982263787,
+      "grad_norm": 1.1554940294215026,
+      "learning_rate": 0.00017900223896932262,
+      "loss": 1.4981,
+      "step": 5075
+    },
+    {
+      "epoch": 0.23384161791127286,
+      "grad_norm": 1.444805404894403,
+      "learning_rate": 0.0001789930904228006,
+      "loss": 1.4599,
+      "step": 5076
+    },
+    {
+      "epoch": 0.23388768599990786,
+      "grad_norm": 2.3592708422032826,
+      "learning_rate": 0.00017898394011763287,
+      "loss": 1.6102,
+      "step": 5077
+    },
+    {
+      "epoch": 0.23393375408854286,
+      "grad_norm": 4.0083483455560565,
+      "learning_rate": 0.00017897478805402317,
+      "loss": 1.6715,
+      "step": 5078
+    },
+    {
+      "epoch": 0.23397982217717786,
+      "grad_norm": 1.280990084717295,
+      "learning_rate": 0.0001789656342321752,
+      "loss": 1.5788,
+      "step": 5079
+    },
+    {
+      "epoch": 0.23402589026581286,
+      "grad_norm": 1.6270607199308305,
+      "learning_rate": 0.0001789564786522928,
+      "loss": 1.3946,
+      "step": 5080
+    },
+    {
+      "epoch": 0.23407195835444788,
+      "grad_norm": 2.031030466992363,
+      "learning_rate": 0.00017894732131457975,
+      "loss": 1.4429,
+      "step": 5081
+    },
+    {
+      "epoch": 0.23411802644308288,
+      "grad_norm": 3.1772146698579298,
+      "learning_rate": 0.00017893816221924,
+      "loss": 1.4941,
+      "step": 5082
+    },
+    {
+      "epoch": 0.23416409453171788,
+      "grad_norm": 4.145246517526111,
+      "learning_rate": 0.0001789290013664774,
+      "loss": 1.5324,
+      "step": 5083
+    },
+    {
+      "epoch": 0.23421016262035288,
+      "grad_norm": 1.6402274192649742,
+      "learning_rate": 0.0001789198387564959,
+      "loss": 1.4836,
+      "step": 5084
+    },
+    {
+      "epoch": 0.23425623070898788,
+      "grad_norm": 2.311913167841554,
+      "learning_rate": 0.00017891067438949955,
+      "loss": 1.4585,
+      "step": 5085
+    },
+    {
+      "epoch": 0.23430229879762288,
+      "grad_norm": 1.4583143939701035,
+      "learning_rate": 0.00017890150826569228,
+      "loss": 1.3398,
+      "step": 5086
+    },
+    {
+      "epoch": 0.2343483668862579,
+      "grad_norm": 2.0327898716380335,
+      "learning_rate": 0.00017889234038527825,
+      "loss": 1.6488,
+      "step": 5087
+    },
+    {
+      "epoch": 0.2343944349748929,
+      "grad_norm": 1.247459524388959,
+      "learning_rate": 0.00017888317074846155,
+      "loss": 1.3851,
+      "step": 5088
+    },
+    {
+      "epoch": 0.2344405030635279,
+      "grad_norm": 3.9797803883061107,
+      "learning_rate": 0.00017887399935544626,
+      "loss": 1.4014,
+      "step": 5089
+    },
+    {
+      "epoch": 0.2344865711521629,
+      "grad_norm": 1.401590693546647,
+      "learning_rate": 0.00017886482620643664,
+      "loss": 1.6504,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2345326392407979,
+      "grad_norm": 17.61489511326625,
+      "learning_rate": 0.0001788556513016369,
+      "loss": 1.4605,
+      "step": 5091
+    },
+    {
+      "epoch": 0.2345787073294329,
+      "grad_norm": 2.6604545377952746,
+      "learning_rate": 0.00017884647464125127,
+      "loss": 1.3839,
+      "step": 5092
+    },
+    {
+      "epoch": 0.2346247754180679,
+      "grad_norm": 2.4084260435393237,
+      "learning_rate": 0.0001788372962254841,
+      "loss": 1.1472,
+      "step": 5093
+    },
+    {
+      "epoch": 0.23467084350670292,
+      "grad_norm": 1.9302084513965314,
+      "learning_rate": 0.00017882811605453972,
+      "loss": 1.6235,
+      "step": 5094
+    },
+    {
+      "epoch": 0.23471691159533792,
+      "grad_norm": 1.1909159383762158,
+      "learning_rate": 0.00017881893412862246,
+      "loss": 1.7694,
+      "step": 5095
+    },
+    {
+      "epoch": 0.23476297968397292,
+      "grad_norm": 1.9891764059957744,
+      "learning_rate": 0.00017880975044793677,
+      "loss": 1.5453,
+      "step": 5096
+    },
+    {
+      "epoch": 0.23480904777260792,
+      "grad_norm": 2.5456161846831873,
+      "learning_rate": 0.00017880056501268715,
+      "loss": 1.5975,
+      "step": 5097
+    },
+    {
+      "epoch": 0.2348551158612429,
+      "grad_norm": 2.7207421149574826,
+      "learning_rate": 0.00017879137782307806,
+      "loss": 1.5324,
+      "step": 5098
+    },
+    {
+      "epoch": 0.2349011839498779,
+      "grad_norm": 4.544602537898767,
+      "learning_rate": 0.00017878218887931401,
+      "loss": 1.3655,
+      "step": 5099
+    },
+    {
+      "epoch": 0.2349472520385129,
+      "grad_norm": 1.3231433260586953,
+      "learning_rate": 0.00017877299818159963,
+      "loss": 1.3484,
+      "step": 5100
+    },
+    {
+      "epoch": 0.23499332012714794,
+      "grad_norm": 2.2404035449990567,
+      "learning_rate": 0.0001787638057301395,
+      "loss": 1.3341,
+      "step": 5101
+    },
+    {
+      "epoch": 0.23503938821578294,
+      "grad_norm": 1.4938005946343278,
+      "learning_rate": 0.0001787546115251383,
+      "loss": 1.3497,
+      "step": 5102
+    },
+    {
+      "epoch": 0.23508545630441793,
+      "grad_norm": 1.2776852183532343,
+      "learning_rate": 0.0001787454155668007,
+      "loss": 1.6895,
+      "step": 5103
+    },
+    {
+      "epoch": 0.23513152439305293,
+      "grad_norm": 2.177589667216448,
+      "learning_rate": 0.00017873621785533146,
+      "loss": 1.0903,
+      "step": 5104
+    },
+    {
+      "epoch": 0.23517759248168793,
+      "grad_norm": 1.399462061204036,
+      "learning_rate": 0.0001787270183909353,
+      "loss": 1.4759,
+      "step": 5105
+    },
+    {
+      "epoch": 0.23522366057032293,
+      "grad_norm": 1.6892785537273158,
+      "learning_rate": 0.00017871781717381712,
+      "loss": 1.631,
+      "step": 5106
+    },
+    {
+      "epoch": 0.23526972865895793,
+      "grad_norm": 1.463898074337145,
+      "learning_rate": 0.00017870861420418165,
+      "loss": 1.3938,
+      "step": 5107
+    },
+    {
+      "epoch": 0.23531579674759295,
+      "grad_norm": 1.1766250267118843,
+      "learning_rate": 0.00017869940948223386,
+      "loss": 1.4535,
+      "step": 5108
+    },
+    {
+      "epoch": 0.23536186483622795,
+      "grad_norm": 2.54149149477661,
+      "learning_rate": 0.00017869020300817867,
+      "loss": 1.3074,
+      "step": 5109
+    },
+    {
+      "epoch": 0.23540793292486295,
+      "grad_norm": 4.444503141247241,
+      "learning_rate": 0.00017868099478222105,
+      "loss": 1.3126,
+      "step": 5110
+    },
+    {
+      "epoch": 0.23545400101349795,
+      "grad_norm": 4.242383386818219,
+      "learning_rate": 0.00017867178480456594,
+      "loss": 1.4976,
+      "step": 5111
+    },
+    {
+      "epoch": 0.23550006910213295,
+      "grad_norm": 1.4162582626314728,
+      "learning_rate": 0.00017866257307541848,
+      "loss": 1.4323,
+      "step": 5112
+    },
+    {
+      "epoch": 0.23554613719076795,
+      "grad_norm": 1.617531623063942,
+      "learning_rate": 0.0001786533595949837,
+      "loss": 1.6152,
+      "step": 5113
+    },
+    {
+      "epoch": 0.23559220527940294,
+      "grad_norm": 1.4070513577368373,
+      "learning_rate": 0.00017864414436346673,
+      "loss": 1.452,
+      "step": 5114
+    },
+    {
+      "epoch": 0.23563827336803797,
+      "grad_norm": 2.16117391903066,
+      "learning_rate": 0.0001786349273810727,
+      "loss": 1.4881,
+      "step": 5115
+    },
+    {
+      "epoch": 0.23568434145667297,
+      "grad_norm": 1.3443607169449474,
+      "learning_rate": 0.00017862570864800688,
+      "loss": 1.3769,
+      "step": 5116
+    },
+    {
+      "epoch": 0.23573040954530797,
+      "grad_norm": 1.8871433587962005,
+      "learning_rate": 0.00017861648816447446,
+      "loss": 1.4618,
+      "step": 5117
+    },
+    {
+      "epoch": 0.23577647763394297,
+      "grad_norm": 4.44879016372698,
+      "learning_rate": 0.00017860726593068072,
+      "loss": 1.3918,
+      "step": 5118
+    },
+    {
+      "epoch": 0.23582254572257796,
+      "grad_norm": 1.3920997314820427,
+      "learning_rate": 0.000178598041946831,
+      "loss": 1.5885,
+      "step": 5119
+    },
+    {
+      "epoch": 0.23586861381121296,
+      "grad_norm": 2.985132481133448,
+      "learning_rate": 0.00017858881621313064,
+      "loss": 1.6398,
+      "step": 5120
+    },
+    {
+      "epoch": 0.235914681899848,
+      "grad_norm": 5.405176293975686,
+      "learning_rate": 0.00017857958872978503,
+      "loss": 1.5669,
+      "step": 5121
+    },
+    {
+      "epoch": 0.235960749988483,
+      "grad_norm": 1.6577440725066501,
+      "learning_rate": 0.00017857035949699963,
+      "loss": 1.5468,
+      "step": 5122
+    },
+    {
+      "epoch": 0.23600681807711799,
+      "grad_norm": 1.1751855588580846,
+      "learning_rate": 0.00017856112851497987,
+      "loss": 1.8791,
+      "step": 5123
+    },
+    {
+      "epoch": 0.23605288616575298,
+      "grad_norm": 1.7049376584719524,
+      "learning_rate": 0.0001785518957839313,
+      "loss": 1.3989,
+      "step": 5124
+    },
+    {
+      "epoch": 0.23609895425438798,
+      "grad_norm": 3.9608293298913284,
+      "learning_rate": 0.00017854266130405946,
+      "loss": 1.5969,
+      "step": 5125
+    },
+    {
+      "epoch": 0.23614502234302298,
+      "grad_norm": 1.4190263397910734,
+      "learning_rate": 0.0001785334250755699,
+      "loss": 1.4819,
+      "step": 5126
+    },
+    {
+      "epoch": 0.23619109043165798,
+      "grad_norm": 2.492662648206682,
+      "learning_rate": 0.00017852418709866835,
+      "loss": 1.4566,
+      "step": 5127
+    },
+    {
+      "epoch": 0.236237158520293,
+      "grad_norm": 2.847517606353952,
+      "learning_rate": 0.00017851494737356038,
+      "loss": 1.4972,
+      "step": 5128
+    },
+    {
+      "epoch": 0.236283226608928,
+      "grad_norm": 8.882613040289387,
+      "learning_rate": 0.00017850570590045175,
+      "loss": 1.5092,
+      "step": 5129
+    },
+    {
+      "epoch": 0.236329294697563,
+      "grad_norm": 1.67665029019791,
+      "learning_rate": 0.00017849646267954815,
+      "loss": 1.3628,
+      "step": 5130
+    },
+    {
+      "epoch": 0.236375362786198,
+      "grad_norm": 4.0339970933677165,
+      "learning_rate": 0.00017848721771105544,
+      "loss": 1.4693,
+      "step": 5131
+    },
+    {
+      "epoch": 0.236421430874833,
+      "grad_norm": 2.8069929188317815,
+      "learning_rate": 0.00017847797099517938,
+      "loss": 1.4974,
+      "step": 5132
+    },
+    {
+      "epoch": 0.236467498963468,
+      "grad_norm": 3.301585847109644,
+      "learning_rate": 0.00017846872253212586,
+      "loss": 1.2913,
+      "step": 5133
+    },
+    {
+      "epoch": 0.236513567052103,
+      "grad_norm": 5.662069683593566,
+      "learning_rate": 0.0001784594723221008,
+      "loss": 1.7835,
+      "step": 5134
+    },
+    {
+      "epoch": 0.23655963514073802,
+      "grad_norm": 2.0697024108889956,
+      "learning_rate": 0.0001784502203653101,
+      "loss": 1.5195,
+      "step": 5135
+    },
+    {
+      "epoch": 0.23660570322937302,
+      "grad_norm": 2.1654170741158443,
+      "learning_rate": 0.00017844096666195974,
+      "loss": 1.3259,
+      "step": 5136
+    },
+    {
+      "epoch": 0.23665177131800802,
+      "grad_norm": 4.137460818548541,
+      "learning_rate": 0.00017843171121225578,
+      "loss": 1.4814,
+      "step": 5137
+    },
+    {
+      "epoch": 0.23669783940664302,
+      "grad_norm": 2.8708140217747014,
+      "learning_rate": 0.00017842245401640425,
+      "loss": 1.4737,
+      "step": 5138
+    },
+    {
+      "epoch": 0.23674390749527802,
+      "grad_norm": 3.8394147231096665,
+      "learning_rate": 0.00017841319507461124,
+      "loss": 1.5964,
+      "step": 5139
+    },
+    {
+      "epoch": 0.23678997558391301,
+      "grad_norm": 1.6529851305709493,
+      "learning_rate": 0.0001784039343870829,
+      "loss": 1.8507,
+      "step": 5140
+    },
+    {
+      "epoch": 0.236836043672548,
+      "grad_norm": 1.845711406234485,
+      "learning_rate": 0.00017839467195402537,
+      "loss": 1.5508,
+      "step": 5141
+    },
+    {
+      "epoch": 0.23688211176118304,
+      "grad_norm": 2.134527028681731,
+      "learning_rate": 0.0001783854077756449,
+      "loss": 1.6429,
+      "step": 5142
+    },
+    {
+      "epoch": 0.23692817984981804,
+      "grad_norm": 1.702659247436571,
+      "learning_rate": 0.00017837614185214773,
+      "loss": 1.2771,
+      "step": 5143
+    },
+    {
+      "epoch": 0.23697424793845304,
+      "grad_norm": 2.0181908636187833,
+      "learning_rate": 0.00017836687418374016,
+      "loss": 1.4838,
+      "step": 5144
+    },
+    {
+      "epoch": 0.23702031602708803,
+      "grad_norm": 15.499452605421736,
+      "learning_rate": 0.0001783576047706285,
+      "loss": 1.5528,
+      "step": 5145
+    },
+    {
+      "epoch": 0.23706638411572303,
+      "grad_norm": 10.315695650480544,
+      "learning_rate": 0.00017834833361301912,
+      "loss": 1.5979,
+      "step": 5146
+    },
+    {
+      "epoch": 0.23711245220435803,
+      "grad_norm": 2.3839639599347695,
+      "learning_rate": 0.00017833906071111844,
+      "loss": 1.5844,
+      "step": 5147
+    },
+    {
+      "epoch": 0.23715852029299303,
+      "grad_norm": 4.962045479165653,
+      "learning_rate": 0.00017832978606513287,
+      "loss": 1.4475,
+      "step": 5148
+    },
+    {
+      "epoch": 0.23720458838162806,
+      "grad_norm": 1.7833773886800768,
+      "learning_rate": 0.00017832050967526894,
+      "loss": 1.5084,
+      "step": 5149
+    },
+    {
+      "epoch": 0.23725065647026305,
+      "grad_norm": 4.271330654290617,
+      "learning_rate": 0.00017831123154173314,
+      "loss": 1.4151,
+      "step": 5150
+    },
+    {
+      "epoch": 0.23729672455889805,
+      "grad_norm": 3.409659596484695,
+      "learning_rate": 0.00017830195166473206,
+      "loss": 1.304,
+      "step": 5151
+    },
+    {
+      "epoch": 0.23734279264753305,
+      "grad_norm": 7.456222859458221,
+      "learning_rate": 0.0001782926700444723,
+      "loss": 1.6122,
+      "step": 5152
+    },
+    {
+      "epoch": 0.23738886073616805,
+      "grad_norm": 4.246964327533174,
+      "learning_rate": 0.00017828338668116045,
+      "loss": 1.7492,
+      "step": 5153
+    },
+    {
+      "epoch": 0.23743492882480305,
+      "grad_norm": 2.343060750969311,
+      "learning_rate": 0.00017827410157500323,
+      "loss": 1.2564,
+      "step": 5154
+    },
+    {
+      "epoch": 0.23748099691343807,
+      "grad_norm": 2.0349881608510185,
+      "learning_rate": 0.00017826481472620737,
+      "loss": 1.5807,
+      "step": 5155
+    },
+    {
+      "epoch": 0.23752706500207307,
+      "grad_norm": 2.8008120693838867,
+      "learning_rate": 0.0001782555261349796,
+      "loss": 1.5545,
+      "step": 5156
+    },
+    {
+      "epoch": 0.23757313309070807,
+      "grad_norm": 2.4523876266062348,
+      "learning_rate": 0.0001782462358015267,
+      "loss": 1.6458,
+      "step": 5157
+    },
+    {
+      "epoch": 0.23761920117934307,
+      "grad_norm": 1.2415815934168903,
+      "learning_rate": 0.00017823694372605555,
+      "loss": 1.8607,
+      "step": 5158
+    },
+    {
+      "epoch": 0.23766526926797807,
+      "grad_norm": 4.0529092010186005,
+      "learning_rate": 0.000178227649908773,
+      "loss": 1.5021,
+      "step": 5159
+    },
+    {
+      "epoch": 0.23771133735661307,
+      "grad_norm": 2.8777337808406047,
+      "learning_rate": 0.00017821835434988592,
+      "loss": 1.6747,
+      "step": 5160
+    },
+    {
+      "epoch": 0.23775740544524807,
+      "grad_norm": 2.6952520648159966,
+      "learning_rate": 0.00017820905704960132,
+      "loss": 1.492,
+      "step": 5161
+    },
+    {
+      "epoch": 0.2378034735338831,
+      "grad_norm": 2.2154446802554437,
+      "learning_rate": 0.00017819975800812614,
+      "loss": 1.3913,
+      "step": 5162
+    },
+    {
+      "epoch": 0.2378495416225181,
+      "grad_norm": 3.299449476440645,
+      "learning_rate": 0.00017819045722566746,
+      "loss": 1.7099,
+      "step": 5163
+    },
+    {
+      "epoch": 0.2378956097111531,
+      "grad_norm": 2.689651959927629,
+      "learning_rate": 0.0001781811547024323,
+      "loss": 1.4323,
+      "step": 5164
+    },
+    {
+      "epoch": 0.2379416777997881,
+      "grad_norm": 3.6389396183636307,
+      "learning_rate": 0.00017817185043862778,
+      "loss": 1.4809,
+      "step": 5165
+    },
+    {
+      "epoch": 0.23798774588842309,
+      "grad_norm": 1.4788166648388767,
+      "learning_rate": 0.00017816254443446108,
+      "loss": 1.7658,
+      "step": 5166
+    },
+    {
+      "epoch": 0.23803381397705808,
+      "grad_norm": 3.768045089431305,
+      "learning_rate": 0.00017815323669013932,
+      "loss": 1.4056,
+      "step": 5167
+    },
+    {
+      "epoch": 0.23807988206569308,
+      "grad_norm": 2.8922788534083765,
+      "learning_rate": 0.00017814392720586977,
+      "loss": 1.4151,
+      "step": 5168
+    },
+    {
+      "epoch": 0.2381259501543281,
+      "grad_norm": 2.3105107361041766,
+      "learning_rate": 0.00017813461598185963,
+      "loss": 1.5155,
+      "step": 5169
+    },
+    {
+      "epoch": 0.2381720182429631,
+      "grad_norm": 3.6833777439250923,
+      "learning_rate": 0.00017812530301831628,
+      "loss": 1.3355,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2382180863315981,
+      "grad_norm": 12.53345200869566,
+      "learning_rate": 0.000178115988315447,
+      "loss": 1.265,
+      "step": 5171
+    },
+    {
+      "epoch": 0.2382641544202331,
+      "grad_norm": 3.8118324706496214,
+      "learning_rate": 0.00017810667187345916,
+      "loss": 1.531,
+      "step": 5172
+    },
+    {
+      "epoch": 0.2383102225088681,
+      "grad_norm": 2.0205701278808563,
+      "learning_rate": 0.0001780973536925602,
+      "loss": 1.4995,
+      "step": 5173
+    },
+    {
+      "epoch": 0.2383562905975031,
+      "grad_norm": 3.6637708076179325,
+      "learning_rate": 0.00017808803377295757,
+      "loss": 1.6711,
+      "step": 5174
+    },
+    {
+      "epoch": 0.2384023586861381,
+      "grad_norm": 5.823495596635684,
+      "learning_rate": 0.00017807871211485878,
+      "loss": 1.7819,
+      "step": 5175
+    },
+    {
+      "epoch": 0.23844842677477313,
+      "grad_norm": 3.4812754373513806,
+      "learning_rate": 0.00017806938871847133,
+      "loss": 1.4965,
+      "step": 5176
+    },
+    {
+      "epoch": 0.23849449486340812,
+      "grad_norm": 2.74965739527072,
+      "learning_rate": 0.0001780600635840028,
+      "loss": 1.4965,
+      "step": 5177
+    },
+    {
+      "epoch": 0.23854056295204312,
+      "grad_norm": 5.0258512464403955,
+      "learning_rate": 0.0001780507367116608,
+      "loss": 0.9811,
+      "step": 5178
+    },
+    {
+      "epoch": 0.23858663104067812,
+      "grad_norm": 3.7012691616205733,
+      "learning_rate": 0.00017804140810165296,
+      "loss": 1.6914,
+      "step": 5179
+    },
+    {
+      "epoch": 0.23863269912931312,
+      "grad_norm": 2.9966806905723584,
+      "learning_rate": 0.00017803207775418702,
+      "loss": 1.4753,
+      "step": 5180
+    },
+    {
+      "epoch": 0.23867876721794812,
+      "grad_norm": 9.197948214141645,
+      "learning_rate": 0.00017802274566947063,
+      "loss": 1.5227,
+      "step": 5181
+    },
+    {
+      "epoch": 0.23872483530658312,
+      "grad_norm": 3.4083380058068635,
+      "learning_rate": 0.00017801341184771162,
+      "loss": 1.7921,
+      "step": 5182
+    },
+    {
+      "epoch": 0.23877090339521814,
+      "grad_norm": 2.6334915339728604,
+      "learning_rate": 0.00017800407628911774,
+      "loss": 1.4795,
+      "step": 5183
+    },
+    {
+      "epoch": 0.23881697148385314,
+      "grad_norm": 4.823467074032504,
+      "learning_rate": 0.00017799473899389681,
+      "loss": 1.855,
+      "step": 5184
+    },
+    {
+      "epoch": 0.23886303957248814,
+      "grad_norm": 4.095673145614503,
+      "learning_rate": 0.0001779853999622568,
+      "loss": 1.633,
+      "step": 5185
+    },
+    {
+      "epoch": 0.23890910766112314,
+      "grad_norm": 2.895830120037296,
+      "learning_rate": 0.00017797605919440555,
+      "loss": 1.4136,
+      "step": 5186
+    },
+    {
+      "epoch": 0.23895517574975814,
+      "grad_norm": 2.886571146847535,
+      "learning_rate": 0.00017796671669055106,
+      "loss": 1.6482,
+      "step": 5187
+    },
+    {
+      "epoch": 0.23900124383839313,
+      "grad_norm": 2.9952690748074198,
+      "learning_rate": 0.00017795737245090133,
+      "loss": 1.9291,
+      "step": 5188
+    },
+    {
+      "epoch": 0.23904731192702816,
+      "grad_norm": 1.8153579911547326,
+      "learning_rate": 0.00017794802647566434,
+      "loss": 1.414,
+      "step": 5189
+    },
+    {
+      "epoch": 0.23909338001566316,
+      "grad_norm": 2.5556005680930016,
+      "learning_rate": 0.0001779386787650482,
+      "loss": 1.5295,
+      "step": 5190
+    },
+    {
+      "epoch": 0.23913944810429816,
+      "grad_norm": 3.4858799984194917,
+      "learning_rate": 0.00017792932931926097,
+      "loss": 1.4129,
+      "step": 5191
+    },
+    {
+      "epoch": 0.23918551619293316,
+      "grad_norm": 1.7315443277901366,
+      "learning_rate": 0.0001779199781385109,
+      "loss": 1.544,
+      "step": 5192
+    },
+    {
+      "epoch": 0.23923158428156815,
+      "grad_norm": 1.741165383287041,
+      "learning_rate": 0.0001779106252230061,
+      "loss": 1.4525,
+      "step": 5193
+    },
+    {
+      "epoch": 0.23927765237020315,
+      "grad_norm": 1.8593329856342422,
+      "learning_rate": 0.00017790127057295486,
+      "loss": 1.4246,
+      "step": 5194
+    },
+    {
+      "epoch": 0.23932372045883815,
+      "grad_norm": 1.4648417470655815,
+      "learning_rate": 0.00017789191418856532,
+      "loss": 1.5967,
+      "step": 5195
+    },
+    {
+      "epoch": 0.23936978854747318,
+      "grad_norm": 1.6741541735141612,
+      "learning_rate": 0.00017788255607004594,
+      "loss": 1.5456,
+      "step": 5196
+    },
+    {
+      "epoch": 0.23941585663610818,
+      "grad_norm": 3.5187951266749518,
+      "learning_rate": 0.00017787319621760493,
+      "loss": 1.6553,
+      "step": 5197
+    },
+    {
+      "epoch": 0.23946192472474317,
+      "grad_norm": 2.555846918148604,
+      "learning_rate": 0.00017786383463145079,
+      "loss": 1.6072,
+      "step": 5198
+    },
+    {
+      "epoch": 0.23950799281337817,
+      "grad_norm": 2.9823032997235535,
+      "learning_rate": 0.00017785447131179184,
+      "loss": 1.412,
+      "step": 5199
+    },
+    {
+      "epoch": 0.23955406090201317,
+      "grad_norm": 1.757908384845788,
+      "learning_rate": 0.00017784510625883657,
+      "loss": 1.4592,
+      "step": 5200
+    },
+    {
+      "epoch": 0.23960012899064817,
+      "grad_norm": 1.5066351451373567,
+      "learning_rate": 0.0001778357394727935,
+      "loss": 1.2895,
+      "step": 5201
+    },
+    {
+      "epoch": 0.23964619707928317,
+      "grad_norm": 7.925723408952124,
+      "learning_rate": 0.00017782637095387115,
+      "loss": 1.6161,
+      "step": 5202
+    },
+    {
+      "epoch": 0.2396922651679182,
+      "grad_norm": 3.7085444807670833,
+      "learning_rate": 0.0001778170007022781,
+      "loss": 1.5099,
+      "step": 5203
+    },
+    {
+      "epoch": 0.2397383332565532,
+      "grad_norm": 3.96168586840727,
+      "learning_rate": 0.00017780762871822295,
+      "loss": 1.6725,
+      "step": 5204
+    },
+    {
+      "epoch": 0.2397844013451882,
+      "grad_norm": 2.1600757124126426,
+      "learning_rate": 0.0001777982550019144,
+      "loss": 1.7876,
+      "step": 5205
+    },
+    {
+      "epoch": 0.2398304694338232,
+      "grad_norm": 2.6504067684818016,
+      "learning_rate": 0.00017778887955356102,
+      "loss": 1.8008,
+      "step": 5206
+    },
+    {
+      "epoch": 0.2398765375224582,
+      "grad_norm": 2.008452088051424,
+      "learning_rate": 0.00017777950237337167,
+      "loss": 1.6123,
+      "step": 5207
+    },
+    {
+      "epoch": 0.2399226056110932,
+      "grad_norm": 2.5200322526185843,
+      "learning_rate": 0.00017777012346155503,
+      "loss": 1.4311,
+      "step": 5208
+    },
+    {
+      "epoch": 0.23996867369972819,
+      "grad_norm": 3.1243746563079307,
+      "learning_rate": 0.00017776074281831996,
+      "loss": 1.4692,
+      "step": 5209
+    },
+    {
+      "epoch": 0.2400147417883632,
+      "grad_norm": 1.6348224406629788,
+      "learning_rate": 0.0001777513604438753,
+      "loss": 1.4065,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2400608098769982,
+      "grad_norm": 1.2943292510918958,
+      "learning_rate": 0.00017774197633842987,
+      "loss": 1.721,
+      "step": 5211
+    },
+    {
+      "epoch": 0.2401068779656332,
+      "grad_norm": 1.954348077127246,
+      "learning_rate": 0.00017773259050219266,
+      "loss": 1.3932,
+      "step": 5212
+    },
+    {
+      "epoch": 0.2401529460542682,
+      "grad_norm": 12.339786653057928,
+      "learning_rate": 0.0001777232029353726,
+      "loss": 1.6421,
+      "step": 5213
+    },
+    {
+      "epoch": 0.2401990141429032,
+      "grad_norm": 3.892336352230601,
+      "learning_rate": 0.00017771381363817872,
+      "loss": 1.7398,
+      "step": 5214
+    },
+    {
+      "epoch": 0.2402450822315382,
+      "grad_norm": 2.423832392406744,
+      "learning_rate": 0.00017770442261081997,
+      "loss": 1.6903,
+      "step": 5215
+    },
+    {
+      "epoch": 0.2402911503201732,
+      "grad_norm": 4.2236723672000185,
+      "learning_rate": 0.00017769502985350556,
+      "loss": 1.4957,
+      "step": 5216
+    },
+    {
+      "epoch": 0.24033721840880823,
+      "grad_norm": 2.0989490013049856,
+      "learning_rate": 0.00017768563536644448,
+      "loss": 1.6467,
+      "step": 5217
+    },
+    {
+      "epoch": 0.24038328649744323,
+      "grad_norm": 3.0692436522297633,
+      "learning_rate": 0.00017767623914984593,
+      "loss": 1.5439,
+      "step": 5218
+    },
+    {
+      "epoch": 0.24042935458607823,
+      "grad_norm": 2.461621146020737,
+      "learning_rate": 0.00017766684120391912,
+      "loss": 1.5942,
+      "step": 5219
+    },
+    {
+      "epoch": 0.24047542267471322,
+      "grad_norm": 3.157936169232742,
+      "learning_rate": 0.00017765744152887326,
+      "loss": 1.5332,
+      "step": 5220
+    },
+    {
+      "epoch": 0.24052149076334822,
+      "grad_norm": 5.919540168301258,
+      "learning_rate": 0.0001776480401249176,
+      "loss": 1.7121,
+      "step": 5221
+    },
+    {
+      "epoch": 0.24056755885198322,
+      "grad_norm": 2.5858517329666824,
+      "learning_rate": 0.0001776386369922615,
+      "loss": 1.6122,
+      "step": 5222
+    },
+    {
+      "epoch": 0.24061362694061825,
+      "grad_norm": 4.0059020502293405,
+      "learning_rate": 0.00017762923213111425,
+      "loss": 1.2499,
+      "step": 5223
+    },
+    {
+      "epoch": 0.24065969502925325,
+      "grad_norm": 1.593637273676087,
+      "learning_rate": 0.0001776198255416852,
+      "loss": 1.5712,
+      "step": 5224
+    },
+    {
+      "epoch": 0.24070576311788824,
+      "grad_norm": 3.9708415292621186,
+      "learning_rate": 0.0001776104172241839,
+      "loss": 1.6009,
+      "step": 5225
+    },
+    {
+      "epoch": 0.24075183120652324,
+      "grad_norm": 2.55737158025025,
+      "learning_rate": 0.0001776010071788197,
+      "loss": 1.3913,
+      "step": 5226
+    },
+    {
+      "epoch": 0.24079789929515824,
+      "grad_norm": 4.316351422613528,
+      "learning_rate": 0.00017759159540580214,
+      "loss": 1.6227,
+      "step": 5227
+    },
+    {
+      "epoch": 0.24084396738379324,
+      "grad_norm": 2.677424139463403,
+      "learning_rate": 0.00017758218190534075,
+      "loss": 1.7455,
+      "step": 5228
+    },
+    {
+      "epoch": 0.24089003547242824,
+      "grad_norm": 4.058303449100364,
+      "learning_rate": 0.00017757276667764512,
+      "loss": 1.3131,
+      "step": 5229
+    },
+    {
+      "epoch": 0.24093610356106326,
+      "grad_norm": 1.9860887645372196,
+      "learning_rate": 0.00017756334972292482,
+      "loss": 1.6561,
+      "step": 5230
+    },
+    {
+      "epoch": 0.24098217164969826,
+      "grad_norm": 1.9133597916178853,
+      "learning_rate": 0.00017755393104138952,
+      "loss": 1.2176,
+      "step": 5231
+    },
+    {
+      "epoch": 0.24102823973833326,
+      "grad_norm": 1.487859425134761,
+      "learning_rate": 0.00017754451063324895,
+      "loss": 1.6172,
+      "step": 5232
+    },
+    {
+      "epoch": 0.24107430782696826,
+      "grad_norm": 2.253566185731917,
+      "learning_rate": 0.00017753508849871282,
+      "loss": 1.5923,
+      "step": 5233
+    },
+    {
+      "epoch": 0.24112037591560326,
+      "grad_norm": 1.703638061705096,
+      "learning_rate": 0.00017752566463799085,
+      "loss": 1.7174,
+      "step": 5234
+    },
+    {
+      "epoch": 0.24116644400423826,
+      "grad_norm": 3.293467789121164,
+      "learning_rate": 0.00017751623905129293,
+      "loss": 1.3439,
+      "step": 5235
+    },
+    {
+      "epoch": 0.24121251209287325,
+      "grad_norm": 2.5229763870683906,
+      "learning_rate": 0.00017750681173882884,
+      "loss": 1.6597,
+      "step": 5236
+    },
+    {
+      "epoch": 0.24125858018150828,
+      "grad_norm": 1.2600452168606553,
+      "learning_rate": 0.0001774973827008085,
+      "loss": 1.3299,
+      "step": 5237
+    },
+    {
+      "epoch": 0.24130464827014328,
+      "grad_norm": 2.1591844101369557,
+      "learning_rate": 0.00017748795193744176,
+      "loss": 1.3502,
+      "step": 5238
+    },
+    {
+      "epoch": 0.24135071635877828,
+      "grad_norm": 1.3992206049693756,
+      "learning_rate": 0.0001774785194489387,
+      "loss": 1.2796,
+      "step": 5239
+    },
+    {
+      "epoch": 0.24139678444741328,
+      "grad_norm": 1.3926922371973995,
+      "learning_rate": 0.0001774690852355092,
+      "loss": 1.5147,
+      "step": 5240
+    },
+    {
+      "epoch": 0.24144285253604827,
+      "grad_norm": 2.0640025713022796,
+      "learning_rate": 0.0001774596492973634,
+      "loss": 1.5187,
+      "step": 5241
+    },
+    {
+      "epoch": 0.24148892062468327,
+      "grad_norm": 1.8887285285003108,
+      "learning_rate": 0.00017745021163471125,
+      "loss": 1.5046,
+      "step": 5242
+    },
+    {
+      "epoch": 0.24153498871331827,
+      "grad_norm": 1.9579210871421118,
+      "learning_rate": 0.00017744077224776298,
+      "loss": 1.635,
+      "step": 5243
+    },
+    {
+      "epoch": 0.2415810568019533,
+      "grad_norm": 2.3903556846454252,
+      "learning_rate": 0.0001774313311367287,
+      "loss": 1.6591,
+      "step": 5244
+    },
+    {
+      "epoch": 0.2416271248905883,
+      "grad_norm": 1.8803889064846706,
+      "learning_rate": 0.0001774218883018186,
+      "loss": 1.5517,
+      "step": 5245
+    },
+    {
+      "epoch": 0.2416731929792233,
+      "grad_norm": 5.628750741671842,
+      "learning_rate": 0.00017741244374324292,
+      "loss": 1.3951,
+      "step": 5246
+    },
+    {
+      "epoch": 0.2417192610678583,
+      "grad_norm": 1.7297923603124759,
+      "learning_rate": 0.00017740299746121186,
+      "loss": 1.8457,
+      "step": 5247
+    },
+    {
+      "epoch": 0.2417653291564933,
+      "grad_norm": 1.691608687999392,
+      "learning_rate": 0.00017739354945593583,
+      "loss": 1.6615,
+      "step": 5248
+    },
+    {
+      "epoch": 0.2418113972451283,
+      "grad_norm": 1.9640967474543942,
+      "learning_rate": 0.00017738409972762508,
+      "loss": 1.5578,
+      "step": 5249
+    },
+    {
+      "epoch": 0.2418574653337633,
+      "grad_norm": 3.22534870069284,
+      "learning_rate": 0.00017737464827649008,
+      "loss": 1.7931,
+      "step": 5250
+    },
+    {
+      "epoch": 0.24190353342239831,
+      "grad_norm": 1.652798348467676,
+      "learning_rate": 0.0001773651951027412,
+      "loss": 1.4009,
+      "step": 5251
+    },
+    {
+      "epoch": 0.2419496015110333,
+      "grad_norm": 1.5532288816995392,
+      "learning_rate": 0.0001773557402065889,
+      "loss": 1.6664,
+      "step": 5252
+    },
+    {
+      "epoch": 0.2419956695996683,
+      "grad_norm": 3.0661522512911805,
+      "learning_rate": 0.00017734628358824365,
+      "loss": 1.421,
+      "step": 5253
+    },
+    {
+      "epoch": 0.2420417376883033,
+      "grad_norm": 2.973259257317361,
+      "learning_rate": 0.00017733682524791603,
+      "loss": 1.6475,
+      "step": 5254
+    },
+    {
+      "epoch": 0.2420878057769383,
+      "grad_norm": 26.947210387592094,
+      "learning_rate": 0.00017732736518581664,
+      "loss": 1.1786,
+      "step": 5255
+    },
+    {
+      "epoch": 0.2421338738655733,
+      "grad_norm": 2.253803508873098,
+      "learning_rate": 0.000177317903402156,
+      "loss": 1.6133,
+      "step": 5256
+    },
+    {
+      "epoch": 0.24217994195420833,
+      "grad_norm": 2.0947840633868715,
+      "learning_rate": 0.00017730843989714482,
+      "loss": 1.4437,
+      "step": 5257
+    },
+    {
+      "epoch": 0.24222601004284333,
+      "grad_norm": 3.1113392971688434,
+      "learning_rate": 0.00017729897467099378,
+      "loss": 1.4094,
+      "step": 5258
+    },
+    {
+      "epoch": 0.24227207813147833,
+      "grad_norm": 8.18471954682258,
+      "learning_rate": 0.00017728950772391366,
+      "loss": 1.6564,
+      "step": 5259
+    },
+    {
+      "epoch": 0.24231814622011333,
+      "grad_norm": 3.1063032482439494,
+      "learning_rate": 0.00017728003905611512,
+      "loss": 1.6512,
+      "step": 5260
+    },
+    {
+      "epoch": 0.24236421430874833,
+      "grad_norm": 2.221689092256146,
+      "learning_rate": 0.000177270568667809,
+      "loss": 1.4552,
+      "step": 5261
+    },
+    {
+      "epoch": 0.24241028239738333,
+      "grad_norm": 2.3101370324227988,
+      "learning_rate": 0.00017726109655920622,
+      "loss": 1.5488,
+      "step": 5262
+    },
+    {
+      "epoch": 0.24245635048601832,
+      "grad_norm": 1.795874693993714,
+      "learning_rate": 0.00017725162273051758,
+      "loss": 1.4817,
+      "step": 5263
+    },
+    {
+      "epoch": 0.24250241857465335,
+      "grad_norm": 7.153381312359439,
+      "learning_rate": 0.000177242147181954,
+      "loss": 1.4559,
+      "step": 5264
+    },
+    {
+      "epoch": 0.24254848666328835,
+      "grad_norm": 79.87907126646971,
+      "learning_rate": 0.00017723266991372647,
+      "loss": 1.6001,
+      "step": 5265
+    },
+    {
+      "epoch": 0.24259455475192335,
+      "grad_norm": 2.4927302613812126,
+      "learning_rate": 0.00017722319092604595,
+      "loss": 1.5073,
+      "step": 5266
+    },
+    {
+      "epoch": 0.24264062284055835,
+      "grad_norm": 5.07086358544952,
+      "learning_rate": 0.00017721371021912354,
+      "loss": 2.5317,
+      "step": 5267
+    },
+    {
+      "epoch": 0.24268669092919334,
+      "grad_norm": 9.142206875720294,
+      "learning_rate": 0.0001772042277931702,
+      "loss": 2.3692,
+      "step": 5268
+    },
+    {
+      "epoch": 0.24273275901782834,
+      "grad_norm": 3.095942653407108,
+      "learning_rate": 0.00017719474364839716,
+      "loss": 1.7885,
+      "step": 5269
+    },
+    {
+      "epoch": 0.24277882710646334,
+      "grad_norm": 8.532150301294399,
+      "learning_rate": 0.00017718525778501548,
+      "loss": 2.2607,
+      "step": 5270
+    },
+    {
+      "epoch": 0.24282489519509837,
+      "grad_norm": 3.602232185413302,
+      "learning_rate": 0.0001771757702032364,
+      "loss": 1.8227,
+      "step": 5271
+    },
+    {
+      "epoch": 0.24287096328373337,
+      "grad_norm": 3.70228274504761,
+      "learning_rate": 0.0001771662809032711,
+      "loss": 2.243,
+      "step": 5272
+    },
+    {
+      "epoch": 0.24291703137236836,
+      "grad_norm": 2.2161455165699357,
+      "learning_rate": 0.0001771567898853309,
+      "loss": 1.8591,
+      "step": 5273
+    },
+    {
+      "epoch": 0.24296309946100336,
+      "grad_norm": 3.81798155442017,
+      "learning_rate": 0.00017714729714962706,
+      "loss": 1.5134,
+      "step": 5274
+    },
+    {
+      "epoch": 0.24300916754963836,
+      "grad_norm": 2.7844235469193075,
+      "learning_rate": 0.00017713780269637091,
+      "loss": 1.8702,
+      "step": 5275
+    },
+    {
+      "epoch": 0.24305523563827336,
+      "grad_norm": 4.085333265505968,
+      "learning_rate": 0.00017712830652577387,
+      "loss": 1.9344,
+      "step": 5276
+    },
+    {
+      "epoch": 0.24310130372690836,
+      "grad_norm": 3.7570446307023873,
+      "learning_rate": 0.00017711880863804732,
+      "loss": 1.8174,
+      "step": 5277
+    },
+    {
+      "epoch": 0.24314737181554338,
+      "grad_norm": 8.780817704098006,
+      "learning_rate": 0.0001771093090334027,
+      "loss": 1.6324,
+      "step": 5278
+    },
+    {
+      "epoch": 0.24319343990417838,
+      "grad_norm": 3.3802473383615412,
+      "learning_rate": 0.00017709980771205157,
+      "loss": 1.8668,
+      "step": 5279
+    },
+    {
+      "epoch": 0.24323950799281338,
+      "grad_norm": 3.5261824155738437,
+      "learning_rate": 0.0001770903046742054,
+      "loss": 1.8106,
+      "step": 5280
+    },
+    {
+      "epoch": 0.24328557608144838,
+      "grad_norm": 3.4068078298511897,
+      "learning_rate": 0.00017708079992007577,
+      "loss": 1.6518,
+      "step": 5281
+    },
+    {
+      "epoch": 0.24333164417008338,
+      "grad_norm": 3.9007662418724665,
+      "learning_rate": 0.00017707129344987432,
+      "loss": 1.7761,
+      "step": 5282
+    },
+    {
+      "epoch": 0.24337771225871838,
+      "grad_norm": 3.155786191881021,
+      "learning_rate": 0.00017706178526381267,
+      "loss": 1.4828,
+      "step": 5283
+    },
+    {
+      "epoch": 0.24342378034735337,
+      "grad_norm": 2.522573678434306,
+      "learning_rate": 0.00017705227536210248,
+      "loss": 1.5935,
+      "step": 5284
+    },
+    {
+      "epoch": 0.2434698484359884,
+      "grad_norm": 4.812871980551906,
+      "learning_rate": 0.0001770427637449555,
+      "loss": 1.622,
+      "step": 5285
+    },
+    {
+      "epoch": 0.2435159165246234,
+      "grad_norm": 3.2923953159087476,
+      "learning_rate": 0.00017703325041258346,
+      "loss": 1.5556,
+      "step": 5286
+    },
+    {
+      "epoch": 0.2435619846132584,
+      "grad_norm": 2.644573069783067,
+      "learning_rate": 0.00017702373536519823,
+      "loss": 1.5702,
+      "step": 5287
+    },
+    {
+      "epoch": 0.2436080527018934,
+      "grad_norm": 2.6835225250674135,
+      "learning_rate": 0.00017701421860301158,
+      "loss": 1.8273,
+      "step": 5288
+    },
+    {
+      "epoch": 0.2436541207905284,
+      "grad_norm": 2.134609545989889,
+      "learning_rate": 0.00017700470012623542,
+      "loss": 1.5597,
+      "step": 5289
+    },
+    {
+      "epoch": 0.2437001888791634,
+      "grad_norm": 2.809843654993567,
+      "learning_rate": 0.00017699517993508161,
+      "loss": 1.6932,
+      "step": 5290
+    },
+    {
+      "epoch": 0.24374625696779842,
+      "grad_norm": 2.859656563703797,
+      "learning_rate": 0.0001769856580297622,
+      "loss": 1.5908,
+      "step": 5291
+    },
+    {
+      "epoch": 0.24379232505643342,
+      "grad_norm": 16.119350566021588,
+      "learning_rate": 0.00017697613441048903,
+      "loss": 1.6416,
+      "step": 5292
+    },
+    {
+      "epoch": 0.24383839314506842,
+      "grad_norm": 3.2689123908943425,
+      "learning_rate": 0.0001769666090774743,
+      "loss": 1.8971,
+      "step": 5293
+    },
+    {
+      "epoch": 0.24388446123370341,
+      "grad_norm": 2.4962501265195853,
+      "learning_rate": 0.00017695708203092992,
+      "loss": 1.9691,
+      "step": 5294
+    },
+    {
+      "epoch": 0.2439305293223384,
+      "grad_norm": 2.9938909314669537,
+      "learning_rate": 0.0001769475532710681,
+      "loss": 1.6123,
+      "step": 5295
+    },
+    {
+      "epoch": 0.2439765974109734,
+      "grad_norm": 4.002750148477527,
+      "learning_rate": 0.0001769380227981009,
+      "loss": 1.6361,
+      "step": 5296
+    },
+    {
+      "epoch": 0.2440226654996084,
+      "grad_norm": 6.260928837353397,
+      "learning_rate": 0.00017692849061224058,
+      "loss": 1.6286,
+      "step": 5297
+    },
+    {
+      "epoch": 0.24406873358824344,
+      "grad_norm": 5.779152848193037,
+      "learning_rate": 0.00017691895671369934,
+      "loss": 1.755,
+      "step": 5298
+    },
+    {
+      "epoch": 0.24411480167687843,
+      "grad_norm": 4.099340034858337,
+      "learning_rate": 0.00017690942110268936,
+      "loss": 1.847,
+      "step": 5299
+    },
+    {
+      "epoch": 0.24416086976551343,
+      "grad_norm": 1.6413963385715626,
+      "learning_rate": 0.00017689988377942307,
+      "loss": 1.6566,
+      "step": 5300
+    },
+    {
+      "epoch": 0.24420693785414843,
+      "grad_norm": 3.2388608421581897,
+      "learning_rate": 0.00017689034474411265,
+      "loss": 1.8979,
+      "step": 5301
+    },
+    {
+      "epoch": 0.24425300594278343,
+      "grad_norm": 1.8083602336127456,
+      "learning_rate": 0.0001768808039969706,
+      "loss": 1.8038,
+      "step": 5302
+    },
+    {
+      "epoch": 0.24429907403141843,
+      "grad_norm": 2.003500119235711,
+      "learning_rate": 0.00017687126153820924,
+      "loss": 1.448,
+      "step": 5303
+    },
+    {
+      "epoch": 0.24434514212005343,
+      "grad_norm": 2.8202365931827407,
+      "learning_rate": 0.00017686171736804103,
+      "loss": 1.6987,
+      "step": 5304
+    },
+    {
+      "epoch": 0.24439121020868845,
+      "grad_norm": 1.4468823740582812,
+      "learning_rate": 0.00017685217148667847,
+      "loss": 1.2861,
+      "step": 5305
+    },
+    {
+      "epoch": 0.24443727829732345,
+      "grad_norm": 2.327778466679373,
+      "learning_rate": 0.00017684262389433412,
+      "loss": 1.3114,
+      "step": 5306
+    },
+    {
+      "epoch": 0.24448334638595845,
+      "grad_norm": 1.9630151969161675,
+      "learning_rate": 0.00017683307459122048,
+      "loss": 1.6075,
+      "step": 5307
+    },
+    {
+      "epoch": 0.24452941447459345,
+      "grad_norm": 1.6865831626599268,
+      "learning_rate": 0.0001768235235775502,
+      "loss": 1.7591,
+      "step": 5308
+    },
+    {
+      "epoch": 0.24457548256322845,
+      "grad_norm": 1.8794336064805963,
+      "learning_rate": 0.00017681397085353586,
+      "loss": 1.7643,
+      "step": 5309
+    },
+    {
+      "epoch": 0.24462155065186345,
+      "grad_norm": 1.264514980521104,
+      "learning_rate": 0.00017680441641939018,
+      "loss": 1.2963,
+      "step": 5310
+    },
+    {
+      "epoch": 0.24466761874049844,
+      "grad_norm": 1.606409752830289,
+      "learning_rate": 0.00017679486027532589,
+      "loss": 1.4337,
+      "step": 5311
+    },
+    {
+      "epoch": 0.24471368682913347,
+      "grad_norm": 1.5098942202179533,
+      "learning_rate": 0.00017678530242155568,
+      "loss": 1.5639,
+      "step": 5312
+    },
+    {
+      "epoch": 0.24475975491776847,
+      "grad_norm": 2.3429568244488634,
+      "learning_rate": 0.00017677574285829238,
+      "loss": 1.791,
+      "step": 5313
+    },
+    {
+      "epoch": 0.24480582300640347,
+      "grad_norm": 3.229239280738775,
+      "learning_rate": 0.0001767661815857488,
+      "loss": 1.5041,
+      "step": 5314
+    },
+    {
+      "epoch": 0.24485189109503847,
+      "grad_norm": 6.624676841657915,
+      "learning_rate": 0.0001767566186041378,
+      "loss": 1.5698,
+      "step": 5315
+    },
+    {
+      "epoch": 0.24489795918367346,
+      "grad_norm": 1.648669059986612,
+      "learning_rate": 0.00017674705391367232,
+      "loss": 1.5889,
+      "step": 5316
+    },
+    {
+      "epoch": 0.24494402727230846,
+      "grad_norm": 1.5816331721577748,
+      "learning_rate": 0.00017673748751456528,
+      "loss": 1.5277,
+      "step": 5317
+    },
+    {
+      "epoch": 0.24499009536094346,
+      "grad_norm": 1.322754841994128,
+      "learning_rate": 0.00017672791940702963,
+      "loss": 1.4594,
+      "step": 5318
+    },
+    {
+      "epoch": 0.2450361634495785,
+      "grad_norm": 2.219041388950614,
+      "learning_rate": 0.00017671834959127845,
+      "loss": 1.5306,
+      "step": 5319
+    },
+    {
+      "epoch": 0.24508223153821349,
+      "grad_norm": 2.3229763072792893,
+      "learning_rate": 0.00017670877806752472,
+      "loss": 1.5987,
+      "step": 5320
+    },
+    {
+      "epoch": 0.24512829962684848,
+      "grad_norm": 2.0904190396345568,
+      "learning_rate": 0.0001766992048359816,
+      "loss": 1.7441,
+      "step": 5321
+    },
+    {
+      "epoch": 0.24517436771548348,
+      "grad_norm": 2.522474115070136,
+      "learning_rate": 0.00017668962989686218,
+      "loss": 1.4509,
+      "step": 5322
+    },
+    {
+      "epoch": 0.24522043580411848,
+      "grad_norm": 2.222228737907333,
+      "learning_rate": 0.00017668005325037966,
+      "loss": 1.7369,
+      "step": 5323
+    },
+    {
+      "epoch": 0.24526650389275348,
+      "grad_norm": 2.537806510344039,
+      "learning_rate": 0.00017667047489674723,
+      "loss": 1.9985,
+      "step": 5324
+    },
+    {
+      "epoch": 0.2453125719813885,
+      "grad_norm": 2.673117393828321,
+      "learning_rate": 0.00017666089483617814,
+      "loss": 1.4799,
+      "step": 5325
+    },
+    {
+      "epoch": 0.2453586400700235,
+      "grad_norm": 2.1783995455061107,
+      "learning_rate": 0.00017665131306888565,
+      "loss": 1.7925,
+      "step": 5326
+    },
+    {
+      "epoch": 0.2454047081586585,
+      "grad_norm": 1.6191475169705354,
+      "learning_rate": 0.0001766417295950831,
+      "loss": 1.6839,
+      "step": 5327
+    },
+    {
+      "epoch": 0.2454507762472935,
+      "grad_norm": 5.592726815993604,
+      "learning_rate": 0.0001766321444149839,
+      "loss": 1.3111,
+      "step": 5328
+    },
+    {
+      "epoch": 0.2454968443359285,
+      "grad_norm": 2.546402609908325,
+      "learning_rate": 0.00017662255752880137,
+      "loss": 1.4534,
+      "step": 5329
+    },
+    {
+      "epoch": 0.2455429124245635,
+      "grad_norm": 2.25711166848989,
+      "learning_rate": 0.00017661296893674896,
+      "loss": 1.7048,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2455889805131985,
+      "grad_norm": 3.8235071246523487,
+      "learning_rate": 0.00017660337863904018,
+      "loss": 1.753,
+      "step": 5331
+    },
+    {
+      "epoch": 0.24563504860183352,
+      "grad_norm": 4.278774560185654,
+      "learning_rate": 0.00017659378663588852,
+      "loss": 1.8109,
+      "step": 5332
+    },
+    {
+      "epoch": 0.24568111669046852,
+      "grad_norm": 2.4918028751409054,
+      "learning_rate": 0.00017658419292750755,
+      "loss": 1.5444,
+      "step": 5333
+    },
+    {
+      "epoch": 0.24572718477910352,
+      "grad_norm": 2.0953709790009425,
+      "learning_rate": 0.00017657459751411084,
+      "loss": 1.33,
+      "step": 5334
+    },
+    {
+      "epoch": 0.24577325286773852,
+      "grad_norm": 3.3836172028695817,
+      "learning_rate": 0.00017656500039591198,
+      "loss": 1.5859,
+      "step": 5335
+    },
+    {
+      "epoch": 0.24581932095637352,
+      "grad_norm": 1.7382546106813004,
+      "learning_rate": 0.00017655540157312467,
+      "loss": 1.3051,
+      "step": 5336
+    },
+    {
+      "epoch": 0.24586538904500851,
+      "grad_norm": 2.7176453102175815,
+      "learning_rate": 0.00017654580104596265,
+      "loss": 1.5707,
+      "step": 5337
+    },
+    {
+      "epoch": 0.2459114571336435,
+      "grad_norm": 2.8692768097332304,
+      "learning_rate": 0.00017653619881463953,
+      "loss": 1.6663,
+      "step": 5338
+    },
+    {
+      "epoch": 0.24595752522227854,
+      "grad_norm": 2.0945913964213125,
+      "learning_rate": 0.00017652659487936924,
+      "loss": 1.4967,
+      "step": 5339
+    },
+    {
+      "epoch": 0.24600359331091354,
+      "grad_norm": 1.9520321064061321,
+      "learning_rate": 0.00017651698924036553,
+      "loss": 1.5851,
+      "step": 5340
+    },
+    {
+      "epoch": 0.24604966139954854,
+      "grad_norm": 1.915421445876786,
+      "learning_rate": 0.00017650738189784227,
+      "loss": 1.4061,
+      "step": 5341
+    },
+    {
+      "epoch": 0.24609572948818353,
+      "grad_norm": 1.9577169319130998,
+      "learning_rate": 0.00017649777285201328,
+      "loss": 1.6858,
+      "step": 5342
+    },
+    {
+      "epoch": 0.24614179757681853,
+      "grad_norm": 2.902775186236754,
+      "learning_rate": 0.00017648816210309257,
+      "loss": 1.6097,
+      "step": 5343
+    },
+    {
+      "epoch": 0.24618786566545353,
+      "grad_norm": 2.1396741979742937,
+      "learning_rate": 0.00017647854965129408,
+      "loss": 1.3885,
+      "step": 5344
+    },
+    {
+      "epoch": 0.24623393375408853,
+      "grad_norm": 1.8779884109577545,
+      "learning_rate": 0.00017646893549683182,
+      "loss": 1.9598,
+      "step": 5345
+    },
+    {
+      "epoch": 0.24628000184272356,
+      "grad_norm": 1.9280633118734154,
+      "learning_rate": 0.00017645931963991985,
+      "loss": 1.6942,
+      "step": 5346
+    },
+    {
+      "epoch": 0.24632606993135855,
+      "grad_norm": 1.831852002038349,
+      "learning_rate": 0.00017644970208077217,
+      "loss": 1.547,
+      "step": 5347
+    },
+    {
+      "epoch": 0.24637213801999355,
+      "grad_norm": 6.831980071460246,
+      "learning_rate": 0.000176440082819603,
+      "loss": 1.326,
+      "step": 5348
+    },
+    {
+      "epoch": 0.24641820610862855,
+      "grad_norm": 1.795913634718264,
+      "learning_rate": 0.00017643046185662645,
+      "loss": 1.5495,
+      "step": 5349
+    },
+    {
+      "epoch": 0.24646427419726355,
+      "grad_norm": 6.660450371475235,
+      "learning_rate": 0.0001764208391920567,
+      "loss": 1.555,
+      "step": 5350
+    },
+    {
+      "epoch": 0.24651034228589855,
+      "grad_norm": 4.71795774227231,
+      "learning_rate": 0.00017641121482610798,
+      "loss": 1.3676,
+      "step": 5351
+    },
+    {
+      "epoch": 0.24655641037453357,
+      "grad_norm": 1.4448673542039696,
+      "learning_rate": 0.00017640158875899462,
+      "loss": 1.6579,
+      "step": 5352
+    },
+    {
+      "epoch": 0.24660247846316857,
+      "grad_norm": 1.9355689872462012,
+      "learning_rate": 0.00017639196099093087,
+      "loss": 1.9008,
+      "step": 5353
+    },
+    {
+      "epoch": 0.24664854655180357,
+      "grad_norm": 2.7919557601503437,
+      "learning_rate": 0.00017638233152213108,
+      "loss": 1.657,
+      "step": 5354
+    },
+    {
+      "epoch": 0.24669461464043857,
+      "grad_norm": 1.7353317025092443,
+      "learning_rate": 0.00017637270035280968,
+      "loss": 1.6415,
+      "step": 5355
+    },
+    {
+      "epoch": 0.24674068272907357,
+      "grad_norm": 2.6878571009661556,
+      "learning_rate": 0.00017636306748318104,
+      "loss": 1.7305,
+      "step": 5356
+    },
+    {
+      "epoch": 0.24678675081770857,
+      "grad_norm": 2.504753300654869,
+      "learning_rate": 0.00017635343291345963,
+      "loss": 1.437,
+      "step": 5357
+    },
+    {
+      "epoch": 0.24683281890634357,
+      "grad_norm": 3.7791548905626713,
+      "learning_rate": 0.00017634379664385996,
+      "loss": 1.6701,
+      "step": 5358
+    },
+    {
+      "epoch": 0.2468788869949786,
+      "grad_norm": 2.6891278374792993,
+      "learning_rate": 0.00017633415867459654,
+      "loss": 1.9389,
+      "step": 5359
+    },
+    {
+      "epoch": 0.2469249550836136,
+      "grad_norm": 3.2977042557297875,
+      "learning_rate": 0.00017632451900588398,
+      "loss": 1.5128,
+      "step": 5360
+    },
+    {
+      "epoch": 0.2469710231722486,
+      "grad_norm": 1.3518427252577134,
+      "learning_rate": 0.00017631487763793689,
+      "loss": 1.3917,
+      "step": 5361
+    },
+    {
+      "epoch": 0.2470170912608836,
+      "grad_norm": 3.5065482963353,
+      "learning_rate": 0.00017630523457096988,
+      "loss": 1.583,
+      "step": 5362
+    },
+    {
+      "epoch": 0.24706315934951859,
+      "grad_norm": 1.999948244266128,
+      "learning_rate": 0.00017629558980519766,
+      "loss": 1.5578,
+      "step": 5363
+    },
+    {
+      "epoch": 0.24710922743815358,
+      "grad_norm": 2.222708247020104,
+      "learning_rate": 0.00017628594334083495,
+      "loss": 1.3754,
+      "step": 5364
+    },
+    {
+      "epoch": 0.24715529552678858,
+      "grad_norm": 2.188360158045076,
+      "learning_rate": 0.00017627629517809652,
+      "loss": 1.5406,
+      "step": 5365
+    },
+    {
+      "epoch": 0.2472013636154236,
+      "grad_norm": 2.072769519177611,
+      "learning_rate": 0.00017626664531719718,
+      "loss": 1.67,
+      "step": 5366
+    },
+    {
+      "epoch": 0.2472474317040586,
+      "grad_norm": 4.2941771485824605,
+      "learning_rate": 0.00017625699375835172,
+      "loss": 1.7941,
+      "step": 5367
+    },
+    {
+      "epoch": 0.2472934997926936,
+      "grad_norm": 1.6542035330375735,
+      "learning_rate": 0.00017624734050177506,
+      "loss": 1.4882,
+      "step": 5368
+    },
+    {
+      "epoch": 0.2473395678813286,
+      "grad_norm": 4.151431253973902,
+      "learning_rate": 0.00017623768554768209,
+      "loss": 1.8099,
+      "step": 5369
+    },
+    {
+      "epoch": 0.2473856359699636,
+      "grad_norm": 2.2311753698636534,
+      "learning_rate": 0.00017622802889628777,
+      "loss": 1.8002,
+      "step": 5370
+    },
+    {
+      "epoch": 0.2474317040585986,
+      "grad_norm": 1.8516655479739375,
+      "learning_rate": 0.00017621837054780713,
+      "loss": 1.6516,
+      "step": 5371
+    },
+    {
+      "epoch": 0.2474777721472336,
+      "grad_norm": 3.927909028186663,
+      "learning_rate": 0.00017620871050245512,
+      "loss": 1.3593,
+      "step": 5372
+    },
+    {
+      "epoch": 0.24752384023586863,
+      "grad_norm": 4.744275774366879,
+      "learning_rate": 0.00017619904876044684,
+      "loss": 1.5384,
+      "step": 5373
+    },
+    {
+      "epoch": 0.24756990832450362,
+      "grad_norm": 1.9521094917024047,
+      "learning_rate": 0.0001761893853219974,
+      "loss": 1.6554,
+      "step": 5374
+    },
+    {
+      "epoch": 0.24761597641313862,
+      "grad_norm": 1.797916071861402,
+      "learning_rate": 0.00017617972018732194,
+      "loss": 1.5595,
+      "step": 5375
+    },
+    {
+      "epoch": 0.24766204450177362,
+      "grad_norm": 2.55311638370794,
+      "learning_rate": 0.00017617005335663564,
+      "loss": 1.1405,
+      "step": 5376
+    },
+    {
+      "epoch": 0.24770811259040862,
+      "grad_norm": 2.235260005532979,
+      "learning_rate": 0.00017616038483015366,
+      "loss": 1.3877,
+      "step": 5377
+    },
+    {
+      "epoch": 0.24775418067904362,
+      "grad_norm": 4.009309960907798,
+      "learning_rate": 0.00017615071460809131,
+      "loss": 1.5269,
+      "step": 5378
+    },
+    {
+      "epoch": 0.24780024876767862,
+      "grad_norm": 2.8374570588454726,
+      "learning_rate": 0.0001761410426906639,
+      "loss": 1.4118,
+      "step": 5379
+    },
+    {
+      "epoch": 0.24784631685631364,
+      "grad_norm": 2.4268279087538884,
+      "learning_rate": 0.00017613136907808672,
+      "loss": 1.6398,
+      "step": 5380
+    },
+    {
+      "epoch": 0.24789238494494864,
+      "grad_norm": 2.0840678176256797,
+      "learning_rate": 0.00017612169377057514,
+      "loss": 1.5082,
+      "step": 5381
+    },
+    {
+      "epoch": 0.24793845303358364,
+      "grad_norm": 2.2472624586198715,
+      "learning_rate": 0.00017611201676834458,
+      "loss": 1.6639,
+      "step": 5382
+    },
+    {
+      "epoch": 0.24798452112221864,
+      "grad_norm": 6.040737000203684,
+      "learning_rate": 0.00017610233807161045,
+      "loss": 1.7081,
+      "step": 5383
+    },
+    {
+      "epoch": 0.24803058921085364,
+      "grad_norm": 2.8756727044121755,
+      "learning_rate": 0.00017609265768058827,
+      "loss": 1.4527,
+      "step": 5384
+    },
+    {
+      "epoch": 0.24807665729948863,
+      "grad_norm": 2.788467229630324,
+      "learning_rate": 0.0001760829755954935,
+      "loss": 1.5773,
+      "step": 5385
+    },
+    {
+      "epoch": 0.24812272538812366,
+      "grad_norm": 1.7069470526042845,
+      "learning_rate": 0.00017607329181654178,
+      "loss": 1.4749,
+      "step": 5386
+    },
+    {
+      "epoch": 0.24816879347675866,
+      "grad_norm": 6.442298012252487,
+      "learning_rate": 0.00017606360634394864,
+      "loss": 1.2468,
+      "step": 5387
+    },
+    {
+      "epoch": 0.24821486156539366,
+      "grad_norm": 2.808894574718007,
+      "learning_rate": 0.0001760539191779297,
+      "loss": 1.524,
+      "step": 5388
+    },
+    {
+      "epoch": 0.24826092965402866,
+      "grad_norm": 7.751289508536604,
+      "learning_rate": 0.0001760442303187007,
+      "loss": 1.6668,
+      "step": 5389
+    },
+    {
+      "epoch": 0.24830699774266365,
+      "grad_norm": 2.5230795820167082,
+      "learning_rate": 0.00017603453976647727,
+      "loss": 1.8869,
+      "step": 5390
+    },
+    {
+      "epoch": 0.24835306583129865,
+      "grad_norm": 4.741935101778723,
+      "learning_rate": 0.0001760248475214752,
+      "loss": 1.4418,
+      "step": 5391
+    },
+    {
+      "epoch": 0.24839913391993365,
+      "grad_norm": 2.5121173863376884,
+      "learning_rate": 0.00017601515358391023,
+      "loss": 1.3876,
+      "step": 5392
+    },
+    {
+      "epoch": 0.24844520200856868,
+      "grad_norm": 1.5970294103195197,
+      "learning_rate": 0.0001760054579539982,
+      "loss": 1.5515,
+      "step": 5393
+    },
+    {
+      "epoch": 0.24849127009720368,
+      "grad_norm": 1.7198042466179106,
+      "learning_rate": 0.00017599576063195502,
+      "loss": 1.1243,
+      "step": 5394
+    },
+    {
+      "epoch": 0.24853733818583867,
+      "grad_norm": 2.0660599164969593,
+      "learning_rate": 0.00017598606161799648,
+      "loss": 1.5653,
+      "step": 5395
+    },
+    {
+      "epoch": 0.24858340627447367,
+      "grad_norm": 2.241458988087777,
+      "learning_rate": 0.0001759763609123386,
+      "loss": 1.3722,
+      "step": 5396
+    },
+    {
+      "epoch": 0.24862947436310867,
+      "grad_norm": 2.988940232318135,
+      "learning_rate": 0.0001759666585151973,
+      "loss": 1.5961,
+      "step": 5397
+    },
+    {
+      "epoch": 0.24867554245174367,
+      "grad_norm": 1.7497185267425601,
+      "learning_rate": 0.00017595695442678859,
+      "loss": 1.7276,
+      "step": 5398
+    },
+    {
+      "epoch": 0.24872161054037867,
+      "grad_norm": 4.801209144966833,
+      "learning_rate": 0.00017594724864732853,
+      "loss": 1.433,
+      "step": 5399
+    },
+    {
+      "epoch": 0.2487676786290137,
+      "grad_norm": 2.6190524566255355,
+      "learning_rate": 0.00017593754117703326,
+      "loss": 1.7166,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2488137467176487,
+      "grad_norm": 4.85521031984858,
+      "learning_rate": 0.00017592783201611877,
+      "loss": 1.397,
+      "step": 5401
+    },
+    {
+      "epoch": 0.2488598148062837,
+      "grad_norm": 2.0882772116313255,
+      "learning_rate": 0.00017591812116480134,
+      "loss": 1.3507,
+      "step": 5402
+    },
+    {
+      "epoch": 0.2489058828949187,
+      "grad_norm": 1.5879961763524932,
+      "learning_rate": 0.00017590840862329707,
+      "loss": 1.6676,
+      "step": 5403
+    },
+    {
+      "epoch": 0.2489519509835537,
+      "grad_norm": 1.6836389149629294,
+      "learning_rate": 0.00017589869439182224,
+      "loss": 1.6534,
+      "step": 5404
+    },
+    {
+      "epoch": 0.2489980190721887,
+      "grad_norm": 4.368167537817708,
+      "learning_rate": 0.00017588897847059314,
+      "loss": 1.5737,
+      "step": 5405
+    },
+    {
+      "epoch": 0.24904408716082369,
+      "grad_norm": 1.4930611979433073,
+      "learning_rate": 0.00017587926085982605,
+      "loss": 1.5637,
+      "step": 5406
+    },
+    {
+      "epoch": 0.2490901552494587,
+      "grad_norm": 1.9794396830278744,
+      "learning_rate": 0.00017586954155973734,
+      "loss": 1.3992,
+      "step": 5407
+    },
+    {
+      "epoch": 0.2491362233380937,
+      "grad_norm": 2.043885923385887,
+      "learning_rate": 0.00017585982057054334,
+      "loss": 1.3671,
+      "step": 5408
+    },
+    {
+      "epoch": 0.2491822914267287,
+      "grad_norm": 1.3998253716887363,
+      "learning_rate": 0.00017585009789246051,
+      "loss": 1.5352,
+      "step": 5409
+    },
+    {
+      "epoch": 0.2492283595153637,
+      "grad_norm": 1.480262698094269,
+      "learning_rate": 0.00017584037352570534,
+      "loss": 1.4851,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2492744276039987,
+      "grad_norm": 3.168586094504591,
+      "learning_rate": 0.00017583064747049427,
+      "loss": 1.5163,
+      "step": 5411
+    },
+    {
+      "epoch": 0.2493204956926337,
+      "grad_norm": 2.0333211253535204,
+      "learning_rate": 0.00017582091972704385,
+      "loss": 1.6407,
+      "step": 5412
+    },
+    {
+      "epoch": 0.2493665637812687,
+      "grad_norm": 1.8283393410778446,
+      "learning_rate": 0.00017581119029557067,
+      "loss": 1.7554,
+      "step": 5413
+    },
+    {
+      "epoch": 0.24941263186990373,
+      "grad_norm": 1.379102026506491,
+      "learning_rate": 0.00017580145917629132,
+      "loss": 1.4229,
+      "step": 5414
+    },
+    {
+      "epoch": 0.24945869995853873,
+      "grad_norm": 1.7988734491896936,
+      "learning_rate": 0.00017579172636942244,
+      "loss": 1.445,
+      "step": 5415
+    },
+    {
+      "epoch": 0.24950476804717373,
+      "grad_norm": 1.1635557518586241,
+      "learning_rate": 0.00017578199187518077,
+      "loss": 1.5136,
+      "step": 5416
+    },
+    {
+      "epoch": 0.24955083613580872,
+      "grad_norm": 2.372024847830814,
+      "learning_rate": 0.00017577225569378294,
+      "loss": 1.5247,
+      "step": 5417
+    },
+    {
+      "epoch": 0.24959690422444372,
+      "grad_norm": 7.900542429184733,
+      "learning_rate": 0.0001757625178254458,
+      "loss": 1.4861,
+      "step": 5418
+    },
+    {
+      "epoch": 0.24964297231307872,
+      "grad_norm": 2.1959283613874256,
+      "learning_rate": 0.00017575277827038606,
+      "loss": 1.3275,
+      "step": 5419
+    },
+    {
+      "epoch": 0.24968904040171375,
+      "grad_norm": 1.530355654200939,
+      "learning_rate": 0.00017574303702882062,
+      "loss": 1.3638,
+      "step": 5420
+    },
+    {
+      "epoch": 0.24973510849034875,
+      "grad_norm": 1.5117251390528041,
+      "learning_rate": 0.00017573329410096634,
+      "loss": 1.5565,
+      "step": 5421
+    },
+    {
+      "epoch": 0.24978117657898374,
+      "grad_norm": 1.2379352662823486,
+      "learning_rate": 0.00017572354948704012,
+      "loss": 1.3732,
+      "step": 5422
+    },
+    {
+      "epoch": 0.24982724466761874,
+      "grad_norm": 2.6367833600637063,
+      "learning_rate": 0.0001757138031872589,
+      "loss": 1.3779,
+      "step": 5423
+    },
+    {
+      "epoch": 0.24987331275625374,
+      "grad_norm": 1.4075314781727732,
+      "learning_rate": 0.00017570405520183968,
+      "loss": 1.4232,
+      "step": 5424
+    },
+    {
+      "epoch": 0.24991938084488874,
+      "grad_norm": 2.2582819482412755,
+      "learning_rate": 0.00017569430553099946,
+      "loss": 1.6949,
+      "step": 5425
+    },
+    {
+      "epoch": 0.24996544893352374,
+      "grad_norm": 2.633573262871944,
+      "learning_rate": 0.00017568455417495536,
+      "loss": 1.6143,
+      "step": 5426
+    },
+    {
+      "epoch": 0.25001151702215874,
+      "grad_norm": 2.1177159332522377,
+      "learning_rate": 0.00017567480113392438,
+      "loss": 1.3138,
+      "step": 5427
+    },
+    {
+      "epoch": 0.25005758511079373,
+      "grad_norm": 8.07331322386614,
+      "learning_rate": 0.00017566504640812374,
+      "loss": 1.3804,
+      "step": 5428
+    },
+    {
+      "epoch": 0.25010365319942873,
+      "grad_norm": 1.7763711424743986,
+      "learning_rate": 0.00017565528999777057,
+      "loss": 1.2743,
+      "step": 5429
+    },
+    {
+      "epoch": 0.25014972128806373,
+      "grad_norm": 2.0155362043782628,
+      "learning_rate": 0.0001756455319030821,
+      "loss": 1.3259,
+      "step": 5430
+    },
+    {
+      "epoch": 0.2501957893766988,
+      "grad_norm": 1.0731301501704065,
+      "learning_rate": 0.00017563577212427554,
+      "loss": 1.5571,
+      "step": 5431
+    },
+    {
+      "epoch": 0.2502418574653338,
+      "grad_norm": 1.5049747506608955,
+      "learning_rate": 0.00017562601066156822,
+      "loss": 1.8258,
+      "step": 5432
+    },
+    {
+      "epoch": 0.2502879255539688,
+      "grad_norm": 1.971062441083597,
+      "learning_rate": 0.00017561624751517743,
+      "loss": 1.6033,
+      "step": 5433
+    },
+    {
+      "epoch": 0.2503339936426038,
+      "grad_norm": 1.553732122373133,
+      "learning_rate": 0.00017560648268532053,
+      "loss": 1.2548,
+      "step": 5434
+    },
+    {
+      "epoch": 0.2503800617312388,
+      "grad_norm": 1.7084758409370417,
+      "learning_rate": 0.00017559671617221496,
+      "loss": 1.6074,
+      "step": 5435
+    },
+    {
+      "epoch": 0.2504261298198738,
+      "grad_norm": 1.8481076483776036,
+      "learning_rate": 0.00017558694797607812,
+      "loss": 1.6376,
+      "step": 5436
+    },
+    {
+      "epoch": 0.2504721979085088,
+      "grad_norm": 2.3163797580635643,
+      "learning_rate": 0.00017557717809712752,
+      "loss": 1.6192,
+      "step": 5437
+    },
+    {
+      "epoch": 0.2505182659971438,
+      "grad_norm": 1.583705097346804,
+      "learning_rate": 0.00017556740653558057,
+      "loss": 1.362,
+      "step": 5438
+    },
+    {
+      "epoch": 0.2505643340857788,
+      "grad_norm": 2.270186354619915,
+      "learning_rate": 0.00017555763329165492,
+      "loss": 1.7367,
+      "step": 5439
+    },
+    {
+      "epoch": 0.25061040217441377,
+      "grad_norm": 7.461364373768367,
+      "learning_rate": 0.00017554785836556812,
+      "loss": 1.5587,
+      "step": 5440
+    },
+    {
+      "epoch": 0.25065647026304877,
+      "grad_norm": 3.8294556658051144,
+      "learning_rate": 0.0001755380817575378,
+      "loss": 1.2841,
+      "step": 5441
+    },
+    {
+      "epoch": 0.25070253835168377,
+      "grad_norm": 2.698983886484633,
+      "learning_rate": 0.00017552830346778157,
+      "loss": 1.4668,
+      "step": 5442
+    },
+    {
+      "epoch": 0.25074860644031877,
+      "grad_norm": 2.207345444510309,
+      "learning_rate": 0.00017551852349651718,
+      "loss": 1.3927,
+      "step": 5443
+    },
+    {
+      "epoch": 0.2507946745289538,
+      "grad_norm": 3.0358194492488586,
+      "learning_rate": 0.00017550874184396235,
+      "loss": 1.4466,
+      "step": 5444
+    },
+    {
+      "epoch": 0.2508407426175888,
+      "grad_norm": 2.5412710491415273,
+      "learning_rate": 0.00017549895851033488,
+      "loss": 1.4933,
+      "step": 5445
+    },
+    {
+      "epoch": 0.2508868107062238,
+      "grad_norm": 1.821929430251437,
+      "learning_rate": 0.00017548917349585253,
+      "loss": 1.204,
+      "step": 5446
+    },
+    {
+      "epoch": 0.2509328787948588,
+      "grad_norm": 2.2294626121787724,
+      "learning_rate": 0.00017547938680073316,
+      "loss": 1.5637,
+      "step": 5447
+    },
+    {
+      "epoch": 0.2509789468834938,
+      "grad_norm": 6.114334046761913,
+      "learning_rate": 0.00017546959842519472,
+      "loss": 1.4805,
+      "step": 5448
+    },
+    {
+      "epoch": 0.2510250149721288,
+      "grad_norm": 1.8891950154688129,
+      "learning_rate": 0.00017545980836945502,
+      "loss": 1.6415,
+      "step": 5449
+    },
+    {
+      "epoch": 0.2510710830607638,
+      "grad_norm": 2.321620718239024,
+      "learning_rate": 0.00017545001663373205,
+      "loss": 1.5643,
+      "step": 5450
+    },
+    {
+      "epoch": 0.2511171511493988,
+      "grad_norm": 6.714749709359901,
+      "learning_rate": 0.0001754402232182439,
+      "loss": 1.6426,
+      "step": 5451
+    },
+    {
+      "epoch": 0.2511632192380338,
+      "grad_norm": 3.5154420014759915,
+      "learning_rate": 0.00017543042812320848,
+      "loss": 1.2032,
+      "step": 5452
+    },
+    {
+      "epoch": 0.2512092873266688,
+      "grad_norm": 3.467046346137944,
+      "learning_rate": 0.00017542063134884392,
+      "loss": 1.5444,
+      "step": 5453
+    },
+    {
+      "epoch": 0.2512553554153038,
+      "grad_norm": 2.8993564138294317,
+      "learning_rate": 0.00017541083289536832,
+      "loss": 1.3849,
+      "step": 5454
+    },
+    {
+      "epoch": 0.2513014235039388,
+      "grad_norm": 5.905698291203194,
+      "learning_rate": 0.00017540103276299983,
+      "loss": 1.6156,
+      "step": 5455
+    },
+    {
+      "epoch": 0.2513474915925738,
+      "grad_norm": 1.6959435412400061,
+      "learning_rate": 0.00017539123095195667,
+      "loss": 1.1813,
+      "step": 5456
+    },
+    {
+      "epoch": 0.2513935596812088,
+      "grad_norm": 2.1659804647149197,
+      "learning_rate": 0.00017538142746245696,
+      "loss": 1.7172,
+      "step": 5457
+    },
+    {
+      "epoch": 0.25143962776984385,
+      "grad_norm": 2.8818176044508292,
+      "learning_rate": 0.00017537162229471905,
+      "loss": 1.5262,
+      "step": 5458
+    },
+    {
+      "epoch": 0.25148569585847885,
+      "grad_norm": 2.2152373720560585,
+      "learning_rate": 0.00017536181544896122,
+      "loss": 1.4736,
+      "step": 5459
+    },
+    {
+      "epoch": 0.25153176394711385,
+      "grad_norm": 2.0511123619741047,
+      "learning_rate": 0.00017535200692540178,
+      "loss": 1.5478,
+      "step": 5460
+    },
+    {
+      "epoch": 0.25157783203574885,
+      "grad_norm": 2.0240549920468105,
+      "learning_rate": 0.00017534219672425908,
+      "loss": 1.458,
+      "step": 5461
+    },
+    {
+      "epoch": 0.25162390012438385,
+      "grad_norm": 4.802346062498671,
+      "learning_rate": 0.00017533238484575153,
+      "loss": 1.4757,
+      "step": 5462
+    },
+    {
+      "epoch": 0.25166996821301885,
+      "grad_norm": 1.4838091938729132,
+      "learning_rate": 0.00017532257129009767,
+      "loss": 1.7622,
+      "step": 5463
+    },
+    {
+      "epoch": 0.25171603630165384,
+      "grad_norm": 5.925849337447367,
+      "learning_rate": 0.00017531275605751586,
+      "loss": 1.5451,
+      "step": 5464
+    },
+    {
+      "epoch": 0.25176210439028884,
+      "grad_norm": 2.70437808285618,
+      "learning_rate": 0.0001753029391482247,
+      "loss": 1.2919,
+      "step": 5465
+    },
+    {
+      "epoch": 0.25180817247892384,
+      "grad_norm": 2.7702301007881402,
+      "learning_rate": 0.00017529312056244268,
+      "loss": 1.2873,
+      "step": 5466
+    },
+    {
+      "epoch": 0.25185424056755884,
+      "grad_norm": 2.1355508618817445,
+      "learning_rate": 0.00017528330030038846,
+      "loss": 1.4131,
+      "step": 5467
+    },
+    {
+      "epoch": 0.25190030865619384,
+      "grad_norm": 9.868979173299397,
+      "learning_rate": 0.00017527347836228066,
+      "loss": 1.5987,
+      "step": 5468
+    },
+    {
+      "epoch": 0.25194637674482884,
+      "grad_norm": 1.6570857072971343,
+      "learning_rate": 0.0001752636547483379,
+      "loss": 1.1424,
+      "step": 5469
+    },
+    {
+      "epoch": 0.25199244483346384,
+      "grad_norm": 4.273809740642672,
+      "learning_rate": 0.00017525382945877892,
+      "loss": 1.4305,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2520385129220989,
+      "grad_norm": 1.9160050557644885,
+      "learning_rate": 0.00017524400249382247,
+      "loss": 1.3521,
+      "step": 5471
+    },
+    {
+      "epoch": 0.2520845810107339,
+      "grad_norm": 1.6552113406747722,
+      "learning_rate": 0.0001752341738536873,
+      "loss": 1.5411,
+      "step": 5472
+    },
+    {
+      "epoch": 0.2521306490993689,
+      "grad_norm": 1.5312097681492658,
+      "learning_rate": 0.00017522434353859227,
+      "loss": 1.6895,
+      "step": 5473
+    },
+    {
+      "epoch": 0.2521767171880039,
+      "grad_norm": 1.2027160078621246,
+      "learning_rate": 0.00017521451154875622,
+      "loss": 1.6106,
+      "step": 5474
+    },
+    {
+      "epoch": 0.2522227852766389,
+      "grad_norm": 1.5084070137236032,
+      "learning_rate": 0.00017520467788439801,
+      "loss": 1.7293,
+      "step": 5475
+    },
+    {
+      "epoch": 0.2522688533652739,
+      "grad_norm": 4.728968238326298,
+      "learning_rate": 0.0001751948425457366,
+      "loss": 1.6832,
+      "step": 5476
+    },
+    {
+      "epoch": 0.2523149214539089,
+      "grad_norm": 1.9064821157003624,
+      "learning_rate": 0.000175185005532991,
+      "loss": 1.5497,
+      "step": 5477
+    },
+    {
+      "epoch": 0.2523609895425439,
+      "grad_norm": 1.5345666988920876,
+      "learning_rate": 0.0001751751668463801,
+      "loss": 1.3243,
+      "step": 5478
+    },
+    {
+      "epoch": 0.2524070576311789,
+      "grad_norm": 1.3580506871551783,
+      "learning_rate": 0.00017516532648612302,
+      "loss": 1.2745,
+      "step": 5479
+    },
+    {
+      "epoch": 0.2524531257198139,
+      "grad_norm": 1.8217338023588578,
+      "learning_rate": 0.00017515548445243883,
+      "loss": 1.5505,
+      "step": 5480
+    },
+    {
+      "epoch": 0.2524991938084489,
+      "grad_norm": 1.4663959312671702,
+      "learning_rate": 0.00017514564074554666,
+      "loss": 1.4335,
+      "step": 5481
+    },
+    {
+      "epoch": 0.2525452618970839,
+      "grad_norm": 3.5881743514509914,
+      "learning_rate": 0.00017513579536566565,
+      "loss": 1.5684,
+      "step": 5482
+    },
+    {
+      "epoch": 0.25259132998571887,
+      "grad_norm": 4.388127366154469,
+      "learning_rate": 0.00017512594831301496,
+      "loss": 1.3821,
+      "step": 5483
+    },
+    {
+      "epoch": 0.25263739807435387,
+      "grad_norm": 1.0164883662664073,
+      "learning_rate": 0.0001751160995878139,
+      "loss": 1.5622,
+      "step": 5484
+    },
+    {
+      "epoch": 0.2526834661629889,
+      "grad_norm": 1.578716626951196,
+      "learning_rate": 0.0001751062491902816,
+      "loss": 1.8683,
+      "step": 5485
+    },
+    {
+      "epoch": 0.2527295342516239,
+      "grad_norm": 1.9907280030347683,
+      "learning_rate": 0.00017509639712063747,
+      "loss": 1.3775,
+      "step": 5486
+    },
+    {
+      "epoch": 0.2527756023402589,
+      "grad_norm": 1.1473639647445504,
+      "learning_rate": 0.00017508654337910085,
+      "loss": 1.5592,
+      "step": 5487
+    },
+    {
+      "epoch": 0.2528216704288939,
+      "grad_norm": 1.388002374740739,
+      "learning_rate": 0.00017507668796589107,
+      "loss": 1.2449,
+      "step": 5488
+    },
+    {
+      "epoch": 0.2528677385175289,
+      "grad_norm": 1.158384872124298,
+      "learning_rate": 0.00017506683088122754,
+      "loss": 1.2533,
+      "step": 5489
+    },
+    {
+      "epoch": 0.2529138066061639,
+      "grad_norm": 3.2213820256797487,
+      "learning_rate": 0.00017505697212532976,
+      "loss": 1.3581,
+      "step": 5490
+    },
+    {
+      "epoch": 0.2529598746947989,
+      "grad_norm": 3.283750206745605,
+      "learning_rate": 0.00017504711169841718,
+      "loss": 1.4088,
+      "step": 5491
+    },
+    {
+      "epoch": 0.2530059427834339,
+      "grad_norm": 1.4197230574708717,
+      "learning_rate": 0.00017503724960070933,
+      "loss": 1.4233,
+      "step": 5492
+    },
+    {
+      "epoch": 0.2530520108720689,
+      "grad_norm": 1.1456447135961116,
+      "learning_rate": 0.00017502738583242579,
+      "loss": 1.2787,
+      "step": 5493
+    },
+    {
+      "epoch": 0.2530980789607039,
+      "grad_norm": 1.175166600060173,
+      "learning_rate": 0.00017501752039378614,
+      "loss": 1.7646,
+      "step": 5494
+    },
+    {
+      "epoch": 0.2531441470493389,
+      "grad_norm": 2.533079640190523,
+      "learning_rate": 0.00017500765328501005,
+      "loss": 1.4114,
+      "step": 5495
+    },
+    {
+      "epoch": 0.2531902151379739,
+      "grad_norm": 2.1359147661853815,
+      "learning_rate": 0.00017499778450631715,
+      "loss": 1.1513,
+      "step": 5496
+    },
+    {
+      "epoch": 0.2532362832266089,
+      "grad_norm": 2.8516578807499173,
+      "learning_rate": 0.00017498791405792718,
+      "loss": 1.241,
+      "step": 5497
+    },
+    {
+      "epoch": 0.2532823513152439,
+      "grad_norm": 2.2998364217881604,
+      "learning_rate": 0.00017497804194005985,
+      "loss": 1.2768,
+      "step": 5498
+    },
+    {
+      "epoch": 0.25332841940387896,
+      "grad_norm": 2.6906088149043845,
+      "learning_rate": 0.00017496816815293503,
+      "loss": 1.3454,
+      "step": 5499
+    },
+    {
+      "epoch": 0.25337448749251396,
+      "grad_norm": 14.098767964710536,
+      "learning_rate": 0.00017495829269677246,
+      "loss": 1.2312,
+      "step": 5500
+    },
+    {
+      "epoch": 0.25342055558114895,
+      "grad_norm": 2.2799804015274576,
+      "learning_rate": 0.00017494841557179203,
+      "loss": 1.4639,
+      "step": 5501
+    },
+    {
+      "epoch": 0.25346662366978395,
+      "grad_norm": 1.1997592397038275,
+      "learning_rate": 0.00017493853677821362,
+      "loss": 1.3986,
+      "step": 5502
+    },
+    {
+      "epoch": 0.25351269175841895,
+      "grad_norm": 2.11097022011053,
+      "learning_rate": 0.00017492865631625718,
+      "loss": 1.5797,
+      "step": 5503
+    },
+    {
+      "epoch": 0.25355875984705395,
+      "grad_norm": 8.384945103556014,
+      "learning_rate": 0.00017491877418614272,
+      "loss": 1.5154,
+      "step": 5504
+    },
+    {
+      "epoch": 0.25360482793568895,
+      "grad_norm": 19.92924744312792,
+      "learning_rate": 0.00017490889038809018,
+      "loss": 1.3745,
+      "step": 5505
+    },
+    {
+      "epoch": 0.25365089602432395,
+      "grad_norm": 2.157184447225862,
+      "learning_rate": 0.00017489900492231964,
+      "loss": 1.4789,
+      "step": 5506
+    },
+    {
+      "epoch": 0.25369696411295894,
+      "grad_norm": 1.9101469726633722,
+      "learning_rate": 0.0001748891177890512,
+      "loss": 1.4674,
+      "step": 5507
+    },
+    {
+      "epoch": 0.25374303220159394,
+      "grad_norm": 3.131596649957299,
+      "learning_rate": 0.00017487922898850493,
+      "loss": 1.4644,
+      "step": 5508
+    },
+    {
+      "epoch": 0.25378910029022894,
+      "grad_norm": 3.442069761515664,
+      "learning_rate": 0.00017486933852090103,
+      "loss": 1.5563,
+      "step": 5509
+    },
+    {
+      "epoch": 0.25383516837886394,
+      "grad_norm": 4.484338027779628,
+      "learning_rate": 0.00017485944638645967,
+      "loss": 1.6919,
+      "step": 5510
+    },
+    {
+      "epoch": 0.25388123646749894,
+      "grad_norm": 2.1192648690022953,
+      "learning_rate": 0.00017484955258540114,
+      "loss": 1.7424,
+      "step": 5511
+    },
+    {
+      "epoch": 0.253927304556134,
+      "grad_norm": 21.515486979698252,
+      "learning_rate": 0.0001748396571179456,
+      "loss": 2.6963,
+      "step": 5512
+    },
+    {
+      "epoch": 0.253973372644769,
+      "grad_norm": 1.9780263420344435,
+      "learning_rate": 0.00017482975998431343,
+      "loss": 1.5695,
+      "step": 5513
+    },
+    {
+      "epoch": 0.254019440733404,
+      "grad_norm": 1.8168517326175802,
+      "learning_rate": 0.000174819861184725,
+      "loss": 1.6378,
+      "step": 5514
+    },
+    {
+      "epoch": 0.254065508822039,
+      "grad_norm": 1.9020213903193257,
+      "learning_rate": 0.0001748099607194006,
+      "loss": 1.7478,
+      "step": 5515
+    },
+    {
+      "epoch": 0.254111576910674,
+      "grad_norm": 2.0220610137704025,
+      "learning_rate": 0.00017480005858856075,
+      "loss": 1.5524,
+      "step": 5516
+    },
+    {
+      "epoch": 0.254157644999309,
+      "grad_norm": 2.5258628946429114,
+      "learning_rate": 0.0001747901547924258,
+      "loss": 1.5311,
+      "step": 5517
+    },
+    {
+      "epoch": 0.254203713087944,
+      "grad_norm": 1.8628181941759578,
+      "learning_rate": 0.0001747802493312163,
+      "loss": 1.8008,
+      "step": 5518
+    },
+    {
+      "epoch": 0.254249781176579,
+      "grad_norm": 2.1948465587111783,
+      "learning_rate": 0.0001747703422051528,
+      "loss": 1.8132,
+      "step": 5519
+    },
+    {
+      "epoch": 0.254295849265214,
+      "grad_norm": 1.6738554370467789,
+      "learning_rate": 0.0001747604334144558,
+      "loss": 1.603,
+      "step": 5520
+    },
+    {
+      "epoch": 0.254341917353849,
+      "grad_norm": 3.051199343794954,
+      "learning_rate": 0.00017475052295934593,
+      "loss": 1.6429,
+      "step": 5521
+    },
+    {
+      "epoch": 0.254387985442484,
+      "grad_norm": 1.3622041617160907,
+      "learning_rate": 0.00017474061084004388,
+      "loss": 1.5002,
+      "step": 5522
+    },
+    {
+      "epoch": 0.254434053531119,
+      "grad_norm": 1.8849195318571195,
+      "learning_rate": 0.00017473069705677024,
+      "loss": 1.6605,
+      "step": 5523
+    },
+    {
+      "epoch": 0.254480121619754,
+      "grad_norm": 2.907207075748759,
+      "learning_rate": 0.00017472078160974582,
+      "loss": 1.5346,
+      "step": 5524
+    },
+    {
+      "epoch": 0.254526189708389,
+      "grad_norm": 2.3076346250497597,
+      "learning_rate": 0.00017471086449919124,
+      "loss": 1.6534,
+      "step": 5525
+    },
+    {
+      "epoch": 0.254572257797024,
+      "grad_norm": 6.5703727122025555,
+      "learning_rate": 0.00017470094572532741,
+      "loss": 1.528,
+      "step": 5526
+    },
+    {
+      "epoch": 0.254618325885659,
+      "grad_norm": 2.3583235560993265,
+      "learning_rate": 0.0001746910252883751,
+      "loss": 1.5073,
+      "step": 5527
+    },
+    {
+      "epoch": 0.254664393974294,
+      "grad_norm": 2.3012132636763596,
+      "learning_rate": 0.00017468110318855518,
+      "loss": 1.5701,
+      "step": 5528
+    },
+    {
+      "epoch": 0.254710462062929,
+      "grad_norm": 1.5791077579419843,
+      "learning_rate": 0.00017467117942608854,
+      "loss": 1.3132,
+      "step": 5529
+    },
+    {
+      "epoch": 0.254756530151564,
+      "grad_norm": 9.404103825698053,
+      "learning_rate": 0.00017466125400119616,
+      "loss": 1.5199,
+      "step": 5530
+    },
+    {
+      "epoch": 0.254802598240199,
+      "grad_norm": 2.916456237596943,
+      "learning_rate": 0.00017465132691409893,
+      "loss": 1.5915,
+      "step": 5531
+    },
+    {
+      "epoch": 0.254848666328834,
+      "grad_norm": 1.5319868895699085,
+      "learning_rate": 0.00017464139816501795,
+      "loss": 1.4646,
+      "step": 5532
+    },
+    {
+      "epoch": 0.254894734417469,
+      "grad_norm": 2.6280944771097645,
+      "learning_rate": 0.0001746314677541742,
+      "loss": 1.4829,
+      "step": 5533
+    },
+    {
+      "epoch": 0.254940802506104,
+      "grad_norm": 2.554480179986707,
+      "learning_rate": 0.00017462153568178877,
+      "loss": 1.6537,
+      "step": 5534
+    },
+    {
+      "epoch": 0.254986870594739,
+      "grad_norm": 2.0882737900784383,
+      "learning_rate": 0.0001746116019480828,
+      "loss": 1.4895,
+      "step": 5535
+    },
+    {
+      "epoch": 0.255032938683374,
+      "grad_norm": 1.47869349901792,
+      "learning_rate": 0.00017460166655327745,
+      "loss": 1.6583,
+      "step": 5536
+    },
+    {
+      "epoch": 0.255079006772009,
+      "grad_norm": 6.374734224498179,
+      "learning_rate": 0.00017459172949759393,
+      "loss": 1.5523,
+      "step": 5537
+    },
+    {
+      "epoch": 0.255125074860644,
+      "grad_norm": 2.740468352468998,
+      "learning_rate": 0.00017458179078125343,
+      "loss": 1.5755,
+      "step": 5538
+    },
+    {
+      "epoch": 0.25517114294927906,
+      "grad_norm": 2.5708937530034564,
+      "learning_rate": 0.00017457185040447724,
+      "loss": 1.428,
+      "step": 5539
+    },
+    {
+      "epoch": 0.25521721103791406,
+      "grad_norm": 3.8077321553249823,
+      "learning_rate": 0.00017456190836748667,
+      "loss": 1.515,
+      "step": 5540
+    },
+    {
+      "epoch": 0.25526327912654906,
+      "grad_norm": 3.0165813224910623,
+      "learning_rate": 0.00017455196467050304,
+      "loss": 1.5495,
+      "step": 5541
+    },
+    {
+      "epoch": 0.25530934721518406,
+      "grad_norm": 1.9830832898605315,
+      "learning_rate": 0.00017454201931374773,
+      "loss": 1.4378,
+      "step": 5542
+    },
+    {
+      "epoch": 0.25535541530381906,
+      "grad_norm": 1.7572277691119393,
+      "learning_rate": 0.0001745320722974422,
+      "loss": 1.2881,
+      "step": 5543
+    },
+    {
+      "epoch": 0.25540148339245405,
+      "grad_norm": 1.5991342556004442,
+      "learning_rate": 0.0001745221236218079,
+      "loss": 1.4085,
+      "step": 5544
+    },
+    {
+      "epoch": 0.25544755148108905,
+      "grad_norm": 2.161973881515504,
+      "learning_rate": 0.00017451217328706627,
+      "loss": 1.8022,
+      "step": 5545
+    },
+    {
+      "epoch": 0.25549361956972405,
+      "grad_norm": 4.211491297875835,
+      "learning_rate": 0.00017450222129343886,
+      "loss": 1.8018,
+      "step": 5546
+    },
+    {
+      "epoch": 0.25553968765835905,
+      "grad_norm": 2.0091661182783347,
+      "learning_rate": 0.00017449226764114725,
+      "loss": 1.6512,
+      "step": 5547
+    },
+    {
+      "epoch": 0.25558575574699405,
+      "grad_norm": 1.466826221262585,
+      "learning_rate": 0.000174482312330413,
+      "loss": 1.4391,
+      "step": 5548
+    },
+    {
+      "epoch": 0.25563182383562905,
+      "grad_norm": 2.1438381032390863,
+      "learning_rate": 0.0001744723553614578,
+      "loss": 1.5497,
+      "step": 5549
+    },
+    {
+      "epoch": 0.25567789192426404,
+      "grad_norm": 2.2373378729554947,
+      "learning_rate": 0.0001744623967345033,
+      "loss": 1.5733,
+      "step": 5550
+    },
+    {
+      "epoch": 0.25572396001289904,
+      "grad_norm": 1.9120502694499497,
+      "learning_rate": 0.00017445243644977122,
+      "loss": 1.5228,
+      "step": 5551
+    },
+    {
+      "epoch": 0.25577002810153404,
+      "grad_norm": 2.844790172738411,
+      "learning_rate": 0.00017444247450748327,
+      "loss": 1.6195,
+      "step": 5552
+    },
+    {
+      "epoch": 0.2558160961901691,
+      "grad_norm": 1.4831899152286008,
+      "learning_rate": 0.0001744325109078613,
+      "loss": 1.5866,
+      "step": 5553
+    },
+    {
+      "epoch": 0.2558621642788041,
+      "grad_norm": 2.1659330783826567,
+      "learning_rate": 0.00017442254565112707,
+      "loss": 1.314,
+      "step": 5554
+    },
+    {
+      "epoch": 0.2559082323674391,
+      "grad_norm": 4.015814136844452,
+      "learning_rate": 0.00017441257873750248,
+      "loss": 1.6123,
+      "step": 5555
+    },
+    {
+      "epoch": 0.2559543004560741,
+      "grad_norm": 1.4133966459179899,
+      "learning_rate": 0.00017440261016720942,
+      "loss": 1.441,
+      "step": 5556
+    },
+    {
+      "epoch": 0.2560003685447091,
+      "grad_norm": 1.5200438553665467,
+      "learning_rate": 0.0001743926399404698,
+      "loss": 1.2349,
+      "step": 5557
+    },
+    {
+      "epoch": 0.2560464366333441,
+      "grad_norm": 3.8026795017232415,
+      "learning_rate": 0.00017438266805750564,
+      "loss": 1.7462,
+      "step": 5558
+    },
+    {
+      "epoch": 0.2560925047219791,
+      "grad_norm": 1.4774016605102605,
+      "learning_rate": 0.0001743726945185389,
+      "loss": 1.392,
+      "step": 5559
+    },
+    {
+      "epoch": 0.2561385728106141,
+      "grad_norm": 2.2415497658048373,
+      "learning_rate": 0.00017436271932379165,
+      "loss": 1.4705,
+      "step": 5560
+    },
+    {
+      "epoch": 0.2561846408992491,
+      "grad_norm": 3.3795310057539094,
+      "learning_rate": 0.00017435274247348592,
+      "loss": 1.4863,
+      "step": 5561
+    },
+    {
+      "epoch": 0.2562307089878841,
+      "grad_norm": 1.4069803075786882,
+      "learning_rate": 0.0001743427639678439,
+      "loss": 1.4597,
+      "step": 5562
+    },
+    {
+      "epoch": 0.2562767770765191,
+      "grad_norm": 2.1012016557182105,
+      "learning_rate": 0.0001743327838070877,
+      "loss": 1.4228,
+      "step": 5563
+    },
+    {
+      "epoch": 0.2563228451651541,
+      "grad_norm": 4.151729414752208,
+      "learning_rate": 0.00017432280199143954,
+      "loss": 1.5897,
+      "step": 5564
+    },
+    {
+      "epoch": 0.2563689132537891,
+      "grad_norm": 1.0025768755541153,
+      "learning_rate": 0.00017431281852112158,
+      "loss": 1.7829,
+      "step": 5565
+    },
+    {
+      "epoch": 0.2564149813424241,
+      "grad_norm": 2.65987371150082,
+      "learning_rate": 0.00017430283339635617,
+      "loss": 1.2488,
+      "step": 5566
+    },
+    {
+      "epoch": 0.25646104943105913,
+      "grad_norm": 1.8701315019533133,
+      "learning_rate": 0.00017429284661736559,
+      "loss": 1.5807,
+      "step": 5567
+    },
+    {
+      "epoch": 0.25650711751969413,
+      "grad_norm": 1.0844246824813972,
+      "learning_rate": 0.00017428285818437214,
+      "loss": 1.3876,
+      "step": 5568
+    },
+    {
+      "epoch": 0.2565531856083291,
+      "grad_norm": 1.6372376697525026,
+      "learning_rate": 0.00017427286809759823,
+      "loss": 1.6262,
+      "step": 5569
+    },
+    {
+      "epoch": 0.2565992536969641,
+      "grad_norm": 5.32839010141331,
+      "learning_rate": 0.00017426287635726625,
+      "loss": 1.4258,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2566453217855991,
+      "grad_norm": 2.378316665882042,
+      "learning_rate": 0.00017425288296359868,
+      "loss": 1.808,
+      "step": 5571
+    },
+    {
+      "epoch": 0.2566913898742341,
+      "grad_norm": 1.5744651537763747,
+      "learning_rate": 0.000174242887916818,
+      "loss": 1.6076,
+      "step": 5572
+    },
+    {
+      "epoch": 0.2567374579628691,
+      "grad_norm": 4.795465721072323,
+      "learning_rate": 0.0001742328912171467,
+      "loss": 1.6269,
+      "step": 5573
+    },
+    {
+      "epoch": 0.2567835260515041,
+      "grad_norm": 1.329613384296055,
+      "learning_rate": 0.00017422289286480734,
+      "loss": 1.5242,
+      "step": 5574
+    },
+    {
+      "epoch": 0.2568295941401391,
+      "grad_norm": 1.536558183746432,
+      "learning_rate": 0.00017421289286002255,
+      "loss": 1.5009,
+      "step": 5575
+    },
+    {
+      "epoch": 0.2568756622287741,
+      "grad_norm": 2.5704978186884158,
+      "learning_rate": 0.00017420289120301496,
+      "loss": 1.4948,
+      "step": 5576
+    },
+    {
+      "epoch": 0.2569217303174091,
+      "grad_norm": 1.4265165723089004,
+      "learning_rate": 0.0001741928878940072,
+      "loss": 1.6642,
+      "step": 5577
+    },
+    {
+      "epoch": 0.2569677984060441,
+      "grad_norm": 2.979591952234197,
+      "learning_rate": 0.00017418288293322205,
+      "loss": 1.3307,
+      "step": 5578
+    },
+    {
+      "epoch": 0.2570138664946791,
+      "grad_norm": 1.2281122882334268,
+      "learning_rate": 0.00017417287632088218,
+      "loss": 1.5348,
+      "step": 5579
+    },
+    {
+      "epoch": 0.25705993458331416,
+      "grad_norm": 1.6853801930851715,
+      "learning_rate": 0.0001741628680572104,
+      "loss": 1.4923,
+      "step": 5580
+    },
+    {
+      "epoch": 0.25710600267194916,
+      "grad_norm": 1.2347486765496034,
+      "learning_rate": 0.00017415285814242952,
+      "loss": 1.4479,
+      "step": 5581
+    },
+    {
+      "epoch": 0.25715207076058416,
+      "grad_norm": 1.036519188323503,
+      "learning_rate": 0.0001741428465767624,
+      "loss": 1.5045,
+      "step": 5582
+    },
+    {
+      "epoch": 0.25719813884921916,
+      "grad_norm": 0.9934910858880949,
+      "learning_rate": 0.00017413283336043195,
+      "loss": 1.5862,
+      "step": 5583
+    },
+    {
+      "epoch": 0.25724420693785416,
+      "grad_norm": 2.0768953203869955,
+      "learning_rate": 0.00017412281849366106,
+      "loss": 1.519,
+      "step": 5584
+    },
+    {
+      "epoch": 0.25729027502648916,
+      "grad_norm": 1.253367363913667,
+      "learning_rate": 0.00017411280197667274,
+      "loss": 1.6289,
+      "step": 5585
+    },
+    {
+      "epoch": 0.25733634311512416,
+      "grad_norm": 1.2679654222688206,
+      "learning_rate": 0.0001741027838096899,
+      "loss": 1.3665,
+      "step": 5586
+    },
+    {
+      "epoch": 0.25738241120375915,
+      "grad_norm": 2.2454072685584947,
+      "learning_rate": 0.00017409276399293566,
+      "loss": 1.53,
+      "step": 5587
+    },
+    {
+      "epoch": 0.25742847929239415,
+      "grad_norm": 1.4190460270731944,
+      "learning_rate": 0.00017408274252663308,
+      "loss": 1.5386,
+      "step": 5588
+    },
+    {
+      "epoch": 0.25747454738102915,
+      "grad_norm": 2.8479275371085424,
+      "learning_rate": 0.00017407271941100526,
+      "loss": 1.3818,
+      "step": 5589
+    },
+    {
+      "epoch": 0.25752061546966415,
+      "grad_norm": 1.4711378392360797,
+      "learning_rate": 0.00017406269464627534,
+      "loss": 1.5313,
+      "step": 5590
+    },
+    {
+      "epoch": 0.25756668355829915,
+      "grad_norm": 2.4576474708789604,
+      "learning_rate": 0.00017405266823266655,
+      "loss": 1.3922,
+      "step": 5591
+    },
+    {
+      "epoch": 0.25761275164693415,
+      "grad_norm": 1.945551743540635,
+      "learning_rate": 0.00017404264017040202,
+      "loss": 1.451,
+      "step": 5592
+    },
+    {
+      "epoch": 0.25765881973556914,
+      "grad_norm": 4.868471783066965,
+      "learning_rate": 0.0001740326104597051,
+      "loss": 1.4414,
+      "step": 5593
+    },
+    {
+      "epoch": 0.2577048878242042,
+      "grad_norm": 1.2945751571223671,
+      "learning_rate": 0.00017402257910079905,
+      "loss": 1.7598,
+      "step": 5594
+    },
+    {
+      "epoch": 0.2577509559128392,
+      "grad_norm": 1.8383788724516164,
+      "learning_rate": 0.00017401254609390718,
+      "loss": 1.1482,
+      "step": 5595
+    },
+    {
+      "epoch": 0.2577970240014742,
+      "grad_norm": 2.376487202632731,
+      "learning_rate": 0.00017400251143925288,
+      "loss": 1.2983,
+      "step": 5596
+    },
+    {
+      "epoch": 0.2578430920901092,
+      "grad_norm": 2.026287618651878,
+      "learning_rate": 0.00017399247513705954,
+      "loss": 1.3386,
+      "step": 5597
+    },
+    {
+      "epoch": 0.2578891601787442,
+      "grad_norm": 1.643866628788842,
+      "learning_rate": 0.00017398243718755064,
+      "loss": 1.344,
+      "step": 5598
+    },
+    {
+      "epoch": 0.2579352282673792,
+      "grad_norm": 1.9636357532716084,
+      "learning_rate": 0.0001739723975909496,
+      "loss": 1.4247,
+      "step": 5599
+    },
+    {
+      "epoch": 0.2579812963560142,
+      "grad_norm": 1.4120278770997812,
+      "learning_rate": 0.00017396235634747998,
+      "loss": 1.4387,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2580273644446492,
+      "grad_norm": 1.193697086724597,
+      "learning_rate": 0.00017395231345736532,
+      "loss": 1.5199,
+      "step": 5601
+    },
+    {
+      "epoch": 0.2580734325332842,
+      "grad_norm": 2.1481975183768682,
+      "learning_rate": 0.00017394226892082916,
+      "loss": 1.4504,
+      "step": 5602
+    },
+    {
+      "epoch": 0.2581195006219192,
+      "grad_norm": 1.3723892984507757,
+      "learning_rate": 0.0001739322227380952,
+      "loss": 1.2166,
+      "step": 5603
+    },
+    {
+      "epoch": 0.2581655687105542,
+      "grad_norm": 1.6546765952673759,
+      "learning_rate": 0.00017392217490938707,
+      "loss": 1.5518,
+      "step": 5604
+    },
+    {
+      "epoch": 0.2582116367991892,
+      "grad_norm": 2.614566396991506,
+      "learning_rate": 0.00017391212543492847,
+      "loss": 1.4931,
+      "step": 5605
+    },
+    {
+      "epoch": 0.2582577048878242,
+      "grad_norm": 1.1273922103461471,
+      "learning_rate": 0.00017390207431494312,
+      "loss": 1.5497,
+      "step": 5606
+    },
+    {
+      "epoch": 0.25830377297645923,
+      "grad_norm": 2.2505373046558805,
+      "learning_rate": 0.0001738920215496548,
+      "loss": 1.5257,
+      "step": 5607
+    },
+    {
+      "epoch": 0.25834984106509423,
+      "grad_norm": 2.1959475416811736,
+      "learning_rate": 0.00017388196713928733,
+      "loss": 1.3512,
+      "step": 5608
+    },
+    {
+      "epoch": 0.25839590915372923,
+      "grad_norm": 1.967509122909494,
+      "learning_rate": 0.00017387191108406448,
+      "loss": 1.2782,
+      "step": 5609
+    },
+    {
+      "epoch": 0.25844197724236423,
+      "grad_norm": 3.148516139744944,
+      "learning_rate": 0.00017386185338421028,
+      "loss": 1.6475,
+      "step": 5610
+    },
+    {
+      "epoch": 0.2584880453309992,
+      "grad_norm": 1.598495352098683,
+      "learning_rate": 0.00017385179403994852,
+      "loss": 1.4308,
+      "step": 5611
+    },
+    {
+      "epoch": 0.2585341134196342,
+      "grad_norm": 1.7742843359587919,
+      "learning_rate": 0.0001738417330515032,
+      "loss": 1.2067,
+      "step": 5612
+    },
+    {
+      "epoch": 0.2585801815082692,
+      "grad_norm": 1.3632680392542917,
+      "learning_rate": 0.0001738316704190983,
+      "loss": 1.6191,
+      "step": 5613
+    },
+    {
+      "epoch": 0.2586262495969042,
+      "grad_norm": 3.806469753768241,
+      "learning_rate": 0.00017382160614295785,
+      "loss": 1.5637,
+      "step": 5614
+    },
+    {
+      "epoch": 0.2586723176855392,
+      "grad_norm": 1.6153680885536132,
+      "learning_rate": 0.0001738115402233059,
+      "loss": 1.4165,
+      "step": 5615
+    },
+    {
+      "epoch": 0.2587183857741742,
+      "grad_norm": 1.7063078785946157,
+      "learning_rate": 0.0001738014726603666,
+      "loss": 1.7032,
+      "step": 5616
+    },
+    {
+      "epoch": 0.2587644538628092,
+      "grad_norm": 1.6566874827064912,
+      "learning_rate": 0.00017379140345436404,
+      "loss": 1.1471,
+      "step": 5617
+    },
+    {
+      "epoch": 0.2588105219514442,
+      "grad_norm": 2.5340226073726133,
+      "learning_rate": 0.0001737813326055224,
+      "loss": 1.5657,
+      "step": 5618
+    },
+    {
+      "epoch": 0.2588565900400792,
+      "grad_norm": 1.8744663874229877,
+      "learning_rate": 0.0001737712601140659,
+      "loss": 1.3989,
+      "step": 5619
+    },
+    {
+      "epoch": 0.2589026581287142,
+      "grad_norm": 2.1674686565469323,
+      "learning_rate": 0.0001737611859802188,
+      "loss": 1.7568,
+      "step": 5620
+    },
+    {
+      "epoch": 0.25894872621734927,
+      "grad_norm": 1.8202329232587036,
+      "learning_rate": 0.00017375111020420536,
+      "loss": 1.0694,
+      "step": 5621
+    },
+    {
+      "epoch": 0.25899479430598427,
+      "grad_norm": 2.025611846937419,
+      "learning_rate": 0.0001737410327862499,
+      "loss": 1.3062,
+      "step": 5622
+    },
+    {
+      "epoch": 0.25904086239461926,
+      "grad_norm": 5.063579223343431,
+      "learning_rate": 0.00017373095372657678,
+      "loss": 1.9327,
+      "step": 5623
+    },
+    {
+      "epoch": 0.25908693048325426,
+      "grad_norm": 1.9466026144718749,
+      "learning_rate": 0.0001737208730254104,
+      "loss": 1.5416,
+      "step": 5624
+    },
+    {
+      "epoch": 0.25913299857188926,
+      "grad_norm": 2.753927815557042,
+      "learning_rate": 0.0001737107906829752,
+      "loss": 1.1143,
+      "step": 5625
+    },
+    {
+      "epoch": 0.25917906666052426,
+      "grad_norm": 1.7938834763589202,
+      "learning_rate": 0.0001737007066994956,
+      "loss": 1.4365,
+      "step": 5626
+    },
+    {
+      "epoch": 0.25922513474915926,
+      "grad_norm": 1.3709381519375161,
+      "learning_rate": 0.00017369062107519615,
+      "loss": 1.7386,
+      "step": 5627
+    },
+    {
+      "epoch": 0.25927120283779426,
+      "grad_norm": 2.390223136990073,
+      "learning_rate": 0.0001736805338103014,
+      "loss": 1.4207,
+      "step": 5628
+    },
+    {
+      "epoch": 0.25931727092642926,
+      "grad_norm": 2.5771906257905672,
+      "learning_rate": 0.0001736704449050359,
+      "loss": 1.5091,
+      "step": 5629
+    },
+    {
+      "epoch": 0.25936333901506425,
+      "grad_norm": 2.5896029676520977,
+      "learning_rate": 0.00017366035435962423,
+      "loss": 1.3917,
+      "step": 5630
+    },
+    {
+      "epoch": 0.25940940710369925,
+      "grad_norm": 3.7501831032529123,
+      "learning_rate": 0.00017365026217429109,
+      "loss": 1.4134,
+      "step": 5631
+    },
+    {
+      "epoch": 0.25945547519233425,
+      "grad_norm": 1.946714478411715,
+      "learning_rate": 0.00017364016834926112,
+      "loss": 1.4386,
+      "step": 5632
+    },
+    {
+      "epoch": 0.25950154328096925,
+      "grad_norm": 4.397369508491472,
+      "learning_rate": 0.0001736300728847591,
+      "loss": 1.5664,
+      "step": 5633
+    },
+    {
+      "epoch": 0.25954761136960425,
+      "grad_norm": 1.8370220141767444,
+      "learning_rate": 0.00017361997578100975,
+      "loss": 1.4778,
+      "step": 5634
+    },
+    {
+      "epoch": 0.2595936794582393,
+      "grad_norm": 2.5897422668294654,
+      "learning_rate": 0.00017360987703823785,
+      "loss": 1.1737,
+      "step": 5635
+    },
+    {
+      "epoch": 0.2596397475468743,
+      "grad_norm": 2.3034768000699697,
+      "learning_rate": 0.00017359977665666826,
+      "loss": 1.4402,
+      "step": 5636
+    },
+    {
+      "epoch": 0.2596858156355093,
+      "grad_norm": 3.5268441442705645,
+      "learning_rate": 0.00017358967463652585,
+      "loss": 1.1548,
+      "step": 5637
+    },
+    {
+      "epoch": 0.2597318837241443,
+      "grad_norm": 1.7857083421231148,
+      "learning_rate": 0.0001735795709780355,
+      "loss": 1.3444,
+      "step": 5638
+    },
+    {
+      "epoch": 0.2597779518127793,
+      "grad_norm": 3.5256583539176205,
+      "learning_rate": 0.00017356946568142218,
+      "loss": 1.386,
+      "step": 5639
+    },
+    {
+      "epoch": 0.2598240199014143,
+      "grad_norm": 1.684144647476405,
+      "learning_rate": 0.00017355935874691084,
+      "loss": 1.4546,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2598700879900493,
+      "grad_norm": 3.05826111010363,
+      "learning_rate": 0.00017354925017472653,
+      "loss": 1.2111,
+      "step": 5641
+    },
+    {
+      "epoch": 0.2599161560786843,
+      "grad_norm": 2.9426403944978383,
+      "learning_rate": 0.00017353913996509424,
+      "loss": 1.5981,
+      "step": 5642
+    },
+    {
+      "epoch": 0.2599622241673193,
+      "grad_norm": 1.403425270491217,
+      "learning_rate": 0.0001735290281182391,
+      "loss": 1.6521,
+      "step": 5643
+    },
+    {
+      "epoch": 0.2600082922559543,
+      "grad_norm": 3.5950137110108376,
+      "learning_rate": 0.00017351891463438624,
+      "loss": 1.3661,
+      "step": 5644
+    },
+    {
+      "epoch": 0.2600543603445893,
+      "grad_norm": 2.814063991453667,
+      "learning_rate": 0.00017350879951376079,
+      "loss": 1.615,
+      "step": 5645
+    },
+    {
+      "epoch": 0.2601004284332243,
+      "grad_norm": 1.4966140238801868,
+      "learning_rate": 0.00017349868275658796,
+      "loss": 1.4525,
+      "step": 5646
+    },
+    {
+      "epoch": 0.2601464965218593,
+      "grad_norm": 1.687797628233121,
+      "learning_rate": 0.00017348856436309298,
+      "loss": 1.7183,
+      "step": 5647
+    },
+    {
+      "epoch": 0.26019256461049434,
+      "grad_norm": 12.151530707919497,
+      "learning_rate": 0.00017347844433350114,
+      "loss": 1.5055,
+      "step": 5648
+    },
+    {
+      "epoch": 0.26023863269912934,
+      "grad_norm": 1.9684154871619297,
+      "learning_rate": 0.0001734683226680377,
+      "loss": 1.3294,
+      "step": 5649
+    },
+    {
+      "epoch": 0.26028470078776433,
+      "grad_norm": 1.487187010284459,
+      "learning_rate": 0.00017345819936692803,
+      "loss": 1.5447,
+      "step": 5650
+    },
+    {
+      "epoch": 0.26033076887639933,
+      "grad_norm": 2.0062749158938944,
+      "learning_rate": 0.0001734480744303975,
+      "loss": 1.4115,
+      "step": 5651
+    },
+    {
+      "epoch": 0.26037683696503433,
+      "grad_norm": 2.017117469839979,
+      "learning_rate": 0.00017343794785867154,
+      "loss": 1.6637,
+      "step": 5652
+    },
+    {
+      "epoch": 0.26042290505366933,
+      "grad_norm": 3.275421091634138,
+      "learning_rate": 0.00017342781965197558,
+      "loss": 1.4356,
+      "step": 5653
+    },
+    {
+      "epoch": 0.2604689731423043,
+      "grad_norm": 1.5552618648091467,
+      "learning_rate": 0.00017341768981053512,
+      "loss": 1.2148,
+      "step": 5654
+    },
+    {
+      "epoch": 0.2605150412309393,
+      "grad_norm": 1.4330499362072333,
+      "learning_rate": 0.0001734075583345757,
+      "loss": 1.1216,
+      "step": 5655
+    },
+    {
+      "epoch": 0.2605611093195743,
+      "grad_norm": 1.6004936192935812,
+      "learning_rate": 0.00017339742522432284,
+      "loss": 1.4466,
+      "step": 5656
+    },
+    {
+      "epoch": 0.2606071774082093,
+      "grad_norm": 3.1562304513391344,
+      "learning_rate": 0.00017338729048000217,
+      "loss": 1.7524,
+      "step": 5657
+    },
+    {
+      "epoch": 0.2606532454968443,
+      "grad_norm": 1.7238931599724274,
+      "learning_rate": 0.0001733771541018393,
+      "loss": 1.3036,
+      "step": 5658
+    },
+    {
+      "epoch": 0.2606993135854793,
+      "grad_norm": 1.7569446314209942,
+      "learning_rate": 0.00017336701609005993,
+      "loss": 1.4966,
+      "step": 5659
+    },
+    {
+      "epoch": 0.2607453816741143,
+      "grad_norm": 1.7679972780812505,
+      "learning_rate": 0.00017335687644488973,
+      "loss": 1.6214,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2607914497627493,
+      "grad_norm": 1.8846095471689093,
+      "learning_rate": 0.00017334673516655445,
+      "loss": 1.4612,
+      "step": 5661
+    },
+    {
+      "epoch": 0.26083751785138437,
+      "grad_norm": 4.3869891696863474,
+      "learning_rate": 0.00017333659225527988,
+      "loss": 1.4494,
+      "step": 5662
+    },
+    {
+      "epoch": 0.26088358594001937,
+      "grad_norm": 2.5744434527664817,
+      "learning_rate": 0.00017332644771129183,
+      "loss": 1.5156,
+      "step": 5663
+    },
+    {
+      "epoch": 0.26092965402865437,
+      "grad_norm": 7.248485302266954,
+      "learning_rate": 0.00017331630153481612,
+      "loss": 1.2657,
+      "step": 5664
+    },
+    {
+      "epoch": 0.26097572211728937,
+      "grad_norm": 1.7619339017516926,
+      "learning_rate": 0.0001733061537260787,
+      "loss": 1.3355,
+      "step": 5665
+    },
+    {
+      "epoch": 0.26102179020592436,
+      "grad_norm": 2.0035939962908027,
+      "learning_rate": 0.00017329600428530546,
+      "loss": 1.6979,
+      "step": 5666
+    },
+    {
+      "epoch": 0.26106785829455936,
+      "grad_norm": 1.3685695726751579,
+      "learning_rate": 0.00017328585321272236,
+      "loss": 1.5354,
+      "step": 5667
+    },
+    {
+      "epoch": 0.26111392638319436,
+      "grad_norm": 4.648380456934328,
+      "learning_rate": 0.00017327570050855537,
+      "loss": 1.5385,
+      "step": 5668
+    },
+    {
+      "epoch": 0.26115999447182936,
+      "grad_norm": 2.308952985466459,
+      "learning_rate": 0.00017326554617303058,
+      "loss": 1.5108,
+      "step": 5669
+    },
+    {
+      "epoch": 0.26120606256046436,
+      "grad_norm": 4.6560724193168515,
+      "learning_rate": 0.00017325539020637403,
+      "loss": 1.4931,
+      "step": 5670
+    },
+    {
+      "epoch": 0.26125213064909936,
+      "grad_norm": 3.9355689846001782,
+      "learning_rate": 0.0001732452326088118,
+      "loss": 1.3335,
+      "step": 5671
+    },
+    {
+      "epoch": 0.26129819873773436,
+      "grad_norm": 1.4766132231083855,
+      "learning_rate": 0.00017323507338057006,
+      "loss": 1.2724,
+      "step": 5672
+    },
+    {
+      "epoch": 0.26134426682636935,
+      "grad_norm": 1.1230343959488505,
+      "learning_rate": 0.000173224912521875,
+      "loss": 1.1242,
+      "step": 5673
+    },
+    {
+      "epoch": 0.26139033491500435,
+      "grad_norm": 1.9358758206029658,
+      "learning_rate": 0.0001732147500329528,
+      "loss": 1.7488,
+      "step": 5674
+    },
+    {
+      "epoch": 0.2614364030036394,
+      "grad_norm": 2.302866490713037,
+      "learning_rate": 0.00017320458591402971,
+      "loss": 1.4286,
+      "step": 5675
+    },
+    {
+      "epoch": 0.2614824710922744,
+      "grad_norm": 2.417331679227632,
+      "learning_rate": 0.00017319442016533207,
+      "loss": 1.6385,
+      "step": 5676
+    },
+    {
+      "epoch": 0.2615285391809094,
+      "grad_norm": 1.70432972054433,
+      "learning_rate": 0.00017318425278708613,
+      "loss": 1.6563,
+      "step": 5677
+    },
+    {
+      "epoch": 0.2615746072695444,
+      "grad_norm": 1.9407180348881747,
+      "learning_rate": 0.00017317408377951835,
+      "loss": 1.3544,
+      "step": 5678
+    },
+    {
+      "epoch": 0.2616206753581794,
+      "grad_norm": 5.827725430953405,
+      "learning_rate": 0.00017316391314285497,
+      "loss": 1.633,
+      "step": 5679
+    },
+    {
+      "epoch": 0.2616667434468144,
+      "grad_norm": 2.2303236325672846,
+      "learning_rate": 0.0001731537408773226,
+      "loss": 1.5496,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2617128115354494,
+      "grad_norm": 1.7314683521373613,
+      "learning_rate": 0.00017314356698314756,
+      "loss": 0.8931,
+      "step": 5681
+    },
+    {
+      "epoch": 0.2617588796240844,
+      "grad_norm": 2.075541971917574,
+      "learning_rate": 0.00017313339146055645,
+      "loss": 1.6358,
+      "step": 5682
+    },
+    {
+      "epoch": 0.2618049477127194,
+      "grad_norm": 1.37933291153702,
+      "learning_rate": 0.00017312321430977575,
+      "loss": 1.5815,
+      "step": 5683
+    },
+    {
+      "epoch": 0.2618510158013544,
+      "grad_norm": 2.0770799631680874,
+      "learning_rate": 0.00017311303553103208,
+      "loss": 1.2914,
+      "step": 5684
+    },
+    {
+      "epoch": 0.2618970838899894,
+      "grad_norm": 2.844404880108005,
+      "learning_rate": 0.00017310285512455202,
+      "loss": 1.3468,
+      "step": 5685
+    },
+    {
+      "epoch": 0.2619431519786244,
+      "grad_norm": 1.4960136016781118,
+      "learning_rate": 0.00017309267309056225,
+      "loss": 1.7525,
+      "step": 5686
+    },
+    {
+      "epoch": 0.2619892200672594,
+      "grad_norm": 3.2248629302654632,
+      "learning_rate": 0.00017308248942928946,
+      "loss": 1.5862,
+      "step": 5687
+    },
+    {
+      "epoch": 0.2620352881558944,
+      "grad_norm": 1.8025729068741232,
+      "learning_rate": 0.00017307230414096033,
+      "loss": 1.5568,
+      "step": 5688
+    },
+    {
+      "epoch": 0.26208135624452944,
+      "grad_norm": 2.1705924890117987,
+      "learning_rate": 0.00017306211722580163,
+      "loss": 1.4639,
+      "step": 5689
+    },
+    {
+      "epoch": 0.26212742433316444,
+      "grad_norm": 2.0569639497443335,
+      "learning_rate": 0.0001730519286840402,
+      "loss": 1.4727,
+      "step": 5690
+    },
+    {
+      "epoch": 0.26217349242179944,
+      "grad_norm": 6.672424035760028,
+      "learning_rate": 0.0001730417385159028,
+      "loss": 1.3081,
+      "step": 5691
+    },
+    {
+      "epoch": 0.26221956051043444,
+      "grad_norm": 2.5227000177327645,
+      "learning_rate": 0.00017303154672161633,
+      "loss": 1.2387,
+      "step": 5692
+    },
+    {
+      "epoch": 0.26226562859906943,
+      "grad_norm": 2.715068855124776,
+      "learning_rate": 0.0001730213533014077,
+      "loss": 1.7011,
+      "step": 5693
+    },
+    {
+      "epoch": 0.26231169668770443,
+      "grad_norm": 1.1402342189598789,
+      "learning_rate": 0.00017301115825550385,
+      "loss": 1.4784,
+      "step": 5694
+    },
+    {
+      "epoch": 0.26235776477633943,
+      "grad_norm": 2.5951415173840195,
+      "learning_rate": 0.00017300096158413174,
+      "loss": 1.7052,
+      "step": 5695
+    },
+    {
+      "epoch": 0.26240383286497443,
+      "grad_norm": 1.3943840893740813,
+      "learning_rate": 0.00017299076328751843,
+      "loss": 1.6857,
+      "step": 5696
+    },
+    {
+      "epoch": 0.2624499009536094,
+      "grad_norm": 1.5719459890090468,
+      "learning_rate": 0.00017298056336589087,
+      "loss": 1.2769,
+      "step": 5697
+    },
+    {
+      "epoch": 0.2624959690422444,
+      "grad_norm": 1.7841878976371954,
+      "learning_rate": 0.00017297036181947623,
+      "loss": 1.5773,
+      "step": 5698
+    },
+    {
+      "epoch": 0.2625420371308794,
+      "grad_norm": 1.2794210577855425,
+      "learning_rate": 0.0001729601586485016,
+      "loss": 1.5204,
+      "step": 5699
+    },
+    {
+      "epoch": 0.2625881052195144,
+      "grad_norm": 1.5934718256498788,
+      "learning_rate": 0.00017294995385319414,
+      "loss": 1.222,
+      "step": 5700
+    },
+    {
+      "epoch": 0.2626341733081494,
+      "grad_norm": 2.4010895298406165,
+      "learning_rate": 0.00017293974743378107,
+      "loss": 1.3398,
+      "step": 5701
+    },
+    {
+      "epoch": 0.2626802413967845,
+      "grad_norm": 2.5745926128562,
+      "learning_rate": 0.00017292953939048952,
+      "loss": 1.5462,
+      "step": 5702
+    },
+    {
+      "epoch": 0.2627263094854195,
+      "grad_norm": 2.031823255429497,
+      "learning_rate": 0.00017291932972354686,
+      "loss": 1.3334,
+      "step": 5703
+    },
+    {
+      "epoch": 0.26277237757405447,
+      "grad_norm": 1.7495442301097595,
+      "learning_rate": 0.00017290911843318037,
+      "loss": 1.1823,
+      "step": 5704
+    },
+    {
+      "epoch": 0.26281844566268947,
+      "grad_norm": 2.2975892965204507,
+      "learning_rate": 0.00017289890551961735,
+      "loss": 1.6205,
+      "step": 5705
+    },
+    {
+      "epoch": 0.26286451375132447,
+      "grad_norm": 3.620776634211489,
+      "learning_rate": 0.00017288869098308518,
+      "loss": 1.4654,
+      "step": 5706
+    },
+    {
+      "epoch": 0.26291058183995947,
+      "grad_norm": 1.673379576671442,
+      "learning_rate": 0.0001728784748238113,
+      "loss": 1.4299,
+      "step": 5707
+    },
+    {
+      "epoch": 0.26295664992859447,
+      "grad_norm": 1.7910671037444161,
+      "learning_rate": 0.00017286825704202312,
+      "loss": 1.3257,
+      "step": 5708
+    },
+    {
+      "epoch": 0.26300271801722946,
+      "grad_norm": 1.2888386905783475,
+      "learning_rate": 0.00017285803763794817,
+      "loss": 1.5004,
+      "step": 5709
+    },
+    {
+      "epoch": 0.26304878610586446,
+      "grad_norm": 2.5309301114244027,
+      "learning_rate": 0.00017284781661181393,
+      "loss": 1.3597,
+      "step": 5710
+    },
+    {
+      "epoch": 0.26309485419449946,
+      "grad_norm": 1.1541611582697207,
+      "learning_rate": 0.00017283759396384793,
+      "loss": 1.5281,
+      "step": 5711
+    },
+    {
+      "epoch": 0.26314092228313446,
+      "grad_norm": 1.8203451292556823,
+      "learning_rate": 0.00017282736969427784,
+      "loss": 1.5961,
+      "step": 5712
+    },
+    {
+      "epoch": 0.26318699037176946,
+      "grad_norm": 3.323162211605503,
+      "learning_rate": 0.00017281714380333122,
+      "loss": 1.5128,
+      "step": 5713
+    },
+    {
+      "epoch": 0.26323305846040446,
+      "grad_norm": 5.826397447205997,
+      "learning_rate": 0.00017280691629123572,
+      "loss": 1.4716,
+      "step": 5714
+    },
+    {
+      "epoch": 0.26327912654903945,
+      "grad_norm": 3.713130400766529,
+      "learning_rate": 0.0001727966871582191,
+      "loss": 1.6333,
+      "step": 5715
+    },
+    {
+      "epoch": 0.2633251946376745,
+      "grad_norm": 7.781719008220836,
+      "learning_rate": 0.00017278645640450902,
+      "loss": 1.4823,
+      "step": 5716
+    },
+    {
+      "epoch": 0.2633712627263095,
+      "grad_norm": 1.8117652561548643,
+      "learning_rate": 0.00017277622403033332,
+      "loss": 1.5775,
+      "step": 5717
+    },
+    {
+      "epoch": 0.2634173308149445,
+      "grad_norm": 1.5619938729842222,
+      "learning_rate": 0.00017276599003591974,
+      "loss": 1.6691,
+      "step": 5718
+    },
+    {
+      "epoch": 0.2634633989035795,
+      "grad_norm": 3.31932594866177,
+      "learning_rate": 0.00017275575442149622,
+      "loss": 1.1785,
+      "step": 5719
+    },
+    {
+      "epoch": 0.2635094669922145,
+      "grad_norm": 2.5544578850685027,
+      "learning_rate": 0.00017274551718729053,
+      "loss": 1.7303,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2635555350808495,
+      "grad_norm": 2.067665457683459,
+      "learning_rate": 0.00017273527833353065,
+      "loss": 1.5065,
+      "step": 5721
+    },
+    {
+      "epoch": 0.2636016031694845,
+      "grad_norm": 1.3833252793880368,
+      "learning_rate": 0.0001727250378604445,
+      "loss": 1.5046,
+      "step": 5722
+    },
+    {
+      "epoch": 0.2636476712581195,
+      "grad_norm": 1.8199732559579151,
+      "learning_rate": 0.0001727147957682601,
+      "loss": 1.6957,
+      "step": 5723
+    },
+    {
+      "epoch": 0.2636937393467545,
+      "grad_norm": 2.9890960029225524,
+      "learning_rate": 0.00017270455205720543,
+      "loss": 1.4578,
+      "step": 5724
+    },
+    {
+      "epoch": 0.2637398074353895,
+      "grad_norm": 1.3317739987570798,
+      "learning_rate": 0.00017269430672750858,
+      "loss": 1.3943,
+      "step": 5725
+    },
+    {
+      "epoch": 0.2637858755240245,
+      "grad_norm": 1.7250742744300631,
+      "learning_rate": 0.00017268405977939762,
+      "loss": 1.4778,
+      "step": 5726
+    },
+    {
+      "epoch": 0.2638319436126595,
+      "grad_norm": 2.0582544673038536,
+      "learning_rate": 0.00017267381121310074,
+      "loss": 1.4791,
+      "step": 5727
+    },
+    {
+      "epoch": 0.2638780117012945,
+      "grad_norm": 1.1013037447231233,
+      "learning_rate": 0.00017266356102884603,
+      "loss": 1.5864,
+      "step": 5728
+    },
+    {
+      "epoch": 0.2639240797899295,
+      "grad_norm": 1.5638841302128808,
+      "learning_rate": 0.00017265330922686172,
+      "loss": 1.7676,
+      "step": 5729
+    },
+    {
+      "epoch": 0.26397014787856454,
+      "grad_norm": 3.3061669781680907,
+      "learning_rate": 0.00017264305580737602,
+      "loss": 1.4157,
+      "step": 5730
+    },
+    {
+      "epoch": 0.26401621596719954,
+      "grad_norm": 1.7932635140261426,
+      "learning_rate": 0.0001726328007706173,
+      "loss": 1.4317,
+      "step": 5731
+    },
+    {
+      "epoch": 0.26406228405583454,
+      "grad_norm": 1.694576365298277,
+      "learning_rate": 0.0001726225441168138,
+      "loss": 1.2945,
+      "step": 5732
+    },
+    {
+      "epoch": 0.26410835214446954,
+      "grad_norm": 1.3855573534364485,
+      "learning_rate": 0.00017261228584619388,
+      "loss": 1.3322,
+      "step": 5733
+    },
+    {
+      "epoch": 0.26415442023310454,
+      "grad_norm": 1.4201685209483688,
+      "learning_rate": 0.00017260202595898588,
+      "loss": 1.3054,
+      "step": 5734
+    },
+    {
+      "epoch": 0.26420048832173954,
+      "grad_norm": 2.074760801190405,
+      "learning_rate": 0.00017259176445541828,
+      "loss": 1.604,
+      "step": 5735
+    },
+    {
+      "epoch": 0.26424655641037453,
+      "grad_norm": 1.5080853552255598,
+      "learning_rate": 0.0001725815013357195,
+      "loss": 1.3633,
+      "step": 5736
+    },
+    {
+      "epoch": 0.26429262449900953,
+      "grad_norm": 1.8757098358271997,
+      "learning_rate": 0.00017257123660011808,
+      "loss": 1.2096,
+      "step": 5737
+    },
+    {
+      "epoch": 0.26433869258764453,
+      "grad_norm": 2.670738411296393,
+      "learning_rate": 0.0001725609702488425,
+      "loss": 1.294,
+      "step": 5738
+    },
+    {
+      "epoch": 0.26438476067627953,
+      "grad_norm": 2.177082279686446,
+      "learning_rate": 0.00017255070228212128,
+      "loss": 1.7308,
+      "step": 5739
+    },
+    {
+      "epoch": 0.2644308287649145,
+      "grad_norm": 30.648889544296992,
+      "learning_rate": 0.0001725404327001831,
+      "loss": 1.121,
+      "step": 5740
+    },
+    {
+      "epoch": 0.2644768968535495,
+      "grad_norm": 1.6245213490983226,
+      "learning_rate": 0.0001725301615032566,
+      "loss": 1.3337,
+      "step": 5741
+    },
+    {
+      "epoch": 0.2645229649421845,
+      "grad_norm": 1.2828808404408547,
+      "learning_rate": 0.00017251988869157039,
+      "loss": 1.5687,
+      "step": 5742
+    },
+    {
+      "epoch": 0.2645690330308196,
+      "grad_norm": 1.4087352433048252,
+      "learning_rate": 0.0001725096142653532,
+      "loss": 1.5824,
+      "step": 5743
+    },
+    {
+      "epoch": 0.2646151011194546,
+      "grad_norm": 2.654734755777661,
+      "learning_rate": 0.0001724993382248338,
+      "loss": 1.5557,
+      "step": 5744
+    },
+    {
+      "epoch": 0.2646611692080896,
+      "grad_norm": 2.047177653105267,
+      "learning_rate": 0.00017248906057024092,
+      "loss": 1.2442,
+      "step": 5745
+    },
+    {
+      "epoch": 0.2647072372967246,
+      "grad_norm": 3.6503320753788056,
+      "learning_rate": 0.0001724787813018034,
+      "loss": 1.8192,
+      "step": 5746
+    },
+    {
+      "epoch": 0.26475330538535957,
+      "grad_norm": 2.592364683717822,
+      "learning_rate": 0.0001724685004197501,
+      "loss": 1.7152,
+      "step": 5747
+    },
+    {
+      "epoch": 0.26479937347399457,
+      "grad_norm": 2.5725878710433547,
+      "learning_rate": 0.0001724582179243099,
+      "loss": 1.6074,
+      "step": 5748
+    },
+    {
+      "epoch": 0.26484544156262957,
+      "grad_norm": 1.6789691231707689,
+      "learning_rate": 0.00017244793381571168,
+      "loss": 1.5982,
+      "step": 5749
+    },
+    {
+      "epoch": 0.26489150965126457,
+      "grad_norm": 1.3799253974612231,
+      "learning_rate": 0.00017243764809418447,
+      "loss": 1.3467,
+      "step": 5750
+    },
+    {
+      "epoch": 0.26493757773989957,
+      "grad_norm": 1.5781523315783361,
+      "learning_rate": 0.00017242736075995723,
+      "loss": 1.4098,
+      "step": 5751
+    },
+    {
+      "epoch": 0.26498364582853456,
+      "grad_norm": 1.542908452218978,
+      "learning_rate": 0.000172417071813259,
+      "loss": 1.7283,
+      "step": 5752
+    },
+    {
+      "epoch": 0.26502971391716956,
+      "grad_norm": 1.534201967529426,
+      "learning_rate": 0.00017240678125431878,
+      "loss": 1.5521,
+      "step": 5753
+    },
+    {
+      "epoch": 0.26507578200580456,
+      "grad_norm": 24.374409758632673,
+      "learning_rate": 0.00017239648908336577,
+      "loss": 1.6962,
+      "step": 5754
+    },
+    {
+      "epoch": 0.26512185009443956,
+      "grad_norm": 1.4992113645028833,
+      "learning_rate": 0.00017238619530062906,
+      "loss": 1.4149,
+      "step": 5755
+    },
+    {
+      "epoch": 0.26516791818307456,
+      "grad_norm": 1.1342094438717718,
+      "learning_rate": 0.00017237589990633782,
+      "loss": 1.5087,
+      "step": 5756
+    },
+    {
+      "epoch": 0.2652139862717096,
+      "grad_norm": 1.5275880566560758,
+      "learning_rate": 0.00017236560290072127,
+      "loss": 1.0546,
+      "step": 5757
+    },
+    {
+      "epoch": 0.2652600543603446,
+      "grad_norm": 2.4888423280029457,
+      "learning_rate": 0.00017235530428400865,
+      "loss": 1.8247,
+      "step": 5758
+    },
+    {
+      "epoch": 0.2653061224489796,
+      "grad_norm": 2.004746632112971,
+      "learning_rate": 0.00017234500405642923,
+      "loss": 1.5653,
+      "step": 5759
+    },
+    {
+      "epoch": 0.2653521905376146,
+      "grad_norm": 2.34700555353948,
+      "learning_rate": 0.00017233470221821238,
+      "loss": 1.7277,
+      "step": 5760
+    },
+    {
+      "epoch": 0.2653982586262496,
+      "grad_norm": 3.114738810992404,
+      "learning_rate": 0.00017232439876958738,
+      "loss": 1.7422,
+      "step": 5761
+    },
+    {
+      "epoch": 0.2654443267148846,
+      "grad_norm": 3.9578366146768422,
+      "learning_rate": 0.00017231409371078367,
+      "loss": 1.6197,
+      "step": 5762
+    },
+    {
+      "epoch": 0.2654903948035196,
+      "grad_norm": 2.3610709665096685,
+      "learning_rate": 0.00017230378704203062,
+      "loss": 1.5853,
+      "step": 5763
+    },
+    {
+      "epoch": 0.2655364628921546,
+      "grad_norm": 5.223157326381685,
+      "learning_rate": 0.00017229347876355773,
+      "loss": 1.4911,
+      "step": 5764
+    },
+    {
+      "epoch": 0.2655825309807896,
+      "grad_norm": 3.290637409711104,
+      "learning_rate": 0.00017228316887559452,
+      "loss": 1.3623,
+      "step": 5765
+    },
+    {
+      "epoch": 0.2656285990694246,
+      "grad_norm": 1.4537934457408925,
+      "learning_rate": 0.00017227285737837048,
+      "loss": 1.7279,
+      "step": 5766
+    },
+    {
+      "epoch": 0.2656746671580596,
+      "grad_norm": 1.5369532984187249,
+      "learning_rate": 0.0001722625442721152,
+      "loss": 1.6453,
+      "step": 5767
+    },
+    {
+      "epoch": 0.2657207352466946,
+      "grad_norm": 1.7345658045412697,
+      "learning_rate": 0.00017225222955705828,
+      "loss": 1.6939,
+      "step": 5768
+    },
+    {
+      "epoch": 0.2657668033353296,
+      "grad_norm": 1.7289990492189964,
+      "learning_rate": 0.00017224191323342934,
+      "loss": 1.6444,
+      "step": 5769
+    },
+    {
+      "epoch": 0.26581287142396465,
+      "grad_norm": 1.567063332637187,
+      "learning_rate": 0.00017223159530145807,
+      "loss": 1.5666,
+      "step": 5770
+    },
+    {
+      "epoch": 0.26585893951259965,
+      "grad_norm": 2.7550693633432464,
+      "learning_rate": 0.0001722212757613742,
+      "loss": 1.7365,
+      "step": 5771
+    },
+    {
+      "epoch": 0.26590500760123464,
+      "grad_norm": 2.4117742668202067,
+      "learning_rate": 0.00017221095461340743,
+      "loss": 1.8538,
+      "step": 5772
+    },
+    {
+      "epoch": 0.26595107568986964,
+      "grad_norm": 2.8037589772200544,
+      "learning_rate": 0.00017220063185778762,
+      "loss": 1.4237,
+      "step": 5773
+    },
+    {
+      "epoch": 0.26599714377850464,
+      "grad_norm": 1.9470290626743878,
+      "learning_rate": 0.0001721903074947445,
+      "loss": 1.3591,
+      "step": 5774
+    },
+    {
+      "epoch": 0.26604321186713964,
+      "grad_norm": 1.5499836608546549,
+      "learning_rate": 0.00017217998152450794,
+      "loss": 1.535,
+      "step": 5775
+    },
+    {
+      "epoch": 0.26608927995577464,
+      "grad_norm": 1.2489708466688403,
+      "learning_rate": 0.00017216965394730788,
+      "loss": 1.21,
+      "step": 5776
+    },
+    {
+      "epoch": 0.26613534804440964,
+      "grad_norm": 1.2605812255556883,
+      "learning_rate": 0.00017215932476337425,
+      "loss": 1.6446,
+      "step": 5777
+    },
+    {
+      "epoch": 0.26618141613304463,
+      "grad_norm": 2.3934793016399336,
+      "learning_rate": 0.00017214899397293692,
+      "loss": 1.4818,
+      "step": 5778
+    },
+    {
+      "epoch": 0.26622748422167963,
+      "grad_norm": 1.6138324864993725,
+      "learning_rate": 0.000172138661576226,
+      "loss": 1.623,
+      "step": 5779
+    },
+    {
+      "epoch": 0.26627355231031463,
+      "grad_norm": 2.7638785848639493,
+      "learning_rate": 0.00017212832757347146,
+      "loss": 1.5337,
+      "step": 5780
+    },
+    {
+      "epoch": 0.26631962039894963,
+      "grad_norm": 1.4813417570148477,
+      "learning_rate": 0.00017211799196490337,
+      "loss": 1.5677,
+      "step": 5781
+    },
+    {
+      "epoch": 0.26636568848758463,
+      "grad_norm": 2.8792316274631364,
+      "learning_rate": 0.00017210765475075186,
+      "loss": 1.5499,
+      "step": 5782
+    },
+    {
+      "epoch": 0.2664117565762196,
+      "grad_norm": 1.8092406628762614,
+      "learning_rate": 0.00017209731593124703,
+      "loss": 1.3827,
+      "step": 5783
+    },
+    {
+      "epoch": 0.2664578246648547,
+      "grad_norm": 4.150320942151844,
+      "learning_rate": 0.0001720869755066191,
+      "loss": 1.2447,
+      "step": 5784
+    },
+    {
+      "epoch": 0.2665038927534897,
+      "grad_norm": 2.043881791734468,
+      "learning_rate": 0.00017207663347709823,
+      "loss": 1.3549,
+      "step": 5785
+    },
+    {
+      "epoch": 0.2665499608421247,
+      "grad_norm": 1.4165571579881964,
+      "learning_rate": 0.00017206628984291474,
+      "loss": 1.3537,
+      "step": 5786
+    },
+    {
+      "epoch": 0.2665960289307597,
+      "grad_norm": 2.17154961021378,
+      "learning_rate": 0.00017205594460429887,
+      "loss": 1.3998,
+      "step": 5787
+    },
+    {
+      "epoch": 0.2666420970193947,
+      "grad_norm": 3.6256774894197785,
+      "learning_rate": 0.00017204559776148092,
+      "loss": 1.4709,
+      "step": 5788
+    },
+    {
+      "epoch": 0.2666881651080297,
+      "grad_norm": 1.2929568426659108,
+      "learning_rate": 0.00017203524931469128,
+      "loss": 1.4212,
+      "step": 5789
+    },
+    {
+      "epoch": 0.26673423319666467,
+      "grad_norm": 2.639253172268066,
+      "learning_rate": 0.00017202489926416035,
+      "loss": 1.604,
+      "step": 5790
+    },
+    {
+      "epoch": 0.26678030128529967,
+      "grad_norm": 1.2031296860254612,
+      "learning_rate": 0.0001720145476101185,
+      "loss": 1.5304,
+      "step": 5791
+    },
+    {
+      "epoch": 0.26682636937393467,
+      "grad_norm": 1.1693622232370284,
+      "learning_rate": 0.00017200419435279624,
+      "loss": 1.5996,
+      "step": 5792
+    },
+    {
+      "epoch": 0.26687243746256967,
+      "grad_norm": 2.4774554364017125,
+      "learning_rate": 0.00017199383949242404,
+      "loss": 1.5254,
+      "step": 5793
+    },
+    {
+      "epoch": 0.26691850555120467,
+      "grad_norm": 1.129011616217464,
+      "learning_rate": 0.00017198348302923248,
+      "loss": 1.7117,
+      "step": 5794
+    },
+    {
+      "epoch": 0.26696457363983966,
+      "grad_norm": 1.2838112047777726,
+      "learning_rate": 0.00017197312496345206,
+      "loss": 1.4474,
+      "step": 5795
+    },
+    {
+      "epoch": 0.26701064172847466,
+      "grad_norm": 1.3562479367918896,
+      "learning_rate": 0.00017196276529531343,
+      "loss": 1.4919,
+      "step": 5796
+    },
+    {
+      "epoch": 0.26705670981710966,
+      "grad_norm": 1.1865306038518122,
+      "learning_rate": 0.00017195240402504723,
+      "loss": 1.6465,
+      "step": 5797
+    },
+    {
+      "epoch": 0.2671027779057447,
+      "grad_norm": 1.0317308090417165,
+      "learning_rate": 0.00017194204115288412,
+      "loss": 1.733,
+      "step": 5798
+    },
+    {
+      "epoch": 0.2671488459943797,
+      "grad_norm": 2.1594440641611756,
+      "learning_rate": 0.00017193167667905484,
+      "loss": 1.6439,
+      "step": 5799
+    },
+    {
+      "epoch": 0.2671949140830147,
+      "grad_norm": 2.269653182775118,
+      "learning_rate": 0.0001719213106037901,
+      "loss": 1.6666,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2672409821716497,
+      "grad_norm": 1.8269534553917326,
+      "learning_rate": 0.0001719109429273207,
+      "loss": 1.9677,
+      "step": 5801
+    },
+    {
+      "epoch": 0.2672870502602847,
+      "grad_norm": 1.6751021703413451,
+      "learning_rate": 0.00017190057364987741,
+      "loss": 1.4397,
+      "step": 5802
+    },
+    {
+      "epoch": 0.2673331183489197,
+      "grad_norm": 2.164506164279533,
+      "learning_rate": 0.00017189020277169118,
+      "loss": 1.5273,
+      "step": 5803
+    },
+    {
+      "epoch": 0.2673791864375547,
+      "grad_norm": 1.8790682855734504,
+      "learning_rate": 0.00017187983029299283,
+      "loss": 1.541,
+      "step": 5804
+    },
+    {
+      "epoch": 0.2674252545261897,
+      "grad_norm": 1.834779130619243,
+      "learning_rate": 0.00017186945621401335,
+      "loss": 1.2401,
+      "step": 5805
+    },
+    {
+      "epoch": 0.2674713226148247,
+      "grad_norm": 1.9834896934894186,
+      "learning_rate": 0.0001718590805349836,
+      "loss": 1.2869,
+      "step": 5806
+    },
+    {
+      "epoch": 0.2675173907034597,
+      "grad_norm": 1.3787600011492012,
+      "learning_rate": 0.00017184870325613467,
+      "loss": 1.3905,
+      "step": 5807
+    },
+    {
+      "epoch": 0.2675634587920947,
+      "grad_norm": 1.3866794364857464,
+      "learning_rate": 0.00017183832437769753,
+      "loss": 1.4813,
+      "step": 5808
+    },
+    {
+      "epoch": 0.2676095268807297,
+      "grad_norm": 3.298945791674417,
+      "learning_rate": 0.0001718279438999033,
+      "loss": 1.215,
+      "step": 5809
+    },
+    {
+      "epoch": 0.2676555949693647,
+      "grad_norm": 1.4469509606496573,
+      "learning_rate": 0.00017181756182298305,
+      "loss": 1.3596,
+      "step": 5810
+    },
+    {
+      "epoch": 0.26770166305799975,
+      "grad_norm": 1.1535119002668877,
+      "learning_rate": 0.00017180717814716792,
+      "loss": 1.4423,
+      "step": 5811
+    },
+    {
+      "epoch": 0.26774773114663475,
+      "grad_norm": 2.700098832733543,
+      "learning_rate": 0.00017179679287268907,
+      "loss": 1.4812,
+      "step": 5812
+    },
+    {
+      "epoch": 0.26779379923526975,
+      "grad_norm": 0.9718283143817531,
+      "learning_rate": 0.00017178640599977777,
+      "loss": 1.5179,
+      "step": 5813
+    },
+    {
+      "epoch": 0.26783986732390475,
+      "grad_norm": 1.956060780793443,
+      "learning_rate": 0.0001717760175286652,
+      "loss": 1.2607,
+      "step": 5814
+    },
+    {
+      "epoch": 0.26788593541253974,
+      "grad_norm": 7.568255212007651,
+      "learning_rate": 0.00017176562745958268,
+      "loss": 1.3484,
+      "step": 5815
+    },
+    {
+      "epoch": 0.26793200350117474,
+      "grad_norm": 1.4913929094070357,
+      "learning_rate": 0.00017175523579276156,
+      "loss": 1.5251,
+      "step": 5816
+    },
+    {
+      "epoch": 0.26797807158980974,
+      "grad_norm": 2.231518763912805,
+      "learning_rate": 0.0001717448425284331,
+      "loss": 1.2796,
+      "step": 5817
+    },
+    {
+      "epoch": 0.26802413967844474,
+      "grad_norm": 1.440238546114865,
+      "learning_rate": 0.00017173444766682873,
+      "loss": 1.2184,
+      "step": 5818
+    },
+    {
+      "epoch": 0.26807020776707974,
+      "grad_norm": 1.5308026434659663,
+      "learning_rate": 0.00017172405120817992,
+      "loss": 1.4856,
+      "step": 5819
+    },
+    {
+      "epoch": 0.26811627585571474,
+      "grad_norm": 2.1125307996771228,
+      "learning_rate": 0.00017171365315271807,
+      "loss": 1.4369,
+      "step": 5820
+    },
+    {
+      "epoch": 0.26816234394434973,
+      "grad_norm": 1.3975614144415542,
+      "learning_rate": 0.00017170325350067465,
+      "loss": 1.5504,
+      "step": 5821
+    },
+    {
+      "epoch": 0.26820841203298473,
+      "grad_norm": 1.7742501899096406,
+      "learning_rate": 0.0001716928522522813,
+      "loss": 1.6388,
+      "step": 5822
+    },
+    {
+      "epoch": 0.26825448012161973,
+      "grad_norm": 2.808364844369087,
+      "learning_rate": 0.00017168244940776948,
+      "loss": 1.293,
+      "step": 5823
+    },
+    {
+      "epoch": 0.26830054821025473,
+      "grad_norm": 1.6396377326397855,
+      "learning_rate": 0.00017167204496737086,
+      "loss": 1.4246,
+      "step": 5824
+    },
+    {
+      "epoch": 0.2683466162988898,
+      "grad_norm": 1.9564801170425625,
+      "learning_rate": 0.00017166163893131704,
+      "loss": 1.4135,
+      "step": 5825
+    },
+    {
+      "epoch": 0.2683926843875248,
+      "grad_norm": 5.757184325590018,
+      "learning_rate": 0.0001716512312998397,
+      "loss": 1.6604,
+      "step": 5826
+    },
+    {
+      "epoch": 0.2684387524761598,
+      "grad_norm": 3.110904665412591,
+      "learning_rate": 0.00017164082207317055,
+      "loss": 1.4509,
+      "step": 5827
+    },
+    {
+      "epoch": 0.2684848205647948,
+      "grad_norm": 1.5980601341684273,
+      "learning_rate": 0.00017163041125154135,
+      "loss": 1.8159,
+      "step": 5828
+    },
+    {
+      "epoch": 0.2685308886534298,
+      "grad_norm": 1.4880542068862026,
+      "learning_rate": 0.00017161999883518386,
+      "loss": 1.7192,
+      "step": 5829
+    },
+    {
+      "epoch": 0.2685769567420648,
+      "grad_norm": 4.372129301404704,
+      "learning_rate": 0.00017160958482432986,
+      "loss": 1.2227,
+      "step": 5830
+    },
+    {
+      "epoch": 0.2686230248306998,
+      "grad_norm": 13.23768894601476,
+      "learning_rate": 0.0001715991692192113,
+      "loss": 1.6576,
+      "step": 5831
+    },
+    {
+      "epoch": 0.2686690929193348,
+      "grad_norm": 6.665577533243173,
+      "learning_rate": 0.00017158875202005996,
+      "loss": 1.488,
+      "step": 5832
+    },
+    {
+      "epoch": 0.26871516100796977,
+      "grad_norm": 2.0762466237425024,
+      "learning_rate": 0.0001715783332271078,
+      "loss": 1.434,
+      "step": 5833
+    },
+    {
+      "epoch": 0.26876122909660477,
+      "grad_norm": 2.7960847114789726,
+      "learning_rate": 0.0001715679128405868,
+      "loss": 1.1416,
+      "step": 5834
+    },
+    {
+      "epoch": 0.26880729718523977,
+      "grad_norm": 2.029824619695915,
+      "learning_rate": 0.00017155749086072893,
+      "loss": 1.6101,
+      "step": 5835
+    },
+    {
+      "epoch": 0.26885336527387477,
+      "grad_norm": 1.917318127860647,
+      "learning_rate": 0.00017154706728776624,
+      "loss": 1.7214,
+      "step": 5836
+    },
+    {
+      "epoch": 0.26889943336250977,
+      "grad_norm": 1.566782079091076,
+      "learning_rate": 0.00017153664212193073,
+      "loss": 1.6374,
+      "step": 5837
+    },
+    {
+      "epoch": 0.2689455014511448,
+      "grad_norm": 2.5063804726095307,
+      "learning_rate": 0.00017152621536345458,
+      "loss": 1.1762,
+      "step": 5838
+    },
+    {
+      "epoch": 0.2689915695397798,
+      "grad_norm": 2.3866478351761704,
+      "learning_rate": 0.00017151578701256987,
+      "loss": 1.6295,
+      "step": 5839
+    },
+    {
+      "epoch": 0.2690376376284148,
+      "grad_norm": 1.5282039587338931,
+      "learning_rate": 0.0001715053570695088,
+      "loss": 1.5981,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2690837057170498,
+      "grad_norm": 1.5192705728468965,
+      "learning_rate": 0.00017149492553450354,
+      "loss": 1.5857,
+      "step": 5841
+    },
+    {
+      "epoch": 0.2691297738056848,
+      "grad_norm": 1.7507992889011492,
+      "learning_rate": 0.00017148449240778635,
+      "loss": 1.5352,
+      "step": 5842
+    },
+    {
+      "epoch": 0.2691758418943198,
+      "grad_norm": 4.543158293648485,
+      "learning_rate": 0.00017147405768958953,
+      "loss": 1.3425,
+      "step": 5843
+    },
+    {
+      "epoch": 0.2692219099829548,
+      "grad_norm": 2.9108149757878534,
+      "learning_rate": 0.00017146362138014534,
+      "loss": 1.4574,
+      "step": 5844
+    },
+    {
+      "epoch": 0.2692679780715898,
+      "grad_norm": 1.703001375612616,
+      "learning_rate": 0.00017145318347968618,
+      "loss": 1.5123,
+      "step": 5845
+    },
+    {
+      "epoch": 0.2693140461602248,
+      "grad_norm": 1.7806983885066163,
+      "learning_rate": 0.00017144274398844437,
+      "loss": 1.573,
+      "step": 5846
+    },
+    {
+      "epoch": 0.2693601142488598,
+      "grad_norm": 1.575883610240662,
+      "learning_rate": 0.0001714323029066524,
+      "loss": 1.6298,
+      "step": 5847
+    },
+    {
+      "epoch": 0.2694061823374948,
+      "grad_norm": 1.2395182565536997,
+      "learning_rate": 0.00017142186023454267,
+      "loss": 1.713,
+      "step": 5848
+    },
+    {
+      "epoch": 0.2694522504261298,
+      "grad_norm": 2.4679301136288996,
+      "learning_rate": 0.0001714114159723477,
+      "loss": 1.3206,
+      "step": 5849
+    },
+    {
+      "epoch": 0.2694983185147648,
+      "grad_norm": 1.4166967270989248,
+      "learning_rate": 0.00017140097012029998,
+      "loss": 1.5307,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2695443866033998,
+      "grad_norm": 1.6611619404807398,
+      "learning_rate": 0.0001713905226786321,
+      "loss": 1.2658,
+      "step": 5851
+    },
+    {
+      "epoch": 0.26959045469203485,
+      "grad_norm": 2.732196176854316,
+      "learning_rate": 0.00017138007364757662,
+      "loss": 1.2919,
+      "step": 5852
+    },
+    {
+      "epoch": 0.26963652278066985,
+      "grad_norm": 1.9033945138857862,
+      "learning_rate": 0.00017136962302736618,
+      "loss": 1.1562,
+      "step": 5853
+    },
+    {
+      "epoch": 0.26968259086930485,
+      "grad_norm": 3.634154695726774,
+      "learning_rate": 0.00017135917081823347,
+      "loss": 1.6766,
+      "step": 5854
+    },
+    {
+      "epoch": 0.26972865895793985,
+      "grad_norm": 1.2133197667655704,
+      "learning_rate": 0.00017134871702041118,
+      "loss": 1.3094,
+      "step": 5855
+    },
+    {
+      "epoch": 0.26977472704657485,
+      "grad_norm": 1.231865812672756,
+      "learning_rate": 0.00017133826163413208,
+      "loss": 1.4901,
+      "step": 5856
+    },
+    {
+      "epoch": 0.26982079513520985,
+      "grad_norm": 5.337025339266096,
+      "learning_rate": 0.00017132780465962887,
+      "loss": 1.4563,
+      "step": 5857
+    },
+    {
+      "epoch": 0.26986686322384484,
+      "grad_norm": 1.1382237682616607,
+      "learning_rate": 0.00017131734609713436,
+      "loss": 1.4885,
+      "step": 5858
+    },
+    {
+      "epoch": 0.26991293131247984,
+      "grad_norm": 5.573157469025341,
+      "learning_rate": 0.00017130688594688148,
+      "loss": 1.3616,
+      "step": 5859
+    },
+    {
+      "epoch": 0.26995899940111484,
+      "grad_norm": 1.2966143596084572,
+      "learning_rate": 0.00017129642420910302,
+      "loss": 1.5578,
+      "step": 5860
+    },
+    {
+      "epoch": 0.27000506748974984,
+      "grad_norm": 1.3298334101999163,
+      "learning_rate": 0.00017128596088403193,
+      "loss": 1.4912,
+      "step": 5861
+    },
+    {
+      "epoch": 0.27005113557838484,
+      "grad_norm": 2.221148482270018,
+      "learning_rate": 0.00017127549597190114,
+      "loss": 1.4268,
+      "step": 5862
+    },
+    {
+      "epoch": 0.27009720366701984,
+      "grad_norm": 1.5992389167945062,
+      "learning_rate": 0.00017126502947294365,
+      "loss": 1.6409,
+      "step": 5863
+    },
+    {
+      "epoch": 0.27014327175565483,
+      "grad_norm": 1.5462187442903326,
+      "learning_rate": 0.00017125456138739246,
+      "loss": 1.6146,
+      "step": 5864
+    },
+    {
+      "epoch": 0.27018933984428983,
+      "grad_norm": 4.953096229493336,
+      "learning_rate": 0.00017124409171548064,
+      "loss": 1.7073,
+      "step": 5865
+    },
+    {
+      "epoch": 0.2702354079329249,
+      "grad_norm": 1.5501629762916884,
+      "learning_rate": 0.0001712336204574413,
+      "loss": 1.585,
+      "step": 5866
+    },
+    {
+      "epoch": 0.2702814760215599,
+      "grad_norm": 1.496245507241256,
+      "learning_rate": 0.00017122314761350751,
+      "loss": 1.4594,
+      "step": 5867
+    },
+    {
+      "epoch": 0.2703275441101949,
+      "grad_norm": 1.1872849800354592,
+      "learning_rate": 0.00017121267318391246,
+      "loss": 1.6187,
+      "step": 5868
+    },
+    {
+      "epoch": 0.2703736121988299,
+      "grad_norm": 2.438328427650502,
+      "learning_rate": 0.00017120219716888936,
+      "loss": 1.3146,
+      "step": 5869
+    },
+    {
+      "epoch": 0.2704196802874649,
+      "grad_norm": 3.0160682063991,
+      "learning_rate": 0.00017119171956867142,
+      "loss": 1.4837,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2704657483760999,
+      "grad_norm": 3.197479286804487,
+      "learning_rate": 0.00017118124038349193,
+      "loss": 0.954,
+      "step": 5871
+    },
+    {
+      "epoch": 0.2705118164647349,
+      "grad_norm": 2.3628339557641063,
+      "learning_rate": 0.00017117075961358414,
+      "loss": 1.5451,
+      "step": 5872
+    },
+    {
+      "epoch": 0.2705578845533699,
+      "grad_norm": 2.688636722904615,
+      "learning_rate": 0.00017116027725918144,
+      "loss": 1.0692,
+      "step": 5873
+    },
+    {
+      "epoch": 0.2706039526420049,
+      "grad_norm": 1.8556260375716722,
+      "learning_rate": 0.00017114979332051719,
+      "loss": 1.6691,
+      "step": 5874
+    },
+    {
+      "epoch": 0.2706500207306399,
+      "grad_norm": 2.378336097856706,
+      "learning_rate": 0.00017113930779782477,
+      "loss": 1.7297,
+      "step": 5875
+    },
+    {
+      "epoch": 0.27069608881927487,
+      "grad_norm": 1.3637974726241935,
+      "learning_rate": 0.00017112882069133763,
+      "loss": 1.201,
+      "step": 5876
+    },
+    {
+      "epoch": 0.27074215690790987,
+      "grad_norm": 2.0373899989066877,
+      "learning_rate": 0.00017111833200128928,
+      "loss": 1.3333,
+      "step": 5877
+    },
+    {
+      "epoch": 0.27078822499654487,
+      "grad_norm": 2.9333780973108867,
+      "learning_rate": 0.0001711078417279132,
+      "loss": 1.0365,
+      "step": 5878
+    },
+    {
+      "epoch": 0.2708342930851799,
+      "grad_norm": 3.8850995047353627,
+      "learning_rate": 0.00017109734987144295,
+      "loss": 1.5249,
+      "step": 5879
+    },
+    {
+      "epoch": 0.2708803611738149,
+      "grad_norm": 1.7546414870739595,
+      "learning_rate": 0.0001710868564321121,
+      "loss": 1.2492,
+      "step": 5880
+    },
+    {
+      "epoch": 0.2709264292624499,
+      "grad_norm": 1.800710078707359,
+      "learning_rate": 0.00017107636141015429,
+      "loss": 1.3009,
+      "step": 5881
+    },
+    {
+      "epoch": 0.2709724973510849,
+      "grad_norm": 1.6652764995421692,
+      "learning_rate": 0.00017106586480580314,
+      "loss": 1.8114,
+      "step": 5882
+    },
+    {
+      "epoch": 0.2710185654397199,
+      "grad_norm": 2.0267013535615073,
+      "learning_rate": 0.00017105536661929241,
+      "loss": 1.2405,
+      "step": 5883
+    },
+    {
+      "epoch": 0.2710646335283549,
+      "grad_norm": 1.7157709728613428,
+      "learning_rate": 0.00017104486685085574,
+      "loss": 1.3888,
+      "step": 5884
+    },
+    {
+      "epoch": 0.2711107016169899,
+      "grad_norm": 4.754263948364722,
+      "learning_rate": 0.00017103436550072694,
+      "loss": 1.3106,
+      "step": 5885
+    },
+    {
+      "epoch": 0.2711567697056249,
+      "grad_norm": 4.875943482224894,
+      "learning_rate": 0.00017102386256913978,
+      "loss": 1.7227,
+      "step": 5886
+    },
+    {
+      "epoch": 0.2712028377942599,
+      "grad_norm": 3.6223684364131943,
+      "learning_rate": 0.0001710133580563281,
+      "loss": 1.5542,
+      "step": 5887
+    },
+    {
+      "epoch": 0.2712489058828949,
+      "grad_norm": 4.936332224098667,
+      "learning_rate": 0.00017100285196252574,
+      "loss": 1.3952,
+      "step": 5888
+    },
+    {
+      "epoch": 0.2712949739715299,
+      "grad_norm": 6.546002189163992,
+      "learning_rate": 0.00017099234428796667,
+      "loss": 1.3202,
+      "step": 5889
+    },
+    {
+      "epoch": 0.2713410420601649,
+      "grad_norm": 4.326356384212586,
+      "learning_rate": 0.00017098183503288475,
+      "loss": 1.5362,
+      "step": 5890
+    },
+    {
+      "epoch": 0.2713871101487999,
+      "grad_norm": 17.005876366551707,
+      "learning_rate": 0.000170971324197514,
+      "loss": 1.6023,
+      "step": 5891
+    },
+    {
+      "epoch": 0.2714331782374349,
+      "grad_norm": 2.02704835992909,
+      "learning_rate": 0.0001709608117820884,
+      "loss": 1.5091,
+      "step": 5892
+    },
+    {
+      "epoch": 0.27147924632606996,
+      "grad_norm": 1.9028801898480618,
+      "learning_rate": 0.00017095029778684201,
+      "loss": 1.6755,
+      "step": 5893
+    },
+    {
+      "epoch": 0.27152531441470495,
+      "grad_norm": 3.217697615836756,
+      "learning_rate": 0.0001709397822120089,
+      "loss": 1.6943,
+      "step": 5894
+    },
+    {
+      "epoch": 0.27157138250333995,
+      "grad_norm": 6.287442785746079,
+      "learning_rate": 0.00017092926505782313,
+      "loss": 1.1719,
+      "step": 5895
+    },
+    {
+      "epoch": 0.27161745059197495,
+      "grad_norm": 4.420024769822394,
+      "learning_rate": 0.00017091874632451891,
+      "loss": 1.8236,
+      "step": 5896
+    },
+    {
+      "epoch": 0.27166351868060995,
+      "grad_norm": 3.479084143898313,
+      "learning_rate": 0.00017090822601233043,
+      "loss": 1.252,
+      "step": 5897
+    },
+    {
+      "epoch": 0.27170958676924495,
+      "grad_norm": 1.6830064013277932,
+      "learning_rate": 0.00017089770412149184,
+      "loss": 1.5228,
+      "step": 5898
+    },
+    {
+      "epoch": 0.27175565485787995,
+      "grad_norm": 8.500123864230778,
+      "learning_rate": 0.00017088718065223746,
+      "loss": 1.3668,
+      "step": 5899
+    },
+    {
+      "epoch": 0.27180172294651495,
+      "grad_norm": 2.2193555902592235,
+      "learning_rate": 0.00017087665560480152,
+      "loss": 1.2582,
+      "step": 5900
+    },
+    {
+      "epoch": 0.27184779103514994,
+      "grad_norm": 2.5884014550448775,
+      "learning_rate": 0.00017086612897941841,
+      "loss": 1.2526,
+      "step": 5901
+    },
+    {
+      "epoch": 0.27189385912378494,
+      "grad_norm": 2.044045929100164,
+      "learning_rate": 0.0001708556007763224,
+      "loss": 1.3615,
+      "step": 5902
+    },
+    {
+      "epoch": 0.27193992721241994,
+      "grad_norm": 1.8703616442349662,
+      "learning_rate": 0.00017084507099574797,
+      "loss": 1.538,
+      "step": 5903
+    },
+    {
+      "epoch": 0.27198599530105494,
+      "grad_norm": 12.04919587025062,
+      "learning_rate": 0.0001708345396379295,
+      "loss": 1.2483,
+      "step": 5904
+    },
+    {
+      "epoch": 0.27203206338968994,
+      "grad_norm": 1.838079283179381,
+      "learning_rate": 0.00017082400670310146,
+      "loss": 1.7363,
+      "step": 5905
+    },
+    {
+      "epoch": 0.272078131478325,
+      "grad_norm": 1.6394087516554012,
+      "learning_rate": 0.00017081347219149832,
+      "loss": 1.8021,
+      "step": 5906
+    },
+    {
+      "epoch": 0.27212419956696,
+      "grad_norm": 2.2661805280952185,
+      "learning_rate": 0.00017080293610335467,
+      "loss": 1.3415,
+      "step": 5907
+    },
+    {
+      "epoch": 0.272170267655595,
+      "grad_norm": 1.2969462973281045,
+      "learning_rate": 0.00017079239843890504,
+      "loss": 1.5822,
+      "step": 5908
+    },
+    {
+      "epoch": 0.27221633574423,
+      "grad_norm": 2.34854351171949,
+      "learning_rate": 0.00017078185919838405,
+      "loss": 1.6571,
+      "step": 5909
+    },
+    {
+      "epoch": 0.272262403832865,
+      "grad_norm": 2.2349696412162565,
+      "learning_rate": 0.00017077131838202632,
+      "loss": 1.7247,
+      "step": 5910
+    },
+    {
+      "epoch": 0.2723084719215,
+      "grad_norm": 3.6944322380177606,
+      "learning_rate": 0.00017076077599006655,
+      "loss": 1.5205,
+      "step": 5911
+    },
+    {
+      "epoch": 0.272354540010135,
+      "grad_norm": 2.5630669983972174,
+      "learning_rate": 0.0001707502320227394,
+      "loss": 1.7146,
+      "step": 5912
+    },
+    {
+      "epoch": 0.27240060809877,
+      "grad_norm": 2.2329149371146655,
+      "learning_rate": 0.00017073968648027968,
+      "loss": 1.6808,
+      "step": 5913
+    },
+    {
+      "epoch": 0.272446676187405,
+      "grad_norm": 2.649123750099176,
+      "learning_rate": 0.0001707291393629221,
+      "loss": 1.6373,
+      "step": 5914
+    },
+    {
+      "epoch": 0.27249274427604,
+      "grad_norm": 4.063117252635169,
+      "learning_rate": 0.00017071859067090153,
+      "loss": 1.487,
+      "step": 5915
+    },
+    {
+      "epoch": 0.272538812364675,
+      "grad_norm": 2.0956241776320614,
+      "learning_rate": 0.00017070804040445275,
+      "loss": 1.6805,
+      "step": 5916
+    },
+    {
+      "epoch": 0.27258488045331,
+      "grad_norm": 2.011559907675439,
+      "learning_rate": 0.00017069748856381075,
+      "loss": 1.4878,
+      "step": 5917
+    },
+    {
+      "epoch": 0.272630948541945,
+      "grad_norm": 3.1008382281732487,
+      "learning_rate": 0.00017068693514921037,
+      "loss": 1.6552,
+      "step": 5918
+    },
+    {
+      "epoch": 0.27267701663057997,
+      "grad_norm": 1.3827322543525464,
+      "learning_rate": 0.00017067638016088655,
+      "loss": 1.5362,
+      "step": 5919
+    },
+    {
+      "epoch": 0.272723084719215,
+      "grad_norm": 3.587421991565585,
+      "learning_rate": 0.00017066582359907435,
+      "loss": 1.8075,
+      "step": 5920
+    },
+    {
+      "epoch": 0.27276915280785,
+      "grad_norm": 3.351399251419602,
+      "learning_rate": 0.00017065526546400874,
+      "loss": 1.8082,
+      "step": 5921
+    },
+    {
+      "epoch": 0.272815220896485,
+      "grad_norm": 5.033059192791988,
+      "learning_rate": 0.0001706447057559248,
+      "loss": 1.6227,
+      "step": 5922
+    },
+    {
+      "epoch": 0.27286128898512,
+      "grad_norm": 4.60575176232868,
+      "learning_rate": 0.0001706341444750576,
+      "loss": 1.2759,
+      "step": 5923
+    },
+    {
+      "epoch": 0.272907357073755,
+      "grad_norm": 2.6093272892931303,
+      "learning_rate": 0.0001706235816216423,
+      "loss": 1.5659,
+      "step": 5924
+    },
+    {
+      "epoch": 0.27295342516239,
+      "grad_norm": 4.938255831749043,
+      "learning_rate": 0.00017061301719591403,
+      "loss": 1.5111,
+      "step": 5925
+    },
+    {
+      "epoch": 0.272999493251025,
+      "grad_norm": 5.931931559746137,
+      "learning_rate": 0.00017060245119810801,
+      "loss": 1.4231,
+      "step": 5926
+    },
+    {
+      "epoch": 0.27304556133966,
+      "grad_norm": 14.415233034196877,
+      "learning_rate": 0.00017059188362845946,
+      "loss": 1.561,
+      "step": 5927
+    },
+    {
+      "epoch": 0.273091629428295,
+      "grad_norm": 4.414261024365465,
+      "learning_rate": 0.0001705813144872037,
+      "loss": 1.7155,
+      "step": 5928
+    },
+    {
+      "epoch": 0.27313769751693,
+      "grad_norm": 3.1595010645892305,
+      "learning_rate": 0.00017057074377457596,
+      "loss": 1.6632,
+      "step": 5929
+    },
+    {
+      "epoch": 0.273183765605565,
+      "grad_norm": 5.538270270812976,
+      "learning_rate": 0.00017056017149081162,
+      "loss": 1.7242,
+      "step": 5930
+    },
+    {
+      "epoch": 0.2732298336942,
+      "grad_norm": 3.150486361884055,
+      "learning_rate": 0.00017054959763614607,
+      "loss": 1.3837,
+      "step": 5931
+    },
+    {
+      "epoch": 0.273275901782835,
+      "grad_norm": 7.847420688315004,
+      "learning_rate": 0.00017053902221081466,
+      "loss": 1.7296,
+      "step": 5932
+    },
+    {
+      "epoch": 0.27332196987147,
+      "grad_norm": 3.038428081088589,
+      "learning_rate": 0.0001705284452150529,
+      "loss": 1.6229,
+      "step": 5933
+    },
+    {
+      "epoch": 0.27336803796010506,
+      "grad_norm": 3.420883347444879,
+      "learning_rate": 0.0001705178666490962,
+      "loss": 1.7722,
+      "step": 5934
+    },
+    {
+      "epoch": 0.27341410604874006,
+      "grad_norm": 5.676639396289702,
+      "learning_rate": 0.00017050728651318012,
+      "loss": 1.3337,
+      "step": 5935
+    },
+    {
+      "epoch": 0.27346017413737506,
+      "grad_norm": 2.6681102903195195,
+      "learning_rate": 0.00017049670480754023,
+      "loss": 1.4444,
+      "step": 5936
+    },
+    {
+      "epoch": 0.27350624222601005,
+      "grad_norm": 2.600490418904845,
+      "learning_rate": 0.00017048612153241208,
+      "loss": 1.6169,
+      "step": 5937
+    },
+    {
+      "epoch": 0.27355231031464505,
+      "grad_norm": 6.420742567527299,
+      "learning_rate": 0.0001704755366880313,
+      "loss": 1.6281,
+      "step": 5938
+    },
+    {
+      "epoch": 0.27359837840328005,
+      "grad_norm": 3.247308739535415,
+      "learning_rate": 0.0001704649502746335,
+      "loss": 1.4823,
+      "step": 5939
+    },
+    {
+      "epoch": 0.27364444649191505,
+      "grad_norm": 2.1282156181467853,
+      "learning_rate": 0.00017045436229245442,
+      "loss": 1.591,
+      "step": 5940
+    },
+    {
+      "epoch": 0.27369051458055005,
+      "grad_norm": 3.3265931669221573,
+      "learning_rate": 0.00017044377274172976,
+      "loss": 1.1185,
+      "step": 5941
+    },
+    {
+      "epoch": 0.27373658266918505,
+      "grad_norm": 3.317059560467031,
+      "learning_rate": 0.00017043318162269526,
+      "loss": 1.6057,
+      "step": 5942
+    },
+    {
+      "epoch": 0.27378265075782005,
+      "grad_norm": 1.8346685030052172,
+      "learning_rate": 0.00017042258893558678,
+      "loss": 1.9841,
+      "step": 5943
+    },
+    {
+      "epoch": 0.27382871884645504,
+      "grad_norm": 3.145259596100008,
+      "learning_rate": 0.00017041199468064008,
+      "loss": 1.5672,
+      "step": 5944
+    },
+    {
+      "epoch": 0.27387478693509004,
+      "grad_norm": 2.446215549458212,
+      "learning_rate": 0.00017040139885809107,
+      "loss": 1.5481,
+      "step": 5945
+    },
+    {
+      "epoch": 0.27392085502372504,
+      "grad_norm": 6.739005366974566,
+      "learning_rate": 0.00017039080146817565,
+      "loss": 1.5201,
+      "step": 5946
+    },
+    {
+      "epoch": 0.2739669231123601,
+      "grad_norm": 2.3919890939440775,
+      "learning_rate": 0.0001703802025111297,
+      "loss": 1.5377,
+      "step": 5947
+    },
+    {
+      "epoch": 0.2740129912009951,
+      "grad_norm": 4.216480588642672,
+      "learning_rate": 0.0001703696019871892,
+      "loss": 1.3777,
+      "step": 5948
+    },
+    {
+      "epoch": 0.2740590592896301,
+      "grad_norm": 3.174629962059197,
+      "learning_rate": 0.00017035899989659018,
+      "loss": 1.5611,
+      "step": 5949
+    },
+    {
+      "epoch": 0.2741051273782651,
+      "grad_norm": 2.3290739893159818,
+      "learning_rate": 0.00017034839623956868,
+      "loss": 1.4906,
+      "step": 5950
+    },
+    {
+      "epoch": 0.2741511954669001,
+      "grad_norm": 2.602268014053496,
+      "learning_rate": 0.00017033779101636076,
+      "loss": 1.5781,
+      "step": 5951
+    },
+    {
+      "epoch": 0.2741972635555351,
+      "grad_norm": 3.166904495850223,
+      "learning_rate": 0.00017032718422720251,
+      "loss": 1.5408,
+      "step": 5952
+    },
+    {
+      "epoch": 0.2742433316441701,
+      "grad_norm": 3.019611966377912,
+      "learning_rate": 0.0001703165758723301,
+      "loss": 1.6165,
+      "step": 5953
+    },
+    {
+      "epoch": 0.2742893997328051,
+      "grad_norm": 5.682323724690759,
+      "learning_rate": 0.00017030596595197966,
+      "loss": 1.3502,
+      "step": 5954
+    },
+    {
+      "epoch": 0.2743354678214401,
+      "grad_norm": 5.247713360067741,
+      "learning_rate": 0.00017029535446638753,
+      "loss": 1.0,
+      "step": 5955
+    },
+    {
+      "epoch": 0.2743815359100751,
+      "grad_norm": 2.869816240928635,
+      "learning_rate": 0.00017028474141578978,
+      "loss": 1.4965,
+      "step": 5956
+    },
+    {
+      "epoch": 0.2744276039987101,
+      "grad_norm": 3.2167988301118977,
+      "learning_rate": 0.0001702741268004228,
+      "loss": 1.2388,
+      "step": 5957
+    },
+    {
+      "epoch": 0.2744736720873451,
+      "grad_norm": 4.06140790402285,
+      "learning_rate": 0.0001702635106205229,
+      "loss": 1.3655,
+      "step": 5958
+    },
+    {
+      "epoch": 0.2745197401759801,
+      "grad_norm": 15.941514015880195,
+      "learning_rate": 0.00017025289287632636,
+      "loss": 1.1273,
+      "step": 5959
+    },
+    {
+      "epoch": 0.2745658082646151,
+      "grad_norm": 4.823671100931513,
+      "learning_rate": 0.00017024227356806967,
+      "loss": 1.744,
+      "step": 5960
+    },
+    {
+      "epoch": 0.27461187635325013,
+      "grad_norm": 4.824606552862611,
+      "learning_rate": 0.00017023165269598918,
+      "loss": 1.4154,
+      "step": 5961
+    },
+    {
+      "epoch": 0.2746579444418851,
+      "grad_norm": 15.960919879889715,
+      "learning_rate": 0.00017022103026032137,
+      "loss": 1.463,
+      "step": 5962
+    },
+    {
+      "epoch": 0.2747040125305201,
+      "grad_norm": 4.359735655166818,
+      "learning_rate": 0.00017021040626130271,
+      "loss": 1.4997,
+      "step": 5963
+    },
+    {
+      "epoch": 0.2747500806191551,
+      "grad_norm": 1.9734470337375658,
+      "learning_rate": 0.00017019978069916975,
+      "loss": 1.4844,
+      "step": 5964
+    },
+    {
+      "epoch": 0.2747961487077901,
+      "grad_norm": 1.9982385109770173,
+      "learning_rate": 0.00017018915357415904,
+      "loss": 1.6441,
+      "step": 5965
+    },
+    {
+      "epoch": 0.2748422167964251,
+      "grad_norm": 2.5223605072112205,
+      "learning_rate": 0.00017017852488650714,
+      "loss": 1.3222,
+      "step": 5966
+    },
+    {
+      "epoch": 0.2748882848850601,
+      "grad_norm": 2.871086457285597,
+      "learning_rate": 0.00017016789463645076,
+      "loss": 1.7952,
+      "step": 5967
+    },
+    {
+      "epoch": 0.2749343529736951,
+      "grad_norm": 4.550686109886635,
+      "learning_rate": 0.0001701572628242265,
+      "loss": 1.4015,
+      "step": 5968
+    },
+    {
+      "epoch": 0.2749804210623301,
+      "grad_norm": 3.6759561436013026,
+      "learning_rate": 0.0001701466294500711,
+      "loss": 1.5672,
+      "step": 5969
+    },
+    {
+      "epoch": 0.2750264891509651,
+      "grad_norm": 10.912165172362915,
+      "learning_rate": 0.00017013599451422122,
+      "loss": 1.6538,
+      "step": 5970
+    },
+    {
+      "epoch": 0.2750725572396001,
+      "grad_norm": 8.086642543875177,
+      "learning_rate": 0.00017012535801691372,
+      "loss": 1.3768,
+      "step": 5971
+    },
+    {
+      "epoch": 0.2751186253282351,
+      "grad_norm": 2.670859063704845,
+      "learning_rate": 0.00017011471995838535,
+      "loss": 1.6208,
+      "step": 5972
+    },
+    {
+      "epoch": 0.2751646934168701,
+      "grad_norm": 4.545943085865412,
+      "learning_rate": 0.00017010408033887297,
+      "loss": 1.2552,
+      "step": 5973
+    },
+    {
+      "epoch": 0.27521076150550516,
+      "grad_norm": 3.151298794721733,
+      "learning_rate": 0.00017009343915861343,
+      "loss": 1.5481,
+      "step": 5974
+    },
+    {
+      "epoch": 0.27525682959414016,
+      "grad_norm": 3.245473443965172,
+      "learning_rate": 0.00017008279641784365,
+      "loss": 1.6438,
+      "step": 5975
+    },
+    {
+      "epoch": 0.27530289768277516,
+      "grad_norm": 7.5630515542814685,
+      "learning_rate": 0.00017007215211680058,
+      "loss": 1.4287,
+      "step": 5976
+    },
+    {
+      "epoch": 0.27534896577141016,
+      "grad_norm": 2.4146106679874912,
+      "learning_rate": 0.0001700615062557212,
+      "loss": 1.6371,
+      "step": 5977
+    },
+    {
+      "epoch": 0.27539503386004516,
+      "grad_norm": 2.536600361801889,
+      "learning_rate": 0.00017005085883484253,
+      "loss": 1.4669,
+      "step": 5978
+    },
+    {
+      "epoch": 0.27544110194868016,
+      "grad_norm": 4.15248942759993,
+      "learning_rate": 0.00017004020985440156,
+      "loss": 1.8205,
+      "step": 5979
+    },
+    {
+      "epoch": 0.27548717003731515,
+      "grad_norm": 3.6765385177007066,
+      "learning_rate": 0.00017002955931463544,
+      "loss": 1.6305,
+      "step": 5980
+    },
+    {
+      "epoch": 0.27553323812595015,
+      "grad_norm": 1.8778067728347811,
+      "learning_rate": 0.00017001890721578126,
+      "loss": 1.3022,
+      "step": 5981
+    },
+    {
+      "epoch": 0.27557930621458515,
+      "grad_norm": 1.8519400695154657,
+      "learning_rate": 0.00017000825355807612,
+      "loss": 1.0019,
+      "step": 5982
+    },
+    {
+      "epoch": 0.27562537430322015,
+      "grad_norm": 2.51760844785422,
+      "learning_rate": 0.0001699975983417573,
+      "loss": 1.5705,
+      "step": 5983
+    },
+    {
+      "epoch": 0.27567144239185515,
+      "grad_norm": 4.966770158587138,
+      "learning_rate": 0.00016998694156706196,
+      "loss": 1.3803,
+      "step": 5984
+    },
+    {
+      "epoch": 0.27571751048049015,
+      "grad_norm": 1.2932916734810458,
+      "learning_rate": 0.00016997628323422737,
+      "loss": 1.6248,
+      "step": 5985
+    },
+    {
+      "epoch": 0.27576357856912515,
+      "grad_norm": 1.6335174948800786,
+      "learning_rate": 0.0001699656233434908,
+      "loss": 1.4066,
+      "step": 5986
+    },
+    {
+      "epoch": 0.27580964665776014,
+      "grad_norm": 4.209163404930454,
+      "learning_rate": 0.00016995496189508961,
+      "loss": 1.6501,
+      "step": 5987
+    },
+    {
+      "epoch": 0.2758557147463952,
+      "grad_norm": 1.2845570885153808,
+      "learning_rate": 0.00016994429888926113,
+      "loss": 1.3475,
+      "step": 5988
+    },
+    {
+      "epoch": 0.2759017828350302,
+      "grad_norm": 9.260994563653771,
+      "learning_rate": 0.00016993363432624275,
+      "loss": 1.3236,
+      "step": 5989
+    },
+    {
+      "epoch": 0.2759478509236652,
+      "grad_norm": 1.1354840125913122,
+      "learning_rate": 0.00016992296820627197,
+      "loss": 1.504,
+      "step": 5990
+    },
+    {
+      "epoch": 0.2759939190123002,
+      "grad_norm": 1.4238933039407569,
+      "learning_rate": 0.00016991230052958614,
+      "loss": 1.5582,
+      "step": 5991
+    },
+    {
+      "epoch": 0.2760399871009352,
+      "grad_norm": 2.417526469838714,
+      "learning_rate": 0.00016990163129642285,
+      "loss": 1.2152,
+      "step": 5992
+    },
+    {
+      "epoch": 0.2760860551895702,
+      "grad_norm": 1.4395864890792793,
+      "learning_rate": 0.00016989096050701958,
+      "loss": 1.3158,
+      "step": 5993
+    },
+    {
+      "epoch": 0.2761321232782052,
+      "grad_norm": 2.500466907713724,
+      "learning_rate": 0.0001698802881616139,
+      "loss": 1.5552,
+      "step": 5994
+    },
+    {
+      "epoch": 0.2761781913668402,
+      "grad_norm": 2.977418371752517,
+      "learning_rate": 0.0001698696142604434,
+      "loss": 1.5894,
+      "step": 5995
+    },
+    {
+      "epoch": 0.2762242594554752,
+      "grad_norm": 2.202406512564053,
+      "learning_rate": 0.0001698589388037458,
+      "loss": 1.4581,
+      "step": 5996
+    },
+    {
+      "epoch": 0.2762703275441102,
+      "grad_norm": 2.808222426666702,
+      "learning_rate": 0.00016984826179175869,
+      "loss": 1.6506,
+      "step": 5997
+    },
+    {
+      "epoch": 0.2763163956327452,
+      "grad_norm": 1.967643182645013,
+      "learning_rate": 0.00016983758322471973,
+      "loss": 1.7444,
+      "step": 5998
+    },
+    {
+      "epoch": 0.2763624637213802,
+      "grad_norm": 2.514698779817875,
+      "learning_rate": 0.0001698269031028668,
+      "loss": 1.4522,
+      "step": 5999
+    },
+    {
+      "epoch": 0.2764085318100152,
+      "grad_norm": 4.834618661873555,
+      "learning_rate": 0.00016981622142643758,
+      "loss": 1.5532,
+      "step": 6000
+    },
+    {
+      "epoch": 0.2764545998986502,
+      "grad_norm": 2.3718727494300564,
+      "learning_rate": 0.00016980553819566992,
+      "loss": 1.7171,
+      "step": 6001
+    },
+    {
+      "epoch": 0.27650066798728523,
+      "grad_norm": 1.3896100838405088,
+      "learning_rate": 0.0001697948534108016,
+      "loss": 1.5348,
+      "step": 6002
+    },
+    {
+      "epoch": 0.27654673607592023,
+      "grad_norm": 2.536834136428905,
+      "learning_rate": 0.00016978416707207056,
+      "loss": 1.2822,
+      "step": 6003
+    },
+    {
+      "epoch": 0.27659280416455523,
+      "grad_norm": 3.3763685737972082,
+      "learning_rate": 0.00016977347917971472,
+      "loss": 1.5346,
+      "step": 6004
+    },
+    {
+      "epoch": 0.2766388722531902,
+      "grad_norm": 3.096026489724172,
+      "learning_rate": 0.00016976278973397196,
+      "loss": 1.497,
+      "step": 6005
+    },
+    {
+      "epoch": 0.2766849403418252,
+      "grad_norm": 1.8891343290288576,
+      "learning_rate": 0.00016975209873508034,
+      "loss": 1.5675,
+      "step": 6006
+    },
+    {
+      "epoch": 0.2767310084304602,
+      "grad_norm": 5.238084410153169,
+      "learning_rate": 0.0001697414061832778,
+      "loss": 1.6335,
+      "step": 6007
+    },
+    {
+      "epoch": 0.2767770765190952,
+      "grad_norm": 1.7112803815856954,
+      "learning_rate": 0.00016973071207880246,
+      "loss": 1.4845,
+      "step": 6008
+    },
+    {
+      "epoch": 0.2768231446077302,
+      "grad_norm": 2.024787989178419,
+      "learning_rate": 0.0001697200164218924,
+      "loss": 1.4099,
+      "step": 6009
+    },
+    {
+      "epoch": 0.2768692126963652,
+      "grad_norm": 2.591342240826514,
+      "learning_rate": 0.0001697093192127857,
+      "loss": 1.4996,
+      "step": 6010
+    },
+    {
+      "epoch": 0.2769152807850002,
+      "grad_norm": 2.0389491516280245,
+      "learning_rate": 0.0001696986204517205,
+      "loss": 1.3636,
+      "step": 6011
+    },
+    {
+      "epoch": 0.2769613488736352,
+      "grad_norm": 1.6103262753555407,
+      "learning_rate": 0.00016968792013893506,
+      "loss": 1.4797,
+      "step": 6012
+    },
+    {
+      "epoch": 0.2770074169622702,
+      "grad_norm": 1.9787681038877192,
+      "learning_rate": 0.00016967721827466754,
+      "loss": 1.4981,
+      "step": 6013
+    },
+    {
+      "epoch": 0.2770534850509052,
+      "grad_norm": 2.0504674574852317,
+      "learning_rate": 0.00016966651485915624,
+      "loss": 1.4142,
+      "step": 6014
+    },
+    {
+      "epoch": 0.27709955313954027,
+      "grad_norm": 2.0950128491014044,
+      "learning_rate": 0.00016965580989263944,
+      "loss": 1.1913,
+      "step": 6015
+    },
+    {
+      "epoch": 0.27714562122817527,
+      "grad_norm": 1.2822770345386363,
+      "learning_rate": 0.00016964510337535548,
+      "loss": 1.2073,
+      "step": 6016
+    },
+    {
+      "epoch": 0.27719168931681026,
+      "grad_norm": 20.39507352053745,
+      "learning_rate": 0.0001696343953075427,
+      "loss": 1.3082,
+      "step": 6017
+    },
+    {
+      "epoch": 0.27723775740544526,
+      "grad_norm": 2.144566529886238,
+      "learning_rate": 0.00016962368568943949,
+      "loss": 1.2987,
+      "step": 6018
+    },
+    {
+      "epoch": 0.27728382549408026,
+      "grad_norm": 1.894786013095599,
+      "learning_rate": 0.0001696129745212843,
+      "loss": 1.2393,
+      "step": 6019
+    },
+    {
+      "epoch": 0.27732989358271526,
+      "grad_norm": 2.4546543892769512,
+      "learning_rate": 0.00016960226180331563,
+      "loss": 1.4305,
+      "step": 6020
+    },
+    {
+      "epoch": 0.27737596167135026,
+      "grad_norm": 2.872855586930183,
+      "learning_rate": 0.0001695915475357719,
+      "loss": 1.4825,
+      "step": 6021
+    },
+    {
+      "epoch": 0.27742202975998526,
+      "grad_norm": 2.5983644478610497,
+      "learning_rate": 0.0001695808317188917,
+      "loss": 1.3417,
+      "step": 6022
+    },
+    {
+      "epoch": 0.27746809784862025,
+      "grad_norm": 2.3680376088325032,
+      "learning_rate": 0.00016957011435291358,
+      "loss": 1.3755,
+      "step": 6023
+    },
+    {
+      "epoch": 0.27751416593725525,
+      "grad_norm": 3.4607809616880805,
+      "learning_rate": 0.00016955939543807617,
+      "loss": 1.4811,
+      "step": 6024
+    },
+    {
+      "epoch": 0.27756023402589025,
+      "grad_norm": 2.0863933283323726,
+      "learning_rate": 0.0001695486749746181,
+      "loss": 1.7339,
+      "step": 6025
+    },
+    {
+      "epoch": 0.27760630211452525,
+      "grad_norm": 1.6931030374357854,
+      "learning_rate": 0.00016953795296277798,
+      "loss": 1.6736,
+      "step": 6026
+    },
+    {
+      "epoch": 0.27765237020316025,
+      "grad_norm": 5.152946551042644,
+      "learning_rate": 0.00016952722940279463,
+      "loss": 1.5276,
+      "step": 6027
+    },
+    {
+      "epoch": 0.27769843829179525,
+      "grad_norm": 2.5593502532195425,
+      "learning_rate": 0.0001695165042949067,
+      "loss": 1.4724,
+      "step": 6028
+    },
+    {
+      "epoch": 0.2777445063804303,
+      "grad_norm": 2.491369380737853,
+      "learning_rate": 0.00016950577763935303,
+      "loss": 1.2996,
+      "step": 6029
+    },
+    {
+      "epoch": 0.2777905744690653,
+      "grad_norm": 1.7941061683080093,
+      "learning_rate": 0.00016949504943637234,
+      "loss": 1.4381,
+      "step": 6030
+    },
+    {
+      "epoch": 0.2778366425577003,
+      "grad_norm": 3.2565881118439015,
+      "learning_rate": 0.00016948431968620358,
+      "loss": 1.6136,
+      "step": 6031
+    },
+    {
+      "epoch": 0.2778827106463353,
+      "grad_norm": 5.563368352419324,
+      "learning_rate": 0.00016947358838908554,
+      "loss": 1.3906,
+      "step": 6032
+    },
+    {
+      "epoch": 0.2779287787349703,
+      "grad_norm": 1.9753258008257337,
+      "learning_rate": 0.00016946285554525722,
+      "loss": 1.5529,
+      "step": 6033
+    },
+    {
+      "epoch": 0.2779748468236053,
+      "grad_norm": 1.1968735426626336,
+      "learning_rate": 0.0001694521211549575,
+      "loss": 1.3546,
+      "step": 6034
+    },
+    {
+      "epoch": 0.2780209149122403,
+      "grad_norm": 39.307936151285105,
+      "learning_rate": 0.00016944138521842537,
+      "loss": 1.4583,
+      "step": 6035
+    },
+    {
+      "epoch": 0.2780669830008753,
+      "grad_norm": 1.5111517528956022,
+      "learning_rate": 0.00016943064773589992,
+      "loss": 1.1407,
+      "step": 6036
+    },
+    {
+      "epoch": 0.2781130510895103,
+      "grad_norm": 2.304011463269863,
+      "learning_rate": 0.00016941990870762014,
+      "loss": 1.6021,
+      "step": 6037
+    },
+    {
+      "epoch": 0.2781591191781453,
+      "grad_norm": 10.697363128681353,
+      "learning_rate": 0.00016940916813382506,
+      "loss": 1.4047,
+      "step": 6038
+    },
+    {
+      "epoch": 0.2782051872667803,
+      "grad_norm": 3.652956949426974,
+      "learning_rate": 0.0001693984260147539,
+      "loss": 1.5432,
+      "step": 6039
+    },
+    {
+      "epoch": 0.2782512553554153,
+      "grad_norm": 2.7358976825037464,
+      "learning_rate": 0.00016938768235064578,
+      "loss": 1.8331,
+      "step": 6040
+    },
+    {
+      "epoch": 0.2782973234440503,
+      "grad_norm": 2.187474977464789,
+      "learning_rate": 0.0001693769371417399,
+      "loss": 1.5472,
+      "step": 6041
+    },
+    {
+      "epoch": 0.27834339153268534,
+      "grad_norm": 3.349059966698314,
+      "learning_rate": 0.0001693661903882754,
+      "loss": 1.698,
+      "step": 6042
+    },
+    {
+      "epoch": 0.27838945962132033,
+      "grad_norm": 2.872827215739443,
+      "learning_rate": 0.00016935544209049168,
+      "loss": 1.3623,
+      "step": 6043
+    },
+    {
+      "epoch": 0.27843552770995533,
+      "grad_norm": 1.6883169755316998,
+      "learning_rate": 0.00016934469224862795,
+      "loss": 1.7707,
+      "step": 6044
+    },
+    {
+      "epoch": 0.27848159579859033,
+      "grad_norm": 2.5608473866822776,
+      "learning_rate": 0.00016933394086292354,
+      "loss": 1.607,
+      "step": 6045
+    },
+    {
+      "epoch": 0.27852766388722533,
+      "grad_norm": 3.676169874102126,
+      "learning_rate": 0.0001693231879336178,
+      "loss": 1.4287,
+      "step": 6046
+    },
+    {
+      "epoch": 0.27857373197586033,
+      "grad_norm": 5.254030612876011,
+      "learning_rate": 0.0001693124334609502,
+      "loss": 1.3854,
+      "step": 6047
+    },
+    {
+      "epoch": 0.2786198000644953,
+      "grad_norm": 2.819588848393247,
+      "learning_rate": 0.00016930167744516007,
+      "loss": 1.6724,
+      "step": 6048
+    },
+    {
+      "epoch": 0.2786658681531303,
+      "grad_norm": 17.743390115228774,
+      "learning_rate": 0.00016929091988648694,
+      "loss": 1.4536,
+      "step": 6049
+    },
+    {
+      "epoch": 0.2787119362417653,
+      "grad_norm": 1.8656576247654464,
+      "learning_rate": 0.00016928016078517026,
+      "loss": 1.9514,
+      "step": 6050
+    },
+    {
+      "epoch": 0.2787580043304003,
+      "grad_norm": 1.8041158813969835,
+      "learning_rate": 0.00016926940014144962,
+      "loss": 1.4101,
+      "step": 6051
+    },
+    {
+      "epoch": 0.2788040724190353,
+      "grad_norm": 2.1953141369692637,
+      "learning_rate": 0.00016925863795556453,
+      "loss": 1.3973,
+      "step": 6052
+    },
+    {
+      "epoch": 0.2788501405076703,
+      "grad_norm": 1.3665197141373073,
+      "learning_rate": 0.00016924787422775462,
+      "loss": 1.5629,
+      "step": 6053
+    },
+    {
+      "epoch": 0.2788962085963053,
+      "grad_norm": 2.5383122210787197,
+      "learning_rate": 0.00016923710895825952,
+      "loss": 1.465,
+      "step": 6054
+    },
+    {
+      "epoch": 0.2789422766849403,
+      "grad_norm": 2.421925647799089,
+      "learning_rate": 0.0001692263421473189,
+      "loss": 1.5653,
+      "step": 6055
+    },
+    {
+      "epoch": 0.27898834477357537,
+      "grad_norm": 6.608755243771632,
+      "learning_rate": 0.00016921557379517252,
+      "loss": 1.7649,
+      "step": 6056
+    },
+    {
+      "epoch": 0.27903441286221037,
+      "grad_norm": 1.7723161052687089,
+      "learning_rate": 0.00016920480390206005,
+      "loss": 1.8101,
+      "step": 6057
+    },
+    {
+      "epoch": 0.27908048095084537,
+      "grad_norm": 5.6488261920645915,
+      "learning_rate": 0.00016919403246822126,
+      "loss": 1.7195,
+      "step": 6058
+    },
+    {
+      "epoch": 0.27912654903948036,
+      "grad_norm": 3.77755069789069,
+      "learning_rate": 0.00016918325949389599,
+      "loss": 1.3181,
+      "step": 6059
+    },
+    {
+      "epoch": 0.27917261712811536,
+      "grad_norm": 2.1300316076641743,
+      "learning_rate": 0.00016917248497932407,
+      "loss": 1.7388,
+      "step": 6060
+    },
+    {
+      "epoch": 0.27921868521675036,
+      "grad_norm": 11.002384164798471,
+      "learning_rate": 0.00016916170892474538,
+      "loss": 1.6554,
+      "step": 6061
+    },
+    {
+      "epoch": 0.27926475330538536,
+      "grad_norm": 4.488838606579916,
+      "learning_rate": 0.00016915093133039984,
+      "loss": 1.8073,
+      "step": 6062
+    },
+    {
+      "epoch": 0.27931082139402036,
+      "grad_norm": 6.4632708021887515,
+      "learning_rate": 0.00016914015219652738,
+      "loss": 1.7837,
+      "step": 6063
+    },
+    {
+      "epoch": 0.27935688948265536,
+      "grad_norm": 14.274366938152214,
+      "learning_rate": 0.00016912937152336794,
+      "loss": 1.4668,
+      "step": 6064
+    },
+    {
+      "epoch": 0.27940295757129036,
+      "grad_norm": 3.8343529913437693,
+      "learning_rate": 0.0001691185893111616,
+      "loss": 1.4099,
+      "step": 6065
+    },
+    {
+      "epoch": 0.27944902565992535,
+      "grad_norm": 7.2998618368111305,
+      "learning_rate": 0.0001691078055601484,
+      "loss": 1.5649,
+      "step": 6066
+    },
+    {
+      "epoch": 0.27949509374856035,
+      "grad_norm": 2.596148396640279,
+      "learning_rate": 0.0001690970202705684,
+      "loss": 1.7711,
+      "step": 6067
+    },
+    {
+      "epoch": 0.27954116183719535,
+      "grad_norm": 4.358513067697143,
+      "learning_rate": 0.0001690862334426617,
+      "loss": 1.7035,
+      "step": 6068
+    },
+    {
+      "epoch": 0.2795872299258304,
+      "grad_norm": 3.1753647554821653,
+      "learning_rate": 0.00016907544507666848,
+      "loss": 1.5769,
+      "step": 6069
+    },
+    {
+      "epoch": 0.2796332980144654,
+      "grad_norm": 3.283785106026819,
+      "learning_rate": 0.0001690646551728289,
+      "loss": 1.6571,
+      "step": 6070
+    },
+    {
+      "epoch": 0.2796793661031004,
+      "grad_norm": 4.996802550682812,
+      "learning_rate": 0.00016905386373138322,
+      "loss": 1.3292,
+      "step": 6071
+    },
+    {
+      "epoch": 0.2797254341917354,
+      "grad_norm": 2.251145974577352,
+      "learning_rate": 0.00016904307075257165,
+      "loss": 1.4656,
+      "step": 6072
+    },
+    {
+      "epoch": 0.2797715022803704,
+      "grad_norm": 5.007183071115417,
+      "learning_rate": 0.00016903227623663448,
+      "loss": 1.5055,
+      "step": 6073
+    },
+    {
+      "epoch": 0.2798175703690054,
+      "grad_norm": 8.089158269048028,
+      "learning_rate": 0.00016902148018381209,
+      "loss": 1.4933,
+      "step": 6074
+    },
+    {
+      "epoch": 0.2798636384576404,
+      "grad_norm": 8.214467181655829,
+      "learning_rate": 0.00016901068259434477,
+      "loss": 1.7573,
+      "step": 6075
+    },
+    {
+      "epoch": 0.2799097065462754,
+      "grad_norm": 2.6395134039313293,
+      "learning_rate": 0.0001689998834684729,
+      "loss": 1.5956,
+      "step": 6076
+    },
+    {
+      "epoch": 0.2799557746349104,
+      "grad_norm": 6.079450690519964,
+      "learning_rate": 0.00016898908280643698,
+      "loss": 1.5461,
+      "step": 6077
+    },
+    {
+      "epoch": 0.2800018427235454,
+      "grad_norm": 3.211115489030544,
+      "learning_rate": 0.00016897828060847737,
+      "loss": 1.3566,
+      "step": 6078
+    },
+    {
+      "epoch": 0.2800479108121804,
+      "grad_norm": 2.742024477369854,
+      "learning_rate": 0.00016896747687483466,
+      "loss": 1.5695,
+      "step": 6079
+    },
+    {
+      "epoch": 0.2800939789008154,
+      "grad_norm": 3.090039683047563,
+      "learning_rate": 0.0001689566716057493,
+      "loss": 1.6523,
+      "step": 6080
+    },
+    {
+      "epoch": 0.2801400469894504,
+      "grad_norm": 6.467945386721781,
+      "learning_rate": 0.0001689458648014619,
+      "loss": 1.4652,
+      "step": 6081
+    },
+    {
+      "epoch": 0.2801861150780854,
+      "grad_norm": 2.4846183120618806,
+      "learning_rate": 0.00016893505646221302,
+      "loss": 1.311,
+      "step": 6082
+    },
+    {
+      "epoch": 0.28023218316672044,
+      "grad_norm": 2.4947245151025554,
+      "learning_rate": 0.00016892424658824333,
+      "loss": 1.555,
+      "step": 6083
+    },
+    {
+      "epoch": 0.28027825125535544,
+      "grad_norm": 2.915437364135993,
+      "learning_rate": 0.00016891343517979346,
+      "loss": 1.55,
+      "step": 6084
+    },
+    {
+      "epoch": 0.28032431934399044,
+      "grad_norm": 2.1398778577270714,
+      "learning_rate": 0.0001689026222371041,
+      "loss": 1.8607,
+      "step": 6085
+    },
+    {
+      "epoch": 0.28037038743262543,
+      "grad_norm": 2.4676510192400185,
+      "learning_rate": 0.00016889180776041602,
+      "loss": 1.5293,
+      "step": 6086
+    },
+    {
+      "epoch": 0.28041645552126043,
+      "grad_norm": 1.6849629124003376,
+      "learning_rate": 0.00016888099174996993,
+      "loss": 1.5168,
+      "step": 6087
+    },
+    {
+      "epoch": 0.28046252360989543,
+      "grad_norm": 2.2841850098939074,
+      "learning_rate": 0.00016887017420600672,
+      "loss": 1.7366,
+      "step": 6088
+    },
+    {
+      "epoch": 0.28050859169853043,
+      "grad_norm": 2.6821787588360713,
+      "learning_rate": 0.00016885935512876715,
+      "loss": 1.4718,
+      "step": 6089
+    },
+    {
+      "epoch": 0.28055465978716543,
+      "grad_norm": 1.8141833998288277,
+      "learning_rate": 0.00016884853451849208,
+      "loss": 1.7572,
+      "step": 6090
+    },
+    {
+      "epoch": 0.2806007278758004,
+      "grad_norm": 3.175187687178657,
+      "learning_rate": 0.00016883771237542246,
+      "loss": 1.7285,
+      "step": 6091
+    },
+    {
+      "epoch": 0.2806467959644354,
+      "grad_norm": 2.960733378513011,
+      "learning_rate": 0.0001688268886997992,
+      "loss": 1.6178,
+      "step": 6092
+    },
+    {
+      "epoch": 0.2806928640530704,
+      "grad_norm": 2.340397781825742,
+      "learning_rate": 0.0001688160634918633,
+      "loss": 1.4966,
+      "step": 6093
+    },
+    {
+      "epoch": 0.2807389321417054,
+      "grad_norm": 2.396958647807904,
+      "learning_rate": 0.0001688052367518557,
+      "loss": 1.4108,
+      "step": 6094
+    },
+    {
+      "epoch": 0.2807850002303404,
+      "grad_norm": 1.518363623292386,
+      "learning_rate": 0.0001687944084800175,
+      "loss": 1.2015,
+      "step": 6095
+    },
+    {
+      "epoch": 0.2808310683189754,
+      "grad_norm": 2.0957315791201006,
+      "learning_rate": 0.00016878357867658973,
+      "loss": 1.5049,
+      "step": 6096
+    },
+    {
+      "epoch": 0.2808771364076105,
+      "grad_norm": 3.087340078848042,
+      "learning_rate": 0.00016877274734181357,
+      "loss": 1.4893,
+      "step": 6097
+    },
+    {
+      "epoch": 0.28092320449624547,
+      "grad_norm": 1.47168506865702,
+      "learning_rate": 0.00016876191447593006,
+      "loss": 1.5552,
+      "step": 6098
+    },
+    {
+      "epoch": 0.28096927258488047,
+      "grad_norm": 1.7799638498872241,
+      "learning_rate": 0.00016875108007918045,
+      "loss": 1.7555,
+      "step": 6099
+    },
+    {
+      "epoch": 0.28101534067351547,
+      "grad_norm": 2.6607055575776117,
+      "learning_rate": 0.0001687402441518059,
+      "loss": 1.4159,
+      "step": 6100
+    },
+    {
+      "epoch": 0.28106140876215047,
+      "grad_norm": 7.2181551043351915,
+      "learning_rate": 0.00016872940669404772,
+      "loss": 1.4763,
+      "step": 6101
+    },
+    {
+      "epoch": 0.28110747685078546,
+      "grad_norm": 1.594013266648116,
+      "learning_rate": 0.00016871856770614714,
+      "loss": 1.6175,
+      "step": 6102
+    },
+    {
+      "epoch": 0.28115354493942046,
+      "grad_norm": 1.9604567530816184,
+      "learning_rate": 0.00016870772718834544,
+      "loss": 1.5196,
+      "step": 6103
+    },
+    {
+      "epoch": 0.28119961302805546,
+      "grad_norm": 2.068618919104326,
+      "learning_rate": 0.000168696885140884,
+      "loss": 1.4066,
+      "step": 6104
+    },
+    {
+      "epoch": 0.28124568111669046,
+      "grad_norm": 4.170485793285133,
+      "learning_rate": 0.00016868604156400425,
+      "loss": 1.8191,
+      "step": 6105
+    },
+    {
+      "epoch": 0.28129174920532546,
+      "grad_norm": 2.6502399578991542,
+      "learning_rate": 0.0001686751964579475,
+      "loss": 1.5062,
+      "step": 6106
+    },
+    {
+      "epoch": 0.28133781729396046,
+      "grad_norm": 6.498859881003199,
+      "learning_rate": 0.0001686643498229553,
+      "loss": 1.5473,
+      "step": 6107
+    },
+    {
+      "epoch": 0.28138388538259546,
+      "grad_norm": 5.772701195939372,
+      "learning_rate": 0.0001686535016592691,
+      "loss": 1.7148,
+      "step": 6108
+    },
+    {
+      "epoch": 0.28142995347123045,
+      "grad_norm": 4.625436443664265,
+      "learning_rate": 0.00016864265196713034,
+      "loss": 1.9038,
+      "step": 6109
+    },
+    {
+      "epoch": 0.2814760215598655,
+      "grad_norm": 1.4001580207801991,
+      "learning_rate": 0.00016863180074678066,
+      "loss": 1.4655,
+      "step": 6110
+    },
+    {
+      "epoch": 0.2815220896485005,
+      "grad_norm": 2.5674714407928816,
+      "learning_rate": 0.00016862094799846165,
+      "loss": 1.3194,
+      "step": 6111
+    },
+    {
+      "epoch": 0.2815681577371355,
+      "grad_norm": 1.7788446144799421,
+      "learning_rate": 0.00016861009372241484,
+      "loss": 1.6458,
+      "step": 6112
+    },
+    {
+      "epoch": 0.2816142258257705,
+      "grad_norm": 1.3104498427491298,
+      "learning_rate": 0.00016859923791888196,
+      "loss": 1.562,
+      "step": 6113
+    },
+    {
+      "epoch": 0.2816602939144055,
+      "grad_norm": 1.2244452471463942,
+      "learning_rate": 0.00016858838058810466,
+      "loss": 1.8022,
+      "step": 6114
+    },
+    {
+      "epoch": 0.2817063620030405,
+      "grad_norm": 2.433625969130612,
+      "learning_rate": 0.00016857752173032466,
+      "loss": 1.6222,
+      "step": 6115
+    },
+    {
+      "epoch": 0.2817524300916755,
+      "grad_norm": 1.411844595369206,
+      "learning_rate": 0.00016856666134578373,
+      "loss": 1.403,
+      "step": 6116
+    },
+    {
+      "epoch": 0.2817984981803105,
+      "grad_norm": 1.9689132947927124,
+      "learning_rate": 0.0001685557994347237,
+      "loss": 1.5017,
+      "step": 6117
+    },
+    {
+      "epoch": 0.2818445662689455,
+      "grad_norm": 5.528892406201897,
+      "learning_rate": 0.00016854493599738628,
+      "loss": 1.3497,
+      "step": 6118
+    },
+    {
+      "epoch": 0.2818906343575805,
+      "grad_norm": 1.6933530221545428,
+      "learning_rate": 0.00016853407103401338,
+      "loss": 1.5742,
+      "step": 6119
+    },
+    {
+      "epoch": 0.2819367024462155,
+      "grad_norm": 1.51341997569067,
+      "learning_rate": 0.00016852320454484695,
+      "loss": 1.6043,
+      "step": 6120
+    },
+    {
+      "epoch": 0.2819827705348505,
+      "grad_norm": 4.944485348453056,
+      "learning_rate": 0.00016851233653012888,
+      "loss": 1.1859,
+      "step": 6121
+    },
+    {
+      "epoch": 0.2820288386234855,
+      "grad_norm": 1.7819951197471426,
+      "learning_rate": 0.00016850146699010104,
+      "loss": 1.3066,
+      "step": 6122
+    },
+    {
+      "epoch": 0.2820749067121205,
+      "grad_norm": 2.087191119793995,
+      "learning_rate": 0.00016849059592500556,
+      "loss": 1.3651,
+      "step": 6123
+    },
+    {
+      "epoch": 0.28212097480075554,
+      "grad_norm": 1.7077915816892681,
+      "learning_rate": 0.00016847972333508435,
+      "loss": 1.4091,
+      "step": 6124
+    },
+    {
+      "epoch": 0.28216704288939054,
+      "grad_norm": 2.360449098166252,
+      "learning_rate": 0.00016846884922057953,
+      "loss": 1.5214,
+      "step": 6125
+    },
+    {
+      "epoch": 0.28221311097802554,
+      "grad_norm": 2.275820402811922,
+      "learning_rate": 0.0001684579735817332,
+      "loss": 1.4171,
+      "step": 6126
+    },
+    {
+      "epoch": 0.28225917906666054,
+      "grad_norm": 2.205615606327344,
+      "learning_rate": 0.00016844709641878743,
+      "loss": 1.3892,
+      "step": 6127
+    },
+    {
+      "epoch": 0.28230524715529554,
+      "grad_norm": 2.4231233924439453,
+      "learning_rate": 0.00016843621773198443,
+      "loss": 1.4077,
+      "step": 6128
+    },
+    {
+      "epoch": 0.28235131524393053,
+      "grad_norm": 5.503279907731756,
+      "learning_rate": 0.00016842533752156643,
+      "loss": 1.4177,
+      "step": 6129
+    },
+    {
+      "epoch": 0.28239738333256553,
+      "grad_norm": 2.6502700683326967,
+      "learning_rate": 0.00016841445578777558,
+      "loss": 1.4561,
+      "step": 6130
+    },
+    {
+      "epoch": 0.28244345142120053,
+      "grad_norm": 1.1201073261816676,
+      "learning_rate": 0.00016840357253085418,
+      "loss": 1.3088,
+      "step": 6131
+    },
+    {
+      "epoch": 0.28248951950983553,
+      "grad_norm": 1.8781563003409274,
+      "learning_rate": 0.00016839268775104452,
+      "loss": 1.4032,
+      "step": 6132
+    },
+    {
+      "epoch": 0.28253558759847053,
+      "grad_norm": 1.3181753189918382,
+      "learning_rate": 0.00016838180144858892,
+      "loss": 1.6925,
+      "step": 6133
+    },
+    {
+      "epoch": 0.2825816556871055,
+      "grad_norm": 1.7413488944268154,
+      "learning_rate": 0.0001683709136237298,
+      "loss": 1.2689,
+      "step": 6134
+    },
+    {
+      "epoch": 0.2826277237757405,
+      "grad_norm": 3.010588015628293,
+      "learning_rate": 0.00016836002427670946,
+      "loss": 1.4621,
+      "step": 6135
+    },
+    {
+      "epoch": 0.2826737918643755,
+      "grad_norm": 1.7756279622063356,
+      "learning_rate": 0.00016834913340777046,
+      "loss": 1.5627,
+      "step": 6136
+    },
+    {
+      "epoch": 0.2827198599530106,
+      "grad_norm": 1.8335545986219708,
+      "learning_rate": 0.00016833824101715515,
+      "loss": 1.6279,
+      "step": 6137
+    },
+    {
+      "epoch": 0.2827659280416456,
+      "grad_norm": 2.6095842436812307,
+      "learning_rate": 0.00016832734710510612,
+      "loss": 1.6502,
+      "step": 6138
+    },
+    {
+      "epoch": 0.2828119961302806,
+      "grad_norm": 1.2806747210756348,
+      "learning_rate": 0.00016831645167186582,
+      "loss": 1.559,
+      "step": 6139
+    },
+    {
+      "epoch": 0.2828580642189156,
+      "grad_norm": 3.207673802278872,
+      "learning_rate": 0.00016830555471767687,
+      "loss": 1.5264,
+      "step": 6140
+    },
+    {
+      "epoch": 0.28290413230755057,
+      "grad_norm": 2.1743950387438824,
+      "learning_rate": 0.0001682946562427819,
+      "loss": 1.6015,
+      "step": 6141
+    },
+    {
+      "epoch": 0.28295020039618557,
+      "grad_norm": 3.0726881663130925,
+      "learning_rate": 0.0001682837562474235,
+      "loss": 1.6467,
+      "step": 6142
+    },
+    {
+      "epoch": 0.28299626848482057,
+      "grad_norm": 1.4639586019901503,
+      "learning_rate": 0.00016827285473184432,
+      "loss": 1.3699,
+      "step": 6143
+    },
+    {
+      "epoch": 0.28304233657345557,
+      "grad_norm": 1.6412186674352112,
+      "learning_rate": 0.0001682619516962871,
+      "loss": 1.7646,
+      "step": 6144
+    },
+    {
+      "epoch": 0.28308840466209056,
+      "grad_norm": 2.4359608007642164,
+      "learning_rate": 0.00016825104714099454,
+      "loss": 1.3317,
+      "step": 6145
+    },
+    {
+      "epoch": 0.28313447275072556,
+      "grad_norm": 1.7882806316084539,
+      "learning_rate": 0.00016824014106620952,
+      "loss": 1.3955,
+      "step": 6146
+    },
+    {
+      "epoch": 0.28318054083936056,
+      "grad_norm": 2.4064993229318024,
+      "learning_rate": 0.00016822923347217472,
+      "loss": 1.5669,
+      "step": 6147
+    },
+    {
+      "epoch": 0.28322660892799556,
+      "grad_norm": 2.0140498523122043,
+      "learning_rate": 0.00016821832435913304,
+      "loss": 1.5226,
+      "step": 6148
+    },
+    {
+      "epoch": 0.28327267701663056,
+      "grad_norm": 1.144487023179655,
+      "learning_rate": 0.00016820741372732732,
+      "loss": 1.6247,
+      "step": 6149
+    },
+    {
+      "epoch": 0.28331874510526556,
+      "grad_norm": 2.364678449399183,
+      "learning_rate": 0.0001681965015770005,
+      "loss": 0.949,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2833648131939006,
+      "grad_norm": 1.695610929452851,
+      "learning_rate": 0.00016818558790839547,
+      "loss": 1.6346,
+      "step": 6151
+    },
+    {
+      "epoch": 0.2834108812825356,
+      "grad_norm": 3.4816090950739316,
+      "learning_rate": 0.00016817467272175525,
+      "loss": 1.3555,
+      "step": 6152
+    },
+    {
+      "epoch": 0.2834569493711706,
+      "grad_norm": 1.9589841247995825,
+      "learning_rate": 0.00016816375601732284,
+      "loss": 1.3679,
+      "step": 6153
+    },
+    {
+      "epoch": 0.2835030174598056,
+      "grad_norm": 4.681226891090876,
+      "learning_rate": 0.00016815283779534127,
+      "loss": 1.7629,
+      "step": 6154
+    },
+    {
+      "epoch": 0.2835490855484406,
+      "grad_norm": 2.9350360479292124,
+      "learning_rate": 0.00016814191805605364,
+      "loss": 1.5004,
+      "step": 6155
+    },
+    {
+      "epoch": 0.2835951536370756,
+      "grad_norm": 2.150917110235284,
+      "learning_rate": 0.0001681309967997031,
+      "loss": 1.4646,
+      "step": 6156
+    },
+    {
+      "epoch": 0.2836412217257106,
+      "grad_norm": 3.2078188430982157,
+      "learning_rate": 0.00016812007402653264,
+      "loss": 1.5154,
+      "step": 6157
+    },
+    {
+      "epoch": 0.2836872898143456,
+      "grad_norm": 1.5222245985223544,
+      "learning_rate": 0.00016810914973678558,
+      "loss": 1.7164,
+      "step": 6158
+    },
+    {
+      "epoch": 0.2837333579029806,
+      "grad_norm": 2.2548185508505703,
+      "learning_rate": 0.00016809822393070508,
+      "loss": 1.5892,
+      "step": 6159
+    },
+    {
+      "epoch": 0.2837794259916156,
+      "grad_norm": 2.3002619128097734,
+      "learning_rate": 0.00016808729660853437,
+      "loss": 1.5704,
+      "step": 6160
+    },
+    {
+      "epoch": 0.2838254940802506,
+      "grad_norm": 2.147805301917523,
+      "learning_rate": 0.00016807636777051675,
+      "loss": 1.5197,
+      "step": 6161
+    },
+    {
+      "epoch": 0.2838715621688856,
+      "grad_norm": 1.6618897943885804,
+      "learning_rate": 0.00016806543741689552,
+      "loss": 1.3411,
+      "step": 6162
+    },
+    {
+      "epoch": 0.2839176302575206,
+      "grad_norm": 2.660597835368012,
+      "learning_rate": 0.00016805450554791405,
+      "loss": 1.0588,
+      "step": 6163
+    },
+    {
+      "epoch": 0.2839636983461556,
+      "grad_norm": 1.996482191136991,
+      "learning_rate": 0.0001680435721638157,
+      "loss": 1.5887,
+      "step": 6164
+    },
+    {
+      "epoch": 0.28400976643479064,
+      "grad_norm": 1.6577934568867292,
+      "learning_rate": 0.00016803263726484387,
+      "loss": 1.2076,
+      "step": 6165
+    },
+    {
+      "epoch": 0.28405583452342564,
+      "grad_norm": 1.507165541540003,
+      "learning_rate": 0.00016802170085124205,
+      "loss": 1.3712,
+      "step": 6166
+    },
+    {
+      "epoch": 0.28410190261206064,
+      "grad_norm": 1.9650763353162048,
+      "learning_rate": 0.0001680107629232537,
+      "loss": 1.3601,
+      "step": 6167
+    },
+    {
+      "epoch": 0.28414797070069564,
+      "grad_norm": 1.7604966632708454,
+      "learning_rate": 0.0001679998234811223,
+      "loss": 1.4262,
+      "step": 6168
+    },
+    {
+      "epoch": 0.28419403878933064,
+      "grad_norm": 1.4322837456290551,
+      "learning_rate": 0.00016798888252509144,
+      "loss": 1.46,
+      "step": 6169
+    },
+    {
+      "epoch": 0.28424010687796564,
+      "grad_norm": 4.925546646692239,
+      "learning_rate": 0.00016797794005540465,
+      "loss": 1.2992,
+      "step": 6170
+    },
+    {
+      "epoch": 0.28428617496660064,
+      "grad_norm": 1.9430382129602257,
+      "learning_rate": 0.0001679669960723056,
+      "loss": 1.3589,
+      "step": 6171
+    },
+    {
+      "epoch": 0.28433224305523563,
+      "grad_norm": 1.1817640779211491,
+      "learning_rate": 0.00016795605057603793,
+      "loss": 1.6262,
+      "step": 6172
+    },
+    {
+      "epoch": 0.28437831114387063,
+      "grad_norm": 1.6646902668859973,
+      "learning_rate": 0.00016794510356684533,
+      "loss": 1.4504,
+      "step": 6173
+    },
+    {
+      "epoch": 0.28442437923250563,
+      "grad_norm": 1.4629752110818244,
+      "learning_rate": 0.00016793415504497147,
+      "loss": 1.3032,
+      "step": 6174
+    },
+    {
+      "epoch": 0.28447044732114063,
+      "grad_norm": 1.4027138565273172,
+      "learning_rate": 0.00016792320501066017,
+      "loss": 1.3563,
+      "step": 6175
+    },
+    {
+      "epoch": 0.2845165154097756,
+      "grad_norm": 1.6121867059520238,
+      "learning_rate": 0.00016791225346415517,
+      "loss": 1.3293,
+      "step": 6176
+    },
+    {
+      "epoch": 0.2845625834984106,
+      "grad_norm": 2.112020788831256,
+      "learning_rate": 0.00016790130040570024,
+      "loss": 1.7087,
+      "step": 6177
+    },
+    {
+      "epoch": 0.2846086515870457,
+      "grad_norm": 1.589237246841405,
+      "learning_rate": 0.00016789034583553932,
+      "loss": 1.2326,
+      "step": 6178
+    },
+    {
+      "epoch": 0.2846547196756807,
+      "grad_norm": 1.973956287445911,
+      "learning_rate": 0.00016787938975391624,
+      "loss": 1.2191,
+      "step": 6179
+    },
+    {
+      "epoch": 0.2847007877643157,
+      "grad_norm": 1.8129199165222802,
+      "learning_rate": 0.00016786843216107494,
+      "loss": 1.3454,
+      "step": 6180
+    },
+    {
+      "epoch": 0.2847468558529507,
+      "grad_norm": 1.2364837756817424,
+      "learning_rate": 0.0001678574730572594,
+      "loss": 1.4791,
+      "step": 6181
+    },
+    {
+      "epoch": 0.2847929239415857,
+      "grad_norm": 3.0957055620401634,
+      "learning_rate": 0.0001678465124427135,
+      "loss": 1.3698,
+      "step": 6182
+    },
+    {
+      "epoch": 0.2848389920302207,
+      "grad_norm": 2.116968670517981,
+      "learning_rate": 0.0001678355503176814,
+      "loss": 1.2973,
+      "step": 6183
+    },
+    {
+      "epoch": 0.28488506011885567,
+      "grad_norm": 2.0568164828216293,
+      "learning_rate": 0.00016782458668240707,
+      "loss": 0.962,
+      "step": 6184
+    },
+    {
+      "epoch": 0.28493112820749067,
+      "grad_norm": 2.2987075294063684,
+      "learning_rate": 0.0001678136215371346,
+      "loss": 1.4594,
+      "step": 6185
+    },
+    {
+      "epoch": 0.28497719629612567,
+      "grad_norm": 4.967115604027073,
+      "learning_rate": 0.0001678026548821081,
+      "loss": 1.262,
+      "step": 6186
+    },
+    {
+      "epoch": 0.28502326438476067,
+      "grad_norm": 12.151880686555751,
+      "learning_rate": 0.00016779168671757178,
+      "loss": 1.5219,
+      "step": 6187
+    },
+    {
+      "epoch": 0.28506933247339566,
+      "grad_norm": 6.47149684211956,
+      "learning_rate": 0.0001677807170437698,
+      "loss": 1.3139,
+      "step": 6188
+    },
+    {
+      "epoch": 0.28511540056203066,
+      "grad_norm": 5.003039208034334,
+      "learning_rate": 0.00016776974586094637,
+      "loss": 1.3073,
+      "step": 6189
+    },
+    {
+      "epoch": 0.28516146865066566,
+      "grad_norm": 4.59748830425291,
+      "learning_rate": 0.0001677587731693457,
+      "loss": 1.3199,
+      "step": 6190
+    },
+    {
+      "epoch": 0.28520753673930066,
+      "grad_norm": 1.768985301639989,
+      "learning_rate": 0.00016774779896921218,
+      "loss": 1.1162,
+      "step": 6191
+    },
+    {
+      "epoch": 0.2852536048279357,
+      "grad_norm": 4.329772929456969,
+      "learning_rate": 0.00016773682326079006,
+      "loss": 1.1312,
+      "step": 6192
+    },
+    {
+      "epoch": 0.2852996729165707,
+      "grad_norm": 1.7428613112419042,
+      "learning_rate": 0.0001677258460443237,
+      "loss": 1.6449,
+      "step": 6193
+    },
+    {
+      "epoch": 0.2853457410052057,
+      "grad_norm": 2.64571205340278,
+      "learning_rate": 0.00016771486732005756,
+      "loss": 1.2011,
+      "step": 6194
+    },
+    {
+      "epoch": 0.2853918090938407,
+      "grad_norm": 2.9615782076564736,
+      "learning_rate": 0.00016770388708823593,
+      "loss": 1.581,
+      "step": 6195
+    },
+    {
+      "epoch": 0.2854378771824757,
+      "grad_norm": 2.8578087359772235,
+      "learning_rate": 0.00016769290534910338,
+      "loss": 1.563,
+      "step": 6196
+    },
+    {
+      "epoch": 0.2854839452711107,
+      "grad_norm": 7.6335499783962755,
+      "learning_rate": 0.00016768192210290438,
+      "loss": 1.4479,
+      "step": 6197
+    },
+    {
+      "epoch": 0.2855300133597457,
+      "grad_norm": 2.498863738121869,
+      "learning_rate": 0.0001676709373498834,
+      "loss": 1.3447,
+      "step": 6198
+    },
+    {
+      "epoch": 0.2855760814483807,
+      "grad_norm": 3.022100813447347,
+      "learning_rate": 0.00016765995109028502,
+      "loss": 1.4359,
+      "step": 6199
+    },
+    {
+      "epoch": 0.2856221495370157,
+      "grad_norm": 6.246853605548486,
+      "learning_rate": 0.00016764896332435386,
+      "loss": 1.4053,
+      "step": 6200
+    },
+    {
+      "epoch": 0.2856682176256507,
+      "grad_norm": 3.1280536783921513,
+      "learning_rate": 0.00016763797405233453,
+      "loss": 1.1464,
+      "step": 6201
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 2.746515580806157,
+      "learning_rate": 0.00016762698327447168,
+      "loss": 1.5778,
+      "step": 6202
+    },
+    {
+      "epoch": 0.2857603538029207,
+      "grad_norm": 2.468031866945024,
+      "learning_rate": 0.00016761599099101,
+      "loss": 1.3102,
+      "step": 6203
+    },
+    {
+      "epoch": 0.2858064218915557,
+      "grad_norm": 3.1419165676025895,
+      "learning_rate": 0.00016760499720219424,
+      "loss": 1.5098,
+      "step": 6204
+    },
+    {
+      "epoch": 0.28585248998019075,
+      "grad_norm": 2.1436291590106604,
+      "learning_rate": 0.00016759400190826908,
+      "loss": 1.6657,
+      "step": 6205
+    },
+    {
+      "epoch": 0.28589855806882575,
+      "grad_norm": 2.987820999775567,
+      "learning_rate": 0.00016758300510947942,
+      "loss": 1.3046,
+      "step": 6206
+    },
+    {
+      "epoch": 0.28594462615746075,
+      "grad_norm": 1.659100638866631,
+      "learning_rate": 0.00016757200680607,
+      "loss": 1.6311,
+      "step": 6207
+    },
+    {
+      "epoch": 0.28599069424609574,
+      "grad_norm": 37.96154921881797,
+      "learning_rate": 0.00016756100699828573,
+      "loss": 1.4352,
+      "step": 6208
+    },
+    {
+      "epoch": 0.28603676233473074,
+      "grad_norm": 2.969356171766531,
+      "learning_rate": 0.00016755000568637147,
+      "loss": 1.4506,
+      "step": 6209
+    },
+    {
+      "epoch": 0.28608283042336574,
+      "grad_norm": 1.9690835694602147,
+      "learning_rate": 0.00016753900287057216,
+      "loss": 1.7047,
+      "step": 6210
+    },
+    {
+      "epoch": 0.28612889851200074,
+      "grad_norm": 1.2605462379716041,
+      "learning_rate": 0.00016752799855113276,
+      "loss": 1.5083,
+      "step": 6211
+    },
+    {
+      "epoch": 0.28617496660063574,
+      "grad_norm": 2.0200277612907196,
+      "learning_rate": 0.00016751699272829823,
+      "loss": 1.528,
+      "step": 6212
+    },
+    {
+      "epoch": 0.28622103468927074,
+      "grad_norm": 3.522034820632736,
+      "learning_rate": 0.00016750598540231365,
+      "loss": 1.597,
+      "step": 6213
+    },
+    {
+      "epoch": 0.28626710277790574,
+      "grad_norm": 1.875678923236213,
+      "learning_rate": 0.00016749497657342406,
+      "loss": 1.5126,
+      "step": 6214
+    },
+    {
+      "epoch": 0.28631317086654073,
+      "grad_norm": 4.124872857653669,
+      "learning_rate": 0.00016748396624187453,
+      "loss": 1.5396,
+      "step": 6215
+    },
+    {
+      "epoch": 0.28635923895517573,
+      "grad_norm": 2.0503512586442283,
+      "learning_rate": 0.0001674729544079102,
+      "loss": 1.6472,
+      "step": 6216
+    },
+    {
+      "epoch": 0.28640530704381073,
+      "grad_norm": 2.6824658910678316,
+      "learning_rate": 0.00016746194107177624,
+      "loss": 1.7568,
+      "step": 6217
+    },
+    {
+      "epoch": 0.28645137513244573,
+      "grad_norm": 3.601699754261787,
+      "learning_rate": 0.00016745092623371782,
+      "loss": 1.3551,
+      "step": 6218
+    },
+    {
+      "epoch": 0.2864974432210808,
+      "grad_norm": 2.448408444601352,
+      "learning_rate": 0.0001674399098939802,
+      "loss": 1.8328,
+      "step": 6219
+    },
+    {
+      "epoch": 0.2865435113097158,
+      "grad_norm": 6.280746835808127,
+      "learning_rate": 0.00016742889205280864,
+      "loss": 1.5658,
+      "step": 6220
+    },
+    {
+      "epoch": 0.2865895793983508,
+      "grad_norm": 3.2348848648509123,
+      "learning_rate": 0.00016741787271044837,
+      "loss": 1.3329,
+      "step": 6221
+    },
+    {
+      "epoch": 0.2866356474869858,
+      "grad_norm": 3.200965518445843,
+      "learning_rate": 0.00016740685186714477,
+      "loss": 1.1472,
+      "step": 6222
+    },
+    {
+      "epoch": 0.2866817155756208,
+      "grad_norm": 5.240856174915069,
+      "learning_rate": 0.00016739582952314322,
+      "loss": 1.4376,
+      "step": 6223
+    },
+    {
+      "epoch": 0.2867277836642558,
+      "grad_norm": 2.5899590550879767,
+      "learning_rate": 0.00016738480567868904,
+      "loss": 1.6123,
+      "step": 6224
+    },
+    {
+      "epoch": 0.2867738517528908,
+      "grad_norm": 2.7590704900968817,
+      "learning_rate": 0.00016737378033402774,
+      "loss": 1.4461,
+      "step": 6225
+    },
+    {
+      "epoch": 0.28681991984152577,
+      "grad_norm": 2.340964893298784,
+      "learning_rate": 0.00016736275348940471,
+      "loss": 1.6864,
+      "step": 6226
+    },
+    {
+      "epoch": 0.28686598793016077,
+      "grad_norm": 2.0537190976741666,
+      "learning_rate": 0.00016735172514506549,
+      "loss": 1.655,
+      "step": 6227
+    },
+    {
+      "epoch": 0.28691205601879577,
+      "grad_norm": 2.1676402133078563,
+      "learning_rate": 0.0001673406953012556,
+      "loss": 1.3038,
+      "step": 6228
+    },
+    {
+      "epoch": 0.28695812410743077,
+      "grad_norm": 2.505718000250684,
+      "learning_rate": 0.0001673296639582206,
+      "loss": 1.6905,
+      "step": 6229
+    },
+    {
+      "epoch": 0.28700419219606577,
+      "grad_norm": 2.731695222273051,
+      "learning_rate": 0.0001673186311162061,
+      "loss": 1.7538,
+      "step": 6230
+    },
+    {
+      "epoch": 0.28705026028470076,
+      "grad_norm": 2.7362066251126462,
+      "learning_rate": 0.00016730759677545764,
+      "loss": 1.4564,
+      "step": 6231
+    },
+    {
+      "epoch": 0.28709632837333576,
+      "grad_norm": 1.616446304041507,
+      "learning_rate": 0.000167296560936221,
+      "loss": 1.4478,
+      "step": 6232
+    },
+    {
+      "epoch": 0.2871423964619708,
+      "grad_norm": 1.1901275378280642,
+      "learning_rate": 0.0001672855235987418,
+      "loss": 1.2286,
+      "step": 6233
+    },
+    {
+      "epoch": 0.2871884645506058,
+      "grad_norm": 2.1810760097347535,
+      "learning_rate": 0.00016727448476326577,
+      "loss": 1.28,
+      "step": 6234
+    },
+    {
+      "epoch": 0.2872345326392408,
+      "grad_norm": 6.099185803954922,
+      "learning_rate": 0.0001672634444300387,
+      "loss": 1.4749,
+      "step": 6235
+    },
+    {
+      "epoch": 0.2872806007278758,
+      "grad_norm": 1.7476937845100315,
+      "learning_rate": 0.00016725240259930638,
+      "loss": 1.5333,
+      "step": 6236
+    },
+    {
+      "epoch": 0.2873266688165108,
+      "grad_norm": 4.209231927538706,
+      "learning_rate": 0.00016724135927131462,
+      "loss": 1.7651,
+      "step": 6237
+    },
+    {
+      "epoch": 0.2873727369051458,
+      "grad_norm": 2.3752703563830835,
+      "learning_rate": 0.0001672303144463093,
+      "loss": 1.3903,
+      "step": 6238
+    },
+    {
+      "epoch": 0.2874188049937808,
+      "grad_norm": 2.323757741675356,
+      "learning_rate": 0.00016721926812453633,
+      "loss": 1.4936,
+      "step": 6239
+    },
+    {
+      "epoch": 0.2874648730824158,
+      "grad_norm": 1.8386809594927365,
+      "learning_rate": 0.00016720822030624155,
+      "loss": 1.3124,
+      "step": 6240
+    },
+    {
+      "epoch": 0.2875109411710508,
+      "grad_norm": 1.3214673877694463,
+      "learning_rate": 0.000167197170991671,
+      "loss": 1.3677,
+      "step": 6241
+    },
+    {
+      "epoch": 0.2875570092596858,
+      "grad_norm": 5.838771121931714,
+      "learning_rate": 0.00016718612018107068,
+      "loss": 1.3813,
+      "step": 6242
+    },
+    {
+      "epoch": 0.2876030773483208,
+      "grad_norm": 1.391895748334631,
+      "learning_rate": 0.00016717506787468657,
+      "loss": 1.4474,
+      "step": 6243
+    },
+    {
+      "epoch": 0.2876491454369558,
+      "grad_norm": 2.141401573138764,
+      "learning_rate": 0.0001671640140727648,
+      "loss": 1.9372,
+      "step": 6244
+    },
+    {
+      "epoch": 0.2876952135255908,
+      "grad_norm": 6.913620052178618,
+      "learning_rate": 0.00016715295877555138,
+      "loss": 1.5787,
+      "step": 6245
+    },
+    {
+      "epoch": 0.28774128161422585,
+      "grad_norm": 2.608205776232339,
+      "learning_rate": 0.00016714190198329246,
+      "loss": 1.3545,
+      "step": 6246
+    },
+    {
+      "epoch": 0.28778734970286085,
+      "grad_norm": 3.8635928202467578,
+      "learning_rate": 0.00016713084369623425,
+      "loss": 1.4431,
+      "step": 6247
+    },
+    {
+      "epoch": 0.28783341779149585,
+      "grad_norm": 4.764763080427777,
+      "learning_rate": 0.0001671197839146229,
+      "loss": 1.6309,
+      "step": 6248
+    },
+    {
+      "epoch": 0.28787948588013085,
+      "grad_norm": 2.546416535062577,
+      "learning_rate": 0.00016710872263870464,
+      "loss": 1.2023,
+      "step": 6249
+    },
+    {
+      "epoch": 0.28792555396876585,
+      "grad_norm": 2.870454647634091,
+      "learning_rate": 0.00016709765986872577,
+      "loss": 1.4838,
+      "step": 6250
+    },
+    {
+      "epoch": 0.28797162205740084,
+      "grad_norm": 1.6999711725081035,
+      "learning_rate": 0.0001670865956049325,
+      "loss": 1.2624,
+      "step": 6251
+    },
+    {
+      "epoch": 0.28801769014603584,
+      "grad_norm": 1.3918789560346077,
+      "learning_rate": 0.0001670755298475712,
+      "loss": 1.6379,
+      "step": 6252
+    },
+    {
+      "epoch": 0.28806375823467084,
+      "grad_norm": 3.5933803732638605,
+      "learning_rate": 0.00016706446259688828,
+      "loss": 1.6963,
+      "step": 6253
+    },
+    {
+      "epoch": 0.28810982632330584,
+      "grad_norm": 2.4292089262705807,
+      "learning_rate": 0.00016705339385313008,
+      "loss": 1.7238,
+      "step": 6254
+    },
+    {
+      "epoch": 0.28815589441194084,
+      "grad_norm": 2.3372918879382465,
+      "learning_rate": 0.00016704232361654305,
+      "loss": 1.5929,
+      "step": 6255
+    },
+    {
+      "epoch": 0.28820196250057584,
+      "grad_norm": 1.527862469677524,
+      "learning_rate": 0.0001670312518873736,
+      "loss": 1.303,
+      "step": 6256
+    },
+    {
+      "epoch": 0.28824803058921084,
+      "grad_norm": 2.034913611905003,
+      "learning_rate": 0.00016702017866586832,
+      "loss": 1.6264,
+      "step": 6257
+    },
+    {
+      "epoch": 0.28829409867784583,
+      "grad_norm": 2.153046378517714,
+      "learning_rate": 0.00016700910395227366,
+      "loss": 1.3482,
+      "step": 6258
+    },
+    {
+      "epoch": 0.28834016676648083,
+      "grad_norm": 1.8812316220265353,
+      "learning_rate": 0.00016699802774683615,
+      "loss": 1.6754,
+      "step": 6259
+    },
+    {
+      "epoch": 0.2883862348551159,
+      "grad_norm": 2.628826986844056,
+      "learning_rate": 0.00016698695004980246,
+      "loss": 1.4083,
+      "step": 6260
+    },
+    {
+      "epoch": 0.2884323029437509,
+      "grad_norm": 1.0070480541488038,
+      "learning_rate": 0.00016697587086141917,
+      "loss": 1.6136,
+      "step": 6261
+    },
+    {
+      "epoch": 0.2884783710323859,
+      "grad_norm": 1.4248552326946704,
+      "learning_rate": 0.00016696479018193298,
+      "loss": 1.4491,
+      "step": 6262
+    },
+    {
+      "epoch": 0.2885244391210209,
+      "grad_norm": 2.0412151086103107,
+      "learning_rate": 0.00016695370801159053,
+      "loss": 1.3732,
+      "step": 6263
+    },
+    {
+      "epoch": 0.2885705072096559,
+      "grad_norm": 2.048814610524227,
+      "learning_rate": 0.0001669426243506386,
+      "loss": 1.6489,
+      "step": 6264
+    },
+    {
+      "epoch": 0.2886165752982909,
+      "grad_norm": 2.2016225846025925,
+      "learning_rate": 0.0001669315391993239,
+      "loss": 1.3914,
+      "step": 6265
+    },
+    {
+      "epoch": 0.2886626433869259,
+      "grad_norm": 1.6179354020036005,
+      "learning_rate": 0.00016692045255789327,
+      "loss": 1.3885,
+      "step": 6266
+    },
+    {
+      "epoch": 0.2887087114755609,
+      "grad_norm": 1.1711550881559272,
+      "learning_rate": 0.00016690936442659347,
+      "loss": 1.6485,
+      "step": 6267
+    },
+    {
+      "epoch": 0.2887547795641959,
+      "grad_norm": 2.7892709513215284,
+      "learning_rate": 0.00016689827480567142,
+      "loss": 1.2635,
+      "step": 6268
+    },
+    {
+      "epoch": 0.28880084765283087,
+      "grad_norm": 2.573698856029921,
+      "learning_rate": 0.00016688718369537394,
+      "loss": 1.5913,
+      "step": 6269
+    },
+    {
+      "epoch": 0.28884691574146587,
+      "grad_norm": 2.4311776062603734,
+      "learning_rate": 0.00016687609109594805,
+      "loss": 1.2812,
+      "step": 6270
+    },
+    {
+      "epoch": 0.28889298383010087,
+      "grad_norm": 4.797369686186717,
+      "learning_rate": 0.00016686499700764066,
+      "loss": 1.2935,
+      "step": 6271
+    },
+    {
+      "epoch": 0.28893905191873587,
+      "grad_norm": 3.074067974538254,
+      "learning_rate": 0.00016685390143069876,
+      "loss": 1.2713,
+      "step": 6272
+    },
+    {
+      "epoch": 0.2889851200073709,
+      "grad_norm": 1.4399089569455623,
+      "learning_rate": 0.00016684280436536937,
+      "loss": 1.3207,
+      "step": 6273
+    },
+    {
+      "epoch": 0.2890311880960059,
+      "grad_norm": 1.6729726745689155,
+      "learning_rate": 0.00016683170581189956,
+      "loss": 1.7503,
+      "step": 6274
+    },
+    {
+      "epoch": 0.2890772561846409,
+      "grad_norm": 1.8224101689648375,
+      "learning_rate": 0.00016682060577053639,
+      "loss": 1.3419,
+      "step": 6275
+    },
+    {
+      "epoch": 0.2891233242732759,
+      "grad_norm": 3.1441183821294403,
+      "learning_rate": 0.00016680950424152703,
+      "loss": 1.296,
+      "step": 6276
+    },
+    {
+      "epoch": 0.2891693923619109,
+      "grad_norm": 2.0553681614421597,
+      "learning_rate": 0.00016679840122511857,
+      "loss": 1.4289,
+      "step": 6277
+    },
+    {
+      "epoch": 0.2892154604505459,
+      "grad_norm": 1.3408478245706736,
+      "learning_rate": 0.0001667872967215583,
+      "loss": 1.5718,
+      "step": 6278
+    },
+    {
+      "epoch": 0.2892615285391809,
+      "grad_norm": 3.9194886360460126,
+      "learning_rate": 0.00016677619073109332,
+      "loss": 1.6219,
+      "step": 6279
+    },
+    {
+      "epoch": 0.2893075966278159,
+      "grad_norm": 1.3890296504105193,
+      "learning_rate": 0.00016676508325397097,
+      "loss": 1.7082,
+      "step": 6280
+    },
+    {
+      "epoch": 0.2893536647164509,
+      "grad_norm": 3.42688004317192,
+      "learning_rate": 0.00016675397429043856,
+      "loss": 1.4062,
+      "step": 6281
+    },
+    {
+      "epoch": 0.2893997328050859,
+      "grad_norm": 1.3624250611496398,
+      "learning_rate": 0.00016674286384074335,
+      "loss": 1.1878,
+      "step": 6282
+    },
+    {
+      "epoch": 0.2894458008937209,
+      "grad_norm": 1.4053205670193742,
+      "learning_rate": 0.0001667317519051327,
+      "loss": 1.4488,
+      "step": 6283
+    },
+    {
+      "epoch": 0.2894918689823559,
+      "grad_norm": 2.186203774580254,
+      "learning_rate": 0.000166720638483854,
+      "loss": 1.5417,
+      "step": 6284
+    },
+    {
+      "epoch": 0.2895379370709909,
+      "grad_norm": 1.4749058666651969,
+      "learning_rate": 0.00016670952357715471,
+      "loss": 1.782,
+      "step": 6285
+    },
+    {
+      "epoch": 0.2895840051596259,
+      "grad_norm": 1.8873383998036093,
+      "learning_rate": 0.00016669840718528227,
+      "loss": 1.6411,
+      "step": 6286
+    },
+    {
+      "epoch": 0.28963007324826096,
+      "grad_norm": 1.5411742930460235,
+      "learning_rate": 0.00016668728930848415,
+      "loss": 1.1496,
+      "step": 6287
+    },
+    {
+      "epoch": 0.28967614133689595,
+      "grad_norm": 2.037794912780328,
+      "learning_rate": 0.00016667616994700787,
+      "loss": 1.4534,
+      "step": 6288
+    },
+    {
+      "epoch": 0.28972220942553095,
+      "grad_norm": 1.1811415188065406,
+      "learning_rate": 0.00016666504910110098,
+      "loss": 0.9905,
+      "step": 6289
+    },
+    {
+      "epoch": 0.28976827751416595,
+      "grad_norm": 2.019560812682984,
+      "learning_rate": 0.0001666539267710111,
+      "loss": 1.4106,
+      "step": 6290
+    },
+    {
+      "epoch": 0.28981434560280095,
+      "grad_norm": 1.33310180717153,
+      "learning_rate": 0.00016664280295698584,
+      "loss": 1.6212,
+      "step": 6291
+    },
+    {
+      "epoch": 0.28986041369143595,
+      "grad_norm": 1.6998496912785188,
+      "learning_rate": 0.0001666316776592728,
+      "loss": 1.3904,
+      "step": 6292
+    },
+    {
+      "epoch": 0.28990648178007095,
+      "grad_norm": 1.966171634278459,
+      "learning_rate": 0.00016662055087811973,
+      "loss": 1.6037,
+      "step": 6293
+    },
+    {
+      "epoch": 0.28995254986870594,
+      "grad_norm": 2.69627021469552,
+      "learning_rate": 0.00016660942261377432,
+      "loss": 1.4411,
+      "step": 6294
+    },
+    {
+      "epoch": 0.28999861795734094,
+      "grad_norm": 2.226169548300298,
+      "learning_rate": 0.00016659829286648435,
+      "loss": 1.3114,
+      "step": 6295
+    },
+    {
+      "epoch": 0.29004468604597594,
+      "grad_norm": 1.3694646942888569,
+      "learning_rate": 0.00016658716163649753,
+      "loss": 1.3839,
+      "step": 6296
+    },
+    {
+      "epoch": 0.29009075413461094,
+      "grad_norm": 1.791179534751053,
+      "learning_rate": 0.00016657602892406178,
+      "loss": 1.5738,
+      "step": 6297
+    },
+    {
+      "epoch": 0.29013682222324594,
+      "grad_norm": 1.3406411712223445,
+      "learning_rate": 0.00016656489472942484,
+      "loss": 1.5543,
+      "step": 6298
+    },
+    {
+      "epoch": 0.29018289031188094,
+      "grad_norm": 1.6273280060171122,
+      "learning_rate": 0.0001665537590528347,
+      "loss": 1.3034,
+      "step": 6299
+    },
+    {
+      "epoch": 0.29022895840051594,
+      "grad_norm": 1.600233900364597,
+      "learning_rate": 0.00016654262189453922,
+      "loss": 1.3136,
+      "step": 6300
+    },
+    {
+      "epoch": 0.290275026489151,
+      "grad_norm": 1.4111374394426353,
+      "learning_rate": 0.00016653148325478635,
+      "loss": 1.4359,
+      "step": 6301
+    },
+    {
+      "epoch": 0.290321094577786,
+      "grad_norm": 7.920984775014324,
+      "learning_rate": 0.0001665203431338241,
+      "loss": 1.2888,
+      "step": 6302
+    },
+    {
+      "epoch": 0.290367162666421,
+      "grad_norm": 4.787571831207814,
+      "learning_rate": 0.00016650920153190048,
+      "loss": 1.3704,
+      "step": 6303
+    },
+    {
+      "epoch": 0.290413230755056,
+      "grad_norm": 2.2840467336732493,
+      "learning_rate": 0.0001664980584492635,
+      "loss": 1.4188,
+      "step": 6304
+    },
+    {
+      "epoch": 0.290459298843691,
+      "grad_norm": 1.37209156102705,
+      "learning_rate": 0.00016648691388616128,
+      "loss": 1.4754,
+      "step": 6305
+    },
+    {
+      "epoch": 0.290505366932326,
+      "grad_norm": 2.717591933830058,
+      "learning_rate": 0.0001664757678428419,
+      "loss": 1.4871,
+      "step": 6306
+    },
+    {
+      "epoch": 0.290551435020961,
+      "grad_norm": 1.8491304616459612,
+      "learning_rate": 0.00016646462031955355,
+      "loss": 1.4128,
+      "step": 6307
+    },
+    {
+      "epoch": 0.290597503109596,
+      "grad_norm": 2.106095629268656,
+      "learning_rate": 0.0001664534713165444,
+      "loss": 1.2514,
+      "step": 6308
+    },
+    {
+      "epoch": 0.290643571198231,
+      "grad_norm": 1.3915773778840785,
+      "learning_rate": 0.00016644232083406262,
+      "loss": 1.4665,
+      "step": 6309
+    },
+    {
+      "epoch": 0.290689639286866,
+      "grad_norm": 2.064865732669326,
+      "learning_rate": 0.0001664311688723565,
+      "loss": 1.4875,
+      "step": 6310
+    },
+    {
+      "epoch": 0.290735707375501,
+      "grad_norm": 2.0003819156926617,
+      "learning_rate": 0.00016642001543167433,
+      "loss": 1.4479,
+      "step": 6311
+    },
+    {
+      "epoch": 0.29078177546413597,
+      "grad_norm": 1.5504427179596956,
+      "learning_rate": 0.00016640886051226441,
+      "loss": 1.4656,
+      "step": 6312
+    },
+    {
+      "epoch": 0.29082784355277097,
+      "grad_norm": 1.9691410705471304,
+      "learning_rate": 0.00016639770411437508,
+      "loss": 1.4714,
+      "step": 6313
+    },
+    {
+      "epoch": 0.290873911641406,
+      "grad_norm": 2.27711730545444,
+      "learning_rate": 0.0001663865462382547,
+      "loss": 1.4261,
+      "step": 6314
+    },
+    {
+      "epoch": 0.290919979730041,
+      "grad_norm": 6.550612212388112,
+      "learning_rate": 0.0001663753868841517,
+      "loss": 1.2107,
+      "step": 6315
+    },
+    {
+      "epoch": 0.290966047818676,
+      "grad_norm": 2.499497164377872,
+      "learning_rate": 0.00016636422605231446,
+      "loss": 1.6133,
+      "step": 6316
+    },
+    {
+      "epoch": 0.291012115907311,
+      "grad_norm": 2.5242234484276187,
+      "learning_rate": 0.0001663530637429916,
+      "loss": 1.3787,
+      "step": 6317
+    },
+    {
+      "epoch": 0.291058183995946,
+      "grad_norm": 1.998738321755596,
+      "learning_rate": 0.00016634189995643153,
+      "loss": 1.5901,
+      "step": 6318
+    },
+    {
+      "epoch": 0.291104252084581,
+      "grad_norm": 2.211135311500112,
+      "learning_rate": 0.00016633073469288277,
+      "loss": 1.5371,
+      "step": 6319
+    },
+    {
+      "epoch": 0.291150320173216,
+      "grad_norm": 2.5451172367754573,
+      "learning_rate": 0.00016631956795259398,
+      "loss": 1.4821,
+      "step": 6320
+    },
+    {
+      "epoch": 0.291196388261851,
+      "grad_norm": 8.74785074254667,
+      "learning_rate": 0.00016630839973581373,
+      "loss": 1.5882,
+      "step": 6321
+    },
+    {
+      "epoch": 0.291242456350486,
+      "grad_norm": 3.8777380259873944,
+      "learning_rate": 0.0001662972300427906,
+      "loss": 1.6038,
+      "step": 6322
+    },
+    {
+      "epoch": 0.291288524439121,
+      "grad_norm": 4.698636054023189,
+      "learning_rate": 0.00016628605887377334,
+      "loss": 1.3112,
+      "step": 6323
+    },
+    {
+      "epoch": 0.291334592527756,
+      "grad_norm": 5.648130812089391,
+      "learning_rate": 0.00016627488622901067,
+      "loss": 1.2464,
+      "step": 6324
+    },
+    {
+      "epoch": 0.291380660616391,
+      "grad_norm": 2.8888375308612804,
+      "learning_rate": 0.00016626371210875127,
+      "loss": 1.6274,
+      "step": 6325
+    },
+    {
+      "epoch": 0.291426728705026,
+      "grad_norm": 1.9926605825470969,
+      "learning_rate": 0.00016625253651324394,
+      "loss": 1.4828,
+      "step": 6326
+    },
+    {
+      "epoch": 0.291472796793661,
+      "grad_norm": 3.635503891019093,
+      "learning_rate": 0.00016624135944273748,
+      "loss": 1.4214,
+      "step": 6327
+    },
+    {
+      "epoch": 0.29151886488229606,
+      "grad_norm": 3.4760679791457343,
+      "learning_rate": 0.00016623018089748074,
+      "loss": 1.438,
+      "step": 6328
+    },
+    {
+      "epoch": 0.29156493297093106,
+      "grad_norm": 1.5989627345323023,
+      "learning_rate": 0.00016621900087772257,
+      "loss": 1.5685,
+      "step": 6329
+    },
+    {
+      "epoch": 0.29161100105956606,
+      "grad_norm": 2.8223527413800413,
+      "learning_rate": 0.0001662078193837119,
+      "loss": 1.3226,
+      "step": 6330
+    },
+    {
+      "epoch": 0.29165706914820105,
+      "grad_norm": 1.5726722209698913,
+      "learning_rate": 0.0001661966364156976,
+      "loss": 1.6617,
+      "step": 6331
+    },
+    {
+      "epoch": 0.29170313723683605,
+      "grad_norm": 6.453054881796527,
+      "learning_rate": 0.00016618545197392875,
+      "loss": 1.3034,
+      "step": 6332
+    },
+    {
+      "epoch": 0.29174920532547105,
+      "grad_norm": 5.70826864492266,
+      "learning_rate": 0.00016617426605865433,
+      "loss": 1.5166,
+      "step": 6333
+    },
+    {
+      "epoch": 0.29179527341410605,
+      "grad_norm": 2.0214383400296096,
+      "learning_rate": 0.00016616307867012327,
+      "loss": 1.409,
+      "step": 6334
+    },
+    {
+      "epoch": 0.29184134150274105,
+      "grad_norm": 3.1619313632909396,
+      "learning_rate": 0.00016615188980858474,
+      "loss": 1.1623,
+      "step": 6335
+    },
+    {
+      "epoch": 0.29188740959137605,
+      "grad_norm": 2.3455568396891566,
+      "learning_rate": 0.0001661406994742878,
+      "loss": 1.5145,
+      "step": 6336
+    },
+    {
+      "epoch": 0.29193347768001104,
+      "grad_norm": 2.644009383015121,
+      "learning_rate": 0.0001661295076674816,
+      "loss": 1.5156,
+      "step": 6337
+    },
+    {
+      "epoch": 0.29197954576864604,
+      "grad_norm": 2.3779739012936334,
+      "learning_rate": 0.0001661183143884153,
+      "loss": 1.622,
+      "step": 6338
+    },
+    {
+      "epoch": 0.29202561385728104,
+      "grad_norm": 2.287455416330234,
+      "learning_rate": 0.0001661071196373381,
+      "loss": 1.6772,
+      "step": 6339
+    },
+    {
+      "epoch": 0.29207168194591604,
+      "grad_norm": 2.2246000028737174,
+      "learning_rate": 0.00016609592341449922,
+      "loss": 1.2797,
+      "step": 6340
+    },
+    {
+      "epoch": 0.2921177500345511,
+      "grad_norm": 3.178156709920901,
+      "learning_rate": 0.00016608472572014792,
+      "loss": 1.4983,
+      "step": 6341
+    },
+    {
+      "epoch": 0.2921638181231861,
+      "grad_norm": 3.242275163372069,
+      "learning_rate": 0.00016607352655453354,
+      "loss": 1.7687,
+      "step": 6342
+    },
+    {
+      "epoch": 0.2922098862118211,
+      "grad_norm": 2.3837696526872607,
+      "learning_rate": 0.00016606232591790537,
+      "loss": 1.2307,
+      "step": 6343
+    },
+    {
+      "epoch": 0.2922559543004561,
+      "grad_norm": 1.5791046959164796,
+      "learning_rate": 0.00016605112381051283,
+      "loss": 1.6196,
+      "step": 6344
+    },
+    {
+      "epoch": 0.2923020223890911,
+      "grad_norm": 2.1682639050168357,
+      "learning_rate": 0.00016603992023260522,
+      "loss": 1.7345,
+      "step": 6345
+    },
+    {
+      "epoch": 0.2923480904777261,
+      "grad_norm": 1.823220819938701,
+      "learning_rate": 0.00016602871518443202,
+      "loss": 1.3805,
+      "step": 6346
+    },
+    {
+      "epoch": 0.2923941585663611,
+      "grad_norm": 1.7709020055208724,
+      "learning_rate": 0.00016601750866624272,
+      "loss": 1.4596,
+      "step": 6347
+    },
+    {
+      "epoch": 0.2924402266549961,
+      "grad_norm": 1.1994689009860175,
+      "learning_rate": 0.00016600630067828678,
+      "loss": 1.4106,
+      "step": 6348
+    },
+    {
+      "epoch": 0.2924862947436311,
+      "grad_norm": 2.0768898403180924,
+      "learning_rate": 0.00016599509122081372,
+      "loss": 1.302,
+      "step": 6349
+    },
+    {
+      "epoch": 0.2925323628322661,
+      "grad_norm": 1.8788943473972912,
+      "learning_rate": 0.0001659838802940731,
+      "loss": 1.2029,
+      "step": 6350
+    },
+    {
+      "epoch": 0.2925784309209011,
+      "grad_norm": 5.46209205872718,
+      "learning_rate": 0.00016597266789831452,
+      "loss": 1.0414,
+      "step": 6351
+    },
+    {
+      "epoch": 0.2926244990095361,
+      "grad_norm": 2.5035270539174275,
+      "learning_rate": 0.00016596145403378768,
+      "loss": 1.4448,
+      "step": 6352
+    },
+    {
+      "epoch": 0.2926705670981711,
+      "grad_norm": 2.3005838697571352,
+      "learning_rate": 0.00016595023870074206,
+      "loss": 1.4571,
+      "step": 6353
+    },
+    {
+      "epoch": 0.2927166351868061,
+      "grad_norm": 2.8848826949100554,
+      "learning_rate": 0.00016593902189942754,
+      "loss": 1.3406,
+      "step": 6354
+    },
+    {
+      "epoch": 0.2927627032754411,
+      "grad_norm": 1.860891893359195,
+      "learning_rate": 0.0001659278036300937,
+      "loss": 1.3298,
+      "step": 6355
+    },
+    {
+      "epoch": 0.2928087713640761,
+      "grad_norm": 2.203858192887368,
+      "learning_rate": 0.0001659165838929904,
+      "loss": 1.4621,
+      "step": 6356
+    },
+    {
+      "epoch": 0.2928548394527111,
+      "grad_norm": 1.526898519339661,
+      "learning_rate": 0.00016590536268836737,
+      "loss": 1.4694,
+      "step": 6357
+    },
+    {
+      "epoch": 0.2929009075413461,
+      "grad_norm": 65.0954449013405,
+      "learning_rate": 0.00016589414001647444,
+      "loss": 0.9449,
+      "step": 6358
+    },
+    {
+      "epoch": 0.2929469756299811,
+      "grad_norm": 3.9820597053588758,
+      "learning_rate": 0.00016588291587756146,
+      "loss": 1.3933,
+      "step": 6359
+    },
+    {
+      "epoch": 0.2929930437186161,
+      "grad_norm": 3.1211343694976192,
+      "learning_rate": 0.0001658716902718783,
+      "loss": 1.4343,
+      "step": 6360
+    },
+    {
+      "epoch": 0.2930391118072511,
+      "grad_norm": 10.80326672327254,
+      "learning_rate": 0.00016586046319967497,
+      "loss": 2.1525,
+      "step": 6361
+    },
+    {
+      "epoch": 0.2930851798958861,
+      "grad_norm": 3.8479268280856886,
+      "learning_rate": 0.00016584923466120128,
+      "loss": 1.5221,
+      "step": 6362
+    },
+    {
+      "epoch": 0.2931312479845211,
+      "grad_norm": 3.4293912255276644,
+      "learning_rate": 0.0001658380046567073,
+      "loss": 1.9271,
+      "step": 6363
+    },
+    {
+      "epoch": 0.2931773160731561,
+      "grad_norm": 2.6242588502752966,
+      "learning_rate": 0.0001658267731864431,
+      "loss": 1.8766,
+      "step": 6364
+    },
+    {
+      "epoch": 0.2932233841617911,
+      "grad_norm": 2.000991336494442,
+      "learning_rate": 0.00016581554025065856,
+      "loss": 1.4381,
+      "step": 6365
+    },
+    {
+      "epoch": 0.2932694522504261,
+      "grad_norm": 2.7281397499803037,
+      "learning_rate": 0.00016580430584960394,
+      "loss": 1.5623,
+      "step": 6366
+    },
+    {
+      "epoch": 0.2933155203390611,
+      "grad_norm": 1.8250716632332347,
+      "learning_rate": 0.00016579306998352927,
+      "loss": 1.6597,
+      "step": 6367
+    },
+    {
+      "epoch": 0.2933615884276961,
+      "grad_norm": 3.8581878536352345,
+      "learning_rate": 0.00016578183265268472,
+      "loss": 1.865,
+      "step": 6368
+    },
+    {
+      "epoch": 0.29340765651633116,
+      "grad_norm": 3.1058920776137606,
+      "learning_rate": 0.00016577059385732042,
+      "loss": 1.8802,
+      "step": 6369
+    },
+    {
+      "epoch": 0.29345372460496616,
+      "grad_norm": 5.256885245262415,
+      "learning_rate": 0.00016575935359768664,
+      "loss": 1.555,
+      "step": 6370
+    },
+    {
+      "epoch": 0.29349979269360116,
+      "grad_norm": 24.574585846273138,
+      "learning_rate": 0.0001657481118740336,
+      "loss": 1.28,
+      "step": 6371
+    },
+    {
+      "epoch": 0.29354586078223616,
+      "grad_norm": 3.0618053820798776,
+      "learning_rate": 0.0001657368686866116,
+      "loss": 1.2967,
+      "step": 6372
+    },
+    {
+      "epoch": 0.29359192887087115,
+      "grad_norm": 3.399920446582835,
+      "learning_rate": 0.0001657256240356709,
+      "loss": 1.824,
+      "step": 6373
+    },
+    {
+      "epoch": 0.29363799695950615,
+      "grad_norm": 4.30508127020545,
+      "learning_rate": 0.0001657143779214619,
+      "loss": 1.6071,
+      "step": 6374
+    },
+    {
+      "epoch": 0.29368406504814115,
+      "grad_norm": 3.8940666579957193,
+      "learning_rate": 0.00016570313034423495,
+      "loss": 1.1058,
+      "step": 6375
+    },
+    {
+      "epoch": 0.29373013313677615,
+      "grad_norm": 2.757987559008761,
+      "learning_rate": 0.00016569188130424048,
+      "loss": 1.5016,
+      "step": 6376
+    },
+    {
+      "epoch": 0.29377620122541115,
+      "grad_norm": 2.492797190625327,
+      "learning_rate": 0.00016568063080172886,
+      "loss": 1.6535,
+      "step": 6377
+    },
+    {
+      "epoch": 0.29382226931404615,
+      "grad_norm": 3.8748392848149074,
+      "learning_rate": 0.00016566937883695065,
+      "loss": 1.8104,
+      "step": 6378
+    },
+    {
+      "epoch": 0.29386833740268115,
+      "grad_norm": 2.882306289512124,
+      "learning_rate": 0.0001656581254101563,
+      "loss": 1.6303,
+      "step": 6379
+    },
+    {
+      "epoch": 0.29391440549131614,
+      "grad_norm": 3.0539374752247728,
+      "learning_rate": 0.00016564687052159638,
+      "loss": 1.4625,
+      "step": 6380
+    },
+    {
+      "epoch": 0.29396047357995114,
+      "grad_norm": 4.6389694234288195,
+      "learning_rate": 0.00016563561417152144,
+      "loss": 1.9281,
+      "step": 6381
+    },
+    {
+      "epoch": 0.2940065416685862,
+      "grad_norm": 1.8191548307084662,
+      "learning_rate": 0.00016562435636018213,
+      "loss": 1.3073,
+      "step": 6382
+    },
+    {
+      "epoch": 0.2940526097572212,
+      "grad_norm": 8.305513772182776,
+      "learning_rate": 0.000165613097087829,
+      "loss": 1.8524,
+      "step": 6383
+    },
+    {
+      "epoch": 0.2940986778458562,
+      "grad_norm": 5.60726454430462,
+      "learning_rate": 0.00016560183635471276,
+      "loss": 1.4911,
+      "step": 6384
+    },
+    {
+      "epoch": 0.2941447459344912,
+      "grad_norm": 2.55994658777776,
+      "learning_rate": 0.00016559057416108415,
+      "loss": 1.1526,
+      "step": 6385
+    },
+    {
+      "epoch": 0.2941908140231262,
+      "grad_norm": 6.934027950844216,
+      "learning_rate": 0.00016557931050719386,
+      "loss": 1.4633,
+      "step": 6386
+    },
+    {
+      "epoch": 0.2942368821117612,
+      "grad_norm": 2.6382087806463623,
+      "learning_rate": 0.00016556804539329267,
+      "loss": 1.6715,
+      "step": 6387
+    },
+    {
+      "epoch": 0.2942829502003962,
+      "grad_norm": 3.8493046061001874,
+      "learning_rate": 0.00016555677881963137,
+      "loss": 1.9147,
+      "step": 6388
+    },
+    {
+      "epoch": 0.2943290182890312,
+      "grad_norm": 2.2694027169947573,
+      "learning_rate": 0.00016554551078646078,
+      "loss": 1.6297,
+      "step": 6389
+    },
+    {
+      "epoch": 0.2943750863776662,
+      "grad_norm": 2.865653469273243,
+      "learning_rate": 0.0001655342412940318,
+      "loss": 1.7314,
+      "step": 6390
+    },
+    {
+      "epoch": 0.2944211544663012,
+      "grad_norm": 2.2092462757005693,
+      "learning_rate": 0.0001655229703425953,
+      "loss": 1.9531,
+      "step": 6391
+    },
+    {
+      "epoch": 0.2944672225549362,
+      "grad_norm": 1.7566137908475952,
+      "learning_rate": 0.0001655116979324022,
+      "loss": 1.4017,
+      "step": 6392
+    },
+    {
+      "epoch": 0.2945132906435712,
+      "grad_norm": 2.288738136508929,
+      "learning_rate": 0.00016550042406370346,
+      "loss": 1.5102,
+      "step": 6393
+    },
+    {
+      "epoch": 0.2945593587322062,
+      "grad_norm": 1.9839096019552978,
+      "learning_rate": 0.00016548914873675013,
+      "loss": 1.6153,
+      "step": 6394
+    },
+    {
+      "epoch": 0.2946054268208412,
+      "grad_norm": 4.122945986844755,
+      "learning_rate": 0.00016547787195179316,
+      "loss": 1.4362,
+      "step": 6395
+    },
+    {
+      "epoch": 0.29465149490947623,
+      "grad_norm": 1.7316963476345917,
+      "learning_rate": 0.00016546659370908365,
+      "loss": 1.7109,
+      "step": 6396
+    },
+    {
+      "epoch": 0.29469756299811123,
+      "grad_norm": 1.5774315271037134,
+      "learning_rate": 0.00016545531400887268,
+      "loss": 1.5537,
+      "step": 6397
+    },
+    {
+      "epoch": 0.2947436310867462,
+      "grad_norm": 1.4320308575120653,
+      "learning_rate": 0.0001654440328514114,
+      "loss": 1.6087,
+      "step": 6398
+    },
+    {
+      "epoch": 0.2947896991753812,
+      "grad_norm": 3.1777882238984185,
+      "learning_rate": 0.0001654327502369509,
+      "loss": 1.2855,
+      "step": 6399
+    },
+    {
+      "epoch": 0.2948357672640162,
+      "grad_norm": 2.5881340832124886,
+      "learning_rate": 0.00016542146616574247,
+      "loss": 1.7548,
+      "step": 6400
+    },
+    {
+      "epoch": 0.2948818353526512,
+      "grad_norm": 5.918602711933264,
+      "learning_rate": 0.00016541018063803722,
+      "loss": 1.5401,
+      "step": 6401
+    },
+    {
+      "epoch": 0.2949279034412862,
+      "grad_norm": 2.7743810699026636,
+      "learning_rate": 0.00016539889365408645,
+      "loss": 1.6427,
+      "step": 6402
+    },
+    {
+      "epoch": 0.2949739715299212,
+      "grad_norm": 6.704712305480911,
+      "learning_rate": 0.00016538760521414146,
+      "loss": 1.4691,
+      "step": 6403
+    },
+    {
+      "epoch": 0.2950200396185562,
+      "grad_norm": 1.4374038802220397,
+      "learning_rate": 0.0001653763153184536,
+      "loss": 1.5841,
+      "step": 6404
+    },
+    {
+      "epoch": 0.2950661077071912,
+      "grad_norm": 2.6406348888620172,
+      "learning_rate": 0.0001653650239672741,
+      "loss": 1.5752,
+      "step": 6405
+    },
+    {
+      "epoch": 0.2951121757958262,
+      "grad_norm": 5.289047561765292,
+      "learning_rate": 0.00016535373116085446,
+      "loss": 1.3693,
+      "step": 6406
+    },
+    {
+      "epoch": 0.2951582438844612,
+      "grad_norm": 5.547709640642844,
+      "learning_rate": 0.00016534243689944603,
+      "loss": 1.6728,
+      "step": 6407
+    },
+    {
+      "epoch": 0.2952043119730962,
+      "grad_norm": 2.5108564292394524,
+      "learning_rate": 0.00016533114118330027,
+      "loss": 1.5145,
+      "step": 6408
+    },
+    {
+      "epoch": 0.29525038006173127,
+      "grad_norm": 2.494158503942236,
+      "learning_rate": 0.00016531984401266874,
+      "loss": 1.6462,
+      "step": 6409
+    },
+    {
+      "epoch": 0.29529644815036626,
+      "grad_norm": 2.2903718579053085,
+      "learning_rate": 0.00016530854538780281,
+      "loss": 1.5191,
+      "step": 6410
+    },
+    {
+      "epoch": 0.29534251623900126,
+      "grad_norm": 4.346535433427088,
+      "learning_rate": 0.00016529724530895415,
+      "loss": 1.5333,
+      "step": 6411
+    },
+    {
+      "epoch": 0.29538858432763626,
+      "grad_norm": 2.439805247910894,
+      "learning_rate": 0.00016528594377637426,
+      "loss": 1.5196,
+      "step": 6412
+    },
+    {
+      "epoch": 0.29543465241627126,
+      "grad_norm": 2.7788077840443184,
+      "learning_rate": 0.00016527464079031478,
+      "loss": 1.3745,
+      "step": 6413
+    },
+    {
+      "epoch": 0.29548072050490626,
+      "grad_norm": 1.9517811690023017,
+      "learning_rate": 0.00016526333635102732,
+      "loss": 1.3758,
+      "step": 6414
+    },
+    {
+      "epoch": 0.29552678859354126,
+      "grad_norm": 2.9483166484623724,
+      "learning_rate": 0.0001652520304587636,
+      "loss": 1.7288,
+      "step": 6415
+    },
+    {
+      "epoch": 0.29557285668217625,
+      "grad_norm": 3.6254185947907054,
+      "learning_rate": 0.00016524072311377533,
+      "loss": 1.3961,
+      "step": 6416
+    },
+    {
+      "epoch": 0.29561892477081125,
+      "grad_norm": 2.0289401392302224,
+      "learning_rate": 0.00016522941431631417,
+      "loss": 1.0277,
+      "step": 6417
+    },
+    {
+      "epoch": 0.29566499285944625,
+      "grad_norm": 2.1201513238780443,
+      "learning_rate": 0.00016521810406663198,
+      "loss": 1.6306,
+      "step": 6418
+    },
+    {
+      "epoch": 0.29571106094808125,
+      "grad_norm": 2.404680515102675,
+      "learning_rate": 0.0001652067923649805,
+      "loss": 1.8059,
+      "step": 6419
+    },
+    {
+      "epoch": 0.29575712903671625,
+      "grad_norm": 2.582671093692728,
+      "learning_rate": 0.00016519547921161162,
+      "loss": 1.4052,
+      "step": 6420
+    },
+    {
+      "epoch": 0.29580319712535125,
+      "grad_norm": 2.165760338130014,
+      "learning_rate": 0.00016518416460677715,
+      "loss": 1.5284,
+      "step": 6421
+    },
+    {
+      "epoch": 0.29584926521398625,
+      "grad_norm": 1.6141083785579755,
+      "learning_rate": 0.00016517284855072908,
+      "loss": 1.6433,
+      "step": 6422
+    },
+    {
+      "epoch": 0.2958953333026213,
+      "grad_norm": 2.721430179815461,
+      "learning_rate": 0.00016516153104371927,
+      "loss": 1.5343,
+      "step": 6423
+    },
+    {
+      "epoch": 0.2959414013912563,
+      "grad_norm": 1.9327898473193716,
+      "learning_rate": 0.00016515021208599964,
+      "loss": 1.4776,
+      "step": 6424
+    },
+    {
+      "epoch": 0.2959874694798913,
+      "grad_norm": 3.763807783732441,
+      "learning_rate": 0.0001651388916778223,
+      "loss": 1.3679,
+      "step": 6425
+    },
+    {
+      "epoch": 0.2960335375685263,
+      "grad_norm": 2.85487376255956,
+      "learning_rate": 0.0001651275698194392,
+      "loss": 1.5392,
+      "step": 6426
+    },
+    {
+      "epoch": 0.2960796056571613,
+      "grad_norm": 2.6009856564079246,
+      "learning_rate": 0.00016511624651110245,
+      "loss": 1.6691,
+      "step": 6427
+    },
+    {
+      "epoch": 0.2961256737457963,
+      "grad_norm": 8.631145281134922,
+      "learning_rate": 0.0001651049217530641,
+      "loss": 1.121,
+      "step": 6428
+    },
+    {
+      "epoch": 0.2961717418344313,
+      "grad_norm": 3.024086405998065,
+      "learning_rate": 0.00016509359554557631,
+      "loss": 1.7108,
+      "step": 6429
+    },
+    {
+      "epoch": 0.2962178099230663,
+      "grad_norm": 2.1796618295535186,
+      "learning_rate": 0.00016508226788889123,
+      "loss": 1.4582,
+      "step": 6430
+    },
+    {
+      "epoch": 0.2962638780117013,
+      "grad_norm": 2.820296591380975,
+      "learning_rate": 0.00016507093878326105,
+      "loss": 1.6781,
+      "step": 6431
+    },
+    {
+      "epoch": 0.2963099461003363,
+      "grad_norm": 1.7267694705314014,
+      "learning_rate": 0.000165059608228938,
+      "loss": 1.6042,
+      "step": 6432
+    },
+    {
+      "epoch": 0.2963560141889713,
+      "grad_norm": 3.4488195956254377,
+      "learning_rate": 0.00016504827622617428,
+      "loss": 1.5603,
+      "step": 6433
+    },
+    {
+      "epoch": 0.2964020822776063,
+      "grad_norm": 2.1582261279406643,
+      "learning_rate": 0.00016503694277522225,
+      "loss": 1.6136,
+      "step": 6434
+    },
+    {
+      "epoch": 0.2964481503662413,
+      "grad_norm": 2.839446345280465,
+      "learning_rate": 0.00016502560787633424,
+      "loss": 1.4015,
+      "step": 6435
+    },
+    {
+      "epoch": 0.29649421845487633,
+      "grad_norm": 2.4667097308574646,
+      "learning_rate": 0.00016501427152976252,
+      "loss": 1.6454,
+      "step": 6436
+    },
+    {
+      "epoch": 0.29654028654351133,
+      "grad_norm": 2.120930679227614,
+      "learning_rate": 0.00016500293373575953,
+      "loss": 1.4994,
+      "step": 6437
+    },
+    {
+      "epoch": 0.29658635463214633,
+      "grad_norm": 1.6890973594660037,
+      "learning_rate": 0.0001649915944945777,
+      "loss": 1.4667,
+      "step": 6438
+    },
+    {
+      "epoch": 0.29663242272078133,
+      "grad_norm": 4.487366948297501,
+      "learning_rate": 0.00016498025380646942,
+      "loss": 1.6724,
+      "step": 6439
+    },
+    {
+      "epoch": 0.29667849080941633,
+      "grad_norm": 3.9845691402486922,
+      "learning_rate": 0.0001649689116716872,
+      "loss": 1.4678,
+      "step": 6440
+    },
+    {
+      "epoch": 0.2967245588980513,
+      "grad_norm": 2.470898061357801,
+      "learning_rate": 0.00016495756809048362,
+      "loss": 1.2807,
+      "step": 6441
+    },
+    {
+      "epoch": 0.2967706269866863,
+      "grad_norm": 2.9779877178557586,
+      "learning_rate": 0.0001649462230631111,
+      "loss": 1.4706,
+      "step": 6442
+    },
+    {
+      "epoch": 0.2968166950753213,
+      "grad_norm": 1.4796894095176139,
+      "learning_rate": 0.0001649348765898223,
+      "loss": 1.6027,
+      "step": 6443
+    },
+    {
+      "epoch": 0.2968627631639563,
+      "grad_norm": 4.8444527054177,
+      "learning_rate": 0.0001649235286708698,
+      "loss": 1.2733,
+      "step": 6444
+    },
+    {
+      "epoch": 0.2969088312525913,
+      "grad_norm": 1.4714110205059288,
+      "learning_rate": 0.0001649121793065063,
+      "loss": 1.532,
+      "step": 6445
+    },
+    {
+      "epoch": 0.2969548993412263,
+      "grad_norm": 1.6251225475639592,
+      "learning_rate": 0.0001649008284969844,
+      "loss": 1.5607,
+      "step": 6446
+    },
+    {
+      "epoch": 0.2970009674298613,
+      "grad_norm": 2.315184718888174,
+      "learning_rate": 0.00016488947624255685,
+      "loss": 1.479,
+      "step": 6447
+    },
+    {
+      "epoch": 0.2970470355184963,
+      "grad_norm": 2.2831686205535817,
+      "learning_rate": 0.00016487812254347635,
+      "loss": 1.4259,
+      "step": 6448
+    },
+    {
+      "epoch": 0.2970931036071313,
+      "grad_norm": 2.5680030477747042,
+      "learning_rate": 0.00016486676739999574,
+      "loss": 1.4422,
+      "step": 6449
+    },
+    {
+      "epoch": 0.29713917169576637,
+      "grad_norm": 2.613589247422224,
+      "learning_rate": 0.0001648554108123677,
+      "loss": 1.3611,
+      "step": 6450
+    },
+    {
+      "epoch": 0.29718523978440137,
+      "grad_norm": 2.1243520781991623,
+      "learning_rate": 0.00016484405278084521,
+      "loss": 1.1515,
+      "step": 6451
+    },
+    {
+      "epoch": 0.29723130787303637,
+      "grad_norm": 1.5061277735399023,
+      "learning_rate": 0.00016483269330568102,
+      "loss": 1.3794,
+      "step": 6452
+    },
+    {
+      "epoch": 0.29727737596167136,
+      "grad_norm": 3.36817260180854,
+      "learning_rate": 0.0001648213323871281,
+      "loss": 1.6457,
+      "step": 6453
+    },
+    {
+      "epoch": 0.29732344405030636,
+      "grad_norm": 3.2407223612790568,
+      "learning_rate": 0.00016480997002543937,
+      "loss": 1.1479,
+      "step": 6454
+    },
+    {
+      "epoch": 0.29736951213894136,
+      "grad_norm": 4.117032287048286,
+      "learning_rate": 0.00016479860622086777,
+      "loss": 1.7466,
+      "step": 6455
+    },
+    {
+      "epoch": 0.29741558022757636,
+      "grad_norm": 1.3701305589081183,
+      "learning_rate": 0.0001647872409736663,
+      "loss": 1.4978,
+      "step": 6456
+    },
+    {
+      "epoch": 0.29746164831621136,
+      "grad_norm": 3.212737710689118,
+      "learning_rate": 0.000164775874284088,
+      "loss": 1.4806,
+      "step": 6457
+    },
+    {
+      "epoch": 0.29750771640484636,
+      "grad_norm": 1.868446927552912,
+      "learning_rate": 0.00016476450615238595,
+      "loss": 1.6379,
+      "step": 6458
+    },
+    {
+      "epoch": 0.29755378449348135,
+      "grad_norm": 1.8734379490018318,
+      "learning_rate": 0.0001647531365788132,
+      "loss": 1.2654,
+      "step": 6459
+    },
+    {
+      "epoch": 0.29759985258211635,
+      "grad_norm": 2.0267807579087633,
+      "learning_rate": 0.00016474176556362287,
+      "loss": 1.1422,
+      "step": 6460
+    },
+    {
+      "epoch": 0.29764592067075135,
+      "grad_norm": 1.526764693528285,
+      "learning_rate": 0.00016473039310706817,
+      "loss": 1.4026,
+      "step": 6461
+    },
+    {
+      "epoch": 0.29769198875938635,
+      "grad_norm": 2.8332400264139097,
+      "learning_rate": 0.00016471901920940224,
+      "loss": 1.3865,
+      "step": 6462
+    },
+    {
+      "epoch": 0.29773805684802135,
+      "grad_norm": 2.020533544910282,
+      "learning_rate": 0.00016470764387087832,
+      "loss": 1.6351,
+      "step": 6463
+    },
+    {
+      "epoch": 0.2977841249366564,
+      "grad_norm": 1.8583761837500266,
+      "learning_rate": 0.00016469626709174966,
+      "loss": 1.3716,
+      "step": 6464
+    },
+    {
+      "epoch": 0.2978301930252914,
+      "grad_norm": 4.847159279131366,
+      "learning_rate": 0.00016468488887226953,
+      "loss": 1.341,
+      "step": 6465
+    },
+    {
+      "epoch": 0.2978762611139264,
+      "grad_norm": 2.9883268359388833,
+      "learning_rate": 0.0001646735092126913,
+      "loss": 1.2101,
+      "step": 6466
+    },
+    {
+      "epoch": 0.2979223292025614,
+      "grad_norm": 2.112193590195283,
+      "learning_rate": 0.00016466212811326822,
+      "loss": 1.5749,
+      "step": 6467
+    },
+    {
+      "epoch": 0.2979683972911964,
+      "grad_norm": 1.27875078081188,
+      "learning_rate": 0.0001646507455742538,
+      "loss": 1.5858,
+      "step": 6468
+    },
+    {
+      "epoch": 0.2980144653798314,
+      "grad_norm": 1.8353363447825188,
+      "learning_rate": 0.00016463936159590134,
+      "loss": 1.6166,
+      "step": 6469
+    },
+    {
+      "epoch": 0.2980605334684664,
+      "grad_norm": 1.5796148994774584,
+      "learning_rate": 0.00016462797617846432,
+      "loss": 1.3631,
+      "step": 6470
+    },
+    {
+      "epoch": 0.2981066015571014,
+      "grad_norm": 2.0639998788024814,
+      "learning_rate": 0.00016461658932219624,
+      "loss": 1.5176,
+      "step": 6471
+    },
+    {
+      "epoch": 0.2981526696457364,
+      "grad_norm": 3.391972754146885,
+      "learning_rate": 0.00016460520102735057,
+      "loss": 1.3531,
+      "step": 6472
+    },
+    {
+      "epoch": 0.2981987377343714,
+      "grad_norm": 2.1257534971506677,
+      "learning_rate": 0.0001645938112941809,
+      "loss": 1.7546,
+      "step": 6473
+    },
+    {
+      "epoch": 0.2982448058230064,
+      "grad_norm": 1.242952281655563,
+      "learning_rate": 0.00016458242012294075,
+      "loss": 1.251,
+      "step": 6474
+    },
+    {
+      "epoch": 0.2982908739116414,
+      "grad_norm": 2.7002073935906377,
+      "learning_rate": 0.00016457102751388377,
+      "loss": 1.2685,
+      "step": 6475
+    },
+    {
+      "epoch": 0.2983369420002764,
+      "grad_norm": 1.208412475000027,
+      "learning_rate": 0.00016455963346726358,
+      "loss": 1.46,
+      "step": 6476
+    },
+    {
+      "epoch": 0.29838301008891144,
+      "grad_norm": 2.9104946088602395,
+      "learning_rate": 0.00016454823798333384,
+      "loss": 1.314,
+      "step": 6477
+    },
+    {
+      "epoch": 0.29842907817754644,
+      "grad_norm": 2.7380224900987677,
+      "learning_rate": 0.00016453684106234828,
+      "loss": 1.3549,
+      "step": 6478
+    },
+    {
+      "epoch": 0.29847514626618143,
+      "grad_norm": 4.606826236997873,
+      "learning_rate": 0.0001645254427045606,
+      "loss": 1.3357,
+      "step": 6479
+    },
+    {
+      "epoch": 0.29852121435481643,
+      "grad_norm": 1.788464680918033,
+      "learning_rate": 0.00016451404291022454,
+      "loss": 1.1735,
+      "step": 6480
+    },
+    {
+      "epoch": 0.29856728244345143,
+      "grad_norm": 1.9848558754217107,
+      "learning_rate": 0.00016450264167959397,
+      "loss": 1.3653,
+      "step": 6481
+    },
+    {
+      "epoch": 0.29861335053208643,
+      "grad_norm": 2.1484850631825716,
+      "learning_rate": 0.00016449123901292266,
+      "loss": 1.6824,
+      "step": 6482
+    },
+    {
+      "epoch": 0.29865941862072143,
+      "grad_norm": 1.420003075841218,
+      "learning_rate": 0.0001644798349104645,
+      "loss": 1.6284,
+      "step": 6483
+    },
+    {
+      "epoch": 0.2987054867093564,
+      "grad_norm": 2.3760344951563526,
+      "learning_rate": 0.00016446842937247337,
+      "loss": 1.2351,
+      "step": 6484
+    },
+    {
+      "epoch": 0.2987515547979914,
+      "grad_norm": 2.0713559694472408,
+      "learning_rate": 0.00016445702239920322,
+      "loss": 1.3498,
+      "step": 6485
+    },
+    {
+      "epoch": 0.2987976228866264,
+      "grad_norm": 2.4395817152771864,
+      "learning_rate": 0.000164445613990908,
+      "loss": 1.0995,
+      "step": 6486
+    },
+    {
+      "epoch": 0.2988436909752614,
+      "grad_norm": 3.5061410879817303,
+      "learning_rate": 0.00016443420414784162,
+      "loss": 1.0949,
+      "step": 6487
+    },
+    {
+      "epoch": 0.2988897590638964,
+      "grad_norm": 1.3342466448895833,
+      "learning_rate": 0.00016442279287025822,
+      "loss": 1.4071,
+      "step": 6488
+    },
+    {
+      "epoch": 0.2989358271525314,
+      "grad_norm": 1.8524707917384202,
+      "learning_rate": 0.00016441138015841177,
+      "loss": 1.4581,
+      "step": 6489
+    },
+    {
+      "epoch": 0.2989818952411664,
+      "grad_norm": 3.577893201999794,
+      "learning_rate": 0.0001643999660125564,
+      "loss": 1.6408,
+      "step": 6490
+    },
+    {
+      "epoch": 0.29902796332980147,
+      "grad_norm": 1.2552193921279857,
+      "learning_rate": 0.0001643885504329462,
+      "loss": 1.498,
+      "step": 6491
+    },
+    {
+      "epoch": 0.29907403141843647,
+      "grad_norm": 1.7740847015166143,
+      "learning_rate": 0.00016437713341983535,
+      "loss": 1.4166,
+      "step": 6492
+    },
+    {
+      "epoch": 0.29912009950707147,
+      "grad_norm": 1.353255289427286,
+      "learning_rate": 0.00016436571497347797,
+      "loss": 1.4092,
+      "step": 6493
+    },
+    {
+      "epoch": 0.29916616759570647,
+      "grad_norm": 1.3475355365285764,
+      "learning_rate": 0.00016435429509412832,
+      "loss": 1.4031,
+      "step": 6494
+    },
+    {
+      "epoch": 0.29921223568434147,
+      "grad_norm": 2.5467090971217736,
+      "learning_rate": 0.00016434287378204067,
+      "loss": 1.6466,
+      "step": 6495
+    },
+    {
+      "epoch": 0.29925830377297646,
+      "grad_norm": 1.55190433067532,
+      "learning_rate": 0.0001643314510374692,
+      "loss": 1.3789,
+      "step": 6496
+    },
+    {
+      "epoch": 0.29930437186161146,
+      "grad_norm": 2.731356965786842,
+      "learning_rate": 0.00016432002686066834,
+      "loss": 1.387,
+      "step": 6497
+    },
+    {
+      "epoch": 0.29935043995024646,
+      "grad_norm": 2.4712634133557163,
+      "learning_rate": 0.00016430860125189232,
+      "loss": 1.3972,
+      "step": 6498
+    },
+    {
+      "epoch": 0.29939650803888146,
+      "grad_norm": 2.2101430107832076,
+      "learning_rate": 0.0001642971742113956,
+      "loss": 1.3225,
+      "step": 6499
+    },
+    {
+      "epoch": 0.29944257612751646,
+      "grad_norm": 1.3547835071389518,
+      "learning_rate": 0.0001642857457394325,
+      "loss": 1.7445,
+      "step": 6500
+    },
+    {
+      "epoch": 0.29948864421615146,
+      "grad_norm": 1.080745607185839,
+      "learning_rate": 0.0001642743158362575,
+      "loss": 1.5654,
+      "step": 6501
+    },
+    {
+      "epoch": 0.29953471230478645,
+      "grad_norm": 1.710794886418954,
+      "learning_rate": 0.0001642628845021251,
+      "loss": 1.5318,
+      "step": 6502
+    },
+    {
+      "epoch": 0.29958078039342145,
+      "grad_norm": 3.5078463518737864,
+      "learning_rate": 0.00016425145173728974,
+      "loss": 1.2709,
+      "step": 6503
+    },
+    {
+      "epoch": 0.2996268484820565,
+      "grad_norm": 1.169878113967953,
+      "learning_rate": 0.00016424001754200597,
+      "loss": 1.5191,
+      "step": 6504
+    },
+    {
+      "epoch": 0.2996729165706915,
+      "grad_norm": 1.6214509457823854,
+      "learning_rate": 0.00016422858191652838,
+      "loss": 1.5409,
+      "step": 6505
+    },
+    {
+      "epoch": 0.2997189846593265,
+      "grad_norm": 1.9442478705260984,
+      "learning_rate": 0.0001642171448611115,
+      "loss": 1.4066,
+      "step": 6506
+    },
+    {
+      "epoch": 0.2997650527479615,
+      "grad_norm": 1.565228319790689,
+      "learning_rate": 0.00016420570637601002,
+      "loss": 1.1821,
+      "step": 6507
+    },
+    {
+      "epoch": 0.2998111208365965,
+      "grad_norm": 1.8407478020037595,
+      "learning_rate": 0.0001641942664614786,
+      "loss": 1.4656,
+      "step": 6508
+    },
+    {
+      "epoch": 0.2998571889252315,
+      "grad_norm": 1.7700726153293418,
+      "learning_rate": 0.00016418282511777187,
+      "loss": 1.4219,
+      "step": 6509
+    },
+    {
+      "epoch": 0.2999032570138665,
+      "grad_norm": 2.8261683735283976,
+      "learning_rate": 0.00016417138234514462,
+      "loss": 1.4061,
+      "step": 6510
+    },
+    {
+      "epoch": 0.2999493251025015,
+      "grad_norm": 6.3253881766210105,
+      "learning_rate": 0.00016415993814385154,
+      "loss": 1.534,
+      "step": 6511
+    },
+    {
+      "epoch": 0.2999953931911365,
+      "grad_norm": 4.270775088161005,
+      "learning_rate": 0.00016414849251414747,
+      "loss": 1.7197,
+      "step": 6512
+    },
+    {
+      "epoch": 0.3000414612797715,
+      "grad_norm": 1.9514917276452257,
+      "learning_rate": 0.00016413704545628724,
+      "loss": 1.4521,
+      "step": 6513
+    },
+    {
+      "epoch": 0.3000875293684065,
+      "grad_norm": 4.831384281199691,
+      "learning_rate": 0.00016412559697052563,
+      "loss": 1.2668,
+      "step": 6514
+    },
+    {
+      "epoch": 0.3001335974570415,
+      "grad_norm": 2.950282548216879,
+      "learning_rate": 0.00016411414705711755,
+      "loss": 1.3432,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3001796655456765,
+      "grad_norm": 1.067895911921343,
+      "learning_rate": 0.00016410269571631792,
+      "loss": 1.4625,
+      "step": 6516
+    },
+    {
+      "epoch": 0.3002257336343115,
+      "grad_norm": 5.033937556550812,
+      "learning_rate": 0.00016409124294838168,
+      "loss": 1.4955,
+      "step": 6517
+    },
+    {
+      "epoch": 0.30027180172294654,
+      "grad_norm": 2.2681706617741813,
+      "learning_rate": 0.0001640797887535638,
+      "loss": 1.1854,
+      "step": 6518
+    },
+    {
+      "epoch": 0.30031786981158154,
+      "grad_norm": 1.823869302487119,
+      "learning_rate": 0.0001640683331321193,
+      "loss": 1.6223,
+      "step": 6519
+    },
+    {
+      "epoch": 0.30036393790021654,
+      "grad_norm": 1.51684531646713,
+      "learning_rate": 0.00016405687608430324,
+      "loss": 1.5264,
+      "step": 6520
+    },
+    {
+      "epoch": 0.30041000598885154,
+      "grad_norm": 1.2585324642988263,
+      "learning_rate": 0.00016404541761037063,
+      "loss": 1.467,
+      "step": 6521
+    },
+    {
+      "epoch": 0.30045607407748653,
+      "grad_norm": 1.2812302099536046,
+      "learning_rate": 0.00016403395771057665,
+      "loss": 1.5867,
+      "step": 6522
+    },
+    {
+      "epoch": 0.30050214216612153,
+      "grad_norm": 1.8169981207396524,
+      "learning_rate": 0.00016402249638517637,
+      "loss": 1.283,
+      "step": 6523
+    },
+    {
+      "epoch": 0.30054821025475653,
+      "grad_norm": 2.4767255435326403,
+      "learning_rate": 0.00016401103363442504,
+      "loss": 1.437,
+      "step": 6524
+    },
+    {
+      "epoch": 0.30059427834339153,
+      "grad_norm": 1.126878124483625,
+      "learning_rate": 0.00016399956945857772,
+      "loss": 1.3783,
+      "step": 6525
+    },
+    {
+      "epoch": 0.30064034643202653,
+      "grad_norm": 1.65809589995881,
+      "learning_rate": 0.00016398810385788977,
+      "loss": 1.6191,
+      "step": 6526
+    },
+    {
+      "epoch": 0.3006864145206615,
+      "grad_norm": 2.304468699096632,
+      "learning_rate": 0.00016397663683261638,
+      "loss": 1.623,
+      "step": 6527
+    },
+    {
+      "epoch": 0.3007324826092965,
+      "grad_norm": 1.228981426811297,
+      "learning_rate": 0.00016396516838301288,
+      "loss": 1.4332,
+      "step": 6528
+    },
+    {
+      "epoch": 0.3007785506979315,
+      "grad_norm": 0.9163933551262216,
+      "learning_rate": 0.00016395369850933457,
+      "loss": 1.354,
+      "step": 6529
+    },
+    {
+      "epoch": 0.3008246187865665,
+      "grad_norm": 1.1782752608436826,
+      "learning_rate": 0.00016394222721183684,
+      "loss": 1.5197,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3008706868752015,
+      "grad_norm": 2.92011188946204,
+      "learning_rate": 0.00016393075449077504,
+      "loss": 1.2121,
+      "step": 6531
+    },
+    {
+      "epoch": 0.3009167549638366,
+      "grad_norm": 2.107049410625295,
+      "learning_rate": 0.00016391928034640462,
+      "loss": 1.6831,
+      "step": 6532
+    },
+    {
+      "epoch": 0.3009628230524716,
+      "grad_norm": 2.2801321988864762,
+      "learning_rate": 0.00016390780477898103,
+      "loss": 1.6654,
+      "step": 6533
+    },
+    {
+      "epoch": 0.30100889114110657,
+      "grad_norm": 1.766930943713217,
+      "learning_rate": 0.00016389632778875975,
+      "loss": 1.6087,
+      "step": 6534
+    },
+    {
+      "epoch": 0.30105495922974157,
+      "grad_norm": 1.3124929326054602,
+      "learning_rate": 0.00016388484937599626,
+      "loss": 1.4447,
+      "step": 6535
+    },
+    {
+      "epoch": 0.30110102731837657,
+      "grad_norm": 1.9992949547235859,
+      "learning_rate": 0.00016387336954094614,
+      "loss": 1.7149,
+      "step": 6536
+    },
+    {
+      "epoch": 0.30114709540701157,
+      "grad_norm": 1.0848965544196636,
+      "learning_rate": 0.00016386188828386498,
+      "loss": 1.3795,
+      "step": 6537
+    },
+    {
+      "epoch": 0.30119316349564657,
+      "grad_norm": 1.1071086105664816,
+      "learning_rate": 0.00016385040560500837,
+      "loss": 1.2982,
+      "step": 6538
+    },
+    {
+      "epoch": 0.30123923158428156,
+      "grad_norm": 1.415796258332044,
+      "learning_rate": 0.00016383892150463196,
+      "loss": 1.3905,
+      "step": 6539
+    },
+    {
+      "epoch": 0.30128529967291656,
+      "grad_norm": 1.6189465016728743,
+      "learning_rate": 0.00016382743598299145,
+      "loss": 1.4079,
+      "step": 6540
+    },
+    {
+      "epoch": 0.30133136776155156,
+      "grad_norm": 2.101055472776722,
+      "learning_rate": 0.0001638159490403425,
+      "loss": 1.3221,
+      "step": 6541
+    },
+    {
+      "epoch": 0.30137743585018656,
+      "grad_norm": 7.888034892062091,
+      "learning_rate": 0.00016380446067694083,
+      "loss": 1.6813,
+      "step": 6542
+    },
+    {
+      "epoch": 0.30142350393882156,
+      "grad_norm": 1.982729913682359,
+      "learning_rate": 0.00016379297089304232,
+      "loss": 1.2184,
+      "step": 6543
+    },
+    {
+      "epoch": 0.30146957202745656,
+      "grad_norm": 1.575429532472022,
+      "learning_rate": 0.00016378147968890264,
+      "loss": 1.5033,
+      "step": 6544
+    },
+    {
+      "epoch": 0.3015156401160916,
+      "grad_norm": 1.2309500474451418,
+      "learning_rate": 0.00016376998706477767,
+      "loss": 1.2402,
+      "step": 6545
+    },
+    {
+      "epoch": 0.3015617082047266,
+      "grad_norm": 2.98804527434677,
+      "learning_rate": 0.0001637584930209233,
+      "loss": 1.4166,
+      "step": 6546
+    },
+    {
+      "epoch": 0.3016077762933616,
+      "grad_norm": 1.4598178434487616,
+      "learning_rate": 0.0001637469975575954,
+      "loss": 1.4471,
+      "step": 6547
+    },
+    {
+      "epoch": 0.3016538443819966,
+      "grad_norm": 1.4136983613025988,
+      "learning_rate": 0.0001637355006750499,
+      "loss": 1.596,
+      "step": 6548
+    },
+    {
+      "epoch": 0.3016999124706316,
+      "grad_norm": 1.7366762518503056,
+      "learning_rate": 0.00016372400237354276,
+      "loss": 1.2035,
+      "step": 6549
+    },
+    {
+      "epoch": 0.3017459805592666,
+      "grad_norm": 1.851564734203986,
+      "learning_rate": 0.00016371250265332998,
+      "loss": 1.6353,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3017920486479016,
+      "grad_norm": 1.1555180617228096,
+      "learning_rate": 0.00016370100151466755,
+      "loss": 1.4373,
+      "step": 6551
+    },
+    {
+      "epoch": 0.3018381167365366,
+      "grad_norm": 0.9799060716935103,
+      "learning_rate": 0.00016368949895781158,
+      "loss": 1.4644,
+      "step": 6552
+    },
+    {
+      "epoch": 0.3018841848251716,
+      "grad_norm": 0.9925524710587766,
+      "learning_rate": 0.0001636779949830181,
+      "loss": 1.4265,
+      "step": 6553
+    },
+    {
+      "epoch": 0.3019302529138066,
+      "grad_norm": 1.4897699380009397,
+      "learning_rate": 0.00016366648959054324,
+      "loss": 1.4339,
+      "step": 6554
+    },
+    {
+      "epoch": 0.3019763210024416,
+      "grad_norm": 1.1334704955683164,
+      "learning_rate": 0.00016365498278064316,
+      "loss": 1.404,
+      "step": 6555
+    },
+    {
+      "epoch": 0.3020223890910766,
+      "grad_norm": 1.263591713513847,
+      "learning_rate": 0.00016364347455357405,
+      "loss": 1.6934,
+      "step": 6556
+    },
+    {
+      "epoch": 0.3020684571797116,
+      "grad_norm": 1.6639375840253363,
+      "learning_rate": 0.0001636319649095921,
+      "loss": 1.4886,
+      "step": 6557
+    },
+    {
+      "epoch": 0.3021145252683466,
+      "grad_norm": 2.0009535871391875,
+      "learning_rate": 0.00016362045384895354,
+      "loss": 1.5344,
+      "step": 6558
+    },
+    {
+      "epoch": 0.30216059335698164,
+      "grad_norm": 1.435176677145707,
+      "learning_rate": 0.00016360894137191467,
+      "loss": 1.4671,
+      "step": 6559
+    },
+    {
+      "epoch": 0.30220666144561664,
+      "grad_norm": 0.959618259465581,
+      "learning_rate": 0.0001635974274787318,
+      "loss": 1.4547,
+      "step": 6560
+    },
+    {
+      "epoch": 0.30225272953425164,
+      "grad_norm": 1.7394694357068559,
+      "learning_rate": 0.00016358591216966123,
+      "loss": 1.4374,
+      "step": 6561
+    },
+    {
+      "epoch": 0.30229879762288664,
+      "grad_norm": 6.0204298733896096,
+      "learning_rate": 0.00016357439544495935,
+      "loss": 1.4098,
+      "step": 6562
+    },
+    {
+      "epoch": 0.30234486571152164,
+      "grad_norm": 1.4997991408476974,
+      "learning_rate": 0.00016356287730488257,
+      "loss": 1.2447,
+      "step": 6563
+    },
+    {
+      "epoch": 0.30239093380015664,
+      "grad_norm": 2.3039930922272456,
+      "learning_rate": 0.0001635513577496873,
+      "loss": 1.3593,
+      "step": 6564
+    },
+    {
+      "epoch": 0.30243700188879163,
+      "grad_norm": 1.0652103109244093,
+      "learning_rate": 0.00016353983677963005,
+      "loss": 1.3668,
+      "step": 6565
+    },
+    {
+      "epoch": 0.30248306997742663,
+      "grad_norm": 1.3631068067914691,
+      "learning_rate": 0.00016352831439496725,
+      "loss": 1.4868,
+      "step": 6566
+    },
+    {
+      "epoch": 0.30252913806606163,
+      "grad_norm": 2.905753043975761,
+      "learning_rate": 0.00016351679059595545,
+      "loss": 1.2698,
+      "step": 6567
+    },
+    {
+      "epoch": 0.30257520615469663,
+      "grad_norm": 4.98550135547752,
+      "learning_rate": 0.00016350526538285126,
+      "loss": 1.5523,
+      "step": 6568
+    },
+    {
+      "epoch": 0.30262127424333163,
+      "grad_norm": 1.0122210478947846,
+      "learning_rate": 0.0001634937387559112,
+      "loss": 1.1734,
+      "step": 6569
+    },
+    {
+      "epoch": 0.3026673423319666,
+      "grad_norm": 1.4452112694163906,
+      "learning_rate": 0.00016348221071539194,
+      "loss": 1.3072,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3027134104206016,
+      "grad_norm": 1.372214042633048,
+      "learning_rate": 0.00016347068126155008,
+      "loss": 1.649,
+      "step": 6571
+    },
+    {
+      "epoch": 0.3027594785092367,
+      "grad_norm": 1.176920226013284,
+      "learning_rate": 0.00016345915039464232,
+      "loss": 1.5077,
+      "step": 6572
+    },
+    {
+      "epoch": 0.3028055465978717,
+      "grad_norm": 2.124237807666364,
+      "learning_rate": 0.0001634476181149254,
+      "loss": 1.1665,
+      "step": 6573
+    },
+    {
+      "epoch": 0.3028516146865067,
+      "grad_norm": 1.4166484796882164,
+      "learning_rate": 0.00016343608442265603,
+      "loss": 1.3583,
+      "step": 6574
+    },
+    {
+      "epoch": 0.3028976827751417,
+      "grad_norm": 1.4983154570933654,
+      "learning_rate": 0.00016342454931809103,
+      "loss": 1.0452,
+      "step": 6575
+    },
+    {
+      "epoch": 0.3029437508637767,
+      "grad_norm": 1.3586910675987225,
+      "learning_rate": 0.0001634130128014872,
+      "loss": 1.5408,
+      "step": 6576
+    },
+    {
+      "epoch": 0.30298981895241167,
+      "grad_norm": 2.3534953713253746,
+      "learning_rate": 0.00016340147487310138,
+      "loss": 1.3255,
+      "step": 6577
+    },
+    {
+      "epoch": 0.30303588704104667,
+      "grad_norm": 1.2019755353070665,
+      "learning_rate": 0.0001633899355331904,
+      "loss": 1.6928,
+      "step": 6578
+    },
+    {
+      "epoch": 0.30308195512968167,
+      "grad_norm": 2.944995888170451,
+      "learning_rate": 0.0001633783947820112,
+      "loss": 1.4871,
+      "step": 6579
+    },
+    {
+      "epoch": 0.30312802321831667,
+      "grad_norm": 3.7912195383261595,
+      "learning_rate": 0.00016336685261982071,
+      "loss": 1.3239,
+      "step": 6580
+    },
+    {
+      "epoch": 0.30317409130695167,
+      "grad_norm": 1.2433340462884108,
+      "learning_rate": 0.00016335530904687594,
+      "loss": 1.2657,
+      "step": 6581
+    },
+    {
+      "epoch": 0.30322015939558666,
+      "grad_norm": 1.6777464898841945,
+      "learning_rate": 0.0001633437640634338,
+      "loss": 1.295,
+      "step": 6582
+    },
+    {
+      "epoch": 0.30326622748422166,
+      "grad_norm": 1.938978052212263,
+      "learning_rate": 0.00016333221766975136,
+      "loss": 1.2019,
+      "step": 6583
+    },
+    {
+      "epoch": 0.30331229557285666,
+      "grad_norm": 1.201939256604944,
+      "learning_rate": 0.00016332066986608573,
+      "loss": 1.5841,
+      "step": 6584
+    },
+    {
+      "epoch": 0.30335836366149166,
+      "grad_norm": 1.3926852265241783,
+      "learning_rate": 0.0001633091206526939,
+      "loss": 1.4748,
+      "step": 6585
+    },
+    {
+      "epoch": 0.3034044317501267,
+      "grad_norm": 0.9374608519355772,
+      "learning_rate": 0.0001632975700298331,
+      "loss": 1.5314,
+      "step": 6586
+    },
+    {
+      "epoch": 0.3034504998387617,
+      "grad_norm": 1.4758528777478055,
+      "learning_rate": 0.0001632860179977604,
+      "loss": 1.2652,
+      "step": 6587
+    },
+    {
+      "epoch": 0.3034965679273967,
+      "grad_norm": 4.807055778730265,
+      "learning_rate": 0.00016327446455673304,
+      "loss": 1.576,
+      "step": 6588
+    },
+    {
+      "epoch": 0.3035426360160317,
+      "grad_norm": 1.1168218143637758,
+      "learning_rate": 0.0001632629097070082,
+      "loss": 1.5242,
+      "step": 6589
+    },
+    {
+      "epoch": 0.3035887041046667,
+      "grad_norm": 1.7057380607915245,
+      "learning_rate": 0.00016325135344884316,
+      "loss": 1.5163,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3036347721933017,
+      "grad_norm": 2.022046838526156,
+      "learning_rate": 0.00016323979578249518,
+      "loss": 1.4382,
+      "step": 6591
+    },
+    {
+      "epoch": 0.3036808402819367,
+      "grad_norm": 1.307063019253419,
+      "learning_rate": 0.00016322823670822162,
+      "loss": 1.2567,
+      "step": 6592
+    },
+    {
+      "epoch": 0.3037269083705717,
+      "grad_norm": 1.3089939980613932,
+      "learning_rate": 0.00016321667622627972,
+      "loss": 1.4396,
+      "step": 6593
+    },
+    {
+      "epoch": 0.3037729764592067,
+      "grad_norm": 9.116669230003732,
+      "learning_rate": 0.00016320511433692695,
+      "loss": 1.3837,
+      "step": 6594
+    },
+    {
+      "epoch": 0.3038190445478417,
+      "grad_norm": 1.5286230057209347,
+      "learning_rate": 0.00016319355104042068,
+      "loss": 1.4621,
+      "step": 6595
+    },
+    {
+      "epoch": 0.3038651126364767,
+      "grad_norm": 1.4830637263540998,
+      "learning_rate": 0.0001631819863370184,
+      "loss": 1.6016,
+      "step": 6596
+    },
+    {
+      "epoch": 0.3039111807251117,
+      "grad_norm": 1.156335874949823,
+      "learning_rate": 0.00016317042022697745,
+      "loss": 1.5683,
+      "step": 6597
+    },
+    {
+      "epoch": 0.3039572488137467,
+      "grad_norm": 1.5364143432849993,
+      "learning_rate": 0.00016315885271055543,
+      "loss": 1.5098,
+      "step": 6598
+    },
+    {
+      "epoch": 0.3040033169023817,
+      "grad_norm": 1.4434887426268967,
+      "learning_rate": 0.00016314728378800987,
+      "loss": 1.266,
+      "step": 6599
+    },
+    {
+      "epoch": 0.30404938499101675,
+      "grad_norm": 1.9445076747985552,
+      "learning_rate": 0.00016313571345959827,
+      "loss": 1.3853,
+      "step": 6600
+    },
+    {
+      "epoch": 0.30409545307965175,
+      "grad_norm": 1.7018890570795009,
+      "learning_rate": 0.0001631241417255783,
+      "loss": 1.6097,
+      "step": 6601
+    },
+    {
+      "epoch": 0.30414152116828674,
+      "grad_norm": 1.814918688120808,
+      "learning_rate": 0.00016311256858620756,
+      "loss": 1.363,
+      "step": 6602
+    },
+    {
+      "epoch": 0.30418758925692174,
+      "grad_norm": 1.3338145111078907,
+      "learning_rate": 0.00016310099404174365,
+      "loss": 1.5609,
+      "step": 6603
+    },
+    {
+      "epoch": 0.30423365734555674,
+      "grad_norm": 1.1436928164945646,
+      "learning_rate": 0.00016308941809244437,
+      "loss": 1.501,
+      "step": 6604
+    },
+    {
+      "epoch": 0.30427972543419174,
+      "grad_norm": 2.4828757959854317,
+      "learning_rate": 0.0001630778407385673,
+      "loss": 1.3721,
+      "step": 6605
+    },
+    {
+      "epoch": 0.30432579352282674,
+      "grad_norm": 1.4658932255779225,
+      "learning_rate": 0.00016306626198037033,
+      "loss": 1.4605,
+      "step": 6606
+    },
+    {
+      "epoch": 0.30437186161146174,
+      "grad_norm": 1.1410486481546487,
+      "learning_rate": 0.00016305468181811113,
+      "loss": 1.4769,
+      "step": 6607
+    },
+    {
+      "epoch": 0.30441792970009673,
+      "grad_norm": 4.886355431004025,
+      "learning_rate": 0.00016304310025204755,
+      "loss": 1.4802,
+      "step": 6608
+    },
+    {
+      "epoch": 0.30446399778873173,
+      "grad_norm": 1.689534486611692,
+      "learning_rate": 0.00016303151728243743,
+      "loss": 1.3268,
+      "step": 6609
+    },
+    {
+      "epoch": 0.30451006587736673,
+      "grad_norm": 2.0249113897940614,
+      "learning_rate": 0.0001630199329095387,
+      "loss": 1.2713,
+      "step": 6610
+    },
+    {
+      "epoch": 0.30455613396600173,
+      "grad_norm": 1.4379029791226816,
+      "learning_rate": 0.00016300834713360918,
+      "loss": 1.0918,
+      "step": 6611
+    },
+    {
+      "epoch": 0.30460220205463673,
+      "grad_norm": 1.1787010754714788,
+      "learning_rate": 0.00016299675995490683,
+      "loss": 1.4642,
+      "step": 6612
+    },
+    {
+      "epoch": 0.3046482701432718,
+      "grad_norm": 1.162211645229413,
+      "learning_rate": 0.00016298517137368967,
+      "loss": 1.391,
+      "step": 6613
+    },
+    {
+      "epoch": 0.3046943382319068,
+      "grad_norm": 1.7306351806969595,
+      "learning_rate": 0.00016297358139021571,
+      "loss": 1.3534,
+      "step": 6614
+    },
+    {
+      "epoch": 0.3047404063205418,
+      "grad_norm": 1.4758494716967994,
+      "learning_rate": 0.00016296199000474286,
+      "loss": 1.6996,
+      "step": 6615
+    },
+    {
+      "epoch": 0.3047864744091768,
+      "grad_norm": 1.4822058388561878,
+      "learning_rate": 0.0001629503972175293,
+      "loss": 1.1114,
+      "step": 6616
+    },
+    {
+      "epoch": 0.3048325424978118,
+      "grad_norm": 0.9738147762868584,
+      "learning_rate": 0.00016293880302883308,
+      "loss": 1.2257,
+      "step": 6617
+    },
+    {
+      "epoch": 0.3048786105864468,
+      "grad_norm": 1.1849171176218702,
+      "learning_rate": 0.0001629272074389123,
+      "loss": 1.7184,
+      "step": 6618
+    },
+    {
+      "epoch": 0.3049246786750818,
+      "grad_norm": 1.5955493795119642,
+      "learning_rate": 0.00016291561044802516,
+      "loss": 1.1963,
+      "step": 6619
+    },
+    {
+      "epoch": 0.30497074676371677,
+      "grad_norm": 1.6557205706457454,
+      "learning_rate": 0.0001629040120564298,
+      "loss": 1.1982,
+      "step": 6620
+    },
+    {
+      "epoch": 0.30501681485235177,
+      "grad_norm": 1.8367269616267292,
+      "learning_rate": 0.0001628924122643845,
+      "loss": 1.5699,
+      "step": 6621
+    },
+    {
+      "epoch": 0.30506288294098677,
+      "grad_norm": 1.5136920846882391,
+      "learning_rate": 0.0001628808110721475,
+      "loss": 1.7201,
+      "step": 6622
+    },
+    {
+      "epoch": 0.30510895102962177,
+      "grad_norm": 1.1353259975570151,
+      "learning_rate": 0.00016286920847997704,
+      "loss": 1.5581,
+      "step": 6623
+    },
+    {
+      "epoch": 0.30515501911825677,
+      "grad_norm": 2.0875572625013734,
+      "learning_rate": 0.00016285760448813146,
+      "loss": 1.4519,
+      "step": 6624
+    },
+    {
+      "epoch": 0.30520108720689176,
+      "grad_norm": 1.0229256288493458,
+      "learning_rate": 0.00016284599909686907,
+      "loss": 1.2845,
+      "step": 6625
+    },
+    {
+      "epoch": 0.30524715529552676,
+      "grad_norm": 1.7120462095353726,
+      "learning_rate": 0.00016283439230644828,
+      "loss": 1.7467,
+      "step": 6626
+    },
+    {
+      "epoch": 0.3052932233841618,
+      "grad_norm": 1.3139916620156724,
+      "learning_rate": 0.00016282278411712748,
+      "loss": 1.4032,
+      "step": 6627
+    },
+    {
+      "epoch": 0.3053392914727968,
+      "grad_norm": 2.0569769224812138,
+      "learning_rate": 0.0001628111745291651,
+      "loss": 1.2782,
+      "step": 6628
+    },
+    {
+      "epoch": 0.3053853595614318,
+      "grad_norm": 1.7354817648739456,
+      "learning_rate": 0.00016279956354281963,
+      "loss": 1.3748,
+      "step": 6629
+    },
+    {
+      "epoch": 0.3054314276500668,
+      "grad_norm": 1.2778276964183874,
+      "learning_rate": 0.00016278795115834958,
+      "loss": 1.3653,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3054774957387018,
+      "grad_norm": 1.0953548880411441,
+      "learning_rate": 0.00016277633737601339,
+      "loss": 1.3779,
+      "step": 6631
+    },
+    {
+      "epoch": 0.3055235638273368,
+      "grad_norm": 1.3742005758551157,
+      "learning_rate": 0.00016276472219606974,
+      "loss": 1.4796,
+      "step": 6632
+    },
+    {
+      "epoch": 0.3055696319159718,
+      "grad_norm": 1.5606370057734877,
+      "learning_rate": 0.00016275310561877717,
+      "loss": 1.296,
+      "step": 6633
+    },
+    {
+      "epoch": 0.3056157000046068,
+      "grad_norm": 1.8272158115079105,
+      "learning_rate": 0.00016274148764439429,
+      "loss": 1.4126,
+      "step": 6634
+    },
+    {
+      "epoch": 0.3056617680932418,
+      "grad_norm": 1.5904750115458512,
+      "learning_rate": 0.00016272986827317975,
+      "loss": 1.3038,
+      "step": 6635
+    },
+    {
+      "epoch": 0.3057078361818768,
+      "grad_norm": 1.3058174589396525,
+      "learning_rate": 0.00016271824750539228,
+      "loss": 1.2787,
+      "step": 6636
+    },
+    {
+      "epoch": 0.3057539042705118,
+      "grad_norm": 1.7070456241234193,
+      "learning_rate": 0.00016270662534129056,
+      "loss": 1.4991,
+      "step": 6637
+    },
+    {
+      "epoch": 0.3057999723591468,
+      "grad_norm": 1.693283919613042,
+      "learning_rate": 0.00016269500178113332,
+      "loss": 1.5751,
+      "step": 6638
+    },
+    {
+      "epoch": 0.3058460404477818,
+      "grad_norm": 1.3990508882691195,
+      "learning_rate": 0.00016268337682517938,
+      "loss": 1.3117,
+      "step": 6639
+    },
+    {
+      "epoch": 0.30589210853641685,
+      "grad_norm": 1.6696872218153318,
+      "learning_rate": 0.00016267175047368754,
+      "loss": 1.4216,
+      "step": 6640
+    },
+    {
+      "epoch": 0.30593817662505185,
+      "grad_norm": 1.7063945676346457,
+      "learning_rate": 0.00016266012272691665,
+      "loss": 1.3611,
+      "step": 6641
+    },
+    {
+      "epoch": 0.30598424471368685,
+      "grad_norm": 2.18405840942332,
+      "learning_rate": 0.00016264849358512554,
+      "loss": 1.281,
+      "step": 6642
+    },
+    {
+      "epoch": 0.30603031280232185,
+      "grad_norm": 1.3999715665041896,
+      "learning_rate": 0.00016263686304857314,
+      "loss": 1.0147,
+      "step": 6643
+    },
+    {
+      "epoch": 0.30607638089095685,
+      "grad_norm": 1.2237051244532593,
+      "learning_rate": 0.0001626252311175184,
+      "loss": 1.616,
+      "step": 6644
+    },
+    {
+      "epoch": 0.30612244897959184,
+      "grad_norm": 2.9104557783247436,
+      "learning_rate": 0.00016261359779222025,
+      "loss": 1.5028,
+      "step": 6645
+    },
+    {
+      "epoch": 0.30616851706822684,
+      "grad_norm": 1.3151065603174537,
+      "learning_rate": 0.0001626019630729377,
+      "loss": 1.584,
+      "step": 6646
+    },
+    {
+      "epoch": 0.30621458515686184,
+      "grad_norm": 1.914694438298577,
+      "learning_rate": 0.0001625903269599298,
+      "loss": 1.2317,
+      "step": 6647
+    },
+    {
+      "epoch": 0.30626065324549684,
+      "grad_norm": 1.4029410521602192,
+      "learning_rate": 0.0001625786894534556,
+      "loss": 1.6768,
+      "step": 6648
+    },
+    {
+      "epoch": 0.30630672133413184,
+      "grad_norm": 2.282316136716876,
+      "learning_rate": 0.00016256705055377416,
+      "loss": 1.4438,
+      "step": 6649
+    },
+    {
+      "epoch": 0.30635278942276684,
+      "grad_norm": 2.2924330122734227,
+      "learning_rate": 0.00016255541026114463,
+      "loss": 1.3198,
+      "step": 6650
+    },
+    {
+      "epoch": 0.30639885751140183,
+      "grad_norm": 2.379293502367048,
+      "learning_rate": 0.00016254376857582616,
+      "loss": 1.1604,
+      "step": 6651
+    },
+    {
+      "epoch": 0.30644492560003683,
+      "grad_norm": 1.0428126228077532,
+      "learning_rate": 0.0001625321254980779,
+      "loss": 1.6935,
+      "step": 6652
+    },
+    {
+      "epoch": 0.30649099368867183,
+      "grad_norm": 2.389182244482783,
+      "learning_rate": 0.0001625204810281591,
+      "loss": 1.3549,
+      "step": 6653
+    },
+    {
+      "epoch": 0.3065370617773069,
+      "grad_norm": 1.1901733046347949,
+      "learning_rate": 0.00016250883516632895,
+      "loss": 1.4124,
+      "step": 6654
+    },
+    {
+      "epoch": 0.3065831298659419,
+      "grad_norm": 2.1638159290364563,
+      "learning_rate": 0.0001624971879128468,
+      "loss": 1.3946,
+      "step": 6655
+    },
+    {
+      "epoch": 0.3066291979545769,
+      "grad_norm": 1.5454945507120672,
+      "learning_rate": 0.0001624855392679719,
+      "loss": 1.6114,
+      "step": 6656
+    },
+    {
+      "epoch": 0.3066752660432119,
+      "grad_norm": 1.4219660515312156,
+      "learning_rate": 0.00016247388923196364,
+      "loss": 1.3178,
+      "step": 6657
+    },
+    {
+      "epoch": 0.3067213341318469,
+      "grad_norm": 1.7992592550906839,
+      "learning_rate": 0.00016246223780508134,
+      "loss": 1.5044,
+      "step": 6658
+    },
+    {
+      "epoch": 0.3067674022204819,
+      "grad_norm": 1.542918611659086,
+      "learning_rate": 0.0001624505849875844,
+      "loss": 1.3439,
+      "step": 6659
+    },
+    {
+      "epoch": 0.3068134703091169,
+      "grad_norm": 2.035183515045591,
+      "learning_rate": 0.0001624389307797323,
+      "loss": 1.5427,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3068595383977519,
+      "grad_norm": 1.398612933622813,
+      "learning_rate": 0.00016242727518178443,
+      "loss": 1.2665,
+      "step": 6661
+    },
+    {
+      "epoch": 0.3069056064863869,
+      "grad_norm": 1.0752622503033422,
+      "learning_rate": 0.00016241561819400034,
+      "loss": 1.4717,
+      "step": 6662
+    },
+    {
+      "epoch": 0.30695167457502187,
+      "grad_norm": 3.6920574698489412,
+      "learning_rate": 0.00016240395981663953,
+      "loss": 1.54,
+      "step": 6663
+    },
+    {
+      "epoch": 0.30699774266365687,
+      "grad_norm": 1.311728693972538,
+      "learning_rate": 0.00016239230004996155,
+      "loss": 1.3379,
+      "step": 6664
+    },
+    {
+      "epoch": 0.30704381075229187,
+      "grad_norm": 1.3913825847631298,
+      "learning_rate": 0.000162380638894226,
+      "loss": 1.3863,
+      "step": 6665
+    },
+    {
+      "epoch": 0.30708987884092687,
+      "grad_norm": 1.5073180809908786,
+      "learning_rate": 0.00016236897634969245,
+      "loss": 1.3978,
+      "step": 6666
+    },
+    {
+      "epoch": 0.30713594692956186,
+      "grad_norm": 1.4969207720064777,
+      "learning_rate": 0.0001623573124166206,
+      "loss": 1.585,
+      "step": 6667
+    },
+    {
+      "epoch": 0.3071820150181969,
+      "grad_norm": 1.574890620744409,
+      "learning_rate": 0.00016234564709527013,
+      "loss": 1.2463,
+      "step": 6668
+    },
+    {
+      "epoch": 0.3072280831068319,
+      "grad_norm": 1.6934505449660433,
+      "learning_rate": 0.00016233398038590074,
+      "loss": 1.4538,
+      "step": 6669
+    },
+    {
+      "epoch": 0.3072741511954669,
+      "grad_norm": 7.149721820534755,
+      "learning_rate": 0.00016232231228877213,
+      "loss": 1.3758,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3073202192841019,
+      "grad_norm": 1.5927656795594842,
+      "learning_rate": 0.00016231064280414407,
+      "loss": 1.5079,
+      "step": 6671
+    },
+    {
+      "epoch": 0.3073662873727369,
+      "grad_norm": 1.3750022171919951,
+      "learning_rate": 0.0001622989719322764,
+      "loss": 1.1968,
+      "step": 6672
+    },
+    {
+      "epoch": 0.3074123554613719,
+      "grad_norm": 1.8140880166934792,
+      "learning_rate": 0.00016228729967342898,
+      "loss": 1.2983,
+      "step": 6673
+    },
+    {
+      "epoch": 0.3074584235500069,
+      "grad_norm": 1.1389889089275425,
+      "learning_rate": 0.0001622756260278616,
+      "loss": 1.6533,
+      "step": 6674
+    },
+    {
+      "epoch": 0.3075044916386419,
+      "grad_norm": 2.82823435892532,
+      "learning_rate": 0.0001622639509958342,
+      "loss": 1.4474,
+      "step": 6675
+    },
+    {
+      "epoch": 0.3075505597272769,
+      "grad_norm": 2.942218892363544,
+      "learning_rate": 0.00016225227457760665,
+      "loss": 1.0683,
+      "step": 6676
+    },
+    {
+      "epoch": 0.3075966278159119,
+      "grad_norm": 0.9533783373418044,
+      "learning_rate": 0.00016224059677343898,
+      "loss": 1.3589,
+      "step": 6677
+    },
+    {
+      "epoch": 0.3076426959045469,
+      "grad_norm": 1.1010568622978019,
+      "learning_rate": 0.00016222891758359116,
+      "loss": 1.414,
+      "step": 6678
+    },
+    {
+      "epoch": 0.3076887639931819,
+      "grad_norm": 1.9773911968653284,
+      "learning_rate": 0.00016221723700832316,
+      "loss": 1.6305,
+      "step": 6679
+    },
+    {
+      "epoch": 0.3077348320818169,
+      "grad_norm": 3.1937146686789486,
+      "learning_rate": 0.00016220555504789508,
+      "loss": 1.4416,
+      "step": 6680
+    },
+    {
+      "epoch": 0.30778090017045195,
+      "grad_norm": 1.8935953784164228,
+      "learning_rate": 0.00016219387170256696,
+      "loss": 1.2799,
+      "step": 6681
+    },
+    {
+      "epoch": 0.30782696825908695,
+      "grad_norm": 2.0429691584964202,
+      "learning_rate": 0.00016218218697259892,
+      "loss": 1.2963,
+      "step": 6682
+    },
+    {
+      "epoch": 0.30787303634772195,
+      "grad_norm": 2.3908965409734377,
+      "learning_rate": 0.00016217050085825111,
+      "loss": 1.1558,
+      "step": 6683
+    },
+    {
+      "epoch": 0.30791910443635695,
+      "grad_norm": 5.074342550391236,
+      "learning_rate": 0.00016215881335978371,
+      "loss": 1.6014,
+      "step": 6684
+    },
+    {
+      "epoch": 0.30796517252499195,
+      "grad_norm": 1.2342759645756396,
+      "learning_rate": 0.00016214712447745692,
+      "loss": 1.4944,
+      "step": 6685
+    },
+    {
+      "epoch": 0.30801124061362695,
+      "grad_norm": 2.4828698223288463,
+      "learning_rate": 0.00016213543421153093,
+      "loss": 1.3941,
+      "step": 6686
+    },
+    {
+      "epoch": 0.30805730870226194,
+      "grad_norm": 1.4920207740453582,
+      "learning_rate": 0.00016212374256226606,
+      "loss": 1.3897,
+      "step": 6687
+    },
+    {
+      "epoch": 0.30810337679089694,
+      "grad_norm": 2.5364679997571082,
+      "learning_rate": 0.00016211204952992257,
+      "loss": 1.3431,
+      "step": 6688
+    },
+    {
+      "epoch": 0.30814944487953194,
+      "grad_norm": 2.6361414887210324,
+      "learning_rate": 0.0001621003551147608,
+      "loss": 1.3003,
+      "step": 6689
+    },
+    {
+      "epoch": 0.30819551296816694,
+      "grad_norm": 1.8126664872713438,
+      "learning_rate": 0.0001620886593170411,
+      "loss": 1.1016,
+      "step": 6690
+    },
+    {
+      "epoch": 0.30824158105680194,
+      "grad_norm": 1.068392054446483,
+      "learning_rate": 0.00016207696213702387,
+      "loss": 1.3146,
+      "step": 6691
+    },
+    {
+      "epoch": 0.30828764914543694,
+      "grad_norm": 2.023813969569466,
+      "learning_rate": 0.00016206526357496955,
+      "loss": 1.5206,
+      "step": 6692
+    },
+    {
+      "epoch": 0.30833371723407194,
+      "grad_norm": 1.9564256346321038,
+      "learning_rate": 0.00016205356363113848,
+      "loss": 1.4622,
+      "step": 6693
+    },
+    {
+      "epoch": 0.30837978532270693,
+      "grad_norm": 2.072876837946252,
+      "learning_rate": 0.00016204186230579125,
+      "loss": 1.4703,
+      "step": 6694
+    },
+    {
+      "epoch": 0.308425853411342,
+      "grad_norm": 1.8733495307936465,
+      "learning_rate": 0.00016203015959918835,
+      "loss": 1.4674,
+      "step": 6695
+    },
+    {
+      "epoch": 0.308471921499977,
+      "grad_norm": 2.4379151921112534,
+      "learning_rate": 0.00016201845551159033,
+      "loss": 1.5088,
+      "step": 6696
+    },
+    {
+      "epoch": 0.308517989588612,
+      "grad_norm": 2.375181172845558,
+      "learning_rate": 0.00016200675004325767,
+      "loss": 1.3783,
+      "step": 6697
+    },
+    {
+      "epoch": 0.308564057677247,
+      "grad_norm": 1.279468487540822,
+      "learning_rate": 0.00016199504319445106,
+      "loss": 1.4918,
+      "step": 6698
+    },
+    {
+      "epoch": 0.308610125765882,
+      "grad_norm": 1.7303230916340515,
+      "learning_rate": 0.0001619833349654311,
+      "loss": 1.0455,
+      "step": 6699
+    },
+    {
+      "epoch": 0.308656193854517,
+      "grad_norm": 1.3957697979973391,
+      "learning_rate": 0.00016197162535645852,
+      "loss": 1.3522,
+      "step": 6700
+    },
+    {
+      "epoch": 0.308702261943152,
+      "grad_norm": 1.0481303946656484,
+      "learning_rate": 0.00016195991436779392,
+      "loss": 1.3214,
+      "step": 6701
+    },
+    {
+      "epoch": 0.308748330031787,
+      "grad_norm": 2.1275864660326156,
+      "learning_rate": 0.00016194820199969806,
+      "loss": 1.5874,
+      "step": 6702
+    },
+    {
+      "epoch": 0.308794398120422,
+      "grad_norm": 5.717206750622285,
+      "learning_rate": 0.00016193648825243173,
+      "loss": 1.4047,
+      "step": 6703
+    },
+    {
+      "epoch": 0.308840466209057,
+      "grad_norm": 0.9721606578375198,
+      "learning_rate": 0.00016192477312625565,
+      "loss": 1.5304,
+      "step": 6704
+    },
+    {
+      "epoch": 0.308886534297692,
+      "grad_norm": 1.3090585015822485,
+      "learning_rate": 0.0001619130566214307,
+      "loss": 1.4979,
+      "step": 6705
+    },
+    {
+      "epoch": 0.30893260238632697,
+      "grad_norm": 1.3396684317307763,
+      "learning_rate": 0.0001619013387382177,
+      "loss": 1.3966,
+      "step": 6706
+    },
+    {
+      "epoch": 0.30897867047496197,
+      "grad_norm": 3.0842981624925745,
+      "learning_rate": 0.00016188961947687753,
+      "loss": 1.2615,
+      "step": 6707
+    },
+    {
+      "epoch": 0.309024738563597,
+      "grad_norm": 1.0920679635807984,
+      "learning_rate": 0.00016187789883767112,
+      "loss": 1.2418,
+      "step": 6708
+    },
+    {
+      "epoch": 0.309070806652232,
+      "grad_norm": 1.8934330312325174,
+      "learning_rate": 0.00016186617682085934,
+      "loss": 1.3765,
+      "step": 6709
+    },
+    {
+      "epoch": 0.309116874740867,
+      "grad_norm": 1.4382586081461892,
+      "learning_rate": 0.00016185445342670323,
+      "loss": 1.1652,
+      "step": 6710
+    },
+    {
+      "epoch": 0.309162942829502,
+      "grad_norm": 1.3178607562643871,
+      "learning_rate": 0.0001618427286554638,
+      "loss": 1.5179,
+      "step": 6711
+    },
+    {
+      "epoch": 0.309209010918137,
+      "grad_norm": 4.151987021645118,
+      "learning_rate": 0.00016183100250740203,
+      "loss": 1.1071,
+      "step": 6712
+    },
+    {
+      "epoch": 0.309255079006772,
+      "grad_norm": 1.2416575909215992,
+      "learning_rate": 0.000161819274982779,
+      "loss": 1.4336,
+      "step": 6713
+    },
+    {
+      "epoch": 0.309301147095407,
+      "grad_norm": 2.903729246648054,
+      "learning_rate": 0.00016180754608185582,
+      "loss": 1.4388,
+      "step": 6714
+    },
+    {
+      "epoch": 0.309347215184042,
+      "grad_norm": 1.797537395486855,
+      "learning_rate": 0.00016179581580489362,
+      "loss": 1.5187,
+      "step": 6715
+    },
+    {
+      "epoch": 0.309393283272677,
+      "grad_norm": 1.553478265071512,
+      "learning_rate": 0.00016178408415215353,
+      "loss": 1.2723,
+      "step": 6716
+    },
+    {
+      "epoch": 0.309439351361312,
+      "grad_norm": 1.935472445292353,
+      "learning_rate": 0.00016177235112389672,
+      "loss": 1.2609,
+      "step": 6717
+    },
+    {
+      "epoch": 0.309485419449947,
+      "grad_norm": 2.724675763913465,
+      "learning_rate": 0.00016176061672038445,
+      "loss": 1.5299,
+      "step": 6718
+    },
+    {
+      "epoch": 0.309531487538582,
+      "grad_norm": 1.4731607802430013,
+      "learning_rate": 0.00016174888094187795,
+      "loss": 1.512,
+      "step": 6719
+    },
+    {
+      "epoch": 0.309577555627217,
+      "grad_norm": 6.030948671350595,
+      "learning_rate": 0.00016173714378863848,
+      "loss": 1.418,
+      "step": 6720
+    },
+    {
+      "epoch": 0.309623623715852,
+      "grad_norm": 1.852446743890169,
+      "learning_rate": 0.00016172540526092734,
+      "loss": 1.3943,
+      "step": 6721
+    },
+    {
+      "epoch": 0.30966969180448706,
+      "grad_norm": 0.9235299883974978,
+      "learning_rate": 0.00016171366535900594,
+      "loss": 1.5396,
+      "step": 6722
+    },
+    {
+      "epoch": 0.30971575989312206,
+      "grad_norm": 2.2966970587629314,
+      "learning_rate": 0.00016170192408313555,
+      "loss": 1.5327,
+      "step": 6723
+    },
+    {
+      "epoch": 0.30976182798175705,
+      "grad_norm": 1.0817721838689387,
+      "learning_rate": 0.00016169018143357767,
+      "loss": 1.5391,
+      "step": 6724
+    },
+    {
+      "epoch": 0.30980789607039205,
+      "grad_norm": 1.1968325986219088,
+      "learning_rate": 0.00016167843741059362,
+      "loss": 1.4821,
+      "step": 6725
+    },
+    {
+      "epoch": 0.30985396415902705,
+      "grad_norm": 2.523159563392436,
+      "learning_rate": 0.00016166669201444494,
+      "loss": 1.5583,
+      "step": 6726
+    },
+    {
+      "epoch": 0.30990003224766205,
+      "grad_norm": 2.083620212942125,
+      "learning_rate": 0.0001616549452453931,
+      "loss": 1.3994,
+      "step": 6727
+    },
+    {
+      "epoch": 0.30994610033629705,
+      "grad_norm": 1.6296083354981605,
+      "learning_rate": 0.00016164319710369962,
+      "loss": 1.3581,
+      "step": 6728
+    },
+    {
+      "epoch": 0.30999216842493205,
+      "grad_norm": 1.1382006777686131,
+      "learning_rate": 0.00016163144758962603,
+      "loss": 1.5664,
+      "step": 6729
+    },
+    {
+      "epoch": 0.31003823651356704,
+      "grad_norm": 1.623929158598128,
+      "learning_rate": 0.00016161969670343396,
+      "loss": 1.6818,
+      "step": 6730
+    },
+    {
+      "epoch": 0.31008430460220204,
+      "grad_norm": 2.2897990629279206,
+      "learning_rate": 0.000161607944445385,
+      "loss": 1.2793,
+      "step": 6731
+    },
+    {
+      "epoch": 0.31013037269083704,
+      "grad_norm": 2.645226117397815,
+      "learning_rate": 0.00016159619081574075,
+      "loss": 1.5082,
+      "step": 6732
+    },
+    {
+      "epoch": 0.31017644077947204,
+      "grad_norm": 1.3780410585504719,
+      "learning_rate": 0.00016158443581476298,
+      "loss": 1.4768,
+      "step": 6733
+    },
+    {
+      "epoch": 0.31022250886810704,
+      "grad_norm": 1.0400067692511779,
+      "learning_rate": 0.00016157267944271335,
+      "loss": 1.8425,
+      "step": 6734
+    },
+    {
+      "epoch": 0.31026857695674204,
+      "grad_norm": 1.2295813489779195,
+      "learning_rate": 0.00016156092169985353,
+      "loss": 1.5089,
+      "step": 6735
+    },
+    {
+      "epoch": 0.3103146450453771,
+      "grad_norm": 1.5595591121306962,
+      "learning_rate": 0.00016154916258644537,
+      "loss": 0.9599,
+      "step": 6736
+    },
+    {
+      "epoch": 0.3103607131340121,
+      "grad_norm": 3.0729015491157314,
+      "learning_rate": 0.00016153740210275065,
+      "loss": 1.3026,
+      "step": 6737
+    },
+    {
+      "epoch": 0.3104067812226471,
+      "grad_norm": 1.3848634476009523,
+      "learning_rate": 0.00016152564024903118,
+      "loss": 1.3477,
+      "step": 6738
+    },
+    {
+      "epoch": 0.3104528493112821,
+      "grad_norm": 2.017606873535569,
+      "learning_rate": 0.00016151387702554882,
+      "loss": 1.6268,
+      "step": 6739
+    },
+    {
+      "epoch": 0.3104989173999171,
+      "grad_norm": 2.051669098631413,
+      "learning_rate": 0.00016150211243256543,
+      "loss": 1.5531,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3105449854885521,
+      "grad_norm": 2.4310445821944735,
+      "learning_rate": 0.000161490346470343,
+      "loss": 1.1293,
+      "step": 6741
+    },
+    {
+      "epoch": 0.3105910535771871,
+      "grad_norm": 2.513004959526791,
+      "learning_rate": 0.00016147857913914346,
+      "loss": 1.3687,
+      "step": 6742
+    },
+    {
+      "epoch": 0.3106371216658221,
+      "grad_norm": 2.48715825617996,
+      "learning_rate": 0.00016146681043922872,
+      "loss": 1.1116,
+      "step": 6743
+    },
+    {
+      "epoch": 0.3106831897544571,
+      "grad_norm": 1.4621263883820739,
+      "learning_rate": 0.00016145504037086086,
+      "loss": 1.5711,
+      "step": 6744
+    },
+    {
+      "epoch": 0.3107292578430921,
+      "grad_norm": 2.29587122611294,
+      "learning_rate": 0.0001614432689343019,
+      "loss": 1.3012,
+      "step": 6745
+    },
+    {
+      "epoch": 0.3107753259317271,
+      "grad_norm": 1.3649215216114052,
+      "learning_rate": 0.0001614314961298139,
+      "loss": 1.3514,
+      "step": 6746
+    },
+    {
+      "epoch": 0.3108213940203621,
+      "grad_norm": 1.4357212197306088,
+      "learning_rate": 0.000161419721957659,
+      "loss": 1.1742,
+      "step": 6747
+    },
+    {
+      "epoch": 0.3108674621089971,
+      "grad_norm": 2.0258758626697775,
+      "learning_rate": 0.00016140794641809927,
+      "loss": 1.3745,
+      "step": 6748
+    },
+    {
+      "epoch": 0.3109135301976321,
+      "grad_norm": 2.120414106708515,
+      "learning_rate": 0.0001613961695113969,
+      "loss": 1.4429,
+      "step": 6749
+    },
+    {
+      "epoch": 0.3109595982862671,
+      "grad_norm": 1.131802102764416,
+      "learning_rate": 0.0001613843912378141,
+      "loss": 1.6967,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3110056663749021,
+      "grad_norm": 1.5932511208739013,
+      "learning_rate": 0.0001613726115976131,
+      "loss": 1.5912,
+      "step": 6751
+    },
+    {
+      "epoch": 0.3110517344635371,
+      "grad_norm": 1.677563163557032,
+      "learning_rate": 0.0001613608305910561,
+      "loss": 1.3799,
+      "step": 6752
+    },
+    {
+      "epoch": 0.3110978025521721,
+      "grad_norm": 1.633180406327301,
+      "learning_rate": 0.00016134904821840545,
+      "loss": 1.4382,
+      "step": 6753
+    },
+    {
+      "epoch": 0.3111438706408071,
+      "grad_norm": 10.556116725826568,
+      "learning_rate": 0.0001613372644799234,
+      "loss": 1.5079,
+      "step": 6754
+    },
+    {
+      "epoch": 0.3111899387294421,
+      "grad_norm": 1.3959017547336467,
+      "learning_rate": 0.00016132547937587236,
+      "loss": 1.3849,
+      "step": 6755
+    },
+    {
+      "epoch": 0.3112360068180771,
+      "grad_norm": 1.7977853971453672,
+      "learning_rate": 0.00016131369290651467,
+      "loss": 1.4892,
+      "step": 6756
+    },
+    {
+      "epoch": 0.3112820749067121,
+      "grad_norm": 5.09132811653042,
+      "learning_rate": 0.0001613019050721127,
+      "loss": 1.4163,
+      "step": 6757
+    },
+    {
+      "epoch": 0.3113281429953471,
+      "grad_norm": 1.3538847099566051,
+      "learning_rate": 0.00016129011587292895,
+      "loss": 1.313,
+      "step": 6758
+    },
+    {
+      "epoch": 0.3113742110839821,
+      "grad_norm": 2.138182363754689,
+      "learning_rate": 0.00016127832530922587,
+      "loss": 1.3458,
+      "step": 6759
+    },
+    {
+      "epoch": 0.3114202791726171,
+      "grad_norm": 1.69894728399441,
+      "learning_rate": 0.00016126653338126594,
+      "loss": 1.5809,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3114663472612521,
+      "grad_norm": 2.250243627428933,
+      "learning_rate": 0.00016125474008931166,
+      "loss": 1.7426,
+      "step": 6761
+    },
+    {
+      "epoch": 0.3115124153498871,
+      "grad_norm": 1.725599869485455,
+      "learning_rate": 0.00016124294543362567,
+      "loss": 1.3076,
+      "step": 6762
+    },
+    {
+      "epoch": 0.31155848343852216,
+      "grad_norm": 1.6937824271955784,
+      "learning_rate": 0.00016123114941447048,
+      "loss": 1.6387,
+      "step": 6763
+    },
+    {
+      "epoch": 0.31160455152715716,
+      "grad_norm": 2.0222360678652347,
+      "learning_rate": 0.00016121935203210877,
+      "loss": 1.2448,
+      "step": 6764
+    },
+    {
+      "epoch": 0.31165061961579216,
+      "grad_norm": 1.8748455885712678,
+      "learning_rate": 0.00016120755328680314,
+      "loss": 1.4,
+      "step": 6765
+    },
+    {
+      "epoch": 0.31169668770442716,
+      "grad_norm": 1.8795335585997375,
+      "learning_rate": 0.00016119575317881626,
+      "loss": 1.3389,
+      "step": 6766
+    },
+    {
+      "epoch": 0.31174275579306215,
+      "grad_norm": 1.5030870945919586,
+      "learning_rate": 0.00016118395170841088,
+      "loss": 1.4011,
+      "step": 6767
+    },
+    {
+      "epoch": 0.31178882388169715,
+      "grad_norm": 1.5128803859389945,
+      "learning_rate": 0.00016117214887584973,
+      "loss": 1.6687,
+      "step": 6768
+    },
+    {
+      "epoch": 0.31183489197033215,
+      "grad_norm": 2.255949433180262,
+      "learning_rate": 0.00016116034468139555,
+      "loss": 1.3423,
+      "step": 6769
+    },
+    {
+      "epoch": 0.31188096005896715,
+      "grad_norm": 1.3984907299651606,
+      "learning_rate": 0.00016114853912531122,
+      "loss": 1.4692,
+      "step": 6770
+    },
+    {
+      "epoch": 0.31192702814760215,
+      "grad_norm": 2.102321252730273,
+      "learning_rate": 0.00016113673220785946,
+      "loss": 1.4092,
+      "step": 6771
+    },
+    {
+      "epoch": 0.31197309623623715,
+      "grad_norm": 1.7623773104639302,
+      "learning_rate": 0.0001611249239293032,
+      "loss": 1.3961,
+      "step": 6772
+    },
+    {
+      "epoch": 0.31201916432487214,
+      "grad_norm": 1.9569326538096843,
+      "learning_rate": 0.00016111311428990529,
+      "loss": 1.412,
+      "step": 6773
+    },
+    {
+      "epoch": 0.31206523241350714,
+      "grad_norm": 1.0662237427494823,
+      "learning_rate": 0.0001611013032899287,
+      "loss": 1.4951,
+      "step": 6774
+    },
+    {
+      "epoch": 0.31211130050214214,
+      "grad_norm": 1.3007252993486176,
+      "learning_rate": 0.00016108949092963635,
+      "loss": 1.575,
+      "step": 6775
+    },
+    {
+      "epoch": 0.3121573685907772,
+      "grad_norm": 1.999028977771617,
+      "learning_rate": 0.00016107767720929128,
+      "loss": 1.593,
+      "step": 6776
+    },
+    {
+      "epoch": 0.3122034366794122,
+      "grad_norm": 1.5337802720017215,
+      "learning_rate": 0.0001610658621291564,
+      "loss": 1.2977,
+      "step": 6777
+    },
+    {
+      "epoch": 0.3122495047680472,
+      "grad_norm": 1.273564785323178,
+      "learning_rate": 0.0001610540456894948,
+      "loss": 1.3193,
+      "step": 6778
+    },
+    {
+      "epoch": 0.3122955728566822,
+      "grad_norm": 1.5459023393893994,
+      "learning_rate": 0.0001610422278905696,
+      "loss": 1.5262,
+      "step": 6779
+    },
+    {
+      "epoch": 0.3123416409453172,
+      "grad_norm": 1.9098153629999528,
+      "learning_rate": 0.00016103040873264384,
+      "loss": 1.5595,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3123877090339522,
+      "grad_norm": 1.7038512625943563,
+      "learning_rate": 0.00016101858821598067,
+      "loss": 1.4209,
+      "step": 6781
+    },
+    {
+      "epoch": 0.3124337771225872,
+      "grad_norm": 2.176759548549679,
+      "learning_rate": 0.00016100676634084326,
+      "loss": 1.5338,
+      "step": 6782
+    },
+    {
+      "epoch": 0.3124798452112222,
+      "grad_norm": 1.5500305485062171,
+      "learning_rate": 0.00016099494310749482,
+      "loss": 1.465,
+      "step": 6783
+    },
+    {
+      "epoch": 0.3125259132998572,
+      "grad_norm": 2.2649808490879355,
+      "learning_rate": 0.00016098311851619852,
+      "loss": 1.3065,
+      "step": 6784
+    },
+    {
+      "epoch": 0.3125719813884922,
+      "grad_norm": 2.510929525175861,
+      "learning_rate": 0.00016097129256721765,
+      "loss": 1.235,
+      "step": 6785
+    },
+    {
+      "epoch": 0.3126180494771272,
+      "grad_norm": 1.4286384426311989,
+      "learning_rate": 0.0001609594652608155,
+      "loss": 1.5141,
+      "step": 6786
+    },
+    {
+      "epoch": 0.3126641175657622,
+      "grad_norm": 1.9591508075793986,
+      "learning_rate": 0.0001609476365972554,
+      "loss": 1.3425,
+      "step": 6787
+    },
+    {
+      "epoch": 0.3127101856543972,
+      "grad_norm": 1.2506207439383088,
+      "learning_rate": 0.0001609358065768007,
+      "loss": 1.3844,
+      "step": 6788
+    },
+    {
+      "epoch": 0.3127562537430322,
+      "grad_norm": 3.353096655814701,
+      "learning_rate": 0.0001609239751997147,
+      "loss": 1.2578,
+      "step": 6789
+    },
+    {
+      "epoch": 0.31280232183166723,
+      "grad_norm": 1.1734083552134267,
+      "learning_rate": 0.00016091214246626088,
+      "loss": 1.4853,
+      "step": 6790
+    },
+    {
+      "epoch": 0.31284838992030223,
+      "grad_norm": 3.1089735424478864,
+      "learning_rate": 0.00016090030837670263,
+      "loss": 1.5835,
+      "step": 6791
+    },
+    {
+      "epoch": 0.3128944580089372,
+      "grad_norm": 1.512837762766996,
+      "learning_rate": 0.00016088847293130342,
+      "loss": 1.3613,
+      "step": 6792
+    },
+    {
+      "epoch": 0.3129405260975722,
+      "grad_norm": 1.9210873357080978,
+      "learning_rate": 0.0001608766361303268,
+      "loss": 1.4193,
+      "step": 6793
+    },
+    {
+      "epoch": 0.3129865941862072,
+      "grad_norm": 1.384682308649665,
+      "learning_rate": 0.00016086479797403622,
+      "loss": 1.4145,
+      "step": 6794
+    },
+    {
+      "epoch": 0.3130326622748422,
+      "grad_norm": 2.3762225694993977,
+      "learning_rate": 0.0001608529584626953,
+      "loss": 1.3589,
+      "step": 6795
+    },
+    {
+      "epoch": 0.3130787303634772,
+      "grad_norm": 2.176612550286576,
+      "learning_rate": 0.0001608411175965676,
+      "loss": 1.4082,
+      "step": 6796
+    },
+    {
+      "epoch": 0.3131247984521122,
+      "grad_norm": 1.187353029679865,
+      "learning_rate": 0.00016082927537591673,
+      "loss": 1.408,
+      "step": 6797
+    },
+    {
+      "epoch": 0.3131708665407472,
+      "grad_norm": 1.3549323388989614,
+      "learning_rate": 0.00016081743180100636,
+      "loss": 1.3229,
+      "step": 6798
+    },
+    {
+      "epoch": 0.3132169346293822,
+      "grad_norm": 2.4487112429998064,
+      "learning_rate": 0.00016080558687210013,
+      "loss": 1.5492,
+      "step": 6799
+    },
+    {
+      "epoch": 0.3132630027180172,
+      "grad_norm": 1.4866556810429894,
+      "learning_rate": 0.00016079374058946178,
+      "loss": 1.2106,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3133090708066522,
+      "grad_norm": 1.3275240876232113,
+      "learning_rate": 0.000160781892953355,
+      "loss": 1.2825,
+      "step": 6801
+    },
+    {
+      "epoch": 0.3133551388952872,
+      "grad_norm": 1.4540148108056672,
+      "learning_rate": 0.00016077004396404364,
+      "loss": 1.4969,
+      "step": 6802
+    },
+    {
+      "epoch": 0.31340120698392226,
+      "grad_norm": 2.5021197647691142,
+      "learning_rate": 0.00016075819362179142,
+      "loss": 1.1984,
+      "step": 6803
+    },
+    {
+      "epoch": 0.31344727507255726,
+      "grad_norm": 1.9292106705035306,
+      "learning_rate": 0.00016074634192686223,
+      "loss": 1.2012,
+      "step": 6804
+    },
+    {
+      "epoch": 0.31349334316119226,
+      "grad_norm": 3.349906012893209,
+      "learning_rate": 0.00016073448887951987,
+      "loss": 1.5278,
+      "step": 6805
+    },
+    {
+      "epoch": 0.31353941124982726,
+      "grad_norm": 1.8874956828104632,
+      "learning_rate": 0.00016072263448002825,
+      "loss": 1.4429,
+      "step": 6806
+    },
+    {
+      "epoch": 0.31358547933846226,
+      "grad_norm": 0.9442312240608485,
+      "learning_rate": 0.0001607107787286513,
+      "loss": 1.5128,
+      "step": 6807
+    },
+    {
+      "epoch": 0.31363154742709726,
+      "grad_norm": 1.8963330123442066,
+      "learning_rate": 0.00016069892162565296,
+      "loss": 1.4574,
+      "step": 6808
+    },
+    {
+      "epoch": 0.31367761551573226,
+      "grad_norm": 1.3442272215684672,
+      "learning_rate": 0.0001606870631712972,
+      "loss": 1.2378,
+      "step": 6809
+    },
+    {
+      "epoch": 0.31372368360436725,
+      "grad_norm": 1.8973121000417765,
+      "learning_rate": 0.00016067520336584802,
+      "loss": 1.5408,
+      "step": 6810
+    },
+    {
+      "epoch": 0.31376975169300225,
+      "grad_norm": 2.365939312053202,
+      "learning_rate": 0.00016066334220956948,
+      "loss": 1.4009,
+      "step": 6811
+    },
+    {
+      "epoch": 0.31381581978163725,
+      "grad_norm": 1.4282229117556409,
+      "learning_rate": 0.00016065147970272564,
+      "loss": 1.7838,
+      "step": 6812
+    },
+    {
+      "epoch": 0.31386188787027225,
+      "grad_norm": 5.302214247161432,
+      "learning_rate": 0.0001606396158455806,
+      "loss": 1.464,
+      "step": 6813
+    },
+    {
+      "epoch": 0.31390795595890725,
+      "grad_norm": 4.19824564757748,
+      "learning_rate": 0.0001606277506383985,
+      "loss": 1.3927,
+      "step": 6814
+    },
+    {
+      "epoch": 0.31395402404754225,
+      "grad_norm": 1.447495672792,
+      "learning_rate": 0.00016061588408144352,
+      "loss": 1.4594,
+      "step": 6815
+    },
+    {
+      "epoch": 0.31400009213617724,
+      "grad_norm": 1.1367826546284772,
+      "learning_rate": 0.00016060401617497977,
+      "loss": 1.663,
+      "step": 6816
+    },
+    {
+      "epoch": 0.3140461602248123,
+      "grad_norm": 3.340025576648705,
+      "learning_rate": 0.0001605921469192715,
+      "loss": 1.4758,
+      "step": 6817
+    },
+    {
+      "epoch": 0.3140922283134473,
+      "grad_norm": 1.0748809587656967,
+      "learning_rate": 0.000160580276314583,
+      "loss": 1.2929,
+      "step": 6818
+    },
+    {
+      "epoch": 0.3141382964020823,
+      "grad_norm": 2.4990783254657103,
+      "learning_rate": 0.0001605684043611785,
+      "loss": 1.3128,
+      "step": 6819
+    },
+    {
+      "epoch": 0.3141843644907173,
+      "grad_norm": 1.873854874058778,
+      "learning_rate": 0.00016055653105932234,
+      "loss": 1.387,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3142304325793523,
+      "grad_norm": 2.4368432171998347,
+      "learning_rate": 0.00016054465640927886,
+      "loss": 1.4217,
+      "step": 6821
+    },
+    {
+      "epoch": 0.3142765006679873,
+      "grad_norm": 3.2945694890885027,
+      "learning_rate": 0.0001605327804113124,
+      "loss": 1.3369,
+      "step": 6822
+    },
+    {
+      "epoch": 0.3143225687566223,
+      "grad_norm": 1.1267966582064264,
+      "learning_rate": 0.00016052090306568738,
+      "loss": 1.7527,
+      "step": 6823
+    },
+    {
+      "epoch": 0.3143686368452573,
+      "grad_norm": 6.092139323068455,
+      "learning_rate": 0.00016050902437266825,
+      "loss": 1.2784,
+      "step": 6824
+    },
+    {
+      "epoch": 0.3144147049338923,
+      "grad_norm": 3.3021034703652594,
+      "learning_rate": 0.00016049714433251944,
+      "loss": 1.2382,
+      "step": 6825
+    },
+    {
+      "epoch": 0.3144607730225273,
+      "grad_norm": 2.1719428343016713,
+      "learning_rate": 0.0001604852629455054,
+      "loss": 1.6463,
+      "step": 6826
+    },
+    {
+      "epoch": 0.3145068411111623,
+      "grad_norm": 1.0662727663326867,
+      "learning_rate": 0.0001604733802118907,
+      "loss": 1.0688,
+      "step": 6827
+    },
+    {
+      "epoch": 0.3145529091997973,
+      "grad_norm": 1.4035911300285866,
+      "learning_rate": 0.0001604614961319399,
+      "loss": 1.2967,
+      "step": 6828
+    },
+    {
+      "epoch": 0.3145989772884323,
+      "grad_norm": 1.7131699548988917,
+      "learning_rate": 0.00016044961070591754,
+      "loss": 1.617,
+      "step": 6829
+    },
+    {
+      "epoch": 0.3146450453770673,
+      "grad_norm": 1.4706568614314364,
+      "learning_rate": 0.00016043772393408828,
+      "loss": 1.0492,
+      "step": 6830
+    },
+    {
+      "epoch": 0.31469111346570233,
+      "grad_norm": 2.1141816961985214,
+      "learning_rate": 0.00016042583581671668,
+      "loss": 1.4372,
+      "step": 6831
+    },
+    {
+      "epoch": 0.31473718155433733,
+      "grad_norm": 1.172404782709179,
+      "learning_rate": 0.00016041394635406747,
+      "loss": 1.6151,
+      "step": 6832
+    },
+    {
+      "epoch": 0.31478324964297233,
+      "grad_norm": 1.562187755011678,
+      "learning_rate": 0.00016040205554640534,
+      "loss": 1.389,
+      "step": 6833
+    },
+    {
+      "epoch": 0.3148293177316073,
+      "grad_norm": 1.681620116345361,
+      "learning_rate": 0.00016039016339399503,
+      "loss": 1.2452,
+      "step": 6834
+    },
+    {
+      "epoch": 0.3148753858202423,
+      "grad_norm": 2.639161825299416,
+      "learning_rate": 0.00016037826989710125,
+      "loss": 1.4825,
+      "step": 6835
+    },
+    {
+      "epoch": 0.3149214539088773,
+      "grad_norm": 3.906418685879224,
+      "learning_rate": 0.00016036637505598882,
+      "loss": 1.4491,
+      "step": 6836
+    },
+    {
+      "epoch": 0.3149675219975123,
+      "grad_norm": 1.3452782702083466,
+      "learning_rate": 0.00016035447887092255,
+      "loss": 1.2205,
+      "step": 6837
+    },
+    {
+      "epoch": 0.3150135900861473,
+      "grad_norm": 1.1639061747140458,
+      "learning_rate": 0.0001603425813421673,
+      "loss": 1.4093,
+      "step": 6838
+    },
+    {
+      "epoch": 0.3150596581747823,
+      "grad_norm": 1.1123375465698209,
+      "learning_rate": 0.00016033068246998794,
+      "loss": 1.5664,
+      "step": 6839
+    },
+    {
+      "epoch": 0.3151057262634173,
+      "grad_norm": 1.8269415437529997,
+      "learning_rate": 0.00016031878225464938,
+      "loss": 1.3517,
+      "step": 6840
+    },
+    {
+      "epoch": 0.3151517943520523,
+      "grad_norm": 2.037206859712878,
+      "learning_rate": 0.00016030688069641657,
+      "loss": 1.6346,
+      "step": 6841
+    },
+    {
+      "epoch": 0.3151978624406873,
+      "grad_norm": 2.2314362705848,
+      "learning_rate": 0.00016029497779555447,
+      "loss": 1.4628,
+      "step": 6842
+    },
+    {
+      "epoch": 0.3152439305293223,
+      "grad_norm": 1.1775799434605756,
+      "learning_rate": 0.00016028307355232804,
+      "loss": 1.2153,
+      "step": 6843
+    },
+    {
+      "epoch": 0.31528999861795737,
+      "grad_norm": 1.578033012978386,
+      "learning_rate": 0.00016027116796700236,
+      "loss": 1.1561,
+      "step": 6844
+    },
+    {
+      "epoch": 0.31533606670659237,
+      "grad_norm": 2.489920524080375,
+      "learning_rate": 0.0001602592610398425,
+      "loss": 1.4506,
+      "step": 6845
+    },
+    {
+      "epoch": 0.31538213479522736,
+      "grad_norm": 1.514548754826939,
+      "learning_rate": 0.00016024735277111348,
+      "loss": 1.6625,
+      "step": 6846
+    },
+    {
+      "epoch": 0.31542820288386236,
+      "grad_norm": 1.12095240093112,
+      "learning_rate": 0.00016023544316108045,
+      "loss": 1.571,
+      "step": 6847
+    },
+    {
+      "epoch": 0.31547427097249736,
+      "grad_norm": 2.253043198069283,
+      "learning_rate": 0.00016022353221000857,
+      "loss": 1.2332,
+      "step": 6848
+    },
+    {
+      "epoch": 0.31552033906113236,
+      "grad_norm": 1.1268475733333192,
+      "learning_rate": 0.000160211619918163,
+      "loss": 1.3614,
+      "step": 6849
+    },
+    {
+      "epoch": 0.31556640714976736,
+      "grad_norm": 1.078898533160077,
+      "learning_rate": 0.00016019970628580895,
+      "loss": 1.3362,
+      "step": 6850
+    },
+    {
+      "epoch": 0.31561247523840236,
+      "grad_norm": 1.4069853538568935,
+      "learning_rate": 0.00016018779131321166,
+      "loss": 1.2681,
+      "step": 6851
+    },
+    {
+      "epoch": 0.31565854332703736,
+      "grad_norm": 1.3652031886459417,
+      "learning_rate": 0.0001601758750006364,
+      "loss": 1.7082,
+      "step": 6852
+    },
+    {
+      "epoch": 0.31570461141567235,
+      "grad_norm": 2.054088378422274,
+      "learning_rate": 0.00016016395734834847,
+      "loss": 1.4211,
+      "step": 6853
+    },
+    {
+      "epoch": 0.31575067950430735,
+      "grad_norm": 1.3481587559898702,
+      "learning_rate": 0.0001601520383566132,
+      "loss": 1.3225,
+      "step": 6854
+    },
+    {
+      "epoch": 0.31579674759294235,
+      "grad_norm": 1.3737694246691667,
+      "learning_rate": 0.0001601401180256959,
+      "loss": 1.3326,
+      "step": 6855
+    },
+    {
+      "epoch": 0.31584281568157735,
+      "grad_norm": 1.072428361210365,
+      "learning_rate": 0.000160128196355862,
+      "loss": 1.7135,
+      "step": 6856
+    },
+    {
+      "epoch": 0.31588888377021235,
+      "grad_norm": 2.4033374902748923,
+      "learning_rate": 0.00016011627334737693,
+      "loss": 1.4899,
+      "step": 6857
+    },
+    {
+      "epoch": 0.3159349518588474,
+      "grad_norm": 1.3584322229430197,
+      "learning_rate": 0.00016010434900050612,
+      "loss": 1.2108,
+      "step": 6858
+    },
+    {
+      "epoch": 0.3159810199474824,
+      "grad_norm": 2.637727586895856,
+      "learning_rate": 0.000160092423315515,
+      "loss": 1.3634,
+      "step": 6859
+    },
+    {
+      "epoch": 0.3160270880361174,
+      "grad_norm": 1.389307202781291,
+      "learning_rate": 0.0001600804962926691,
+      "loss": 1.3655,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3160731561247524,
+      "grad_norm": 1.288276953823126,
+      "learning_rate": 0.000160068567932234,
+      "loss": 1.293,
+      "step": 6861
+    },
+    {
+      "epoch": 0.3161192242133874,
+      "grad_norm": 1.8893438976443975,
+      "learning_rate": 0.00016005663823447523,
+      "loss": 1.4369,
+      "step": 6862
+    },
+    {
+      "epoch": 0.3161652923020224,
+      "grad_norm": 1.2443516968228514,
+      "learning_rate": 0.00016004470719965837,
+      "loss": 1.6692,
+      "step": 6863
+    },
+    {
+      "epoch": 0.3162113603906574,
+      "grad_norm": 1.5529861240216634,
+      "learning_rate": 0.00016003277482804905,
+      "loss": 1.4894,
+      "step": 6864
+    },
+    {
+      "epoch": 0.3162574284792924,
+      "grad_norm": 0.9732694163002336,
+      "learning_rate": 0.00016002084111991298,
+      "loss": 0.9915,
+      "step": 6865
+    },
+    {
+      "epoch": 0.3163034965679274,
+      "grad_norm": 1.5947179382699452,
+      "learning_rate": 0.00016000890607551575,
+      "loss": 1.4091,
+      "step": 6866
+    },
+    {
+      "epoch": 0.3163495646565624,
+      "grad_norm": 1.335876473810033,
+      "learning_rate": 0.0001599969696951231,
+      "loss": 1.5692,
+      "step": 6867
+    },
+    {
+      "epoch": 0.3163956327451974,
+      "grad_norm": 1.3315917331345375,
+      "learning_rate": 0.00015998503197900085,
+      "loss": 1.2022,
+      "step": 6868
+    },
+    {
+      "epoch": 0.3164417008338324,
+      "grad_norm": 1.6154567027857756,
+      "learning_rate": 0.0001599730929274147,
+      "loss": 1.714,
+      "step": 6869
+    },
+    {
+      "epoch": 0.3164877689224674,
+      "grad_norm": 1.333703799664107,
+      "learning_rate": 0.00015996115254063046,
+      "loss": 1.242,
+      "step": 6870
+    },
+    {
+      "epoch": 0.31653383701110244,
+      "grad_norm": 2.0036993284563165,
+      "learning_rate": 0.00015994921081891397,
+      "loss": 1.5066,
+      "step": 6871
+    },
+    {
+      "epoch": 0.31657990509973744,
+      "grad_norm": 1.7400273567447617,
+      "learning_rate": 0.00015993726776253108,
+      "loss": 1.4555,
+      "step": 6872
+    },
+    {
+      "epoch": 0.31662597318837243,
+      "grad_norm": 2.3130904913936003,
+      "learning_rate": 0.0001599253233717477,
+      "loss": 1.4233,
+      "step": 6873
+    },
+    {
+      "epoch": 0.31667204127700743,
+      "grad_norm": 2.293411727774617,
+      "learning_rate": 0.00015991337764682974,
+      "loss": 1.3303,
+      "step": 6874
+    },
+    {
+      "epoch": 0.31671810936564243,
+      "grad_norm": 1.5933929088574739,
+      "learning_rate": 0.00015990143058804317,
+      "loss": 1.6357,
+      "step": 6875
+    },
+    {
+      "epoch": 0.31676417745427743,
+      "grad_norm": 1.1228808686802638,
+      "learning_rate": 0.00015988948219565393,
+      "loss": 1.1016,
+      "step": 6876
+    },
+    {
+      "epoch": 0.3168102455429124,
+      "grad_norm": 1.767822111822696,
+      "learning_rate": 0.00015987753246992812,
+      "loss": 1.2699,
+      "step": 6877
+    },
+    {
+      "epoch": 0.3168563136315474,
+      "grad_norm": 1.2914034456547367,
+      "learning_rate": 0.00015986558141113167,
+      "loss": 1.2559,
+      "step": 6878
+    },
+    {
+      "epoch": 0.3169023817201824,
+      "grad_norm": 1.6163551767161441,
+      "learning_rate": 0.00015985362901953072,
+      "loss": 1.5456,
+      "step": 6879
+    },
+    {
+      "epoch": 0.3169484498088174,
+      "grad_norm": 1.8574080543322333,
+      "learning_rate": 0.0001598416752953913,
+      "loss": 1.3212,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3169945178974524,
+      "grad_norm": 1.380127863381503,
+      "learning_rate": 0.00015982972023897963,
+      "loss": 1.502,
+      "step": 6881
+    },
+    {
+      "epoch": 0.3170405859860874,
+      "grad_norm": 1.275662802068764,
+      "learning_rate": 0.0001598177638505618,
+      "loss": 1.3146,
+      "step": 6882
+    },
+    {
+      "epoch": 0.3170866540747224,
+      "grad_norm": 1.7574615427705371,
+      "learning_rate": 0.00015980580613040405,
+      "loss": 1.5883,
+      "step": 6883
+    },
+    {
+      "epoch": 0.3171327221633574,
+      "grad_norm": 1.2275445087886887,
+      "learning_rate": 0.00015979384707877253,
+      "loss": 1.2552,
+      "step": 6884
+    },
+    {
+      "epoch": 0.31717879025199247,
+      "grad_norm": 1.5566399253067904,
+      "learning_rate": 0.00015978188669593356,
+      "loss": 1.5788,
+      "step": 6885
+    },
+    {
+      "epoch": 0.31722485834062747,
+      "grad_norm": 2.1675786932909067,
+      "learning_rate": 0.00015976992498215338,
+      "loss": 1.4448,
+      "step": 6886
+    },
+    {
+      "epoch": 0.31727092642926247,
+      "grad_norm": 1.2248471861889287,
+      "learning_rate": 0.00015975796193769831,
+      "loss": 1.5369,
+      "step": 6887
+    },
+    {
+      "epoch": 0.31731699451789747,
+      "grad_norm": 1.1268007361839352,
+      "learning_rate": 0.00015974599756283467,
+      "loss": 1.3064,
+      "step": 6888
+    },
+    {
+      "epoch": 0.31736306260653246,
+      "grad_norm": 1.538044000278205,
+      "learning_rate": 0.00015973403185782887,
+      "loss": 1.4013,
+      "step": 6889
+    },
+    {
+      "epoch": 0.31740913069516746,
+      "grad_norm": 1.1389175589141285,
+      "learning_rate": 0.00015972206482294725,
+      "loss": 1.2531,
+      "step": 6890
+    },
+    {
+      "epoch": 0.31745519878380246,
+      "grad_norm": 1.399923660183488,
+      "learning_rate": 0.00015971009645845624,
+      "loss": 1.6186,
+      "step": 6891
+    },
+    {
+      "epoch": 0.31750126687243746,
+      "grad_norm": 1.8856300476991357,
+      "learning_rate": 0.00015969812676462235,
+      "loss": 1.2505,
+      "step": 6892
+    },
+    {
+      "epoch": 0.31754733496107246,
+      "grad_norm": 1.457104883034752,
+      "learning_rate": 0.00015968615574171198,
+      "loss": 1.4568,
+      "step": 6893
+    },
+    {
+      "epoch": 0.31759340304970746,
+      "grad_norm": 1.3503727454100611,
+      "learning_rate": 0.00015967418338999173,
+      "loss": 1.3528,
+      "step": 6894
+    },
+    {
+      "epoch": 0.31763947113834246,
+      "grad_norm": 1.223103479809643,
+      "learning_rate": 0.0001596622097097281,
+      "loss": 0.9426,
+      "step": 6895
+    },
+    {
+      "epoch": 0.31768553922697745,
+      "grad_norm": 1.5538412197740492,
+      "learning_rate": 0.00015965023470118763,
+      "loss": 1.8029,
+      "step": 6896
+    },
+    {
+      "epoch": 0.31773160731561245,
+      "grad_norm": 3.9477764071269936,
+      "learning_rate": 0.000159638258364637,
+      "loss": 1.3659,
+      "step": 6897
+    },
+    {
+      "epoch": 0.31777767540424745,
+      "grad_norm": 1.6486424022013877,
+      "learning_rate": 0.0001596262807003428,
+      "loss": 1.3124,
+      "step": 6898
+    },
+    {
+      "epoch": 0.3178237434928825,
+      "grad_norm": 1.5748400040291073,
+      "learning_rate": 0.0001596143017085717,
+      "loss": 1.308,
+      "step": 6899
+    },
+    {
+      "epoch": 0.3178698115815175,
+      "grad_norm": 2.580642112924322,
+      "learning_rate": 0.0001596023213895904,
+      "loss": 1.3334,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3179158796701525,
+      "grad_norm": 1.945753792464881,
+      "learning_rate": 0.00015959033974366558,
+      "loss": 1.4431,
+      "step": 6901
+    },
+    {
+      "epoch": 0.3179619477587875,
+      "grad_norm": 1.3224945588985855,
+      "learning_rate": 0.00015957835677106406,
+      "loss": 1.3515,
+      "step": 6902
+    },
+    {
+      "epoch": 0.3180080158474225,
+      "grad_norm": 1.1894734334294734,
+      "learning_rate": 0.00015956637247205256,
+      "loss": 1.3527,
+      "step": 6903
+    },
+    {
+      "epoch": 0.3180540839360575,
+      "grad_norm": 1.4569856419939726,
+      "learning_rate": 0.0001595543868468979,
+      "loss": 1.3538,
+      "step": 6904
+    },
+    {
+      "epoch": 0.3181001520246925,
+      "grad_norm": 1.7809156010546234,
+      "learning_rate": 0.00015954239989586695,
+      "loss": 1.1432,
+      "step": 6905
+    },
+    {
+      "epoch": 0.3181462201133275,
+      "grad_norm": 1.3210619610884384,
+      "learning_rate": 0.00015953041161922654,
+      "loss": 1.3641,
+      "step": 6906
+    },
+    {
+      "epoch": 0.3181922882019625,
+      "grad_norm": 1.2394329465005471,
+      "learning_rate": 0.00015951842201724362,
+      "loss": 1.2145,
+      "step": 6907
+    },
+    {
+      "epoch": 0.3182383562905975,
+      "grad_norm": 2.4079403871464624,
+      "learning_rate": 0.00015950643109018505,
+      "loss": 1.3753,
+      "step": 6908
+    },
+    {
+      "epoch": 0.3182844243792325,
+      "grad_norm": 2.794757040687761,
+      "learning_rate": 0.00015949443883831786,
+      "loss": 1.5529,
+      "step": 6909
+    },
+    {
+      "epoch": 0.3183304924678675,
+      "grad_norm": 1.5919160949509024,
+      "learning_rate": 0.00015948244526190896,
+      "loss": 1.2766,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3183765605565025,
+      "grad_norm": 1.9133010151213259,
+      "learning_rate": 0.00015947045036122543,
+      "loss": 1.5063,
+      "step": 6911
+    },
+    {
+      "epoch": 0.31842262864513754,
+      "grad_norm": 1.553682944181397,
+      "learning_rate": 0.00015945845413653428,
+      "loss": 1.0843,
+      "step": 6912
+    },
+    {
+      "epoch": 0.31846869673377254,
+      "grad_norm": 2.374107679086441,
+      "learning_rate": 0.0001594464565881026,
+      "loss": 1.3897,
+      "step": 6913
+    },
+    {
+      "epoch": 0.31851476482240754,
+      "grad_norm": 3.03961167676597,
+      "learning_rate": 0.0001594344577161975,
+      "loss": 1.4625,
+      "step": 6914
+    },
+    {
+      "epoch": 0.31856083291104254,
+      "grad_norm": 1.4687090750062,
+      "learning_rate": 0.0001594224575210861,
+      "loss": 1.556,
+      "step": 6915
+    },
+    {
+      "epoch": 0.31860690099967753,
+      "grad_norm": 1.2492253612098128,
+      "learning_rate": 0.00015941045600303562,
+      "loss": 1.45,
+      "step": 6916
+    },
+    {
+      "epoch": 0.31865296908831253,
+      "grad_norm": 1.0555505059086667,
+      "learning_rate": 0.00015939845316231314,
+      "loss": 1.2978,
+      "step": 6917
+    },
+    {
+      "epoch": 0.31869903717694753,
+      "grad_norm": 1.0569304130792927,
+      "learning_rate": 0.00015938644899918594,
+      "loss": 1.3731,
+      "step": 6918
+    },
+    {
+      "epoch": 0.31874510526558253,
+      "grad_norm": 3.7910226337017088,
+      "learning_rate": 0.00015937444351392133,
+      "loss": 1.178,
+      "step": 6919
+    },
+    {
+      "epoch": 0.3187911733542175,
+      "grad_norm": 7.361592499434576,
+      "learning_rate": 0.00015936243670678653,
+      "loss": 1.2936,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3188372414428525,
+      "grad_norm": 1.4918998940140982,
+      "learning_rate": 0.00015935042857804886,
+      "loss": 1.6156,
+      "step": 6921
+    },
+    {
+      "epoch": 0.3188833095314875,
+      "grad_norm": 1.7283046053730124,
+      "learning_rate": 0.00015933841912797565,
+      "loss": 1.2802,
+      "step": 6922
+    },
+    {
+      "epoch": 0.3189293776201225,
+      "grad_norm": 2.2492025381941647,
+      "learning_rate": 0.00015932640835683427,
+      "loss": 1.2966,
+      "step": 6923
+    },
+    {
+      "epoch": 0.3189754457087575,
+      "grad_norm": 1.0045541482652351,
+      "learning_rate": 0.00015931439626489216,
+      "loss": 1.6293,
+      "step": 6924
+    },
+    {
+      "epoch": 0.3190215137973925,
+      "grad_norm": 1.0824131104113321,
+      "learning_rate": 0.0001593023828524167,
+      "loss": 1.6549,
+      "step": 6925
+    },
+    {
+      "epoch": 0.3190675818860276,
+      "grad_norm": 1.0788319244896607,
+      "learning_rate": 0.00015929036811967537,
+      "loss": 1.2902,
+      "step": 6926
+    },
+    {
+      "epoch": 0.31911364997466257,
+      "grad_norm": 1.0733932839490399,
+      "learning_rate": 0.00015927835206693564,
+      "loss": 1.35,
+      "step": 6927
+    },
+    {
+      "epoch": 0.31915971806329757,
+      "grad_norm": 4.872106316944894,
+      "learning_rate": 0.00015926633469446507,
+      "loss": 1.3163,
+      "step": 6928
+    },
+    {
+      "epoch": 0.31920578615193257,
+      "grad_norm": 1.414183180587148,
+      "learning_rate": 0.00015925431600253117,
+      "loss": 1.4501,
+      "step": 6929
+    },
+    {
+      "epoch": 0.31925185424056757,
+      "grad_norm": 1.4454971080155945,
+      "learning_rate": 0.0001592422959914015,
+      "loss": 1.4029,
+      "step": 6930
+    },
+    {
+      "epoch": 0.31929792232920257,
+      "grad_norm": 1.5811999131160337,
+      "learning_rate": 0.0001592302746613437,
+      "loss": 1.3781,
+      "step": 6931
+    },
+    {
+      "epoch": 0.31934399041783756,
+      "grad_norm": 1.0014299452846098,
+      "learning_rate": 0.0001592182520126254,
+      "loss": 1.7479,
+      "step": 6932
+    },
+    {
+      "epoch": 0.31939005850647256,
+      "grad_norm": 1.9272287828608923,
+      "learning_rate": 0.00015920622804551426,
+      "loss": 1.7227,
+      "step": 6933
+    },
+    {
+      "epoch": 0.31943612659510756,
+      "grad_norm": 1.867098603282342,
+      "learning_rate": 0.00015919420276027798,
+      "loss": 1.3727,
+      "step": 6934
+    },
+    {
+      "epoch": 0.31948219468374256,
+      "grad_norm": 1.7434676524410073,
+      "learning_rate": 0.00015918217615718428,
+      "loss": 1.2718,
+      "step": 6935
+    },
+    {
+      "epoch": 0.31952826277237756,
+      "grad_norm": 1.1509084230690871,
+      "learning_rate": 0.00015917014823650088,
+      "loss": 1.7041,
+      "step": 6936
+    },
+    {
+      "epoch": 0.31957433086101256,
+      "grad_norm": 3.1291847329158236,
+      "learning_rate": 0.00015915811899849558,
+      "loss": 1.2366,
+      "step": 6937
+    },
+    {
+      "epoch": 0.31962039894964756,
+      "grad_norm": 2.239449925033144,
+      "learning_rate": 0.00015914608844343618,
+      "loss": 1.3513,
+      "step": 6938
+    },
+    {
+      "epoch": 0.3196664670382826,
+      "grad_norm": 2.372508919818166,
+      "learning_rate": 0.00015913405657159058,
+      "loss": 1.2476,
+      "step": 6939
+    },
+    {
+      "epoch": 0.3197125351269176,
+      "grad_norm": 3.2211964276774796,
+      "learning_rate": 0.00015912202338322657,
+      "loss": 1.4706,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3197586032155526,
+      "grad_norm": 1.3791101726831223,
+      "learning_rate": 0.0001591099888786121,
+      "loss": 1.543,
+      "step": 6941
+    },
+    {
+      "epoch": 0.3198046713041876,
+      "grad_norm": 1.549405328059204,
+      "learning_rate": 0.00015909795305801505,
+      "loss": 1.3357,
+      "step": 6942
+    },
+    {
+      "epoch": 0.3198507393928226,
+      "grad_norm": 1.4201289341981582,
+      "learning_rate": 0.00015908591592170344,
+      "loss": 1.6446,
+      "step": 6943
+    },
+    {
+      "epoch": 0.3198968074814576,
+      "grad_norm": 2.6701040603502566,
+      "learning_rate": 0.00015907387746994523,
+      "loss": 1.256,
+      "step": 6944
+    },
+    {
+      "epoch": 0.3199428755700926,
+      "grad_norm": 3.036866717428883,
+      "learning_rate": 0.0001590618377030084,
+      "loss": 1.42,
+      "step": 6945
+    },
+    {
+      "epoch": 0.3199889436587276,
+      "grad_norm": 1.73644445598107,
+      "learning_rate": 0.00015904979662116105,
+      "loss": 1.5374,
+      "step": 6946
+    },
+    {
+      "epoch": 0.3200350117473626,
+      "grad_norm": 2.8729653880510977,
+      "learning_rate": 0.00015903775422467116,
+      "loss": 1.2787,
+      "step": 6947
+    },
+    {
+      "epoch": 0.3200810798359976,
+      "grad_norm": 1.7918724713833263,
+      "learning_rate": 0.00015902571051380697,
+      "loss": 1.488,
+      "step": 6948
+    },
+    {
+      "epoch": 0.3201271479246326,
+      "grad_norm": 1.0541870427654934,
+      "learning_rate": 0.0001590136654888365,
+      "loss": 1.4522,
+      "step": 6949
+    },
+    {
+      "epoch": 0.3201732160132676,
+      "grad_norm": 0.9127702161578698,
+      "learning_rate": 0.000159001619150028,
+      "loss": 1.3949,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3202192841019026,
+      "grad_norm": 1.857566209029725,
+      "learning_rate": 0.00015898957149764958,
+      "loss": 1.4449,
+      "step": 6951
+    },
+    {
+      "epoch": 0.3202653521905376,
+      "grad_norm": 1.8037745931555074,
+      "learning_rate": 0.00015897752253196952,
+      "loss": 1.441,
+      "step": 6952
+    },
+    {
+      "epoch": 0.32031142027917264,
+      "grad_norm": 1.0936696021506753,
+      "learning_rate": 0.00015896547225325604,
+      "loss": 1.1752,
+      "step": 6953
+    },
+    {
+      "epoch": 0.32035748836780764,
+      "grad_norm": 1.1036990054170246,
+      "learning_rate": 0.00015895342066177742,
+      "loss": 1.5748,
+      "step": 6954
+    },
+    {
+      "epoch": 0.32040355645644264,
+      "grad_norm": 4.014706971685277,
+      "learning_rate": 0.00015894136775780195,
+      "loss": 1.4629,
+      "step": 6955
+    },
+    {
+      "epoch": 0.32044962454507764,
+      "grad_norm": 0.9271449802666203,
+      "learning_rate": 0.00015892931354159802,
+      "loss": 1.5774,
+      "step": 6956
+    },
+    {
+      "epoch": 0.32049569263371264,
+      "grad_norm": 1.5564374197363349,
+      "learning_rate": 0.00015891725801343395,
+      "loss": 1.7325,
+      "step": 6957
+    },
+    {
+      "epoch": 0.32054176072234764,
+      "grad_norm": 1.7889356145418898,
+      "learning_rate": 0.00015890520117357816,
+      "loss": 1.3671,
+      "step": 6958
+    },
+    {
+      "epoch": 0.32058782881098263,
+      "grad_norm": 2.2276782252048526,
+      "learning_rate": 0.00015889314302229907,
+      "loss": 1.3445,
+      "step": 6959
+    },
+    {
+      "epoch": 0.32063389689961763,
+      "grad_norm": 2.0954711912701374,
+      "learning_rate": 0.00015888108355986512,
+      "loss": 1.3633,
+      "step": 6960
+    },
+    {
+      "epoch": 0.32067996498825263,
+      "grad_norm": 2.2171759155149813,
+      "learning_rate": 0.0001588690227865448,
+      "loss": 1.4878,
+      "step": 6961
+    },
+    {
+      "epoch": 0.32072603307688763,
+      "grad_norm": 3.761951930218843,
+      "learning_rate": 0.00015885696070260666,
+      "loss": 1.3016,
+      "step": 6962
+    },
+    {
+      "epoch": 0.3207721011655226,
+      "grad_norm": 1.6321888332794257,
+      "learning_rate": 0.00015884489730831917,
+      "loss": 1.5003,
+      "step": 6963
+    },
+    {
+      "epoch": 0.3208181692541576,
+      "grad_norm": 3.1755861243594876,
+      "learning_rate": 0.00015883283260395095,
+      "loss": 1.5303,
+      "step": 6964
+    },
+    {
+      "epoch": 0.3208642373427926,
+      "grad_norm": 2.386069392033615,
+      "learning_rate": 0.0001588207665897706,
+      "loss": 1.1436,
+      "step": 6965
+    },
+    {
+      "epoch": 0.3209103054314276,
+      "grad_norm": 1.1167565489709006,
+      "learning_rate": 0.00015880869926604673,
+      "loss": 1.204,
+      "step": 6966
+    },
+    {
+      "epoch": 0.3209563735200627,
+      "grad_norm": 1.4951679427385314,
+      "learning_rate": 0.00015879663063304798,
+      "loss": 1.5398,
+      "step": 6967
+    },
+    {
+      "epoch": 0.3210024416086977,
+      "grad_norm": 1.6141226488813714,
+      "learning_rate": 0.0001587845606910431,
+      "loss": 1.4059,
+      "step": 6968
+    },
+    {
+      "epoch": 0.3210485096973327,
+      "grad_norm": 2.0253719526987335,
+      "learning_rate": 0.00015877248944030076,
+      "loss": 1.3902,
+      "step": 6969
+    },
+    {
+      "epoch": 0.32109457778596767,
+      "grad_norm": 1.5717233496292964,
+      "learning_rate": 0.0001587604168810897,
+      "loss": 1.104,
+      "step": 6970
+    },
+    {
+      "epoch": 0.32114064587460267,
+      "grad_norm": 1.5683895788065274,
+      "learning_rate": 0.00015874834301367875,
+      "loss": 1.267,
+      "step": 6971
+    },
+    {
+      "epoch": 0.32118671396323767,
+      "grad_norm": 1.1467939934019893,
+      "learning_rate": 0.00015873626783833667,
+      "loss": 1.6126,
+      "step": 6972
+    },
+    {
+      "epoch": 0.32123278205187267,
+      "grad_norm": 2.828609981711123,
+      "learning_rate": 0.00015872419135533225,
+      "loss": 1.5344,
+      "step": 6973
+    },
+    {
+      "epoch": 0.32127885014050767,
+      "grad_norm": 1.3687098818506316,
+      "learning_rate": 0.00015871211356493443,
+      "loss": 1.6632,
+      "step": 6974
+    },
+    {
+      "epoch": 0.32132491822914266,
+      "grad_norm": 1.5977792156481587,
+      "learning_rate": 0.00015870003446741206,
+      "loss": 1.2552,
+      "step": 6975
+    },
+    {
+      "epoch": 0.32137098631777766,
+      "grad_norm": 3.011502817555293,
+      "learning_rate": 0.00015868795406303408,
+      "loss": 1.3333,
+      "step": 6976
+    },
+    {
+      "epoch": 0.32141705440641266,
+      "grad_norm": 1.3274931749992431,
+      "learning_rate": 0.00015867587235206943,
+      "loss": 1.3188,
+      "step": 6977
+    },
+    {
+      "epoch": 0.32146312249504766,
+      "grad_norm": 1.1111605945172873,
+      "learning_rate": 0.0001586637893347871,
+      "loss": 1.3877,
+      "step": 6978
+    },
+    {
+      "epoch": 0.32150919058368266,
+      "grad_norm": 1.55582749051472,
+      "learning_rate": 0.00015865170501145608,
+      "loss": 1.1432,
+      "step": 6979
+    },
+    {
+      "epoch": 0.3215552586723177,
+      "grad_norm": 1.6101598147019474,
+      "learning_rate": 0.0001586396193823454,
+      "loss": 1.543,
+      "step": 6980
+    },
+    {
+      "epoch": 0.3216013267609527,
+      "grad_norm": 0.89584147747135,
+      "learning_rate": 0.00015862753244772417,
+      "loss": 1.6097,
+      "step": 6981
+    },
+    {
+      "epoch": 0.3216473948495877,
+      "grad_norm": 1.2183648842469381,
+      "learning_rate": 0.00015861544420786144,
+      "loss": 1.7504,
+      "step": 6982
+    },
+    {
+      "epoch": 0.3216934629382227,
+      "grad_norm": 1.4270139788980254,
+      "learning_rate": 0.00015860335466302635,
+      "loss": 1.3701,
+      "step": 6983
+    },
+    {
+      "epoch": 0.3217395310268577,
+      "grad_norm": 2.2190362792039124,
+      "learning_rate": 0.00015859126381348804,
+      "loss": 1.5848,
+      "step": 6984
+    },
+    {
+      "epoch": 0.3217855991154927,
+      "grad_norm": 4.905795063066986,
+      "learning_rate": 0.0001585791716595157,
+      "loss": 1.2894,
+      "step": 6985
+    },
+    {
+      "epoch": 0.3218316672041277,
+      "grad_norm": 1.9383887928024992,
+      "learning_rate": 0.00015856707820137856,
+      "loss": 1.4132,
+      "step": 6986
+    },
+    {
+      "epoch": 0.3218777352927627,
+      "grad_norm": 1.9827800509103033,
+      "learning_rate": 0.00015855498343934581,
+      "loss": 1.2917,
+      "step": 6987
+    },
+    {
+      "epoch": 0.3219238033813977,
+      "grad_norm": 2.8295165108846994,
+      "learning_rate": 0.0001585428873736868,
+      "loss": 1.104,
+      "step": 6988
+    },
+    {
+      "epoch": 0.3219698714700327,
+      "grad_norm": 1.0034489138531524,
+      "learning_rate": 0.0001585307900046708,
+      "loss": 1.0579,
+      "step": 6989
+    },
+    {
+      "epoch": 0.3220159395586677,
+      "grad_norm": 1.457271362620135,
+      "learning_rate": 0.00015851869133256708,
+      "loss": 1.565,
+      "step": 6990
+    },
+    {
+      "epoch": 0.3220620076473027,
+      "grad_norm": 5.426759229802432,
+      "learning_rate": 0.00015850659135764503,
+      "loss": 1.2515,
+      "step": 6991
+    },
+    {
+      "epoch": 0.3221080757359377,
+      "grad_norm": 1.278797007170771,
+      "learning_rate": 0.00015849449008017405,
+      "loss": 1.2859,
+      "step": 6992
+    },
+    {
+      "epoch": 0.3221541438245727,
+      "grad_norm": 0.832992208265677,
+      "learning_rate": 0.00015848238750042356,
+      "loss": 1.4632,
+      "step": 6993
+    },
+    {
+      "epoch": 0.32220021191320775,
+      "grad_norm": 1.1934143011153173,
+      "learning_rate": 0.00015847028361866297,
+      "loss": 1.452,
+      "step": 6994
+    },
+    {
+      "epoch": 0.32224628000184274,
+      "grad_norm": 1.1821193824774732,
+      "learning_rate": 0.00015845817843516174,
+      "loss": 1.1269,
+      "step": 6995
+    },
+    {
+      "epoch": 0.32229234809047774,
+      "grad_norm": 2.0087441271800617,
+      "learning_rate": 0.00015844607195018945,
+      "loss": 1.3632,
+      "step": 6996
+    },
+    {
+      "epoch": 0.32233841617911274,
+      "grad_norm": 2.7794853388880387,
+      "learning_rate": 0.00015843396416401556,
+      "loss": 1.3717,
+      "step": 6997
+    },
+    {
+      "epoch": 0.32238448426774774,
+      "grad_norm": 2.10485106361769,
+      "learning_rate": 0.0001584218550769097,
+      "loss": 1.3786,
+      "step": 6998
+    },
+    {
+      "epoch": 0.32243055235638274,
+      "grad_norm": 1.5923190852622815,
+      "learning_rate": 0.00015840974468914136,
+      "loss": 1.4106,
+      "step": 6999
+    },
+    {
+      "epoch": 0.32247662044501774,
+      "grad_norm": 12.746023249318405,
+      "learning_rate": 0.0001583976330009802,
+      "loss": 1.3209,
+      "step": 7000
+    },
+    {
+      "epoch": 0.32252268853365273,
+      "grad_norm": 1.0153782099656155,
+      "learning_rate": 0.0001583855200126959,
+      "loss": 1.6103,
+      "step": 7001
+    },
+    {
+      "epoch": 0.32256875662228773,
+      "grad_norm": 2.0653629982562514,
+      "learning_rate": 0.00015837340572455807,
+      "loss": 1.272,
+      "step": 7002
+    },
+    {
+      "epoch": 0.32261482471092273,
+      "grad_norm": 1.7484336606465265,
+      "learning_rate": 0.0001583612901368365,
+      "loss": 1.5594,
+      "step": 7003
+    },
+    {
+      "epoch": 0.32266089279955773,
+      "grad_norm": 1.9657066246650339,
+      "learning_rate": 0.0001583491732498008,
+      "loss": 1.3473,
+      "step": 7004
+    },
+    {
+      "epoch": 0.32270696088819273,
+      "grad_norm": 1.4871547616911127,
+      "learning_rate": 0.00015833705506372087,
+      "loss": 1.2271,
+      "step": 7005
+    },
+    {
+      "epoch": 0.3227530289768277,
+      "grad_norm": 2.9561733330082207,
+      "learning_rate": 0.00015832493557886644,
+      "loss": 1.157,
+      "step": 7006
+    },
+    {
+      "epoch": 0.3227990970654628,
+      "grad_norm": 2.7867408338348607,
+      "learning_rate": 0.0001583128147955073,
+      "loss": 1.1306,
+      "step": 7007
+    },
+    {
+      "epoch": 0.3228451651540978,
+      "grad_norm": 1.4007982663522032,
+      "learning_rate": 0.00015830069271391333,
+      "loss": 1.683,
+      "step": 7008
+    },
+    {
+      "epoch": 0.3228912332427328,
+      "grad_norm": 1.7412649159636728,
+      "learning_rate": 0.00015828856933435437,
+      "loss": 1.6242,
+      "step": 7009
+    },
+    {
+      "epoch": 0.3229373013313678,
+      "grad_norm": 1.4505612996294375,
+      "learning_rate": 0.0001582764446571004,
+      "loss": 1.415,
+      "step": 7010
+    },
+    {
+      "epoch": 0.3229833694200028,
+      "grad_norm": 3.6008640840124837,
+      "learning_rate": 0.00015826431868242127,
+      "loss": 1.3893,
+      "step": 7011
+    },
+    {
+      "epoch": 0.3230294375086378,
+      "grad_norm": 2.0995533008006584,
+      "learning_rate": 0.00015825219141058704,
+      "loss": 1.3632,
+      "step": 7012
+    },
+    {
+      "epoch": 0.32307550559727277,
+      "grad_norm": 1.095923017789099,
+      "learning_rate": 0.00015824006284186762,
+      "loss": 1.3247,
+      "step": 7013
+    },
+    {
+      "epoch": 0.32312157368590777,
+      "grad_norm": 1.932758908659886,
+      "learning_rate": 0.00015822793297653304,
+      "loss": 1.4357,
+      "step": 7014
+    },
+    {
+      "epoch": 0.32316764177454277,
+      "grad_norm": 1.7025917862862185,
+      "learning_rate": 0.0001582158018148534,
+      "loss": 1.5972,
+      "step": 7015
+    },
+    {
+      "epoch": 0.32321370986317777,
+      "grad_norm": 2.3058955986120884,
+      "learning_rate": 0.00015820366935709874,
+      "loss": 1.2665,
+      "step": 7016
+    },
+    {
+      "epoch": 0.32325977795181277,
+      "grad_norm": 2.369789297063595,
+      "learning_rate": 0.00015819153560353919,
+      "loss": 1.3987,
+      "step": 7017
+    },
+    {
+      "epoch": 0.32330584604044776,
+      "grad_norm": 1.2986896680185918,
+      "learning_rate": 0.00015817940055444488,
+      "loss": 1.5069,
+      "step": 7018
+    },
+    {
+      "epoch": 0.32335191412908276,
+      "grad_norm": 2.308192163329067,
+      "learning_rate": 0.00015816726421008593,
+      "loss": 1.6361,
+      "step": 7019
+    },
+    {
+      "epoch": 0.32339798221771776,
+      "grad_norm": 1.5926751617386006,
+      "learning_rate": 0.00015815512657073265,
+      "loss": 0.9861,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3234440503063528,
+      "grad_norm": 2.052297524878773,
+      "learning_rate": 0.00015814298763665513,
+      "loss": 1.2242,
+      "step": 7021
+    },
+    {
+      "epoch": 0.3234901183949878,
+      "grad_norm": 0.858191410425847,
+      "learning_rate": 0.0001581308474081237,
+      "loss": 1.2871,
+      "step": 7022
+    },
+    {
+      "epoch": 0.3235361864836228,
+      "grad_norm": 1.3623382472676788,
+      "learning_rate": 0.00015811870588540864,
+      "loss": 1.4845,
+      "step": 7023
+    },
+    {
+      "epoch": 0.3235822545722578,
+      "grad_norm": 2.248491185935069,
+      "learning_rate": 0.00015810656306878025,
+      "loss": 1.3776,
+      "step": 7024
+    },
+    {
+      "epoch": 0.3236283226608928,
+      "grad_norm": 18.73060932531364,
+      "learning_rate": 0.00015809441895850884,
+      "loss": 1.148,
+      "step": 7025
+    },
+    {
+      "epoch": 0.3236743907495278,
+      "grad_norm": 1.248059787681006,
+      "learning_rate": 0.00015808227355486486,
+      "loss": 1.2146,
+      "step": 7026
+    },
+    {
+      "epoch": 0.3237204588381628,
+      "grad_norm": 6.267432386783657,
+      "learning_rate": 0.00015807012685811858,
+      "loss": 1.2535,
+      "step": 7027
+    },
+    {
+      "epoch": 0.3237665269267978,
+      "grad_norm": 2.1556026684084997,
+      "learning_rate": 0.00015805797886854053,
+      "loss": 1.1167,
+      "step": 7028
+    },
+    {
+      "epoch": 0.3238125950154328,
+      "grad_norm": 1.925157412105725,
+      "learning_rate": 0.00015804582958640112,
+      "loss": 1.3652,
+      "step": 7029
+    },
+    {
+      "epoch": 0.3238586631040678,
+      "grad_norm": 1.2497266820067967,
+      "learning_rate": 0.00015803367901197083,
+      "loss": 1.4908,
+      "step": 7030
+    },
+    {
+      "epoch": 0.3239047311927028,
+      "grad_norm": 2.1397664539073102,
+      "learning_rate": 0.00015802152714552018,
+      "loss": 1.2693,
+      "step": 7031
+    },
+    {
+      "epoch": 0.3239507992813378,
+      "grad_norm": 1.2051095249034494,
+      "learning_rate": 0.0001580093739873197,
+      "loss": 1.4365,
+      "step": 7032
+    },
+    {
+      "epoch": 0.3239968673699728,
+      "grad_norm": 1.4747372167527724,
+      "learning_rate": 0.00015799721953764002,
+      "loss": 1.5426,
+      "step": 7033
+    },
+    {
+      "epoch": 0.3240429354586078,
+      "grad_norm": 1.1265178774983025,
+      "learning_rate": 0.00015798506379675168,
+      "loss": 1.5248,
+      "step": 7034
+    },
+    {
+      "epoch": 0.32408900354724285,
+      "grad_norm": 1.4922362235629496,
+      "learning_rate": 0.00015797290676492534,
+      "loss": 1.2839,
+      "step": 7035
+    },
+    {
+      "epoch": 0.32413507163587785,
+      "grad_norm": 1.3056967286829455,
+      "learning_rate": 0.00015796074844243158,
+      "loss": 1.499,
+      "step": 7036
+    },
+    {
+      "epoch": 0.32418113972451285,
+      "grad_norm": 1.784475200055992,
+      "learning_rate": 0.00015794858882954116,
+      "loss": 1.3001,
+      "step": 7037
+    },
+    {
+      "epoch": 0.32422720781314784,
+      "grad_norm": 1.6604238613306461,
+      "learning_rate": 0.0001579364279265248,
+      "loss": 1.2415,
+      "step": 7038
+    },
+    {
+      "epoch": 0.32427327590178284,
+      "grad_norm": 3.312672554011102,
+      "learning_rate": 0.00015792426573365314,
+      "loss": 1.5834,
+      "step": 7039
+    },
+    {
+      "epoch": 0.32431934399041784,
+      "grad_norm": 4.328278774984645,
+      "learning_rate": 0.00015791210225119708,
+      "loss": 1.2172,
+      "step": 7040
+    },
+    {
+      "epoch": 0.32436541207905284,
+      "grad_norm": 5.303235630573727,
+      "learning_rate": 0.00015789993747942734,
+      "loss": 1.2239,
+      "step": 7041
+    },
+    {
+      "epoch": 0.32441148016768784,
+      "grad_norm": 5.410953977733492,
+      "learning_rate": 0.00015788777141861475,
+      "loss": 1.454,
+      "step": 7042
+    },
+    {
+      "epoch": 0.32445754825632284,
+      "grad_norm": 3.667097206234571,
+      "learning_rate": 0.0001578756040690302,
+      "loss": 1.5809,
+      "step": 7043
+    },
+    {
+      "epoch": 0.32450361634495783,
+      "grad_norm": 1.0720507054569268,
+      "learning_rate": 0.00015786343543094459,
+      "loss": 1.3368,
+      "step": 7044
+    },
+    {
+      "epoch": 0.32454968443359283,
+      "grad_norm": 3.010059872833518,
+      "learning_rate": 0.00015785126550462878,
+      "loss": 1.3314,
+      "step": 7045
+    },
+    {
+      "epoch": 0.32459575252222783,
+      "grad_norm": 1.748798241079762,
+      "learning_rate": 0.00015783909429035377,
+      "loss": 1.3328,
+      "step": 7046
+    },
+    {
+      "epoch": 0.32464182061086283,
+      "grad_norm": 1.8146770665327503,
+      "learning_rate": 0.00015782692178839045,
+      "loss": 1.2731,
+      "step": 7047
+    },
+    {
+      "epoch": 0.3246878886994979,
+      "grad_norm": 2.1069171537854423,
+      "learning_rate": 0.0001578147479990099,
+      "loss": 1.4113,
+      "step": 7048
+    },
+    {
+      "epoch": 0.3247339567881329,
+      "grad_norm": 4.562880286052587,
+      "learning_rate": 0.0001578025729224831,
+      "loss": 1.3175,
+      "step": 7049
+    },
+    {
+      "epoch": 0.3247800248767679,
+      "grad_norm": 2.8979008786836444,
+      "learning_rate": 0.00015779039655908114,
+      "loss": 1.7077,
+      "step": 7050
+    },
+    {
+      "epoch": 0.3248260929654029,
+      "grad_norm": 1.639983619130271,
+      "learning_rate": 0.00015777821890907508,
+      "loss": 1.6373,
+      "step": 7051
+    },
+    {
+      "epoch": 0.3248721610540379,
+      "grad_norm": 2.5483107821355544,
+      "learning_rate": 0.00015776603997273605,
+      "loss": 1.4327,
+      "step": 7052
+    },
+    {
+      "epoch": 0.3249182291426729,
+      "grad_norm": 2.4192624171635706,
+      "learning_rate": 0.0001577538597503352,
+      "loss": 1.2874,
+      "step": 7053
+    },
+    {
+      "epoch": 0.3249642972313079,
+      "grad_norm": 1.2706813205210543,
+      "learning_rate": 0.00015774167824214366,
+      "loss": 1.3999,
+      "step": 7054
+    },
+    {
+      "epoch": 0.3250103653199429,
+      "grad_norm": 2.491509694554473,
+      "learning_rate": 0.00015772949544843264,
+      "loss": 1.4361,
+      "step": 7055
+    },
+    {
+      "epoch": 0.32505643340857787,
+      "grad_norm": 1.2885163936574913,
+      "learning_rate": 0.00015771731136947343,
+      "loss": 1.1621,
+      "step": 7056
+    },
+    {
+      "epoch": 0.32510250149721287,
+      "grad_norm": 4.65022918991676,
+      "learning_rate": 0.00015770512600553723,
+      "loss": 1.5585,
+      "step": 7057
+    },
+    {
+      "epoch": 0.32514856958584787,
+      "grad_norm": 1.6696711557136161,
+      "learning_rate": 0.00015769293935689535,
+      "loss": 1.0521,
+      "step": 7058
+    },
+    {
+      "epoch": 0.32519463767448287,
+      "grad_norm": 2.0587075408138706,
+      "learning_rate": 0.00015768075142381908,
+      "loss": 1.416,
+      "step": 7059
+    },
+    {
+      "epoch": 0.32524070576311787,
+      "grad_norm": 2.4348309358700906,
+      "learning_rate": 0.0001576685622065798,
+      "loss": 1.339,
+      "step": 7060
+    },
+    {
+      "epoch": 0.32528677385175286,
+      "grad_norm": 1.2881292931868513,
+      "learning_rate": 0.00015765637170544886,
+      "loss": 1.7624,
+      "step": 7061
+    },
+    {
+      "epoch": 0.3253328419403879,
+      "grad_norm": 4.336953449267513,
+      "learning_rate": 0.00015764417992069767,
+      "loss": 1.1312,
+      "step": 7062
+    },
+    {
+      "epoch": 0.3253789100290229,
+      "grad_norm": 1.4780820260283134,
+      "learning_rate": 0.00015763198685259764,
+      "loss": 1.634,
+      "step": 7063
+    },
+    {
+      "epoch": 0.3254249781176579,
+      "grad_norm": 4.0400634678023755,
+      "learning_rate": 0.00015761979250142024,
+      "loss": 1.7001,
+      "step": 7064
+    },
+    {
+      "epoch": 0.3254710462062929,
+      "grad_norm": 3.49873527379676,
+      "learning_rate": 0.00015760759686743697,
+      "loss": 1.2515,
+      "step": 7065
+    },
+    {
+      "epoch": 0.3255171142949279,
+      "grad_norm": 1.2136927032697196,
+      "learning_rate": 0.00015759539995091935,
+      "loss": 1.3873,
+      "step": 7066
+    },
+    {
+      "epoch": 0.3255631823835629,
+      "grad_norm": 1.2433349498579818,
+      "learning_rate": 0.00015758320175213885,
+      "loss": 1.7403,
+      "step": 7067
+    },
+    {
+      "epoch": 0.3256092504721979,
+      "grad_norm": 2.3264128970755507,
+      "learning_rate": 0.00015757100227136714,
+      "loss": 1.6756,
+      "step": 7068
+    },
+    {
+      "epoch": 0.3256553185608329,
+      "grad_norm": 2.637769218519418,
+      "learning_rate": 0.00015755880150887576,
+      "loss": 1.3526,
+      "step": 7069
+    },
+    {
+      "epoch": 0.3257013866494679,
+      "grad_norm": 1.4253139835850086,
+      "learning_rate": 0.00015754659946493635,
+      "loss": 1.4317,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3257474547381029,
+      "grad_norm": 3.068839997282559,
+      "learning_rate": 0.0001575343961398206,
+      "loss": 1.4055,
+      "step": 7071
+    },
+    {
+      "epoch": 0.3257935228267379,
+      "grad_norm": 0.9447741730492809,
+      "learning_rate": 0.00015752219153380013,
+      "loss": 1.6482,
+      "step": 7072
+    },
+    {
+      "epoch": 0.3258395909153729,
+      "grad_norm": 1.799955751007006,
+      "learning_rate": 0.00015750998564714673,
+      "loss": 1.0708,
+      "step": 7073
+    },
+    {
+      "epoch": 0.3258856590040079,
+      "grad_norm": 2.1682860341623456,
+      "learning_rate": 0.0001574977784801321,
+      "loss": 1.2961,
+      "step": 7074
+    },
+    {
+      "epoch": 0.32593172709264295,
+      "grad_norm": 1.7601237181895668,
+      "learning_rate": 0.000157485570033028,
+      "loss": 1.5734,
+      "step": 7075
+    },
+    {
+      "epoch": 0.32597779518127795,
+      "grad_norm": 2.287423479688956,
+      "learning_rate": 0.00015747336030610625,
+      "loss": 1.2873,
+      "step": 7076
+    },
+    {
+      "epoch": 0.32602386326991295,
+      "grad_norm": 2.4840058157485925,
+      "learning_rate": 0.00015746114929963872,
+      "loss": 1.4209,
+      "step": 7077
+    },
+    {
+      "epoch": 0.32606993135854795,
+      "grad_norm": 0.9700041836835488,
+      "learning_rate": 0.0001574489370138972,
+      "loss": 1.4053,
+      "step": 7078
+    },
+    {
+      "epoch": 0.32611599944718295,
+      "grad_norm": 1.8718549963574274,
+      "learning_rate": 0.0001574367234491536,
+      "loss": 1.5753,
+      "step": 7079
+    },
+    {
+      "epoch": 0.32616206753581795,
+      "grad_norm": 5.608608767191779,
+      "learning_rate": 0.00015742450860567985,
+      "loss": 1.3914,
+      "step": 7080
+    },
+    {
+      "epoch": 0.32620813562445294,
+      "grad_norm": 2.0727582544301004,
+      "learning_rate": 0.00015741229248374786,
+      "loss": 1.3906,
+      "step": 7081
+    },
+    {
+      "epoch": 0.32625420371308794,
+      "grad_norm": 1.1387369342625624,
+      "learning_rate": 0.00015740007508362964,
+      "loss": 1.5829,
+      "step": 7082
+    },
+    {
+      "epoch": 0.32630027180172294,
+      "grad_norm": 2.6841556705385816,
+      "learning_rate": 0.00015738785640559714,
+      "loss": 1.1671,
+      "step": 7083
+    },
+    {
+      "epoch": 0.32634633989035794,
+      "grad_norm": 1.5366718445687273,
+      "learning_rate": 0.0001573756364499224,
+      "loss": 1.3442,
+      "step": 7084
+    },
+    {
+      "epoch": 0.32639240797899294,
+      "grad_norm": 4.137994117042853,
+      "learning_rate": 0.00015736341521687755,
+      "loss": 1.4946,
+      "step": 7085
+    },
+    {
+      "epoch": 0.32643847606762794,
+      "grad_norm": 1.7127670791031924,
+      "learning_rate": 0.0001573511927067346,
+      "loss": 1.5,
+      "step": 7086
+    },
+    {
+      "epoch": 0.32648454415626293,
+      "grad_norm": 2.6703036436647816,
+      "learning_rate": 0.00015733896891976566,
+      "loss": 1.4546,
+      "step": 7087
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 2.3625752514105383,
+      "learning_rate": 0.00015732674385624288,
+      "loss": 1.2368,
+      "step": 7088
+    },
+    {
+      "epoch": 0.326576680333533,
+      "grad_norm": 1.9865260000850364,
+      "learning_rate": 0.00015731451751643845,
+      "loss": 1.5233,
+      "step": 7089
+    },
+    {
+      "epoch": 0.326622748422168,
+      "grad_norm": 2.827581654824354,
+      "learning_rate": 0.00015730228990062462,
+      "loss": 1.322,
+      "step": 7090
+    },
+    {
+      "epoch": 0.326668816510803,
+      "grad_norm": 2.3396268685622883,
+      "learning_rate": 0.0001572900610090735,
+      "loss": 1.2732,
+      "step": 7091
+    },
+    {
+      "epoch": 0.326714884599438,
+      "grad_norm": 1.2402818334549903,
+      "learning_rate": 0.00015727783084205742,
+      "loss": 1.5178,
+      "step": 7092
+    },
+    {
+      "epoch": 0.326760952688073,
+      "grad_norm": 1.708884224759119,
+      "learning_rate": 0.00015726559939984865,
+      "loss": 1.4343,
+      "step": 7093
+    },
+    {
+      "epoch": 0.326807020776708,
+      "grad_norm": 4.008998481433817,
+      "learning_rate": 0.00015725336668271947,
+      "loss": 1.2125,
+      "step": 7094
+    },
+    {
+      "epoch": 0.326853088865343,
+      "grad_norm": 18.395401794602854,
+      "learning_rate": 0.0001572411326909423,
+      "loss": 1.2458,
+      "step": 7095
+    },
+    {
+      "epoch": 0.326899156953978,
+      "grad_norm": 1.5181957435710902,
+      "learning_rate": 0.00015722889742478938,
+      "loss": 1.6801,
+      "step": 7096
+    },
+    {
+      "epoch": 0.326945225042613,
+      "grad_norm": 2.0977521005176705,
+      "learning_rate": 0.00015721666088453325,
+      "loss": 1.3142,
+      "step": 7097
+    },
+    {
+      "epoch": 0.326991293131248,
+      "grad_norm": 1.7253596419613035,
+      "learning_rate": 0.00015720442307044625,
+      "loss": 1.5579,
+      "step": 7098
+    },
+    {
+      "epoch": 0.32703736121988297,
+      "grad_norm": 1.5142615738683753,
+      "learning_rate": 0.00015719218398280087,
+      "loss": 1.4509,
+      "step": 7099
+    },
+    {
+      "epoch": 0.32708342930851797,
+      "grad_norm": 2.995818559947178,
+      "learning_rate": 0.00015717994362186957,
+      "loss": 1.2733,
+      "step": 7100
+    },
+    {
+      "epoch": 0.32712949739715297,
+      "grad_norm": 3.407837288831582,
+      "learning_rate": 0.00015716770198792487,
+      "loss": 1.5523,
+      "step": 7101
+    },
+    {
+      "epoch": 0.32717556548578797,
+      "grad_norm": 2.1256140729239124,
+      "learning_rate": 0.0001571554590812393,
+      "loss": 1.5517,
+      "step": 7102
+    },
+    {
+      "epoch": 0.327221633574423,
+      "grad_norm": 2.628384410759441,
+      "learning_rate": 0.00015714321490208544,
+      "loss": 1.1954,
+      "step": 7103
+    },
+    {
+      "epoch": 0.327267701663058,
+      "grad_norm": 1.6786689630223004,
+      "learning_rate": 0.00015713096945073587,
+      "loss": 1.386,
+      "step": 7104
+    },
+    {
+      "epoch": 0.327313769751693,
+      "grad_norm": 1.0338068434363816,
+      "learning_rate": 0.00015711872272746323,
+      "loss": 1.4171,
+      "step": 7105
+    },
+    {
+      "epoch": 0.327359837840328,
+      "grad_norm": 2.6837568011220965,
+      "learning_rate": 0.00015710647473254015,
+      "loss": 1.3165,
+      "step": 7106
+    },
+    {
+      "epoch": 0.327405905928963,
+      "grad_norm": 11.585194379759182,
+      "learning_rate": 0.00015709422546623936,
+      "loss": 1.5665,
+      "step": 7107
+    },
+    {
+      "epoch": 0.327451974017598,
+      "grad_norm": 1.3272685630436327,
+      "learning_rate": 0.00015708197492883351,
+      "loss": 1.6737,
+      "step": 7108
+    },
+    {
+      "epoch": 0.327498042106233,
+      "grad_norm": 2.2395960939726463,
+      "learning_rate": 0.00015706972312059535,
+      "loss": 1.3929,
+      "step": 7109
+    },
+    {
+      "epoch": 0.327544110194868,
+      "grad_norm": 1.0534782412365264,
+      "learning_rate": 0.0001570574700417977,
+      "loss": 1.3646,
+      "step": 7110
+    },
+    {
+      "epoch": 0.327590178283503,
+      "grad_norm": 1.9516145194915169,
+      "learning_rate": 0.00015704521569271328,
+      "loss": 1.4198,
+      "step": 7111
+    },
+    {
+      "epoch": 0.327636246372138,
+      "grad_norm": 3.0724434038103743,
+      "learning_rate": 0.00015703296007361497,
+      "loss": 1.2609,
+      "step": 7112
+    },
+    {
+      "epoch": 0.327682314460773,
+      "grad_norm": 2.278541972444565,
+      "learning_rate": 0.00015702070318477558,
+      "loss": 1.542,
+      "step": 7113
+    },
+    {
+      "epoch": 0.327728382549408,
+      "grad_norm": 3.0487874761073646,
+      "learning_rate": 0.000157008445026468,
+      "loss": 1.4808,
+      "step": 7114
+    },
+    {
+      "epoch": 0.327774450638043,
+      "grad_norm": 1.0997350662110787,
+      "learning_rate": 0.00015699618559896514,
+      "loss": 1.692,
+      "step": 7115
+    },
+    {
+      "epoch": 0.32782051872667806,
+      "grad_norm": 1.6045324566552057,
+      "learning_rate": 0.00015698392490253992,
+      "loss": 1.2881,
+      "step": 7116
+    },
+    {
+      "epoch": 0.32786658681531305,
+      "grad_norm": 1.1540206128297745,
+      "learning_rate": 0.00015697166293746534,
+      "loss": 1.2739,
+      "step": 7117
+    },
+    {
+      "epoch": 0.32791265490394805,
+      "grad_norm": 1.3198609715000025,
+      "learning_rate": 0.00015695939970401435,
+      "loss": 1.3366,
+      "step": 7118
+    },
+    {
+      "epoch": 0.32795872299258305,
+      "grad_norm": 11.599717447547244,
+      "learning_rate": 0.00015694713520246004,
+      "loss": 1.3488,
+      "step": 7119
+    },
+    {
+      "epoch": 0.32800479108121805,
+      "grad_norm": 4.450339096819857,
+      "learning_rate": 0.00015693486943307535,
+      "loss": 1.423,
+      "step": 7120
+    },
+    {
+      "epoch": 0.32805085916985305,
+      "grad_norm": 3.8411122535600746,
+      "learning_rate": 0.00015692260239613345,
+      "loss": 1.3424,
+      "step": 7121
+    },
+    {
+      "epoch": 0.32809692725848805,
+      "grad_norm": 7.2830469873203265,
+      "learning_rate": 0.00015691033409190742,
+      "loss": 1.7888,
+      "step": 7122
+    },
+    {
+      "epoch": 0.32814299534712305,
+      "grad_norm": 1.382022460588002,
+      "learning_rate": 0.00015689806452067034,
+      "loss": 1.4439,
+      "step": 7123
+    },
+    {
+      "epoch": 0.32818906343575804,
+      "grad_norm": 2.746117431075737,
+      "learning_rate": 0.00015688579368269544,
+      "loss": 1.9881,
+      "step": 7124
+    },
+    {
+      "epoch": 0.32823513152439304,
+      "grad_norm": 1.651997791330978,
+      "learning_rate": 0.00015687352157825587,
+      "loss": 1.4207,
+      "step": 7125
+    },
+    {
+      "epoch": 0.32828119961302804,
+      "grad_norm": 1.747330983738893,
+      "learning_rate": 0.00015686124820762486,
+      "loss": 1.4093,
+      "step": 7126
+    },
+    {
+      "epoch": 0.32832726770166304,
+      "grad_norm": 2.285479712330408,
+      "learning_rate": 0.00015684897357107566,
+      "loss": 1.4918,
+      "step": 7127
+    },
+    {
+      "epoch": 0.32837333579029804,
+      "grad_norm": 1.3581059817088468,
+      "learning_rate": 0.00015683669766888153,
+      "loss": 1.4938,
+      "step": 7128
+    },
+    {
+      "epoch": 0.32841940387893304,
+      "grad_norm": 2.364969009518959,
+      "learning_rate": 0.0001568244205013158,
+      "loss": 1.5609,
+      "step": 7129
+    },
+    {
+      "epoch": 0.3284654719675681,
+      "grad_norm": 1.5743163669052225,
+      "learning_rate": 0.00015681214206865175,
+      "loss": 1.1721,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3285115400562031,
+      "grad_norm": 0.9873479002995142,
+      "learning_rate": 0.0001567998623711628,
+      "loss": 1.2937,
+      "step": 7131
+    },
+    {
+      "epoch": 0.3285576081448381,
+      "grad_norm": 1.22909802094867,
+      "learning_rate": 0.0001567875814091223,
+      "loss": 1.4696,
+      "step": 7132
+    },
+    {
+      "epoch": 0.3286036762334731,
+      "grad_norm": 1.07633960931415,
+      "learning_rate": 0.00015677529918280364,
+      "loss": 1.3915,
+      "step": 7133
+    },
+    {
+      "epoch": 0.3286497443221081,
+      "grad_norm": 1.287387666919053,
+      "learning_rate": 0.00015676301569248033,
+      "loss": 1.5277,
+      "step": 7134
+    },
+    {
+      "epoch": 0.3286958124107431,
+      "grad_norm": 4.634189318402101,
+      "learning_rate": 0.0001567507309384258,
+      "loss": 1.4187,
+      "step": 7135
+    },
+    {
+      "epoch": 0.3287418804993781,
+      "grad_norm": 1.6944939573154523,
+      "learning_rate": 0.00015673844492091355,
+      "loss": 1.1422,
+      "step": 7136
+    },
+    {
+      "epoch": 0.3287879485880131,
+      "grad_norm": 1.7863480865434502,
+      "learning_rate": 0.00015672615764021709,
+      "loss": 1.442,
+      "step": 7137
+    },
+    {
+      "epoch": 0.3288340166766481,
+      "grad_norm": 1.0697868411536635,
+      "learning_rate": 0.00015671386909661,
+      "loss": 1.5713,
+      "step": 7138
+    },
+    {
+      "epoch": 0.3288800847652831,
+      "grad_norm": 1.6262658049364485,
+      "learning_rate": 0.00015670157929036584,
+      "loss": 1.2124,
+      "step": 7139
+    },
+    {
+      "epoch": 0.3289261528539181,
+      "grad_norm": 4.231867028950593,
+      "learning_rate": 0.00015668928822175826,
+      "loss": 1.4608,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3289722209425531,
+      "grad_norm": 2.7967815993232246,
+      "learning_rate": 0.00015667699589106085,
+      "loss": 1.4936,
+      "step": 7141
+    },
+    {
+      "epoch": 0.32901828903118807,
+      "grad_norm": 1.1607285969652341,
+      "learning_rate": 0.00015666470229854738,
+      "loss": 1.3758,
+      "step": 7142
+    },
+    {
+      "epoch": 0.3290643571198231,
+      "grad_norm": 1.6825254465638473,
+      "learning_rate": 0.0001566524074444914,
+      "loss": 1.2659,
+      "step": 7143
+    },
+    {
+      "epoch": 0.3291104252084581,
+      "grad_norm": 1.548893132825869,
+      "learning_rate": 0.00015664011132916674,
+      "loss": 1.3905,
+      "step": 7144
+    },
+    {
+      "epoch": 0.3291564932970931,
+      "grad_norm": 1.2344554738294506,
+      "learning_rate": 0.00015662781395284713,
+      "loss": 1.6397,
+      "step": 7145
+    },
+    {
+      "epoch": 0.3292025613857281,
+      "grad_norm": 1.079864162013653,
+      "learning_rate": 0.00015661551531580633,
+      "loss": 1.5696,
+      "step": 7146
+    },
+    {
+      "epoch": 0.3292486294743631,
+      "grad_norm": 1.8215270372872492,
+      "learning_rate": 0.00015660321541831812,
+      "loss": 1.3409,
+      "step": 7147
+    },
+    {
+      "epoch": 0.3292946975629981,
+      "grad_norm": 4.1483520703908106,
+      "learning_rate": 0.00015659091426065643,
+      "loss": 1.603,
+      "step": 7148
+    },
+    {
+      "epoch": 0.3293407656516331,
+      "grad_norm": 1.8780508781104492,
+      "learning_rate": 0.000156578611843095,
+      "loss": 1.3901,
+      "step": 7149
+    },
+    {
+      "epoch": 0.3293868337402681,
+      "grad_norm": 1.1151425097644987,
+      "learning_rate": 0.00015656630816590785,
+      "loss": 1.4622,
+      "step": 7150
+    },
+    {
+      "epoch": 0.3294329018289031,
+      "grad_norm": 1.517468961249966,
+      "learning_rate": 0.0001565540032293688,
+      "loss": 1.3932,
+      "step": 7151
+    },
+    {
+      "epoch": 0.3294789699175381,
+      "grad_norm": 1.0351779670289936,
+      "learning_rate": 0.00015654169703375188,
+      "loss": 1.5493,
+      "step": 7152
+    },
+    {
+      "epoch": 0.3295250380061731,
+      "grad_norm": 1.3180033199308363,
+      "learning_rate": 0.00015652938957933099,
+      "loss": 1.4659,
+      "step": 7153
+    },
+    {
+      "epoch": 0.3295711060948081,
+      "grad_norm": 2.6842580831809255,
+      "learning_rate": 0.00015651708086638017,
+      "loss": 1.6168,
+      "step": 7154
+    },
+    {
+      "epoch": 0.3296171741834431,
+      "grad_norm": 1.537905699939841,
+      "learning_rate": 0.00015650477089517345,
+      "loss": 1.1885,
+      "step": 7155
+    },
+    {
+      "epoch": 0.3296632422720781,
+      "grad_norm": 1.2194831032896847,
+      "learning_rate": 0.00015649245966598493,
+      "loss": 1.5372,
+      "step": 7156
+    },
+    {
+      "epoch": 0.32970931036071316,
+      "grad_norm": 1.2964337357419065,
+      "learning_rate": 0.0001564801471790886,
+      "loss": 1.4973,
+      "step": 7157
+    },
+    {
+      "epoch": 0.32975537844934816,
+      "grad_norm": 1.3854104176208237,
+      "learning_rate": 0.00015646783343475866,
+      "loss": 1.4364,
+      "step": 7158
+    },
+    {
+      "epoch": 0.32980144653798316,
+      "grad_norm": 1.0027703642840329,
+      "learning_rate": 0.00015645551843326922,
+      "loss": 1.4044,
+      "step": 7159
+    },
+    {
+      "epoch": 0.32984751462661815,
+      "grad_norm": 9.050563476696006,
+      "learning_rate": 0.00015644320217489448,
+      "loss": 1.5199,
+      "step": 7160
+    },
+    {
+      "epoch": 0.32989358271525315,
+      "grad_norm": 1.085663722040599,
+      "learning_rate": 0.0001564308846599086,
+      "loss": 1.238,
+      "step": 7161
+    },
+    {
+      "epoch": 0.32993965080388815,
+      "grad_norm": 1.5217834859089967,
+      "learning_rate": 0.00015641856588858585,
+      "loss": 1.4133,
+      "step": 7162
+    },
+    {
+      "epoch": 0.32998571889252315,
+      "grad_norm": 2.498610074129514,
+      "learning_rate": 0.00015640624586120048,
+      "loss": 1.412,
+      "step": 7163
+    },
+    {
+      "epoch": 0.33003178698115815,
+      "grad_norm": 14.123161253245872,
+      "learning_rate": 0.00015639392457802675,
+      "loss": 1.3275,
+      "step": 7164
+    },
+    {
+      "epoch": 0.33007785506979315,
+      "grad_norm": 1.566604655466108,
+      "learning_rate": 0.000156381602039339,
+      "loss": 1.4377,
+      "step": 7165
+    },
+    {
+      "epoch": 0.33012392315842815,
+      "grad_norm": 1.9202573369474423,
+      "learning_rate": 0.0001563692782454115,
+      "loss": 1.1485,
+      "step": 7166
+    },
+    {
+      "epoch": 0.33016999124706314,
+      "grad_norm": 1.244094039221612,
+      "learning_rate": 0.00015635695319651872,
+      "loss": 1.4969,
+      "step": 7167
+    },
+    {
+      "epoch": 0.33021605933569814,
+      "grad_norm": 1.753609904932423,
+      "learning_rate": 0.000156344626892935,
+      "loss": 1.4712,
+      "step": 7168
+    },
+    {
+      "epoch": 0.33026212742433314,
+      "grad_norm": 2.0358352167235125,
+      "learning_rate": 0.00015633229933493476,
+      "loss": 1.2786,
+      "step": 7169
+    },
+    {
+      "epoch": 0.3303081955129682,
+      "grad_norm": 2.9965284842916606,
+      "learning_rate": 0.00015631997052279248,
+      "loss": 1.2,
+      "step": 7170
+    },
+    {
+      "epoch": 0.3303542636016032,
+      "grad_norm": 1.5785530874473155,
+      "learning_rate": 0.0001563076404567826,
+      "loss": 1.3785,
+      "step": 7171
+    },
+    {
+      "epoch": 0.3304003316902382,
+      "grad_norm": 1.753359715721062,
+      "learning_rate": 0.0001562953091371797,
+      "loss": 1.1645,
+      "step": 7172
+    },
+    {
+      "epoch": 0.3304463997788732,
+      "grad_norm": 2.3377259970676603,
+      "learning_rate": 0.00015628297656425824,
+      "loss": 1.4399,
+      "step": 7173
+    },
+    {
+      "epoch": 0.3304924678675082,
+      "grad_norm": 2.2669998344328546,
+      "learning_rate": 0.00015627064273829278,
+      "loss": 1.4942,
+      "step": 7174
+    },
+    {
+      "epoch": 0.3305385359561432,
+      "grad_norm": 1.408660519193226,
+      "learning_rate": 0.00015625830765955798,
+      "loss": 1.4726,
+      "step": 7175
+    },
+    {
+      "epoch": 0.3305846040447782,
+      "grad_norm": 2.8808447959552845,
+      "learning_rate": 0.0001562459713283284,
+      "loss": 1.5746,
+      "step": 7176
+    },
+    {
+      "epoch": 0.3306306721334132,
+      "grad_norm": 2.8272004638068875,
+      "learning_rate": 0.00015623363374487874,
+      "loss": 1.6407,
+      "step": 7177
+    },
+    {
+      "epoch": 0.3306767402220482,
+      "grad_norm": 2.1996926835080783,
+      "learning_rate": 0.00015622129490948364,
+      "loss": 1.219,
+      "step": 7178
+    },
+    {
+      "epoch": 0.3307228083106832,
+      "grad_norm": 3.800973154416764,
+      "learning_rate": 0.0001562089548224178,
+      "loss": 1.3336,
+      "step": 7179
+    },
+    {
+      "epoch": 0.3307688763993182,
+      "grad_norm": 2.3161751028264828,
+      "learning_rate": 0.00015619661348395596,
+      "loss": 1.4833,
+      "step": 7180
+    },
+    {
+      "epoch": 0.3308149444879532,
+      "grad_norm": 2.048085600784921,
+      "learning_rate": 0.00015618427089437287,
+      "loss": 1.4392,
+      "step": 7181
+    },
+    {
+      "epoch": 0.3308610125765882,
+      "grad_norm": 1.341614634502616,
+      "learning_rate": 0.00015617192705394333,
+      "loss": 1.3161,
+      "step": 7182
+    },
+    {
+      "epoch": 0.3309070806652232,
+      "grad_norm": 3.61943055783747,
+      "learning_rate": 0.00015615958196294214,
+      "loss": 1.6262,
+      "step": 7183
+    },
+    {
+      "epoch": 0.33095314875385823,
+      "grad_norm": 1.4746041468452487,
+      "learning_rate": 0.00015614723562164414,
+      "loss": 1.55,
+      "step": 7184
+    },
+    {
+      "epoch": 0.3309992168424932,
+      "grad_norm": 1.527636050438229,
+      "learning_rate": 0.00015613488803032424,
+      "loss": 1.3734,
+      "step": 7185
+    },
+    {
+      "epoch": 0.3310452849311282,
+      "grad_norm": 2.5624791377112497,
+      "learning_rate": 0.00015612253918925728,
+      "loss": 1.6015,
+      "step": 7186
+    },
+    {
+      "epoch": 0.3310913530197632,
+      "grad_norm": 2.044477298099,
+      "learning_rate": 0.0001561101890987182,
+      "loss": 1.1845,
+      "step": 7187
+    },
+    {
+      "epoch": 0.3311374211083982,
+      "grad_norm": 13.06351969900875,
+      "learning_rate": 0.00015609783775898204,
+      "loss": 1.4414,
+      "step": 7188
+    },
+    {
+      "epoch": 0.3311834891970332,
+      "grad_norm": 1.8534221860923528,
+      "learning_rate": 0.00015608548517032363,
+      "loss": 1.3327,
+      "step": 7189
+    },
+    {
+      "epoch": 0.3312295572856682,
+      "grad_norm": 1.6589144306852404,
+      "learning_rate": 0.00015607313133301812,
+      "loss": 1.0584,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3312756253743032,
+      "grad_norm": 1.2815479936642369,
+      "learning_rate": 0.00015606077624734046,
+      "loss": 1.6844,
+      "step": 7191
+    },
+    {
+      "epoch": 0.3313216934629382,
+      "grad_norm": 1.7142794664098724,
+      "learning_rate": 0.00015604841991356575,
+      "loss": 1.5211,
+      "step": 7192
+    },
+    {
+      "epoch": 0.3313677615515732,
+      "grad_norm": 4.890988142553321,
+      "learning_rate": 0.0001560360623319691,
+      "loss": 1.5844,
+      "step": 7193
+    },
+    {
+      "epoch": 0.3314138296402082,
+      "grad_norm": 3.095578544984074,
+      "learning_rate": 0.00015602370350282553,
+      "loss": 1.2308,
+      "step": 7194
+    },
+    {
+      "epoch": 0.3314598977288432,
+      "grad_norm": 2.3091094406034407,
+      "learning_rate": 0.0001560113434264103,
+      "loss": 1.3672,
+      "step": 7195
+    },
+    {
+      "epoch": 0.3315059658174782,
+      "grad_norm": 1.808595473753934,
+      "learning_rate": 0.00015599898210299853,
+      "loss": 1.6195,
+      "step": 7196
+    },
+    {
+      "epoch": 0.3315520339061132,
+      "grad_norm": 1.5339005982727714,
+      "learning_rate": 0.00015598661953286546,
+      "loss": 1.5568,
+      "step": 7197
+    },
+    {
+      "epoch": 0.33159810199474826,
+      "grad_norm": 1.528370066717773,
+      "learning_rate": 0.0001559742557162863,
+      "loss": 1.5126,
+      "step": 7198
+    },
+    {
+      "epoch": 0.33164417008338326,
+      "grad_norm": 1.047503207568482,
+      "learning_rate": 0.00015596189065353634,
+      "loss": 1.3783,
+      "step": 7199
+    },
+    {
+      "epoch": 0.33169023817201826,
+      "grad_norm": 3.376796051790754,
+      "learning_rate": 0.0001559495243448908,
+      "loss": 1.5796,
+      "step": 7200
+    },
+    {
+      "epoch": 0.33173630626065326,
+      "grad_norm": 1.985088713952778,
+      "learning_rate": 0.00015593715679062506,
+      "loss": 1.3741,
+      "step": 7201
+    },
+    {
+      "epoch": 0.33178237434928826,
+      "grad_norm": 1.4383308625558358,
+      "learning_rate": 0.0001559247879910144,
+      "loss": 1.4147,
+      "step": 7202
+    },
+    {
+      "epoch": 0.33182844243792325,
+      "grad_norm": 11.005266729639535,
+      "learning_rate": 0.00015591241794633427,
+      "loss": 1.0665,
+      "step": 7203
+    },
+    {
+      "epoch": 0.33187451052655825,
+      "grad_norm": 1.2785185648495845,
+      "learning_rate": 0.00015590004665685998,
+      "loss": 1.2776,
+      "step": 7204
+    },
+    {
+      "epoch": 0.33192057861519325,
+      "grad_norm": 1.763677696351982,
+      "learning_rate": 0.00015588767412286702,
+      "loss": 1.3868,
+      "step": 7205
+    },
+    {
+      "epoch": 0.33196664670382825,
+      "grad_norm": 1.8992113685604481,
+      "learning_rate": 0.0001558753003446308,
+      "loss": 1.519,
+      "step": 7206
+    },
+    {
+      "epoch": 0.33201271479246325,
+      "grad_norm": 4.937328867059511,
+      "learning_rate": 0.00015586292532242686,
+      "loss": 1.4297,
+      "step": 7207
+    },
+    {
+      "epoch": 0.33205878288109825,
+      "grad_norm": 2.2437161660552567,
+      "learning_rate": 0.00015585054905653065,
+      "loss": 1.396,
+      "step": 7208
+    },
+    {
+      "epoch": 0.33210485096973325,
+      "grad_norm": 1.1088734342362974,
+      "learning_rate": 0.00015583817154721775,
+      "loss": 1.2965,
+      "step": 7209
+    },
+    {
+      "epoch": 0.33215091905836824,
+      "grad_norm": 1.404632588481426,
+      "learning_rate": 0.00015582579279476368,
+      "loss": 1.3757,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3321969871470033,
+      "grad_norm": 142.74247887982497,
+      "learning_rate": 0.00015581341279944408,
+      "loss": 1.4735,
+      "step": 7211
+    },
+    {
+      "epoch": 0.3322430552356383,
+      "grad_norm": 6.247884292017664,
+      "learning_rate": 0.00015580103156153448,
+      "loss": 2.0444,
+      "step": 7212
+    },
+    {
+      "epoch": 0.3322891233242733,
+      "grad_norm": 13.787880067586125,
+      "learning_rate": 0.00015578864908131067,
+      "loss": 2.9314,
+      "step": 7213
+    },
+    {
+      "epoch": 0.3323351914129083,
+      "grad_norm": 5.150645981882962,
+      "learning_rate": 0.0001557762653590482,
+      "loss": 2.1166,
+      "step": 7214
+    },
+    {
+      "epoch": 0.3323812595015433,
+      "grad_norm": 1.9198144004133388,
+      "learning_rate": 0.00015576388039502284,
+      "loss": 1.8548,
+      "step": 7215
+    },
+    {
+      "epoch": 0.3324273275901783,
+      "grad_norm": 5.635712626259693,
+      "learning_rate": 0.00015575149418951028,
+      "loss": 2.1129,
+      "step": 7216
+    },
+    {
+      "epoch": 0.3324733956788133,
+      "grad_norm": 1.79836135279427,
+      "learning_rate": 0.0001557391067427863,
+      "loss": 2.1429,
+      "step": 7217
+    },
+    {
+      "epoch": 0.3325194637674483,
+      "grad_norm": 2.5936588352183736,
+      "learning_rate": 0.0001557267180551267,
+      "loss": 1.4834,
+      "step": 7218
+    },
+    {
+      "epoch": 0.3325655318560833,
+      "grad_norm": 3.245841142610312,
+      "learning_rate": 0.00015571432812680724,
+      "loss": 2.0787,
+      "step": 7219
+    },
+    {
+      "epoch": 0.3326115999447183,
+      "grad_norm": 2.3860972533537663,
+      "learning_rate": 0.00015570193695810383,
+      "loss": 2.0613,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3326576680333533,
+      "grad_norm": 2.0213244490833033,
+      "learning_rate": 0.00015568954454929225,
+      "loss": 1.7067,
+      "step": 7221
+    },
+    {
+      "epoch": 0.3327037361219883,
+      "grad_norm": 2.1039434393735648,
+      "learning_rate": 0.00015567715090064848,
+      "loss": 1.7714,
+      "step": 7222
+    },
+    {
+      "epoch": 0.3327498042106233,
+      "grad_norm": 2.1585424633298786,
+      "learning_rate": 0.0001556647560124484,
+      "loss": 1.5314,
+      "step": 7223
+    },
+    {
+      "epoch": 0.3327958722992583,
+      "grad_norm": 1.4713922740181982,
+      "learning_rate": 0.000155652359884968,
+      "loss": 1.8169,
+      "step": 7224
+    },
+    {
+      "epoch": 0.33284194038789333,
+      "grad_norm": 1.8168763387456344,
+      "learning_rate": 0.00015563996251848322,
+      "loss": 1.5856,
+      "step": 7225
+    },
+    {
+      "epoch": 0.33288800847652833,
+      "grad_norm": 1.8773797602655677,
+      "learning_rate": 0.00015562756391327004,
+      "loss": 1.4375,
+      "step": 7226
+    },
+    {
+      "epoch": 0.33293407656516333,
+      "grad_norm": 3.1578108098798277,
+      "learning_rate": 0.0001556151640696046,
+      "loss": 1.6017,
+      "step": 7227
+    },
+    {
+      "epoch": 0.3329801446537983,
+      "grad_norm": 2.584470202016077,
+      "learning_rate": 0.0001556027629877628,
+      "loss": 1.5122,
+      "step": 7228
+    },
+    {
+      "epoch": 0.3330262127424333,
+      "grad_norm": 3.1085226175941876,
+      "learning_rate": 0.00015559036066802091,
+      "loss": 1.8948,
+      "step": 7229
+    },
+    {
+      "epoch": 0.3330722808310683,
+      "grad_norm": 1.2218478253699805,
+      "learning_rate": 0.0001555779571106549,
+      "loss": 1.7518,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3331183489197033,
+      "grad_norm": 1.294611413892417,
+      "learning_rate": 0.00015556555231594098,
+      "loss": 1.3991,
+      "step": 7231
+    },
+    {
+      "epoch": 0.3331644170083383,
+      "grad_norm": 2.006561257116689,
+      "learning_rate": 0.00015555314628415534,
+      "loss": 1.4189,
+      "step": 7232
+    },
+    {
+      "epoch": 0.3332104850969733,
+      "grad_norm": 1.779776423880052,
+      "learning_rate": 0.00015554073901557412,
+      "loss": 1.4703,
+      "step": 7233
+    },
+    {
+      "epoch": 0.3332565531856083,
+      "grad_norm": 1.4710141041587965,
+      "learning_rate": 0.0001555283305104736,
+      "loss": 1.756,
+      "step": 7234
+    },
+    {
+      "epoch": 0.3333026212742433,
+      "grad_norm": 2.854015217416846,
+      "learning_rate": 0.00015551592076913,
+      "loss": 1.6728,
+      "step": 7235
+    },
+    {
+      "epoch": 0.3333486893628783,
+      "grad_norm": 1.9092052934809247,
+      "learning_rate": 0.00015550350979181963,
+      "loss": 1.7252,
+      "step": 7236
+    },
+    {
+      "epoch": 0.3333947574515133,
+      "grad_norm": 2.1901172628248946,
+      "learning_rate": 0.00015549109757881876,
+      "loss": 1.7359,
+      "step": 7237
+    },
+    {
+      "epoch": 0.33344082554014837,
+      "grad_norm": 2.7388972135638365,
+      "learning_rate": 0.00015547868413040377,
+      "loss": 1.6588,
+      "step": 7238
+    },
+    {
+      "epoch": 0.33348689362878337,
+      "grad_norm": 2.40571112006173,
+      "learning_rate": 0.00015546626944685097,
+      "loss": 1.5617,
+      "step": 7239
+    },
+    {
+      "epoch": 0.33353296171741836,
+      "grad_norm": 4.3172985972562765,
+      "learning_rate": 0.0001554538535284368,
+      "loss": 1.4501,
+      "step": 7240
+    },
+    {
+      "epoch": 0.33357902980605336,
+      "grad_norm": 2.073440733164429,
+      "learning_rate": 0.00015544143637543766,
+      "loss": 1.5812,
+      "step": 7241
+    },
+    {
+      "epoch": 0.33362509789468836,
+      "grad_norm": 1.4516134821543347,
+      "learning_rate": 0.00015542901798813,
+      "loss": 1.7313,
+      "step": 7242
+    },
+    {
+      "epoch": 0.33367116598332336,
+      "grad_norm": 2.600831294800713,
+      "learning_rate": 0.00015541659836679032,
+      "loss": 1.1154,
+      "step": 7243
+    },
+    {
+      "epoch": 0.33371723407195836,
+      "grad_norm": 2.4929242152802358,
+      "learning_rate": 0.00015540417751169509,
+      "loss": 1.588,
+      "step": 7244
+    },
+    {
+      "epoch": 0.33376330216059336,
+      "grad_norm": 1.8870257605917076,
+      "learning_rate": 0.00015539175542312084,
+      "loss": 1.6354,
+      "step": 7245
+    },
+    {
+      "epoch": 0.33380937024922835,
+      "grad_norm": 6.456046813592851,
+      "learning_rate": 0.00015537933210134415,
+      "loss": 1.5054,
+      "step": 7246
+    },
+    {
+      "epoch": 0.33385543833786335,
+      "grad_norm": 1.9482390939845255,
+      "learning_rate": 0.00015536690754664156,
+      "loss": 1.6141,
+      "step": 7247
+    },
+    {
+      "epoch": 0.33390150642649835,
+      "grad_norm": 2.373001632217907,
+      "learning_rate": 0.0001553544817592897,
+      "loss": 1.5213,
+      "step": 7248
+    },
+    {
+      "epoch": 0.33394757451513335,
+      "grad_norm": 4.489740297830298,
+      "learning_rate": 0.00015534205473956522,
+      "loss": 1.3854,
+      "step": 7249
+    },
+    {
+      "epoch": 0.33399364260376835,
+      "grad_norm": 1.8258470848602641,
+      "learning_rate": 0.0001553296264877448,
+      "loss": 1.5567,
+      "step": 7250
+    },
+    {
+      "epoch": 0.33403971069240335,
+      "grad_norm": 1.9241484994821576,
+      "learning_rate": 0.0001553171970041051,
+      "loss": 1.5928,
+      "step": 7251
+    },
+    {
+      "epoch": 0.3340857787810384,
+      "grad_norm": 1.5373122863364537,
+      "learning_rate": 0.00015530476628892285,
+      "loss": 1.6185,
+      "step": 7252
+    },
+    {
+      "epoch": 0.3341318468696734,
+      "grad_norm": 2.220488828873502,
+      "learning_rate": 0.0001552923343424748,
+      "loss": 1.582,
+      "step": 7253
+    },
+    {
+      "epoch": 0.3341779149583084,
+      "grad_norm": 3.4811547629142834,
+      "learning_rate": 0.00015527990116503778,
+      "loss": 1.6765,
+      "step": 7254
+    },
+    {
+      "epoch": 0.3342239830469434,
+      "grad_norm": 2.016438367782537,
+      "learning_rate": 0.0001552674667568885,
+      "loss": 1.4946,
+      "step": 7255
+    },
+    {
+      "epoch": 0.3342700511355784,
+      "grad_norm": 3.3765753094268147,
+      "learning_rate": 0.00015525503111830387,
+      "loss": 1.2942,
+      "step": 7256
+    },
+    {
+      "epoch": 0.3343161192242134,
+      "grad_norm": 1.485751584314414,
+      "learning_rate": 0.0001552425942495607,
+      "loss": 1.7356,
+      "step": 7257
+    },
+    {
+      "epoch": 0.3343621873128484,
+      "grad_norm": 2.1152994838672257,
+      "learning_rate": 0.0001552301561509359,
+      "loss": 1.5554,
+      "step": 7258
+    },
+    {
+      "epoch": 0.3344082554014834,
+      "grad_norm": 2.442139008189935,
+      "learning_rate": 0.00015521771682270634,
+      "loss": 1.3928,
+      "step": 7259
+    },
+    {
+      "epoch": 0.3344543234901184,
+      "grad_norm": 2.137612046493582,
+      "learning_rate": 0.00015520527626514902,
+      "loss": 1.4688,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3345003915787534,
+      "grad_norm": 1.6844769120932803,
+      "learning_rate": 0.0001551928344785409,
+      "loss": 1.5718,
+      "step": 7261
+    },
+    {
+      "epoch": 0.3345464596673884,
+      "grad_norm": 2.8071630920312485,
+      "learning_rate": 0.00015518039146315891,
+      "loss": 1.2757,
+      "step": 7262
+    },
+    {
+      "epoch": 0.3345925277560234,
+      "grad_norm": 1.9874845664650167,
+      "learning_rate": 0.00015516794721928014,
+      "loss": 1.5155,
+      "step": 7263
+    },
+    {
+      "epoch": 0.3346385958446584,
+      "grad_norm": 2.554389698203475,
+      "learning_rate": 0.00015515550174718162,
+      "loss": 1.4296,
+      "step": 7264
+    },
+    {
+      "epoch": 0.3346846639332934,
+      "grad_norm": 2.3338227031942083,
+      "learning_rate": 0.0001551430550471404,
+      "loss": 1.4532,
+      "step": 7265
+    },
+    {
+      "epoch": 0.33473073202192843,
+      "grad_norm": 4.046837239329892,
+      "learning_rate": 0.00015513060711943365,
+      "loss": 1.4528,
+      "step": 7266
+    },
+    {
+      "epoch": 0.33477680011056343,
+      "grad_norm": 11.931313200203883,
+      "learning_rate": 0.00015511815796433843,
+      "loss": 1.6676,
+      "step": 7267
+    },
+    {
+      "epoch": 0.33482286819919843,
+      "grad_norm": 1.948159223852034,
+      "learning_rate": 0.00015510570758213197,
+      "loss": 1.6589,
+      "step": 7268
+    },
+    {
+      "epoch": 0.33486893628783343,
+      "grad_norm": 2.6926379705619303,
+      "learning_rate": 0.00015509325597309136,
+      "loss": 1.5022,
+      "step": 7269
+    },
+    {
+      "epoch": 0.33491500437646843,
+      "grad_norm": 2.1152330027018555,
+      "learning_rate": 0.00015508080313749391,
+      "loss": 1.329,
+      "step": 7270
+    },
+    {
+      "epoch": 0.3349610724651034,
+      "grad_norm": 5.057165924709499,
+      "learning_rate": 0.00015506834907561681,
+      "loss": 1.4193,
+      "step": 7271
+    },
+    {
+      "epoch": 0.3350071405537384,
+      "grad_norm": 2.1980850927206257,
+      "learning_rate": 0.00015505589378773732,
+      "loss": 1.416,
+      "step": 7272
+    },
+    {
+      "epoch": 0.3350532086423734,
+      "grad_norm": 4.235001453713233,
+      "learning_rate": 0.0001550434372741328,
+      "loss": 1.505,
+      "step": 7273
+    },
+    {
+      "epoch": 0.3350992767310084,
+      "grad_norm": 2.3491385855458393,
+      "learning_rate": 0.0001550309795350805,
+      "loss": 1.842,
+      "step": 7274
+    },
+    {
+      "epoch": 0.3351453448196434,
+      "grad_norm": 3.161296398692195,
+      "learning_rate": 0.0001550185205708578,
+      "loss": 1.6777,
+      "step": 7275
+    },
+    {
+      "epoch": 0.3351914129082784,
+      "grad_norm": 4.88297903808044,
+      "learning_rate": 0.00015500606038174207,
+      "loss": 1.596,
+      "step": 7276
+    },
+    {
+      "epoch": 0.3352374809969134,
+      "grad_norm": 1.69899578536991,
+      "learning_rate": 0.00015499359896801074,
+      "loss": 1.5195,
+      "step": 7277
+    },
+    {
+      "epoch": 0.3352835490855484,
+      "grad_norm": 2.3507669501264585,
+      "learning_rate": 0.0001549811363299412,
+      "loss": 1.5153,
+      "step": 7278
+    },
+    {
+      "epoch": 0.33532961717418347,
+      "grad_norm": 1.7416353160139477,
+      "learning_rate": 0.00015496867246781094,
+      "loss": 1.5922,
+      "step": 7279
+    },
+    {
+      "epoch": 0.33537568526281847,
+      "grad_norm": 2.383982054294897,
+      "learning_rate": 0.00015495620738189745,
+      "loss": 1.3383,
+      "step": 7280
+    },
+    {
+      "epoch": 0.33542175335145347,
+      "grad_norm": 1.9304319688310445,
+      "learning_rate": 0.00015494374107247823,
+      "loss": 1.475,
+      "step": 7281
+    },
+    {
+      "epoch": 0.33546782144008847,
+      "grad_norm": 2.2495908045479363,
+      "learning_rate": 0.0001549312735398308,
+      "loss": 1.3534,
+      "step": 7282
+    },
+    {
+      "epoch": 0.33551388952872346,
+      "grad_norm": 3.4672919927093098,
+      "learning_rate": 0.0001549188047842328,
+      "loss": 1.6371,
+      "step": 7283
+    },
+    {
+      "epoch": 0.33555995761735846,
+      "grad_norm": 1.771610724835763,
+      "learning_rate": 0.00015490633480596175,
+      "loss": 1.6002,
+      "step": 7284
+    },
+    {
+      "epoch": 0.33560602570599346,
+      "grad_norm": 2.5000581533879416,
+      "learning_rate": 0.00015489386360529526,
+      "loss": 1.3937,
+      "step": 7285
+    },
+    {
+      "epoch": 0.33565209379462846,
+      "grad_norm": 2.0657316189752186,
+      "learning_rate": 0.00015488139118251108,
+      "loss": 1.5922,
+      "step": 7286
+    },
+    {
+      "epoch": 0.33569816188326346,
+      "grad_norm": 1.7288905285755098,
+      "learning_rate": 0.0001548689175378868,
+      "loss": 1.4071,
+      "step": 7287
+    },
+    {
+      "epoch": 0.33574422997189846,
+      "grad_norm": 1.45459058413561,
+      "learning_rate": 0.00015485644267170017,
+      "loss": 1.6676,
+      "step": 7288
+    },
+    {
+      "epoch": 0.33579029806053345,
+      "grad_norm": 1.645130015550744,
+      "learning_rate": 0.00015484396658422886,
+      "loss": 1.4095,
+      "step": 7289
+    },
+    {
+      "epoch": 0.33583636614916845,
+      "grad_norm": 1.2827652548744384,
+      "learning_rate": 0.00015483148927575072,
+      "loss": 1.6749,
+      "step": 7290
+    },
+    {
+      "epoch": 0.33588243423780345,
+      "grad_norm": 3.600738657806925,
+      "learning_rate": 0.0001548190107465435,
+      "loss": 1.5779,
+      "step": 7291
+    },
+    {
+      "epoch": 0.33592850232643845,
+      "grad_norm": 3.6622725734139343,
+      "learning_rate": 0.00015480653099688493,
+      "loss": 1.5391,
+      "step": 7292
+    },
+    {
+      "epoch": 0.3359745704150735,
+      "grad_norm": 1.5923723384328274,
+      "learning_rate": 0.00015479405002705296,
+      "loss": 1.52,
+      "step": 7293
+    },
+    {
+      "epoch": 0.3360206385037085,
+      "grad_norm": 1.367371479035501,
+      "learning_rate": 0.0001547815678373254,
+      "loss": 1.641,
+      "step": 7294
+    },
+    {
+      "epoch": 0.3360667065923435,
+      "grad_norm": 1.1027308396771935,
+      "learning_rate": 0.00015476908442798017,
+      "loss": 1.6473,
+      "step": 7295
+    },
+    {
+      "epoch": 0.3361127746809785,
+      "grad_norm": 1.76530366220652,
+      "learning_rate": 0.0001547565997992952,
+      "loss": 1.3078,
+      "step": 7296
+    },
+    {
+      "epoch": 0.3361588427696135,
+      "grad_norm": 1.082151810380404,
+      "learning_rate": 0.00015474411395154843,
+      "loss": 1.4252,
+      "step": 7297
+    },
+    {
+      "epoch": 0.3362049108582485,
+      "grad_norm": 1.8711348031942912,
+      "learning_rate": 0.0001547316268850178,
+      "loss": 1.3732,
+      "step": 7298
+    },
+    {
+      "epoch": 0.3362509789468835,
+      "grad_norm": 1.7841204201273795,
+      "learning_rate": 0.00015471913859998134,
+      "loss": 1.5758,
+      "step": 7299
+    },
+    {
+      "epoch": 0.3362970470355185,
+      "grad_norm": 1.4673125473567923,
+      "learning_rate": 0.0001547066490967171,
+      "loss": 1.2803,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3363431151241535,
+      "grad_norm": 1.5502564898007372,
+      "learning_rate": 0.00015469415837550315,
+      "loss": 1.3251,
+      "step": 7301
+    },
+    {
+      "epoch": 0.3363891832127885,
+      "grad_norm": 1.3874987373591476,
+      "learning_rate": 0.00015468166643661747,
+      "loss": 1.5799,
+      "step": 7302
+    },
+    {
+      "epoch": 0.3364352513014235,
+      "grad_norm": 4.717627589837179,
+      "learning_rate": 0.00015466917328033829,
+      "loss": 1.5023,
+      "step": 7303
+    },
+    {
+      "epoch": 0.3364813193900585,
+      "grad_norm": 2.0443800866545088,
+      "learning_rate": 0.00015465667890694372,
+      "loss": 1.3878,
+      "step": 7304
+    },
+    {
+      "epoch": 0.3365273874786935,
+      "grad_norm": 1.1540698148189872,
+      "learning_rate": 0.00015464418331671186,
+      "loss": 1.6832,
+      "step": 7305
+    },
+    {
+      "epoch": 0.33657345556732854,
+      "grad_norm": 1.1962307550846336,
+      "learning_rate": 0.000154631686509921,
+      "loss": 1.3829,
+      "step": 7306
+    },
+    {
+      "epoch": 0.33661952365596354,
+      "grad_norm": 2.3930055045366014,
+      "learning_rate": 0.00015461918848684928,
+      "loss": 1.259,
+      "step": 7307
+    },
+    {
+      "epoch": 0.33666559174459854,
+      "grad_norm": 1.9521509237697152,
+      "learning_rate": 0.000154606689247775,
+      "loss": 1.2682,
+      "step": 7308
+    },
+    {
+      "epoch": 0.33671165983323353,
+      "grad_norm": 1.6449650029641267,
+      "learning_rate": 0.00015459418879297643,
+      "loss": 1.3978,
+      "step": 7309
+    },
+    {
+      "epoch": 0.33675772792186853,
+      "grad_norm": 1.477660824424239,
+      "learning_rate": 0.00015458168712273183,
+      "loss": 1.0519,
+      "step": 7310
+    },
+    {
+      "epoch": 0.33680379601050353,
+      "grad_norm": 1.8622142274718763,
+      "learning_rate": 0.00015456918423731956,
+      "loss": 1.465,
+      "step": 7311
+    },
+    {
+      "epoch": 0.33684986409913853,
+      "grad_norm": 1.1953048493428107,
+      "learning_rate": 0.00015455668013701797,
+      "loss": 1.4779,
+      "step": 7312
+    },
+    {
+      "epoch": 0.33689593218777353,
+      "grad_norm": 1.6251292060018903,
+      "learning_rate": 0.00015454417482210544,
+      "loss": 1.6442,
+      "step": 7313
+    },
+    {
+      "epoch": 0.3369420002764085,
+      "grad_norm": 1.2729711608938832,
+      "learning_rate": 0.00015453166829286036,
+      "loss": 1.5968,
+      "step": 7314
+    },
+    {
+      "epoch": 0.3369880683650435,
+      "grad_norm": 1.7265280022095517,
+      "learning_rate": 0.00015451916054956124,
+      "loss": 0.9461,
+      "step": 7315
+    },
+    {
+      "epoch": 0.3370341364536785,
+      "grad_norm": 1.7390245175101637,
+      "learning_rate": 0.00015450665159248646,
+      "loss": 1.7051,
+      "step": 7316
+    },
+    {
+      "epoch": 0.3370802045423135,
+      "grad_norm": 1.728508908115874,
+      "learning_rate": 0.00015449414142191452,
+      "loss": 1.6939,
+      "step": 7317
+    },
+    {
+      "epoch": 0.3371262726309485,
+      "grad_norm": 3.0631730810576427,
+      "learning_rate": 0.00015448163003812398,
+      "loss": 1.5406,
+      "step": 7318
+    },
+    {
+      "epoch": 0.3371723407195835,
+      "grad_norm": 2.1439116308398516,
+      "learning_rate": 0.00015446911744139338,
+      "loss": 1.5186,
+      "step": 7319
+    },
+    {
+      "epoch": 0.3372184088082186,
+      "grad_norm": 1.563259248876094,
+      "learning_rate": 0.00015445660363200126,
+      "loss": 1.1759,
+      "step": 7320
+    },
+    {
+      "epoch": 0.33726447689685357,
+      "grad_norm": 2.0148288027331183,
+      "learning_rate": 0.0001544440886102262,
+      "loss": 1.3751,
+      "step": 7321
+    },
+    {
+      "epoch": 0.33731054498548857,
+      "grad_norm": 1.7553765717655907,
+      "learning_rate": 0.00015443157237634688,
+      "loss": 1.7427,
+      "step": 7322
+    },
+    {
+      "epoch": 0.33735661307412357,
+      "grad_norm": 6.160612792457722,
+      "learning_rate": 0.00015441905493064194,
+      "loss": 1.547,
+      "step": 7323
+    },
+    {
+      "epoch": 0.33740268116275857,
+      "grad_norm": 2.3931354689105717,
+      "learning_rate": 0.00015440653627339006,
+      "loss": 1.392,
+      "step": 7324
+    },
+    {
+      "epoch": 0.33744874925139356,
+      "grad_norm": 3.531636659730506,
+      "learning_rate": 0.00015439401640486992,
+      "loss": 1.4355,
+      "step": 7325
+    },
+    {
+      "epoch": 0.33749481734002856,
+      "grad_norm": 20.162289790670137,
+      "learning_rate": 0.0001543814953253603,
+      "loss": 1.1379,
+      "step": 7326
+    },
+    {
+      "epoch": 0.33754088542866356,
+      "grad_norm": 1.8182452072203092,
+      "learning_rate": 0.0001543689730351399,
+      "loss": 1.1797,
+      "step": 7327
+    },
+    {
+      "epoch": 0.33758695351729856,
+      "grad_norm": 8.475156182886494,
+      "learning_rate": 0.00015435644953448757,
+      "loss": 1.63,
+      "step": 7328
+    },
+    {
+      "epoch": 0.33763302160593356,
+      "grad_norm": 2.9511762371848027,
+      "learning_rate": 0.00015434392482368206,
+      "loss": 1.5355,
+      "step": 7329
+    },
+    {
+      "epoch": 0.33767908969456856,
+      "grad_norm": 4.450795158871542,
+      "learning_rate": 0.00015433139890300222,
+      "loss": 1.4592,
+      "step": 7330
+    },
+    {
+      "epoch": 0.33772515778320356,
+      "grad_norm": 7.750823818576616,
+      "learning_rate": 0.000154318871772727,
+      "loss": 1.6812,
+      "step": 7331
+    },
+    {
+      "epoch": 0.33777122587183855,
+      "grad_norm": 2.4612203553929417,
+      "learning_rate": 0.00015430634343313523,
+      "loss": 1.7013,
+      "step": 7332
+    },
+    {
+      "epoch": 0.33781729396047355,
+      "grad_norm": 3.5061553853204788,
+      "learning_rate": 0.0001542938138845058,
+      "loss": 1.4757,
+      "step": 7333
+    },
+    {
+      "epoch": 0.3378633620491086,
+      "grad_norm": 3.3612376096144274,
+      "learning_rate": 0.0001542812831271177,
+      "loss": 1.6562,
+      "step": 7334
+    },
+    {
+      "epoch": 0.3379094301377436,
+      "grad_norm": 2.288790175324985,
+      "learning_rate": 0.00015426875116124995,
+      "loss": 1.6443,
+      "step": 7335
+    },
+    {
+      "epoch": 0.3379554982263786,
+      "grad_norm": 6.089420104921099,
+      "learning_rate": 0.00015425621798718147,
+      "loss": 1.6159,
+      "step": 7336
+    },
+    {
+      "epoch": 0.3380015663150136,
+      "grad_norm": 3.249354696096646,
+      "learning_rate": 0.00015424368360519134,
+      "loss": 1.4842,
+      "step": 7337
+    },
+    {
+      "epoch": 0.3380476344036486,
+      "grad_norm": 2.6295316214110804,
+      "learning_rate": 0.0001542311480155586,
+      "loss": 1.4608,
+      "step": 7338
+    },
+    {
+      "epoch": 0.3380937024922836,
+      "grad_norm": 6.227994163556446,
+      "learning_rate": 0.0001542186112185623,
+      "loss": 1.5869,
+      "step": 7339
+    },
+    {
+      "epoch": 0.3381397705809186,
+      "grad_norm": 1.9138051328139134,
+      "learning_rate": 0.00015420607321448167,
+      "loss": 1.6355,
+      "step": 7340
+    },
+    {
+      "epoch": 0.3381858386695536,
+      "grad_norm": 1.2520584115810198,
+      "learning_rate": 0.0001541935340035957,
+      "loss": 1.6251,
+      "step": 7341
+    },
+    {
+      "epoch": 0.3382319067581886,
+      "grad_norm": 2.132618498262247,
+      "learning_rate": 0.00015418099358618364,
+      "loss": 1.3957,
+      "step": 7342
+    },
+    {
+      "epoch": 0.3382779748468236,
+      "grad_norm": 4.938400873982275,
+      "learning_rate": 0.00015416845196252463,
+      "loss": 1.6944,
+      "step": 7343
+    },
+    {
+      "epoch": 0.3383240429354586,
+      "grad_norm": 2.2081603387117648,
+      "learning_rate": 0.00015415590913289793,
+      "loss": 1.5164,
+      "step": 7344
+    },
+    {
+      "epoch": 0.3383701110240936,
+      "grad_norm": 1.7832256917516982,
+      "learning_rate": 0.00015414336509758276,
+      "loss": 1.6604,
+      "step": 7345
+    },
+    {
+      "epoch": 0.3384161791127286,
+      "grad_norm": 2.195125421602363,
+      "learning_rate": 0.00015413081985685844,
+      "loss": 1.6865,
+      "step": 7346
+    },
+    {
+      "epoch": 0.33846224720136364,
+      "grad_norm": 2.5627118326898994,
+      "learning_rate": 0.00015411827341100422,
+      "loss": 1.4715,
+      "step": 7347
+    },
+    {
+      "epoch": 0.33850831528999864,
+      "grad_norm": 1.097655530024425,
+      "learning_rate": 0.0001541057257602994,
+      "loss": 1.8621,
+      "step": 7348
+    },
+    {
+      "epoch": 0.33855438337863364,
+      "grad_norm": 5.460801936019664,
+      "learning_rate": 0.0001540931769050234,
+      "loss": 1.5016,
+      "step": 7349
+    },
+    {
+      "epoch": 0.33860045146726864,
+      "grad_norm": 2.766191966746114,
+      "learning_rate": 0.00015408062684545555,
+      "loss": 1.4924,
+      "step": 7350
+    },
+    {
+      "epoch": 0.33864651955590364,
+      "grad_norm": 37.40555874801331,
+      "learning_rate": 0.00015406807558187525,
+      "loss": 1.8288,
+      "step": 7351
+    },
+    {
+      "epoch": 0.33869258764453863,
+      "grad_norm": 2.2235984886960747,
+      "learning_rate": 0.000154055523114562,
+      "loss": 1.0898,
+      "step": 7352
+    },
+    {
+      "epoch": 0.33873865573317363,
+      "grad_norm": 4.634134163088489,
+      "learning_rate": 0.0001540429694437952,
+      "loss": 1.5355,
+      "step": 7353
+    },
+    {
+      "epoch": 0.33878472382180863,
+      "grad_norm": 5.164429064463904,
+      "learning_rate": 0.00015403041456985434,
+      "loss": 2.111,
+      "step": 7354
+    },
+    {
+      "epoch": 0.33883079191044363,
+      "grad_norm": 4.185677793296945,
+      "learning_rate": 0.00015401785849301894,
+      "loss": 1.5849,
+      "step": 7355
+    },
+    {
+      "epoch": 0.33887685999907863,
+      "grad_norm": 5.131910370928645,
+      "learning_rate": 0.00015400530121356857,
+      "loss": 2.1065,
+      "step": 7356
+    },
+    {
+      "epoch": 0.3389229280877136,
+      "grad_norm": 1.8405439235211702,
+      "learning_rate": 0.00015399274273178275,
+      "loss": 1.6282,
+      "step": 7357
+    },
+    {
+      "epoch": 0.3389689961763486,
+      "grad_norm": 1.7140083398511317,
+      "learning_rate": 0.00015398018304794108,
+      "loss": 1.8909,
+      "step": 7358
+    },
+    {
+      "epoch": 0.3390150642649836,
+      "grad_norm": 3.95556344742628,
+      "learning_rate": 0.00015396762216232318,
+      "loss": 1.6034,
+      "step": 7359
+    },
+    {
+      "epoch": 0.3390611323536186,
+      "grad_norm": 3.613461296191798,
+      "learning_rate": 0.00015395506007520875,
+      "loss": 1.6308,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3391072004422537,
+      "grad_norm": 1.9171784968859,
+      "learning_rate": 0.00015394249678687738,
+      "loss": 1.8085,
+      "step": 7361
+    },
+    {
+      "epoch": 0.3391532685308887,
+      "grad_norm": 1.8089595997014034,
+      "learning_rate": 0.00015392993229760885,
+      "loss": 1.7856,
+      "step": 7362
+    },
+    {
+      "epoch": 0.3391993366195237,
+      "grad_norm": 4.813298551693985,
+      "learning_rate": 0.0001539173666076828,
+      "loss": 1.8157,
+      "step": 7363
+    },
+    {
+      "epoch": 0.33924540470815867,
+      "grad_norm": 2.064188491776976,
+      "learning_rate": 0.00015390479971737908,
+      "loss": 1.8227,
+      "step": 7364
+    },
+    {
+      "epoch": 0.33929147279679367,
+      "grad_norm": 2.025204994709925,
+      "learning_rate": 0.00015389223162697737,
+      "loss": 1.7818,
+      "step": 7365
+    },
+    {
+      "epoch": 0.33933754088542867,
+      "grad_norm": 2.2196165853697223,
+      "learning_rate": 0.00015387966233675754,
+      "loss": 1.7999,
+      "step": 7366
+    },
+    {
+      "epoch": 0.33938360897406367,
+      "grad_norm": 2.295727928911321,
+      "learning_rate": 0.00015386709184699943,
+      "loss": 1.5056,
+      "step": 7367
+    },
+    {
+      "epoch": 0.33942967706269866,
+      "grad_norm": 1.7456782318053736,
+      "learning_rate": 0.00015385452015798288,
+      "loss": 1.1,
+      "step": 7368
+    },
+    {
+      "epoch": 0.33947574515133366,
+      "grad_norm": 3.049431815533223,
+      "learning_rate": 0.00015384194726998778,
+      "loss": 1.7249,
+      "step": 7369
+    },
+    {
+      "epoch": 0.33952181323996866,
+      "grad_norm": 2.26486370116023,
+      "learning_rate": 0.00015382937318329402,
+      "loss": 1.5739,
+      "step": 7370
+    },
+    {
+      "epoch": 0.33956788132860366,
+      "grad_norm": 2.360049340803223,
+      "learning_rate": 0.0001538167978981816,
+      "loss": 1.489,
+      "step": 7371
+    },
+    {
+      "epoch": 0.33961394941723866,
+      "grad_norm": 7.380169069321951,
+      "learning_rate": 0.00015380422141493042,
+      "loss": 1.499,
+      "step": 7372
+    },
+    {
+      "epoch": 0.33966001750587366,
+      "grad_norm": 1.9334561696270718,
+      "learning_rate": 0.00015379164373382053,
+      "loss": 1.5586,
+      "step": 7373
+    },
+    {
+      "epoch": 0.3397060855945087,
+      "grad_norm": 2.5488736462509745,
+      "learning_rate": 0.00015377906485513192,
+      "loss": 1.8156,
+      "step": 7374
+    },
+    {
+      "epoch": 0.3397521536831437,
+      "grad_norm": 1.7636434492410629,
+      "learning_rate": 0.00015376648477914465,
+      "loss": 1.2978,
+      "step": 7375
+    },
+    {
+      "epoch": 0.3397982217717787,
+      "grad_norm": 1.5386868429798939,
+      "learning_rate": 0.00015375390350613874,
+      "loss": 1.6888,
+      "step": 7376
+    },
+    {
+      "epoch": 0.3398442898604137,
+      "grad_norm": 2.4757297008862285,
+      "learning_rate": 0.0001537413210363944,
+      "loss": 1.5321,
+      "step": 7377
+    },
+    {
+      "epoch": 0.3398903579490487,
+      "grad_norm": 2.3116159739628483,
+      "learning_rate": 0.00015372873737019166,
+      "loss": 1.7693,
+      "step": 7378
+    },
+    {
+      "epoch": 0.3399364260376837,
+      "grad_norm": 1.924970793593141,
+      "learning_rate": 0.0001537161525078107,
+      "loss": 1.4461,
+      "step": 7379
+    },
+    {
+      "epoch": 0.3399824941263187,
+      "grad_norm": 1.784762085135455,
+      "learning_rate": 0.0001537035664495317,
+      "loss": 1.3993,
+      "step": 7380
+    },
+    {
+      "epoch": 0.3400285622149537,
+      "grad_norm": 1.6808209816724347,
+      "learning_rate": 0.00015369097919563492,
+      "loss": 1.4503,
+      "step": 7381
+    },
+    {
+      "epoch": 0.3400746303035887,
+      "grad_norm": 2.1652048940679425,
+      "learning_rate": 0.00015367839074640052,
+      "loss": 1.5755,
+      "step": 7382
+    },
+    {
+      "epoch": 0.3401206983922237,
+      "grad_norm": 1.7045319602329605,
+      "learning_rate": 0.00015366580110210876,
+      "loss": 1.2659,
+      "step": 7383
+    },
+    {
+      "epoch": 0.3401667664808587,
+      "grad_norm": 1.4613535172570564,
+      "learning_rate": 0.00015365321026304,
+      "loss": 1.4184,
+      "step": 7384
+    },
+    {
+      "epoch": 0.3402128345694937,
+      "grad_norm": 2.54809230729114,
+      "learning_rate": 0.0001536406182294745,
+      "loss": 1.3299,
+      "step": 7385
+    },
+    {
+      "epoch": 0.3402589026581287,
+      "grad_norm": 1.4545521587209622,
+      "learning_rate": 0.0001536280250016926,
+      "loss": 1.6494,
+      "step": 7386
+    },
+    {
+      "epoch": 0.3403049707467637,
+      "grad_norm": 2.95262090747183,
+      "learning_rate": 0.00015361543057997468,
+      "loss": 1.4005,
+      "step": 7387
+    },
+    {
+      "epoch": 0.34035103883539874,
+      "grad_norm": 1.4950774668896354,
+      "learning_rate": 0.00015360283496460112,
+      "loss": 1.479,
+      "step": 7388
+    },
+    {
+      "epoch": 0.34039710692403374,
+      "grad_norm": 2.1093080563491777,
+      "learning_rate": 0.00015359023815585235,
+      "loss": 1.6876,
+      "step": 7389
+    },
+    {
+      "epoch": 0.34044317501266874,
+      "grad_norm": 1.0013297666599856,
+      "learning_rate": 0.00015357764015400883,
+      "loss": 1.1993,
+      "step": 7390
+    },
+    {
+      "epoch": 0.34048924310130374,
+      "grad_norm": 3.014869799941568,
+      "learning_rate": 0.00015356504095935102,
+      "loss": 1.4651,
+      "step": 7391
+    },
+    {
+      "epoch": 0.34053531118993874,
+      "grad_norm": 1.9784333998772474,
+      "learning_rate": 0.0001535524405721594,
+      "loss": 1.8655,
+      "step": 7392
+    },
+    {
+      "epoch": 0.34058137927857374,
+      "grad_norm": 2.935980380624018,
+      "learning_rate": 0.00015353983899271453,
+      "loss": 1.5507,
+      "step": 7393
+    },
+    {
+      "epoch": 0.34062744736720874,
+      "grad_norm": 1.2764033289965049,
+      "learning_rate": 0.00015352723622129695,
+      "loss": 1.5759,
+      "step": 7394
+    },
+    {
+      "epoch": 0.34067351545584373,
+      "grad_norm": 0.8601499229113371,
+      "learning_rate": 0.0001535146322581872,
+      "loss": 1.4706,
+      "step": 7395
+    },
+    {
+      "epoch": 0.34071958354447873,
+      "grad_norm": 2.3354282544206555,
+      "learning_rate": 0.00015350202710366596,
+      "loss": 1.3295,
+      "step": 7396
+    },
+    {
+      "epoch": 0.34076565163311373,
+      "grad_norm": 2.2573036119307686,
+      "learning_rate": 0.0001534894207580138,
+      "loss": 1.6882,
+      "step": 7397
+    },
+    {
+      "epoch": 0.34081171972174873,
+      "grad_norm": 3.153284976711108,
+      "learning_rate": 0.0001534768132215114,
+      "loss": 1.3623,
+      "step": 7398
+    },
+    {
+      "epoch": 0.34085778781038373,
+      "grad_norm": 2.6692774949841076,
+      "learning_rate": 0.00015346420449443947,
+      "loss": 1.1314,
+      "step": 7399
+    },
+    {
+      "epoch": 0.3409038558990187,
+      "grad_norm": 1.7982540872970745,
+      "learning_rate": 0.00015345159457707866,
+      "loss": 1.579,
+      "step": 7400
+    },
+    {
+      "epoch": 0.3409499239876537,
+      "grad_norm": 3.5156230280899607,
+      "learning_rate": 0.00015343898346970977,
+      "loss": 1.3645,
+      "step": 7401
+    },
+    {
+      "epoch": 0.3409959920762888,
+      "grad_norm": 1.7825521717840205,
+      "learning_rate": 0.00015342637117261355,
+      "loss": 1.7943,
+      "step": 7402
+    },
+    {
+      "epoch": 0.3410420601649238,
+      "grad_norm": 4.8110866680224165,
+      "learning_rate": 0.00015341375768607076,
+      "loss": 1.404,
+      "step": 7403
+    },
+    {
+      "epoch": 0.3410881282535588,
+      "grad_norm": 2.5132258716519313,
+      "learning_rate": 0.00015340114301036226,
+      "loss": 1.444,
+      "step": 7404
+    },
+    {
+      "epoch": 0.3411341963421938,
+      "grad_norm": 2.756904793817449,
+      "learning_rate": 0.00015338852714576887,
+      "loss": 1.4913,
+      "step": 7405
+    },
+    {
+      "epoch": 0.3411802644308288,
+      "grad_norm": 1.6973344668364236,
+      "learning_rate": 0.00015337591009257146,
+      "loss": 1.675,
+      "step": 7406
+    },
+    {
+      "epoch": 0.34122633251946377,
+      "grad_norm": 5.518582349651393,
+      "learning_rate": 0.00015336329185105092,
+      "loss": 1.381,
+      "step": 7407
+    },
+    {
+      "epoch": 0.34127240060809877,
+      "grad_norm": 5.306682210734146,
+      "learning_rate": 0.0001533506724214882,
+      "loss": 1.573,
+      "step": 7408
+    },
+    {
+      "epoch": 0.34131846869673377,
+      "grad_norm": 2.9988734755329647,
+      "learning_rate": 0.00015333805180416422,
+      "loss": 1.5378,
+      "step": 7409
+    },
+    {
+      "epoch": 0.34136453678536877,
+      "grad_norm": 6.468399441339033,
+      "learning_rate": 0.00015332542999936,
+      "loss": 1.6181,
+      "step": 7410
+    },
+    {
+      "epoch": 0.34141060487400376,
+      "grad_norm": 1.5271102726139818,
+      "learning_rate": 0.00015331280700735651,
+      "loss": 1.3292,
+      "step": 7411
+    },
+    {
+      "epoch": 0.34145667296263876,
+      "grad_norm": 1.3372952267553093,
+      "learning_rate": 0.00015330018282843478,
+      "loss": 1.4492,
+      "step": 7412
+    },
+    {
+      "epoch": 0.34150274105127376,
+      "grad_norm": 2.2552467885183085,
+      "learning_rate": 0.00015328755746287585,
+      "loss": 1.5922,
+      "step": 7413
+    },
+    {
+      "epoch": 0.34154880913990876,
+      "grad_norm": 2.79484988953258,
+      "learning_rate": 0.00015327493091096082,
+      "loss": 1.4941,
+      "step": 7414
+    },
+    {
+      "epoch": 0.3415948772285438,
+      "grad_norm": 2.5500531135028215,
+      "learning_rate": 0.00015326230317297084,
+      "loss": 1.498,
+      "step": 7415
+    },
+    {
+      "epoch": 0.3416409453171788,
+      "grad_norm": 1.5910486482210244,
+      "learning_rate": 0.00015324967424918694,
+      "loss": 1.6061,
+      "step": 7416
+    },
+    {
+      "epoch": 0.3416870134058138,
+      "grad_norm": 1.7886175905023938,
+      "learning_rate": 0.00015323704413989042,
+      "loss": 1.559,
+      "step": 7417
+    },
+    {
+      "epoch": 0.3417330814944488,
+      "grad_norm": 1.3088265000970187,
+      "learning_rate": 0.00015322441284536234,
+      "loss": 1.6525,
+      "step": 7418
+    },
+    {
+      "epoch": 0.3417791495830838,
+      "grad_norm": 2.1472733298080406,
+      "learning_rate": 0.00015321178036588405,
+      "loss": 1.1262,
+      "step": 7419
+    },
+    {
+      "epoch": 0.3418252176717188,
+      "grad_norm": 1.1407089652015887,
+      "learning_rate": 0.00015319914670173664,
+      "loss": 1.7147,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3418712857603538,
+      "grad_norm": 2.0893639783971545,
+      "learning_rate": 0.00015318651185320143,
+      "loss": 1.4307,
+      "step": 7421
+    },
+    {
+      "epoch": 0.3419173538489888,
+      "grad_norm": 3.1420905980798266,
+      "learning_rate": 0.00015317387582055973,
+      "loss": 1.3454,
+      "step": 7422
+    },
+    {
+      "epoch": 0.3419634219376238,
+      "grad_norm": 1.3648777455121655,
+      "learning_rate": 0.00015316123860409288,
+      "loss": 1.4193,
+      "step": 7423
+    },
+    {
+      "epoch": 0.3420094900262588,
+      "grad_norm": 0.9264500802817496,
+      "learning_rate": 0.00015314860020408224,
+      "loss": 1.7311,
+      "step": 7424
+    },
+    {
+      "epoch": 0.3420555581148938,
+      "grad_norm": 1.3473256659053228,
+      "learning_rate": 0.00015313596062080907,
+      "loss": 1.2859,
+      "step": 7425
+    },
+    {
+      "epoch": 0.3421016262035288,
+      "grad_norm": 1.0804333258837846,
+      "learning_rate": 0.00015312331985455492,
+      "loss": 1.5853,
+      "step": 7426
+    },
+    {
+      "epoch": 0.3421476942921638,
+      "grad_norm": 1.9232447549750575,
+      "learning_rate": 0.00015311067790560107,
+      "loss": 1.227,
+      "step": 7427
+    },
+    {
+      "epoch": 0.3421937623807988,
+      "grad_norm": 1.5341960852654932,
+      "learning_rate": 0.00015309803477422913,
+      "loss": 1.5694,
+      "step": 7428
+    },
+    {
+      "epoch": 0.34223983046943385,
+      "grad_norm": 2.939801845490508,
+      "learning_rate": 0.00015308539046072044,
+      "loss": 1.4027,
+      "step": 7429
+    },
+    {
+      "epoch": 0.34228589855806885,
+      "grad_norm": 3.507510848505349,
+      "learning_rate": 0.00015307274496535652,
+      "loss": 1.4358,
+      "step": 7430
+    },
+    {
+      "epoch": 0.34233196664670384,
+      "grad_norm": 4.085734562502328,
+      "learning_rate": 0.000153060098288419,
+      "loss": 1.3887,
+      "step": 7431
+    },
+    {
+      "epoch": 0.34237803473533884,
+      "grad_norm": 2.7998644031266244,
+      "learning_rate": 0.0001530474504301893,
+      "loss": 1.6938,
+      "step": 7432
+    },
+    {
+      "epoch": 0.34242410282397384,
+      "grad_norm": 1.7894886563459622,
+      "learning_rate": 0.00015303480139094912,
+      "loss": 1.2381,
+      "step": 7433
+    },
+    {
+      "epoch": 0.34247017091260884,
+      "grad_norm": 1.0771429913486794,
+      "learning_rate": 0.00015302215117098003,
+      "loss": 1.667,
+      "step": 7434
+    },
+    {
+      "epoch": 0.34251623900124384,
+      "grad_norm": 2.1070604728796787,
+      "learning_rate": 0.00015300949977056362,
+      "loss": 1.4626,
+      "step": 7435
+    },
+    {
+      "epoch": 0.34256230708987884,
+      "grad_norm": 1.68979984597854,
+      "learning_rate": 0.0001529968471899816,
+      "loss": 1.3114,
+      "step": 7436
+    },
+    {
+      "epoch": 0.34260837517851384,
+      "grad_norm": 2.71584495967845,
+      "learning_rate": 0.00015298419342951567,
+      "loss": 1.6573,
+      "step": 7437
+    },
+    {
+      "epoch": 0.34265444326714883,
+      "grad_norm": 2.237619924967308,
+      "learning_rate": 0.0001529715384894475,
+      "loss": 1.4921,
+      "step": 7438
+    },
+    {
+      "epoch": 0.34270051135578383,
+      "grad_norm": 1.4202371977639574,
+      "learning_rate": 0.00015295888237005888,
+      "loss": 1.5918,
+      "step": 7439
+    },
+    {
+      "epoch": 0.34274657944441883,
+      "grad_norm": 1.672955252963371,
+      "learning_rate": 0.0001529462250716315,
+      "loss": 1.4778,
+      "step": 7440
+    },
+    {
+      "epoch": 0.34279264753305383,
+      "grad_norm": 3.6482718433364587,
+      "learning_rate": 0.00015293356659444726,
+      "loss": 1.3788,
+      "step": 7441
+    },
+    {
+      "epoch": 0.3428387156216889,
+      "grad_norm": 1.4895819542816597,
+      "learning_rate": 0.0001529209069387879,
+      "loss": 1.6427,
+      "step": 7442
+    },
+    {
+      "epoch": 0.3428847837103239,
+      "grad_norm": 1.329192801400228,
+      "learning_rate": 0.00015290824610493525,
+      "loss": 1.4002,
+      "step": 7443
+    },
+    {
+      "epoch": 0.3429308517989589,
+      "grad_norm": 4.461571223263933,
+      "learning_rate": 0.00015289558409317127,
+      "loss": 1.4858,
+      "step": 7444
+    },
+    {
+      "epoch": 0.3429769198875939,
+      "grad_norm": 9.043439010330674,
+      "learning_rate": 0.00015288292090377779,
+      "loss": 1.3058,
+      "step": 7445
+    },
+    {
+      "epoch": 0.3430229879762289,
+      "grad_norm": 2.7687497832137806,
+      "learning_rate": 0.0001528702565370367,
+      "loss": 1.3798,
+      "step": 7446
+    },
+    {
+      "epoch": 0.3430690560648639,
+      "grad_norm": 0.8627090036536051,
+      "learning_rate": 0.00015285759099323008,
+      "loss": 1.6394,
+      "step": 7447
+    },
+    {
+      "epoch": 0.3431151241534989,
+      "grad_norm": 2.952992005679153,
+      "learning_rate": 0.00015284492427263978,
+      "loss": 1.4678,
+      "step": 7448
+    },
+    {
+      "epoch": 0.3431611922421339,
+      "grad_norm": 1.4632021607020087,
+      "learning_rate": 0.00015283225637554787,
+      "loss": 1.7231,
+      "step": 7449
+    },
+    {
+      "epoch": 0.34320726033076887,
+      "grad_norm": 3.0425441347524513,
+      "learning_rate": 0.00015281958730223636,
+      "loss": 1.3593,
+      "step": 7450
+    },
+    {
+      "epoch": 0.34325332841940387,
+      "grad_norm": 5.32936155187653,
+      "learning_rate": 0.00015280691705298727,
+      "loss": 1.5446,
+      "step": 7451
+    },
+    {
+      "epoch": 0.34329939650803887,
+      "grad_norm": 2.3748348728863142,
+      "learning_rate": 0.00015279424562808278,
+      "loss": 1.5794,
+      "step": 7452
+    },
+    {
+      "epoch": 0.34334546459667387,
+      "grad_norm": 1.9030713823463177,
+      "learning_rate": 0.0001527815730278049,
+      "loss": 1.5991,
+      "step": 7453
+    },
+    {
+      "epoch": 0.34339153268530886,
+      "grad_norm": 2.4342321417958668,
+      "learning_rate": 0.0001527688992524358,
+      "loss": 1.4355,
+      "step": 7454
+    },
+    {
+      "epoch": 0.34343760077394386,
+      "grad_norm": 1.400060778924179,
+      "learning_rate": 0.00015275622430225763,
+      "loss": 1.6309,
+      "step": 7455
+    },
+    {
+      "epoch": 0.3434836688625789,
+      "grad_norm": 1.7167924498627465,
+      "learning_rate": 0.00015274354817755258,
+      "loss": 1.5536,
+      "step": 7456
+    },
+    {
+      "epoch": 0.3435297369512139,
+      "grad_norm": 7.747043403163837,
+      "learning_rate": 0.00015273087087860288,
+      "loss": 1.4953,
+      "step": 7457
+    },
+    {
+      "epoch": 0.3435758050398489,
+      "grad_norm": 1.393724492787998,
+      "learning_rate": 0.00015271819240569073,
+      "loss": 1.3375,
+      "step": 7458
+    },
+    {
+      "epoch": 0.3436218731284839,
+      "grad_norm": 1.8471229491131445,
+      "learning_rate": 0.00015270551275909847,
+      "loss": 1.5344,
+      "step": 7459
+    },
+    {
+      "epoch": 0.3436679412171189,
+      "grad_norm": 4.920216150759131,
+      "learning_rate": 0.0001526928319391083,
+      "loss": 1.229,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3437140093057539,
+      "grad_norm": 2.3159388630338054,
+      "learning_rate": 0.00015268014994600256,
+      "loss": 1.6359,
+      "step": 7461
+    },
+    {
+      "epoch": 0.3437600773943889,
+      "grad_norm": 2.6429971122362343,
+      "learning_rate": 0.00015266746678006362,
+      "loss": 1.3841,
+      "step": 7462
+    },
+    {
+      "epoch": 0.3438061454830239,
+      "grad_norm": 1.6233025335950277,
+      "learning_rate": 0.00015265478244157389,
+      "loss": 1.5192,
+      "step": 7463
+    },
+    {
+      "epoch": 0.3438522135716589,
+      "grad_norm": 1.3883873088793173,
+      "learning_rate": 0.00015264209693081565,
+      "loss": 1.615,
+      "step": 7464
+    },
+    {
+      "epoch": 0.3438982816602939,
+      "grad_norm": 3.8925095805848056,
+      "learning_rate": 0.0001526294102480714,
+      "loss": 1.2699,
+      "step": 7465
+    },
+    {
+      "epoch": 0.3439443497489289,
+      "grad_norm": 5.849119482766462,
+      "learning_rate": 0.0001526167223936236,
+      "loss": 1.5048,
+      "step": 7466
+    },
+    {
+      "epoch": 0.3439904178375639,
+      "grad_norm": 2.033093615072383,
+      "learning_rate": 0.00015260403336775464,
+      "loss": 1.1867,
+      "step": 7467
+    },
+    {
+      "epoch": 0.3440364859261989,
+      "grad_norm": 2.492481230615138,
+      "learning_rate": 0.00015259134317074713,
+      "loss": 1.5818,
+      "step": 7468
+    },
+    {
+      "epoch": 0.3440825540148339,
+      "grad_norm": 5.582767568986525,
+      "learning_rate": 0.00015257865180288348,
+      "loss": 1.3853,
+      "step": 7469
+    },
+    {
+      "epoch": 0.34412862210346895,
+      "grad_norm": 10.3317232074844,
+      "learning_rate": 0.00015256595926444635,
+      "loss": 1.5299,
+      "step": 7470
+    },
+    {
+      "epoch": 0.34417469019210395,
+      "grad_norm": 4.474901038213008,
+      "learning_rate": 0.00015255326555571822,
+      "loss": 1.421,
+      "step": 7471
+    },
+    {
+      "epoch": 0.34422075828073895,
+      "grad_norm": 2.430032024599024,
+      "learning_rate": 0.0001525405706769818,
+      "loss": 1.5254,
+      "step": 7472
+    },
+    {
+      "epoch": 0.34426682636937395,
+      "grad_norm": 8.843636873594779,
+      "learning_rate": 0.0001525278746285196,
+      "loss": 1.4819,
+      "step": 7473
+    },
+    {
+      "epoch": 0.34431289445800894,
+      "grad_norm": 6.173829994132924,
+      "learning_rate": 0.0001525151774106144,
+      "loss": 1.5414,
+      "step": 7474
+    },
+    {
+      "epoch": 0.34435896254664394,
+      "grad_norm": 1.971979638762282,
+      "learning_rate": 0.0001525024790235488,
+      "loss": 1.4925,
+      "step": 7475
+    },
+    {
+      "epoch": 0.34440503063527894,
+      "grad_norm": 2.9248785016778474,
+      "learning_rate": 0.00015248977946760552,
+      "loss": 1.4853,
+      "step": 7476
+    },
+    {
+      "epoch": 0.34445109872391394,
+      "grad_norm": 3.321069798953843,
+      "learning_rate": 0.00015247707874306728,
+      "loss": 1.6237,
+      "step": 7477
+    },
+    {
+      "epoch": 0.34449716681254894,
+      "grad_norm": 2.713645555113895,
+      "learning_rate": 0.00015246437685021688,
+      "loss": 1.415,
+      "step": 7478
+    },
+    {
+      "epoch": 0.34454323490118394,
+      "grad_norm": 4.0238980628273096,
+      "learning_rate": 0.0001524516737893371,
+      "loss": 1.5637,
+      "step": 7479
+    },
+    {
+      "epoch": 0.34458930298981894,
+      "grad_norm": 2.840417365383303,
+      "learning_rate": 0.00015243896956071068,
+      "loss": 1.3767,
+      "step": 7480
+    },
+    {
+      "epoch": 0.34463537107845393,
+      "grad_norm": 2.520037444012201,
+      "learning_rate": 0.0001524262641646206,
+      "loss": 1.4969,
+      "step": 7481
+    },
+    {
+      "epoch": 0.34468143916708893,
+      "grad_norm": 2.6103534305099165,
+      "learning_rate": 0.0001524135576013496,
+      "loss": 1.5584,
+      "step": 7482
+    },
+    {
+      "epoch": 0.344727507255724,
+      "grad_norm": 10.907905505903477,
+      "learning_rate": 0.0001524008498711806,
+      "loss": 1.5835,
+      "step": 7483
+    },
+    {
+      "epoch": 0.344773575344359,
+      "grad_norm": 2.709174724040091,
+      "learning_rate": 0.00015238814097439654,
+      "loss": 1.5393,
+      "step": 7484
+    },
+    {
+      "epoch": 0.344819643432994,
+      "grad_norm": 2.257725710876407,
+      "learning_rate": 0.00015237543091128035,
+      "loss": 1.4551,
+      "step": 7485
+    },
+    {
+      "epoch": 0.344865711521629,
+      "grad_norm": 5.108493090044843,
+      "learning_rate": 0.000152362719682115,
+      "loss": 1.8083,
+      "step": 7486
+    },
+    {
+      "epoch": 0.344911779610264,
+      "grad_norm": 3.5239416236596717,
+      "learning_rate": 0.00015235000728718346,
+      "loss": 1.562,
+      "step": 7487
+    },
+    {
+      "epoch": 0.344957847698899,
+      "grad_norm": 2.142946898957716,
+      "learning_rate": 0.00015233729372676877,
+      "loss": 1.5451,
+      "step": 7488
+    },
+    {
+      "epoch": 0.345003915787534,
+      "grad_norm": 2.5769179177554156,
+      "learning_rate": 0.00015232457900115398,
+      "loss": 1.3897,
+      "step": 7489
+    },
+    {
+      "epoch": 0.345049983876169,
+      "grad_norm": 2.5657453779921657,
+      "learning_rate": 0.00015231186311062214,
+      "loss": 1.6321,
+      "step": 7490
+    },
+    {
+      "epoch": 0.345096051964804,
+      "grad_norm": 4.796554725577912,
+      "learning_rate": 0.00015229914605545636,
+      "loss": 1.662,
+      "step": 7491
+    },
+    {
+      "epoch": 0.34514212005343897,
+      "grad_norm": 20.86723152912615,
+      "learning_rate": 0.00015228642783593984,
+      "loss": 1.2624,
+      "step": 7492
+    },
+    {
+      "epoch": 0.34518818814207397,
+      "grad_norm": 2.4085787286080063,
+      "learning_rate": 0.0001522737084523556,
+      "loss": 1.5418,
+      "step": 7493
+    },
+    {
+      "epoch": 0.34523425623070897,
+      "grad_norm": 1.5403246868345497,
+      "learning_rate": 0.00015226098790498686,
+      "loss": 1.4988,
+      "step": 7494
+    },
+    {
+      "epoch": 0.34528032431934397,
+      "grad_norm": 2.1078321865383356,
+      "learning_rate": 0.00015224826619411686,
+      "loss": 1.5771,
+      "step": 7495
+    },
+    {
+      "epoch": 0.34532639240797897,
+      "grad_norm": 1.0859449163739447,
+      "learning_rate": 0.00015223554332002877,
+      "loss": 1.7632,
+      "step": 7496
+    },
+    {
+      "epoch": 0.345372460496614,
+      "grad_norm": 2.748585029712029,
+      "learning_rate": 0.0001522228192830059,
+      "loss": 1.6227,
+      "step": 7497
+    },
+    {
+      "epoch": 0.345418528585249,
+      "grad_norm": 1.8143365195856576,
+      "learning_rate": 0.0001522100940833315,
+      "loss": 1.6775,
+      "step": 7498
+    },
+    {
+      "epoch": 0.345464596673884,
+      "grad_norm": 1.8789257859063389,
+      "learning_rate": 0.00015219736772128885,
+      "loss": 1.5183,
+      "step": 7499
+    },
+    {
+      "epoch": 0.345510664762519,
+      "grad_norm": 1.6737996962525088,
+      "learning_rate": 0.00015218464019716135,
+      "loss": 1.6035,
+      "step": 7500
+    },
+    {
+      "epoch": 0.345556732851154,
+      "grad_norm": 1.3526900732971427,
+      "learning_rate": 0.00015217191151123226,
+      "loss": 1.6104,
+      "step": 7501
+    },
+    {
+      "epoch": 0.345602800939789,
+      "grad_norm": 1.4235399873825763,
+      "learning_rate": 0.00015215918166378505,
+      "loss": 1.399,
+      "step": 7502
+    },
+    {
+      "epoch": 0.345648869028424,
+      "grad_norm": 1.4139152992733395,
+      "learning_rate": 0.00015214645065510308,
+      "loss": 1.4741,
+      "step": 7503
+    },
+    {
+      "epoch": 0.345694937117059,
+      "grad_norm": 2.363113993682504,
+      "learning_rate": 0.0001521337184854698,
+      "loss": 1.3547,
+      "step": 7504
+    },
+    {
+      "epoch": 0.345741005205694,
+      "grad_norm": 1.3146916696987816,
+      "learning_rate": 0.0001521209851551687,
+      "loss": 1.2698,
+      "step": 7505
+    },
+    {
+      "epoch": 0.345787073294329,
+      "grad_norm": 1.7456884190839286,
+      "learning_rate": 0.0001521082506644832,
+      "loss": 1.3755,
+      "step": 7506
+    },
+    {
+      "epoch": 0.345833141382964,
+      "grad_norm": 1.7305309840152623,
+      "learning_rate": 0.00015209551501369685,
+      "loss": 1.2714,
+      "step": 7507
+    },
+    {
+      "epoch": 0.345879209471599,
+      "grad_norm": 1.616266173910984,
+      "learning_rate": 0.0001520827782030932,
+      "loss": 1.7487,
+      "step": 7508
+    },
+    {
+      "epoch": 0.345925277560234,
+      "grad_norm": 3.4770222465883167,
+      "learning_rate": 0.0001520700402329558,
+      "loss": 1.3561,
+      "step": 7509
+    },
+    {
+      "epoch": 0.34597134564886906,
+      "grad_norm": 1.881526494663083,
+      "learning_rate": 0.0001520573011035682,
+      "loss": 1.3989,
+      "step": 7510
+    },
+    {
+      "epoch": 0.34601741373750405,
+      "grad_norm": 7.107894322869939,
+      "learning_rate": 0.0001520445608152141,
+      "loss": 1.2593,
+      "step": 7511
+    },
+    {
+      "epoch": 0.34606348182613905,
+      "grad_norm": 2.941599392027124,
+      "learning_rate": 0.00015203181936817702,
+      "loss": 1.4534,
+      "step": 7512
+    },
+    {
+      "epoch": 0.34610954991477405,
+      "grad_norm": 3.2858425325930556,
+      "learning_rate": 0.00015201907676274073,
+      "loss": 1.5624,
+      "step": 7513
+    },
+    {
+      "epoch": 0.34615561800340905,
+      "grad_norm": 2.139702152927542,
+      "learning_rate": 0.00015200633299918893,
+      "loss": 1.7298,
+      "step": 7514
+    },
+    {
+      "epoch": 0.34620168609204405,
+      "grad_norm": 1.7644413713073512,
+      "learning_rate": 0.00015199358807780524,
+      "loss": 1.4275,
+      "step": 7515
+    },
+    {
+      "epoch": 0.34624775418067905,
+      "grad_norm": 2.2658918431519735,
+      "learning_rate": 0.00015198084199887348,
+      "loss": 1.3998,
+      "step": 7516
+    },
+    {
+      "epoch": 0.34629382226931404,
+      "grad_norm": 2.1706597503024185,
+      "learning_rate": 0.0001519680947626774,
+      "loss": 1.258,
+      "step": 7517
+    },
+    {
+      "epoch": 0.34633989035794904,
+      "grad_norm": 1.4293367735266203,
+      "learning_rate": 0.00015195534636950082,
+      "loss": 1.5558,
+      "step": 7518
+    },
+    {
+      "epoch": 0.34638595844658404,
+      "grad_norm": 1.008559281423595,
+      "learning_rate": 0.0001519425968196275,
+      "loss": 1.5209,
+      "step": 7519
+    },
+    {
+      "epoch": 0.34643202653521904,
+      "grad_norm": 1.1881929345430708,
+      "learning_rate": 0.00015192984611334137,
+      "loss": 1.3694,
+      "step": 7520
+    },
+    {
+      "epoch": 0.34647809462385404,
+      "grad_norm": 2.8750040920860847,
+      "learning_rate": 0.0001519170942509262,
+      "loss": 1.5507,
+      "step": 7521
+    },
+    {
+      "epoch": 0.34652416271248904,
+      "grad_norm": 2.5790143197088136,
+      "learning_rate": 0.000151904341232666,
+      "loss": 1.3819,
+      "step": 7522
+    },
+    {
+      "epoch": 0.34657023080112404,
+      "grad_norm": 1.3902386924982515,
+      "learning_rate": 0.0001518915870588446,
+      "loss": 1.4175,
+      "step": 7523
+    },
+    {
+      "epoch": 0.3466162988897591,
+      "grad_norm": 2.1804313875109402,
+      "learning_rate": 0.00015187883172974601,
+      "loss": 1.8105,
+      "step": 7524
+    },
+    {
+      "epoch": 0.3466623669783941,
+      "grad_norm": 2.931268825737152,
+      "learning_rate": 0.00015186607524565418,
+      "loss": 1.521,
+      "step": 7525
+    },
+    {
+      "epoch": 0.3467084350670291,
+      "grad_norm": 1.0179796855243155,
+      "learning_rate": 0.00015185331760685311,
+      "loss": 1.3853,
+      "step": 7526
+    },
+    {
+      "epoch": 0.3467545031556641,
+      "grad_norm": 1.8076952512068847,
+      "learning_rate": 0.00015184055881362684,
+      "loss": 1.4417,
+      "step": 7527
+    },
+    {
+      "epoch": 0.3468005712442991,
+      "grad_norm": 2.462343838180077,
+      "learning_rate": 0.0001518277988662594,
+      "loss": 1.4804,
+      "step": 7528
+    },
+    {
+      "epoch": 0.3468466393329341,
+      "grad_norm": 4.261276416561141,
+      "learning_rate": 0.0001518150377650349,
+      "loss": 1.4618,
+      "step": 7529
+    },
+    {
+      "epoch": 0.3468927074215691,
+      "grad_norm": 5.311313784895153,
+      "learning_rate": 0.00015180227551023743,
+      "loss": 1.3452,
+      "step": 7530
+    },
+    {
+      "epoch": 0.3469387755102041,
+      "grad_norm": 2.324023611698947,
+      "learning_rate": 0.00015178951210215112,
+      "loss": 1.3909,
+      "step": 7531
+    },
+    {
+      "epoch": 0.3469848435988391,
+      "grad_norm": 2.6772333235807975,
+      "learning_rate": 0.00015177674754106012,
+      "loss": 1.5662,
+      "step": 7532
+    },
+    {
+      "epoch": 0.3470309116874741,
+      "grad_norm": 1.9836101777115176,
+      "learning_rate": 0.00015176398182724862,
+      "loss": 1.3716,
+      "step": 7533
+    },
+    {
+      "epoch": 0.3470769797761091,
+      "grad_norm": 2.6840883295719338,
+      "learning_rate": 0.00015175121496100079,
+      "loss": 1.6812,
+      "step": 7534
+    },
+    {
+      "epoch": 0.34712304786474407,
+      "grad_norm": 6.684402327974339,
+      "learning_rate": 0.0001517384469426009,
+      "loss": 1.2941,
+      "step": 7535
+    },
+    {
+      "epoch": 0.34716911595337907,
+      "grad_norm": 5.018774507808715,
+      "learning_rate": 0.00015172567777233323,
+      "loss": 1.275,
+      "step": 7536
+    },
+    {
+      "epoch": 0.3472151840420141,
+      "grad_norm": 1.8432971795919044,
+      "learning_rate": 0.000151712907450482,
+      "loss": 1.6085,
+      "step": 7537
+    },
+    {
+      "epoch": 0.3472612521306491,
+      "grad_norm": 1.8137104824092514,
+      "learning_rate": 0.0001517001359773316,
+      "loss": 1.4192,
+      "step": 7538
+    },
+    {
+      "epoch": 0.3473073202192841,
+      "grad_norm": 3.9244391038228517,
+      "learning_rate": 0.0001516873633531663,
+      "loss": 1.5301,
+      "step": 7539
+    },
+    {
+      "epoch": 0.3473533883079191,
+      "grad_norm": 1.370943463678089,
+      "learning_rate": 0.00015167458957827048,
+      "loss": 1.7305,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3473994563965541,
+      "grad_norm": 1.6183208356594554,
+      "learning_rate": 0.00015166181465292851,
+      "loss": 1.6295,
+      "step": 7541
+    },
+    {
+      "epoch": 0.3474455244851891,
+      "grad_norm": 2.7151137238877254,
+      "learning_rate": 0.00015164903857742482,
+      "loss": 1.524,
+      "step": 7542
+    },
+    {
+      "epoch": 0.3474915925738241,
+      "grad_norm": 5.764324001553472,
+      "learning_rate": 0.00015163626135204384,
+      "loss": 1.3446,
+      "step": 7543
+    },
+    {
+      "epoch": 0.3475376606624591,
+      "grad_norm": 1.461994961432252,
+      "learning_rate": 0.00015162348297707008,
+      "loss": 1.3885,
+      "step": 7544
+    },
+    {
+      "epoch": 0.3475837287510941,
+      "grad_norm": 2.374719373668958,
+      "learning_rate": 0.00015161070345278793,
+      "loss": 1.4464,
+      "step": 7545
+    },
+    {
+      "epoch": 0.3476297968397291,
+      "grad_norm": 3.2781777581170446,
+      "learning_rate": 0.000151597922779482,
+      "loss": 1.3974,
+      "step": 7546
+    },
+    {
+      "epoch": 0.3476758649283641,
+      "grad_norm": 3.7639295672448516,
+      "learning_rate": 0.00015158514095743678,
+      "loss": 1.5195,
+      "step": 7547
+    },
+    {
+      "epoch": 0.3477219330169991,
+      "grad_norm": 4.415706519835531,
+      "learning_rate": 0.00015157235798693685,
+      "loss": 1.2906,
+      "step": 7548
+    },
+    {
+      "epoch": 0.3477680011056341,
+      "grad_norm": 4.796214469946825,
+      "learning_rate": 0.00015155957386826677,
+      "loss": 1.5552,
+      "step": 7549
+    },
+    {
+      "epoch": 0.3478140691942691,
+      "grad_norm": 1.0109689871808287,
+      "learning_rate": 0.00015154678860171124,
+      "loss": 1.299,
+      "step": 7550
+    },
+    {
+      "epoch": 0.34786013728290416,
+      "grad_norm": 4.347096358931975,
+      "learning_rate": 0.00015153400218755481,
+      "loss": 1.3436,
+      "step": 7551
+    },
+    {
+      "epoch": 0.34790620537153916,
+      "grad_norm": 1.2961687024223987,
+      "learning_rate": 0.00015152121462608218,
+      "loss": 1.6711,
+      "step": 7552
+    },
+    {
+      "epoch": 0.34795227346017416,
+      "grad_norm": 1.5152614763762533,
+      "learning_rate": 0.00015150842591757802,
+      "loss": 1.635,
+      "step": 7553
+    },
+    {
+      "epoch": 0.34799834154880915,
+      "grad_norm": 2.142881911442924,
+      "learning_rate": 0.0001514956360623271,
+      "loss": 1.4605,
+      "step": 7554
+    },
+    {
+      "epoch": 0.34804440963744415,
+      "grad_norm": 4.021217357616041,
+      "learning_rate": 0.00015148284506061416,
+      "loss": 1.582,
+      "step": 7555
+    },
+    {
+      "epoch": 0.34809047772607915,
+      "grad_norm": 9.585727930937416,
+      "learning_rate": 0.00015147005291272394,
+      "loss": 1.4726,
+      "step": 7556
+    },
+    {
+      "epoch": 0.34813654581471415,
+      "grad_norm": 3.647232066601959,
+      "learning_rate": 0.00015145725961894125,
+      "loss": 1.4814,
+      "step": 7557
+    },
+    {
+      "epoch": 0.34818261390334915,
+      "grad_norm": 2.7246636452474866,
+      "learning_rate": 0.0001514444651795509,
+      "loss": 1.6718,
+      "step": 7558
+    },
+    {
+      "epoch": 0.34822868199198415,
+      "grad_norm": 2.257274589238962,
+      "learning_rate": 0.00015143166959483771,
+      "loss": 1.307,
+      "step": 7559
+    },
+    {
+      "epoch": 0.34827475008061914,
+      "grad_norm": 4.503148980801874,
+      "learning_rate": 0.0001514188728650866,
+      "loss": 1.6659,
+      "step": 7560
+    },
+    {
+      "epoch": 0.34832081816925414,
+      "grad_norm": 2.166699838298507,
+      "learning_rate": 0.00015140607499058246,
+      "loss": 1.6152,
+      "step": 7561
+    },
+    {
+      "epoch": 0.34836688625788914,
+      "grad_norm": 1.2301844575913747,
+      "learning_rate": 0.00015139327597161022,
+      "loss": 1.3502,
+      "step": 7562
+    },
+    {
+      "epoch": 0.34841295434652414,
+      "grad_norm": 3.926396528587672,
+      "learning_rate": 0.00015138047580845476,
+      "loss": 1.661,
+      "step": 7563
+    },
+    {
+      "epoch": 0.34845902243515914,
+      "grad_norm": 1.5436495943266912,
+      "learning_rate": 0.00015136767450140117,
+      "loss": 1.6564,
+      "step": 7564
+    },
+    {
+      "epoch": 0.3485050905237942,
+      "grad_norm": 3.1529159045960595,
+      "learning_rate": 0.00015135487205073433,
+      "loss": 1.7047,
+      "step": 7565
+    },
+    {
+      "epoch": 0.3485511586124292,
+      "grad_norm": 7.239546412228251,
+      "learning_rate": 0.00015134206845673936,
+      "loss": 1.598,
+      "step": 7566
+    },
+    {
+      "epoch": 0.3485972267010642,
+      "grad_norm": 2.2626871706881255,
+      "learning_rate": 0.00015132926371970123,
+      "loss": 1.5271,
+      "step": 7567
+    },
+    {
+      "epoch": 0.3486432947896992,
+      "grad_norm": 3.6615260439739936,
+      "learning_rate": 0.00015131645783990507,
+      "loss": 1.5753,
+      "step": 7568
+    },
+    {
+      "epoch": 0.3486893628783342,
+      "grad_norm": 1.460602802838276,
+      "learning_rate": 0.000151303650817636,
+      "loss": 1.7159,
+      "step": 7569
+    },
+    {
+      "epoch": 0.3487354309669692,
+      "grad_norm": 5.4992751260096835,
+      "learning_rate": 0.00015129084265317903,
+      "loss": 1.4964,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3487814990556042,
+      "grad_norm": 4.063709564300868,
+      "learning_rate": 0.00015127803334681944,
+      "loss": 1.3897,
+      "step": 7571
+    },
+    {
+      "epoch": 0.3488275671442392,
+      "grad_norm": 4.661040272957364,
+      "learning_rate": 0.00015126522289884235,
+      "loss": 1.3032,
+      "step": 7572
+    },
+    {
+      "epoch": 0.3488736352328742,
+      "grad_norm": 4.096085102158213,
+      "learning_rate": 0.000151252411309533,
+      "loss": 1.775,
+      "step": 7573
+    },
+    {
+      "epoch": 0.3489197033215092,
+      "grad_norm": 3.032052120683641,
+      "learning_rate": 0.00015123959857917657,
+      "loss": 1.3809,
+      "step": 7574
+    },
+    {
+      "epoch": 0.3489657714101442,
+      "grad_norm": 2.074989632481881,
+      "learning_rate": 0.00015122678470805834,
+      "loss": 1.4244,
+      "step": 7575
+    },
+    {
+      "epoch": 0.3490118394987792,
+      "grad_norm": 1.6724589763238271,
+      "learning_rate": 0.0001512139696964636,
+      "loss": 1.5741,
+      "step": 7576
+    },
+    {
+      "epoch": 0.3490579075874142,
+      "grad_norm": 1.9457792706881696,
+      "learning_rate": 0.0001512011535446776,
+      "loss": 1.1635,
+      "step": 7577
+    },
+    {
+      "epoch": 0.3491039756760492,
+      "grad_norm": 2.248763765368277,
+      "learning_rate": 0.00015118833625298573,
+      "loss": 1.5277,
+      "step": 7578
+    },
+    {
+      "epoch": 0.3491500437646842,
+      "grad_norm": 1.6685440618323313,
+      "learning_rate": 0.00015117551782167332,
+      "loss": 1.5314,
+      "step": 7579
+    },
+    {
+      "epoch": 0.3491961118533192,
+      "grad_norm": 8.303316868011832,
+      "learning_rate": 0.00015116269825102578,
+      "loss": 1.4436,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3492421799419542,
+      "grad_norm": 5.343250506986125,
+      "learning_rate": 0.00015114987754132842,
+      "loss": 1.3036,
+      "step": 7581
+    },
+    {
+      "epoch": 0.3492882480305892,
+      "grad_norm": 1.3750783854810709,
+      "learning_rate": 0.0001511370556928668,
+      "loss": 1.5321,
+      "step": 7582
+    },
+    {
+      "epoch": 0.3493343161192242,
+      "grad_norm": 2.325729627084082,
+      "learning_rate": 0.00015112423270592632,
+      "loss": 1.4617,
+      "step": 7583
+    },
+    {
+      "epoch": 0.3493803842078592,
+      "grad_norm": 13.428241616791503,
+      "learning_rate": 0.00015111140858079243,
+      "loss": 1.601,
+      "step": 7584
+    },
+    {
+      "epoch": 0.3494264522964942,
+      "grad_norm": 6.731196103678626,
+      "learning_rate": 0.0001510985833177507,
+      "loss": 1.402,
+      "step": 7585
+    },
+    {
+      "epoch": 0.3494725203851292,
+      "grad_norm": 1.0881562678296517,
+      "learning_rate": 0.00015108575691708657,
+      "loss": 1.5586,
+      "step": 7586
+    },
+    {
+      "epoch": 0.3495185884737642,
+      "grad_norm": 21.89802451889706,
+      "learning_rate": 0.0001510729293790857,
+      "loss": 1.3236,
+      "step": 7587
+    },
+    {
+      "epoch": 0.3495646565623992,
+      "grad_norm": 7.276285210336559,
+      "learning_rate": 0.0001510601007040336,
+      "loss": 1.3956,
+      "step": 7588
+    },
+    {
+      "epoch": 0.3496107246510342,
+      "grad_norm": 3.4588991480006404,
+      "learning_rate": 0.00015104727089221593,
+      "loss": 1.4478,
+      "step": 7589
+    },
+    {
+      "epoch": 0.3496567927396692,
+      "grad_norm": 1.5656051890865836,
+      "learning_rate": 0.00015103443994391827,
+      "loss": 1.5428,
+      "step": 7590
+    },
+    {
+      "epoch": 0.3497028608283042,
+      "grad_norm": 2.700170058208463,
+      "learning_rate": 0.00015102160785942633,
+      "loss": 1.4688,
+      "step": 7591
+    },
+    {
+      "epoch": 0.34974892891693926,
+      "grad_norm": 3.736449181068952,
+      "learning_rate": 0.00015100877463902575,
+      "loss": 1.3845,
+      "step": 7592
+    },
+    {
+      "epoch": 0.34979499700557426,
+      "grad_norm": 1.3678543651629311,
+      "learning_rate": 0.00015099594028300235,
+      "loss": 1.6877,
+      "step": 7593
+    },
+    {
+      "epoch": 0.34984106509420926,
+      "grad_norm": 2.4305519788721033,
+      "learning_rate": 0.0001509831047916417,
+      "loss": 1.2735,
+      "step": 7594
+    },
+    {
+      "epoch": 0.34988713318284426,
+      "grad_norm": 1.7035522599783002,
+      "learning_rate": 0.00015097026816522969,
+      "loss": 1.4178,
+      "step": 7595
+    },
+    {
+      "epoch": 0.34993320127147926,
+      "grad_norm": 1.5450394254830524,
+      "learning_rate": 0.00015095743040405196,
+      "loss": 1.6985,
+      "step": 7596
+    },
+    {
+      "epoch": 0.34997926936011425,
+      "grad_norm": 1.159579904718869,
+      "learning_rate": 0.0001509445915083945,
+      "loss": 1.372,
+      "step": 7597
+    },
+    {
+      "epoch": 0.35002533744874925,
+      "grad_norm": 2.7099933823795137,
+      "learning_rate": 0.00015093175147854303,
+      "loss": 1.6085,
+      "step": 7598
+    },
+    {
+      "epoch": 0.35007140553738425,
+      "grad_norm": 3.1496111082547387,
+      "learning_rate": 0.00015091891031478348,
+      "loss": 1.576,
+      "step": 7599
+    },
+    {
+      "epoch": 0.35011747362601925,
+      "grad_norm": 2.324043699224164,
+      "learning_rate": 0.00015090606801740166,
+      "loss": 1.583,
+      "step": 7600
+    },
+    {
+      "epoch": 0.35016354171465425,
+      "grad_norm": 2.3276784235703807,
+      "learning_rate": 0.00015089322458668354,
+      "loss": 1.653,
+      "step": 7601
+    },
+    {
+      "epoch": 0.35020960980328925,
+      "grad_norm": 4.136758271468211,
+      "learning_rate": 0.000150880380022915,
+      "loss": 1.3452,
+      "step": 7602
+    },
+    {
+      "epoch": 0.35025567789192424,
+      "grad_norm": 1.7165555209473964,
+      "learning_rate": 0.00015086753432638208,
+      "loss": 1.3033,
+      "step": 7603
+    },
+    {
+      "epoch": 0.35030174598055924,
+      "grad_norm": 1.719271706506566,
+      "learning_rate": 0.00015085468749737074,
+      "loss": 1.3316,
+      "step": 7604
+    },
+    {
+      "epoch": 0.3503478140691943,
+      "grad_norm": 2.328740313163438,
+      "learning_rate": 0.00015084183953616696,
+      "loss": 1.3238,
+      "step": 7605
+    },
+    {
+      "epoch": 0.3503938821578293,
+      "grad_norm": 3.1736334943063165,
+      "learning_rate": 0.0001508289904430568,
+      "loss": 1.4218,
+      "step": 7606
+    },
+    {
+      "epoch": 0.3504399502464643,
+      "grad_norm": 2.3062443284958802,
+      "learning_rate": 0.00015081614021832628,
+      "loss": 1.2697,
+      "step": 7607
+    },
+    {
+      "epoch": 0.3504860183350993,
+      "grad_norm": 2.3886682748272126,
+      "learning_rate": 0.00015080328886226158,
+      "loss": 1.4915,
+      "step": 7608
+    },
+    {
+      "epoch": 0.3505320864237343,
+      "grad_norm": 1.883825674442341,
+      "learning_rate": 0.00015079043637514873,
+      "loss": 1.8273,
+      "step": 7609
+    },
+    {
+      "epoch": 0.3505781545123693,
+      "grad_norm": 1.3365553898081748,
+      "learning_rate": 0.00015077758275727394,
+      "loss": 1.5762,
+      "step": 7610
+    },
+    {
+      "epoch": 0.3506242226010043,
+      "grad_norm": 2.019236588736867,
+      "learning_rate": 0.0001507647280089233,
+      "loss": 1.3503,
+      "step": 7611
+    },
+    {
+      "epoch": 0.3506702906896393,
+      "grad_norm": 1.951866274671235,
+      "learning_rate": 0.00015075187213038304,
+      "loss": 1.6085,
+      "step": 7612
+    },
+    {
+      "epoch": 0.3507163587782743,
+      "grad_norm": 2.944683411075365,
+      "learning_rate": 0.0001507390151219394,
+      "loss": 1.3012,
+      "step": 7613
+    },
+    {
+      "epoch": 0.3507624268669093,
+      "grad_norm": 1.4810831595711473,
+      "learning_rate": 0.00015072615698387854,
+      "loss": 1.6257,
+      "step": 7614
+    },
+    {
+      "epoch": 0.3508084949555443,
+      "grad_norm": 1.4955256520557882,
+      "learning_rate": 0.00015071329771648682,
+      "loss": 1.5356,
+      "step": 7615
+    },
+    {
+      "epoch": 0.3508545630441793,
+      "grad_norm": 1.4983753692081765,
+      "learning_rate": 0.0001507004373200504,
+      "loss": 1.6385,
+      "step": 7616
+    },
+    {
+      "epoch": 0.3509006311328143,
+      "grad_norm": 1.3263144634945538,
+      "learning_rate": 0.00015068757579485577,
+      "loss": 1.3899,
+      "step": 7617
+    },
+    {
+      "epoch": 0.3509466992214493,
+      "grad_norm": 2.264295364757836,
+      "learning_rate": 0.00015067471314118911,
+      "loss": 1.6149,
+      "step": 7618
+    },
+    {
+      "epoch": 0.35099276731008433,
+      "grad_norm": 1.1310664025144197,
+      "learning_rate": 0.0001506618493593369,
+      "loss": 1.4417,
+      "step": 7619
+    },
+    {
+      "epoch": 0.35103883539871933,
+      "grad_norm": 3.0647542487875374,
+      "learning_rate": 0.00015064898444958543,
+      "loss": 1.3513,
+      "step": 7620
+    },
+    {
+      "epoch": 0.3510849034873543,
+      "grad_norm": 1.6676766132701364,
+      "learning_rate": 0.00015063611841222122,
+      "loss": 1.4975,
+      "step": 7621
+    },
+    {
+      "epoch": 0.3511309715759893,
+      "grad_norm": 1.5218074968755784,
+      "learning_rate": 0.0001506232512475306,
+      "loss": 1.5812,
+      "step": 7622
+    },
+    {
+      "epoch": 0.3511770396646243,
+      "grad_norm": 3.1522360596333425,
+      "learning_rate": 0.00015061038295580012,
+      "loss": 1.5278,
+      "step": 7623
+    },
+    {
+      "epoch": 0.3512231077532593,
+      "grad_norm": 1.7219359590820025,
+      "learning_rate": 0.00015059751353731625,
+      "loss": 1.547,
+      "step": 7624
+    },
+    {
+      "epoch": 0.3512691758418943,
+      "grad_norm": 1.9768235580580888,
+      "learning_rate": 0.00015058464299236548,
+      "loss": 1.6034,
+      "step": 7625
+    },
+    {
+      "epoch": 0.3513152439305293,
+      "grad_norm": 2.938134381023097,
+      "learning_rate": 0.0001505717713212344,
+      "loss": 1.4631,
+      "step": 7626
+    },
+    {
+      "epoch": 0.3513613120191643,
+      "grad_norm": 2.6344626224331136,
+      "learning_rate": 0.0001505588985242095,
+      "loss": 1.4104,
+      "step": 7627
+    },
+    {
+      "epoch": 0.3514073801077993,
+      "grad_norm": 1.8780315690217722,
+      "learning_rate": 0.00015054602460157744,
+      "loss": 1.3184,
+      "step": 7628
+    },
+    {
+      "epoch": 0.3514534481964343,
+      "grad_norm": 6.995126607710581,
+      "learning_rate": 0.00015053314955362478,
+      "loss": 1.4214,
+      "step": 7629
+    },
+    {
+      "epoch": 0.3514995162850693,
+      "grad_norm": 2.3061178056738685,
+      "learning_rate": 0.00015052027338063827,
+      "loss": 1.3077,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3515455843737043,
+      "grad_norm": 2.207396109700973,
+      "learning_rate": 0.0001505073960829044,
+      "loss": 1.5909,
+      "step": 7631
+    },
+    {
+      "epoch": 0.3515916524623393,
+      "grad_norm": 1.868855652764806,
+      "learning_rate": 0.00015049451766071002,
+      "loss": 1.2025,
+      "step": 7632
+    },
+    {
+      "epoch": 0.35163772055097436,
+      "grad_norm": 2.417774911871874,
+      "learning_rate": 0.00015048163811434175,
+      "loss": 1.4398,
+      "step": 7633
+    },
+    {
+      "epoch": 0.35168378863960936,
+      "grad_norm": 1.453722214423492,
+      "learning_rate": 0.00015046875744408636,
+      "loss": 1.4703,
+      "step": 7634
+    },
+    {
+      "epoch": 0.35172985672824436,
+      "grad_norm": 1.4276382011541018,
+      "learning_rate": 0.00015045587565023065,
+      "loss": 1.6533,
+      "step": 7635
+    },
+    {
+      "epoch": 0.35177592481687936,
+      "grad_norm": 1.2129645943263094,
+      "learning_rate": 0.00015044299273306137,
+      "loss": 1.4332,
+      "step": 7636
+    },
+    {
+      "epoch": 0.35182199290551436,
+      "grad_norm": 22.654299270384833,
+      "learning_rate": 0.00015043010869286534,
+      "loss": 1.508,
+      "step": 7637
+    },
+    {
+      "epoch": 0.35186806099414936,
+      "grad_norm": 1.511079497537471,
+      "learning_rate": 0.0001504172235299294,
+      "loss": 1.521,
+      "step": 7638
+    },
+    {
+      "epoch": 0.35191412908278435,
+      "grad_norm": 2.15926874886527,
+      "learning_rate": 0.00015040433724454049,
+      "loss": 1.2559,
+      "step": 7639
+    },
+    {
+      "epoch": 0.35196019717141935,
+      "grad_norm": 1.4976476691021723,
+      "learning_rate": 0.00015039144983698536,
+      "loss": 1.133,
+      "step": 7640
+    },
+    {
+      "epoch": 0.35200626526005435,
+      "grad_norm": 11.804314678458146,
+      "learning_rate": 0.00015037856130755106,
+      "loss": 1.4758,
+      "step": 7641
+    },
+    {
+      "epoch": 0.35205233334868935,
+      "grad_norm": 6.087661467136737,
+      "learning_rate": 0.0001503656716565244,
+      "loss": 1.2828,
+      "step": 7642
+    },
+    {
+      "epoch": 0.35209840143732435,
+      "grad_norm": 3.2989992844491836,
+      "learning_rate": 0.00015035278088419249,
+      "loss": 1.3414,
+      "step": 7643
+    },
+    {
+      "epoch": 0.35214446952595935,
+      "grad_norm": 6.009384296075156,
+      "learning_rate": 0.00015033988899084222,
+      "loss": 1.3727,
+      "step": 7644
+    },
+    {
+      "epoch": 0.35219053761459435,
+      "grad_norm": 1.668438537420532,
+      "learning_rate": 0.00015032699597676064,
+      "loss": 1.4835,
+      "step": 7645
+    },
+    {
+      "epoch": 0.3522366057032294,
+      "grad_norm": 2.1558703976687186,
+      "learning_rate": 0.00015031410184223478,
+      "loss": 1.42,
+      "step": 7646
+    },
+    {
+      "epoch": 0.3522826737918644,
+      "grad_norm": 3.8443516045552593,
+      "learning_rate": 0.0001503012065875517,
+      "loss": 1.2809,
+      "step": 7647
+    },
+    {
+      "epoch": 0.3523287418804994,
+      "grad_norm": 2.3985831208330066,
+      "learning_rate": 0.00015028831021299857,
+      "loss": 1.6436,
+      "step": 7648
+    },
+    {
+      "epoch": 0.3523748099691344,
+      "grad_norm": 2.38667564606342,
+      "learning_rate": 0.00015027541271886238,
+      "loss": 1.6282,
+      "step": 7649
+    },
+    {
+      "epoch": 0.3524208780577694,
+      "grad_norm": 1.6904582267734063,
+      "learning_rate": 0.00015026251410543034,
+      "loss": 1.4569,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3524669461464044,
+      "grad_norm": 1.8114516403918042,
+      "learning_rate": 0.00015024961437298958,
+      "loss": 1.5484,
+      "step": 7651
+    },
+    {
+      "epoch": 0.3525130142350394,
+      "grad_norm": 1.7205666283786942,
+      "learning_rate": 0.00015023671352182736,
+      "loss": 1.5984,
+      "step": 7652
+    },
+    {
+      "epoch": 0.3525590823236744,
+      "grad_norm": 1.5708240353515088,
+      "learning_rate": 0.00015022381155223083,
+      "loss": 1.5882,
+      "step": 7653
+    },
+    {
+      "epoch": 0.3526051504123094,
+      "grad_norm": 2.2774491938560772,
+      "learning_rate": 0.00015021090846448724,
+      "loss": 1.4854,
+      "step": 7654
+    },
+    {
+      "epoch": 0.3526512185009444,
+      "grad_norm": 2.0969793881467593,
+      "learning_rate": 0.0001501980042588839,
+      "loss": 1.4576,
+      "step": 7655
+    },
+    {
+      "epoch": 0.3526972865895794,
+      "grad_norm": 4.308698170668681,
+      "learning_rate": 0.00015018509893570804,
+      "loss": 1.4366,
+      "step": 7656
+    },
+    {
+      "epoch": 0.3527433546782144,
+      "grad_norm": 2.259183319406709,
+      "learning_rate": 0.00015017219249524702,
+      "loss": 1.3916,
+      "step": 7657
+    },
+    {
+      "epoch": 0.3527894227668494,
+      "grad_norm": 2.0582465579718745,
+      "learning_rate": 0.00015015928493778815,
+      "loss": 1.4144,
+      "step": 7658
+    },
+    {
+      "epoch": 0.3528354908554844,
+      "grad_norm": 2.279242048856981,
+      "learning_rate": 0.0001501463762636188,
+      "loss": 1.1481,
+      "step": 7659
+    },
+    {
+      "epoch": 0.35288155894411943,
+      "grad_norm": 2.290449250544178,
+      "learning_rate": 0.00015013346647302635,
+      "loss": 1.3984,
+      "step": 7660
+    },
+    {
+      "epoch": 0.35292762703275443,
+      "grad_norm": 2.0709392930099684,
+      "learning_rate": 0.00015012055556629824,
+      "loss": 1.2798,
+      "step": 7661
+    },
+    {
+      "epoch": 0.35297369512138943,
+      "grad_norm": 1.415155541220382,
+      "learning_rate": 0.0001501076435437219,
+      "loss": 1.3227,
+      "step": 7662
+    },
+    {
+      "epoch": 0.35301976321002443,
+      "grad_norm": 1.92861851873316,
+      "learning_rate": 0.0001500947304055848,
+      "loss": 1.5634,
+      "step": 7663
+    },
+    {
+      "epoch": 0.3530658312986594,
+      "grad_norm": 2.2973758017738253,
+      "learning_rate": 0.00015008181615217442,
+      "loss": 1.3091,
+      "step": 7664
+    },
+    {
+      "epoch": 0.3531118993872944,
+      "grad_norm": 1.7373515032217164,
+      "learning_rate": 0.0001500689007837783,
+      "loss": 1.4829,
+      "step": 7665
+    },
+    {
+      "epoch": 0.3531579674759294,
+      "grad_norm": 4.45153064487976,
+      "learning_rate": 0.00015005598430068394,
+      "loss": 1.1049,
+      "step": 7666
+    },
+    {
+      "epoch": 0.3532040355645644,
+      "grad_norm": 1.4736101371063686,
+      "learning_rate": 0.00015004306670317888,
+      "loss": 1.5687,
+      "step": 7667
+    },
+    {
+      "epoch": 0.3532501036531994,
+      "grad_norm": 2.0178029448463892,
+      "learning_rate": 0.0001500301479915508,
+      "loss": 1.3931,
+      "step": 7668
+    },
+    {
+      "epoch": 0.3532961717418344,
+      "grad_norm": 1.9318160258819046,
+      "learning_rate": 0.00015001722816608722,
+      "loss": 1.7255,
+      "step": 7669
+    },
+    {
+      "epoch": 0.3533422398304694,
+      "grad_norm": 1.5647506410672034,
+      "learning_rate": 0.00015000430722707583,
+      "loss": 1.2419,
+      "step": 7670
+    },
+    {
+      "epoch": 0.3533883079191044,
+      "grad_norm": 1.293203504526779,
+      "learning_rate": 0.00014999138517480428,
+      "loss": 1.3882,
+      "step": 7671
+    },
+    {
+      "epoch": 0.3534343760077394,
+      "grad_norm": 1.353393425960712,
+      "learning_rate": 0.00014997846200956024,
+      "loss": 1.2897,
+      "step": 7672
+    },
+    {
+      "epoch": 0.35348044409637447,
+      "grad_norm": 2.6730256802554004,
+      "learning_rate": 0.00014996553773163144,
+      "loss": 1.3631,
+      "step": 7673
+    },
+    {
+      "epoch": 0.35352651218500947,
+      "grad_norm": 2.2836696264047522,
+      "learning_rate": 0.00014995261234130564,
+      "loss": 1.709,
+      "step": 7674
+    },
+    {
+      "epoch": 0.35357258027364447,
+      "grad_norm": 1.7883811141123993,
+      "learning_rate": 0.00014993968583887056,
+      "loss": 1.2843,
+      "step": 7675
+    },
+    {
+      "epoch": 0.35361864836227946,
+      "grad_norm": 3.3791735253353616,
+      "learning_rate": 0.000149926758224614,
+      "loss": 1.339,
+      "step": 7676
+    },
+    {
+      "epoch": 0.35366471645091446,
+      "grad_norm": 1.5541561579880487,
+      "learning_rate": 0.00014991382949882377,
+      "loss": 1.353,
+      "step": 7677
+    },
+    {
+      "epoch": 0.35371078453954946,
+      "grad_norm": 1.7556433365232618,
+      "learning_rate": 0.00014990089966178773,
+      "loss": 1.1948,
+      "step": 7678
+    },
+    {
+      "epoch": 0.35375685262818446,
+      "grad_norm": 1.5967885020988986,
+      "learning_rate": 0.00014988796871379372,
+      "loss": 1.3981,
+      "step": 7679
+    },
+    {
+      "epoch": 0.35380292071681946,
+      "grad_norm": 2.135914480612778,
+      "learning_rate": 0.00014987503665512959,
+      "loss": 1.7301,
+      "step": 7680
+    },
+    {
+      "epoch": 0.35384898880545446,
+      "grad_norm": 6.150601545831394,
+      "learning_rate": 0.00014986210348608335,
+      "loss": 1.5064,
+      "step": 7681
+    },
+    {
+      "epoch": 0.35389505689408945,
+      "grad_norm": 1.8803871069508216,
+      "learning_rate": 0.0001498491692069428,
+      "loss": 1.4366,
+      "step": 7682
+    },
+    {
+      "epoch": 0.35394112498272445,
+      "grad_norm": 1.2539880012665447,
+      "learning_rate": 0.000149836233817996,
+      "loss": 1.4924,
+      "step": 7683
+    },
+    {
+      "epoch": 0.35398719307135945,
+      "grad_norm": 1.3613839190195243,
+      "learning_rate": 0.0001498232973195309,
+      "loss": 1.5084,
+      "step": 7684
+    },
+    {
+      "epoch": 0.35403326115999445,
+      "grad_norm": 1.1664249278666812,
+      "learning_rate": 0.00014981035971183557,
+      "loss": 1.3174,
+      "step": 7685
+    },
+    {
+      "epoch": 0.35407932924862945,
+      "grad_norm": 1.2030248093476827,
+      "learning_rate": 0.00014979742099519792,
+      "loss": 1.3717,
+      "step": 7686
+    },
+    {
+      "epoch": 0.3541253973372645,
+      "grad_norm": 2.682807224247457,
+      "learning_rate": 0.0001497844811699061,
+      "loss": 1.537,
+      "step": 7687
+    },
+    {
+      "epoch": 0.3541714654258995,
+      "grad_norm": 1.432580176012761,
+      "learning_rate": 0.00014977154023624817,
+      "loss": 1.4025,
+      "step": 7688
+    },
+    {
+      "epoch": 0.3542175335145345,
+      "grad_norm": 1.244540131784371,
+      "learning_rate": 0.0001497585981945122,
+      "loss": 1.2373,
+      "step": 7689
+    },
+    {
+      "epoch": 0.3542636016031695,
+      "grad_norm": 2.5541388253454165,
+      "learning_rate": 0.00014974565504498642,
+      "loss": 1.2075,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3543096696918045,
+      "grad_norm": 2.225872430742455,
+      "learning_rate": 0.00014973271078795887,
+      "loss": 1.4598,
+      "step": 7691
+    },
+    {
+      "epoch": 0.3543557377804395,
+      "grad_norm": 1.77962201279175,
+      "learning_rate": 0.00014971976542371782,
+      "loss": 1.4549,
+      "step": 7692
+    },
+    {
+      "epoch": 0.3544018058690745,
+      "grad_norm": 1.9933503431783908,
+      "learning_rate": 0.00014970681895255144,
+      "loss": 1.4995,
+      "step": 7693
+    },
+    {
+      "epoch": 0.3544478739577095,
+      "grad_norm": 4.421407980548896,
+      "learning_rate": 0.00014969387137474797,
+      "loss": 1.1939,
+      "step": 7694
+    },
+    {
+      "epoch": 0.3544939420463445,
+      "grad_norm": 3.530252217269373,
+      "learning_rate": 0.00014968092269059566,
+      "loss": 1.4946,
+      "step": 7695
+    },
+    {
+      "epoch": 0.3545400101349795,
+      "grad_norm": 2.7730100968842004,
+      "learning_rate": 0.0001496679729003828,
+      "loss": 1.4183,
+      "step": 7696
+    },
+    {
+      "epoch": 0.3545860782236145,
+      "grad_norm": 1.5875990951512449,
+      "learning_rate": 0.00014965502200439766,
+      "loss": 1.6424,
+      "step": 7697
+    },
+    {
+      "epoch": 0.3546321463122495,
+      "grad_norm": 2.2238757511055027,
+      "learning_rate": 0.00014964207000292857,
+      "loss": 1.4555,
+      "step": 7698
+    },
+    {
+      "epoch": 0.3546782144008845,
+      "grad_norm": 3.065775080736559,
+      "learning_rate": 0.00014962911689626396,
+      "loss": 1.4814,
+      "step": 7699
+    },
+    {
+      "epoch": 0.3547242824895195,
+      "grad_norm": 8.120721920591759,
+      "learning_rate": 0.00014961616268469215,
+      "loss": 1.8175,
+      "step": 7700
+    },
+    {
+      "epoch": 0.35477035057815454,
+      "grad_norm": 3.9155834511717873,
+      "learning_rate": 0.00014960320736850152,
+      "loss": 1.7065,
+      "step": 7701
+    },
+    {
+      "epoch": 0.35481641866678953,
+      "grad_norm": 1.8841201525041804,
+      "learning_rate": 0.00014959025094798056,
+      "loss": 1.3365,
+      "step": 7702
+    },
+    {
+      "epoch": 0.35486248675542453,
+      "grad_norm": 1.450143119257742,
+      "learning_rate": 0.0001495772934234177,
+      "loss": 1.4539,
+      "step": 7703
+    },
+    {
+      "epoch": 0.35490855484405953,
+      "grad_norm": 1.583778005576858,
+      "learning_rate": 0.00014956433479510137,
+      "loss": 1.447,
+      "step": 7704
+    },
+    {
+      "epoch": 0.35495462293269453,
+      "grad_norm": 2.720767236160874,
+      "learning_rate": 0.00014955137506332016,
+      "loss": 1.3041,
+      "step": 7705
+    },
+    {
+      "epoch": 0.35500069102132953,
+      "grad_norm": 1.4753495618470411,
+      "learning_rate": 0.0001495384142283625,
+      "loss": 1.5339,
+      "step": 7706
+    },
+    {
+      "epoch": 0.3550467591099645,
+      "grad_norm": 4.45095973082483,
+      "learning_rate": 0.000149525452290517,
+      "loss": 1.595,
+      "step": 7707
+    },
+    {
+      "epoch": 0.3550928271985995,
+      "grad_norm": 1.1580639821911476,
+      "learning_rate": 0.00014951248925007223,
+      "loss": 1.4705,
+      "step": 7708
+    },
+    {
+      "epoch": 0.3551388952872345,
+      "grad_norm": 1.809088411076334,
+      "learning_rate": 0.0001494995251073168,
+      "loss": 1.4103,
+      "step": 7709
+    },
+    {
+      "epoch": 0.3551849633758695,
+      "grad_norm": 1.7509278493503164,
+      "learning_rate": 0.00014948655986253928,
+      "loss": 1.5582,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3552310314645045,
+      "grad_norm": 1.7476193637448614,
+      "learning_rate": 0.00014947359351602838,
+      "loss": 1.4806,
+      "step": 7711
+    },
+    {
+      "epoch": 0.3552770995531395,
+      "grad_norm": 5.290515607915056,
+      "learning_rate": 0.00014946062606807276,
+      "loss": 1.235,
+      "step": 7712
+    },
+    {
+      "epoch": 0.3553231676417745,
+      "grad_norm": 2.0070588041872877,
+      "learning_rate": 0.00014944765751896108,
+      "loss": 1.4613,
+      "step": 7713
+    },
+    {
+      "epoch": 0.35536923573040957,
+      "grad_norm": 5.928112444800271,
+      "learning_rate": 0.00014943468786898212,
+      "loss": 1.5892,
+      "step": 7714
+    },
+    {
+      "epoch": 0.35541530381904457,
+      "grad_norm": 2.0714933019602237,
+      "learning_rate": 0.00014942171711842457,
+      "loss": 1.4138,
+      "step": 7715
+    },
+    {
+      "epoch": 0.35546137190767957,
+      "grad_norm": 1.7499001659419138,
+      "learning_rate": 0.00014940874526757724,
+      "loss": 1.3738,
+      "step": 7716
+    },
+    {
+      "epoch": 0.35550743999631457,
+      "grad_norm": 3.0542613614051297,
+      "learning_rate": 0.0001493957723167289,
+      "loss": 1.1112,
+      "step": 7717
+    },
+    {
+      "epoch": 0.35555350808494957,
+      "grad_norm": 1.183470363868738,
+      "learning_rate": 0.00014938279826616838,
+      "loss": 1.3417,
+      "step": 7718
+    },
+    {
+      "epoch": 0.35559957617358456,
+      "grad_norm": 3.0918169562332065,
+      "learning_rate": 0.00014936982311618456,
+      "loss": 1.4241,
+      "step": 7719
+    },
+    {
+      "epoch": 0.35564564426221956,
+      "grad_norm": 1.8915782843744788,
+      "learning_rate": 0.00014935684686706626,
+      "loss": 1.3773,
+      "step": 7720
+    },
+    {
+      "epoch": 0.35569171235085456,
+      "grad_norm": 1.327194632807982,
+      "learning_rate": 0.00014934386951910241,
+      "loss": 1.2108,
+      "step": 7721
+    },
+    {
+      "epoch": 0.35573778043948956,
+      "grad_norm": 2.482485296053542,
+      "learning_rate": 0.00014933089107258189,
+      "loss": 1.493,
+      "step": 7722
+    },
+    {
+      "epoch": 0.35578384852812456,
+      "grad_norm": 3.8024738333534103,
+      "learning_rate": 0.0001493179115277937,
+      "loss": 1.5693,
+      "step": 7723
+    },
+    {
+      "epoch": 0.35582991661675956,
+      "grad_norm": 1.7712977610787115,
+      "learning_rate": 0.00014930493088502673,
+      "loss": 1.5029,
+      "step": 7724
+    },
+    {
+      "epoch": 0.35587598470539455,
+      "grad_norm": 3.8612927619321202,
+      "learning_rate": 0.00014929194914457003,
+      "loss": 1.1778,
+      "step": 7725
+    },
+    {
+      "epoch": 0.35592205279402955,
+      "grad_norm": 2.041911184107146,
+      "learning_rate": 0.00014927896630671257,
+      "loss": 1.2992,
+      "step": 7726
+    },
+    {
+      "epoch": 0.35596812088266455,
+      "grad_norm": 1.6531693321546428,
+      "learning_rate": 0.00014926598237174348,
+      "loss": 1.444,
+      "step": 7727
+    },
+    {
+      "epoch": 0.3560141889712996,
+      "grad_norm": 7.485036501348614,
+      "learning_rate": 0.00014925299733995173,
+      "loss": 1.4195,
+      "step": 7728
+    },
+    {
+      "epoch": 0.3560602570599346,
+      "grad_norm": 1.9011831532535226,
+      "learning_rate": 0.00014924001121162646,
+      "loss": 1.3353,
+      "step": 7729
+    },
+    {
+      "epoch": 0.3561063251485696,
+      "grad_norm": 1.2346720751115083,
+      "learning_rate": 0.00014922702398705677,
+      "loss": 1.5463,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3561523932372046,
+      "grad_norm": 1.0945528128779485,
+      "learning_rate": 0.0001492140356665318,
+      "loss": 1.2908,
+      "step": 7731
+    },
+    {
+      "epoch": 0.3561984613258396,
+      "grad_norm": 2.7810599377328313,
+      "learning_rate": 0.00014920104625034067,
+      "loss": 1.5798,
+      "step": 7732
+    },
+    {
+      "epoch": 0.3562445294144746,
+      "grad_norm": 1.6047565495678646,
+      "learning_rate": 0.00014918805573877264,
+      "loss": 1.3828,
+      "step": 7733
+    },
+    {
+      "epoch": 0.3562905975031096,
+      "grad_norm": 4.329632257241812,
+      "learning_rate": 0.0001491750641321169,
+      "loss": 1.2543,
+      "step": 7734
+    },
+    {
+      "epoch": 0.3563366655917446,
+      "grad_norm": 1.5232752935937044,
+      "learning_rate": 0.00014916207143066262,
+      "loss": 1.4456,
+      "step": 7735
+    },
+    {
+      "epoch": 0.3563827336803796,
+      "grad_norm": 1.964227243981464,
+      "learning_rate": 0.00014914907763469918,
+      "loss": 1.1715,
+      "step": 7736
+    },
+    {
+      "epoch": 0.3564288017690146,
+      "grad_norm": 2.963625081232281,
+      "learning_rate": 0.00014913608274451576,
+      "loss": 1.2581,
+      "step": 7737
+    },
+    {
+      "epoch": 0.3564748698576496,
+      "grad_norm": 2.218777336658401,
+      "learning_rate": 0.0001491230867604017,
+      "loss": 1.5196,
+      "step": 7738
+    },
+    {
+      "epoch": 0.3565209379462846,
+      "grad_norm": 1.7001892754206926,
+      "learning_rate": 0.0001491100896826464,
+      "loss": 1.2591,
+      "step": 7739
+    },
+    {
+      "epoch": 0.3565670060349196,
+      "grad_norm": 2.1023812647987556,
+      "learning_rate": 0.00014909709151153912,
+      "loss": 1.5108,
+      "step": 7740
+    },
+    {
+      "epoch": 0.35661307412355464,
+      "grad_norm": 2.168323308027782,
+      "learning_rate": 0.00014908409224736926,
+      "loss": 1.339,
+      "step": 7741
+    },
+    {
+      "epoch": 0.35665914221218964,
+      "grad_norm": 2.2996835821809003,
+      "learning_rate": 0.0001490710918904263,
+      "loss": 1.3975,
+      "step": 7742
+    },
+    {
+      "epoch": 0.35670521030082464,
+      "grad_norm": 5.618487771683483,
+      "learning_rate": 0.00014905809044099956,
+      "loss": 1.3598,
+      "step": 7743
+    },
+    {
+      "epoch": 0.35675127838945964,
+      "grad_norm": 3.4259654224623706,
+      "learning_rate": 0.00014904508789937858,
+      "loss": 1.5684,
+      "step": 7744
+    },
+    {
+      "epoch": 0.35679734647809463,
+      "grad_norm": 1.3643137107811254,
+      "learning_rate": 0.0001490320842658528,
+      "loss": 1.6595,
+      "step": 7745
+    },
+    {
+      "epoch": 0.35684341456672963,
+      "grad_norm": 4.695877678230561,
+      "learning_rate": 0.00014901907954071174,
+      "loss": 1.5155,
+      "step": 7746
+    },
+    {
+      "epoch": 0.35688948265536463,
+      "grad_norm": 2.019006260512246,
+      "learning_rate": 0.00014900607372424495,
+      "loss": 1.8297,
+      "step": 7747
+    },
+    {
+      "epoch": 0.35693555074399963,
+      "grad_norm": 2.5491279942633467,
+      "learning_rate": 0.00014899306681674193,
+      "loss": 1.2744,
+      "step": 7748
+    },
+    {
+      "epoch": 0.35698161883263463,
+      "grad_norm": 3.8222599337304377,
+      "learning_rate": 0.0001489800588184923,
+      "loss": 1.2558,
+      "step": 7749
+    },
+    {
+      "epoch": 0.3570276869212696,
+      "grad_norm": 1.8241510725031822,
+      "learning_rate": 0.00014896704972978562,
+      "loss": 1.3183,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3570737550099046,
+      "grad_norm": 3.1421864617470554,
+      "learning_rate": 0.00014895403955091156,
+      "loss": 1.6421,
+      "step": 7751
+    },
+    {
+      "epoch": 0.3571198230985396,
+      "grad_norm": 2.7311572273110443,
+      "learning_rate": 0.00014894102828215972,
+      "loss": 1.3933,
+      "step": 7752
+    },
+    {
+      "epoch": 0.3571658911871746,
+      "grad_norm": 1.176727326141504,
+      "learning_rate": 0.00014892801592381985,
+      "loss": 1.4643,
+      "step": 7753
+    },
+    {
+      "epoch": 0.3572119592758096,
+      "grad_norm": 2.7669489367260836,
+      "learning_rate": 0.00014891500247618154,
+      "loss": 1.3752,
+      "step": 7754
+    },
+    {
+      "epoch": 0.3572580273644447,
+      "grad_norm": 3.024059375491631,
+      "learning_rate": 0.0001489019879395346,
+      "loss": 1.28,
+      "step": 7755
+    },
+    {
+      "epoch": 0.3573040954530797,
+      "grad_norm": 1.9621900191016093,
+      "learning_rate": 0.00014888897231416873,
+      "loss": 1.6864,
+      "step": 7756
+    },
+    {
+      "epoch": 0.35735016354171467,
+      "grad_norm": 2.3101675985456356,
+      "learning_rate": 0.00014887595560037376,
+      "loss": 1.274,
+      "step": 7757
+    },
+    {
+      "epoch": 0.35739623163034967,
+      "grad_norm": 1.6841244047238353,
+      "learning_rate": 0.00014886293779843942,
+      "loss": 1.4611,
+      "step": 7758
+    },
+    {
+      "epoch": 0.35744229971898467,
+      "grad_norm": 1.4069198989975613,
+      "learning_rate": 0.00014884991890865554,
+      "loss": 1.5695,
+      "step": 7759
+    },
+    {
+      "epoch": 0.35748836780761967,
+      "grad_norm": 1.926302993887033,
+      "learning_rate": 0.000148836898931312,
+      "loss": 1.5738,
+      "step": 7760
+    },
+    {
+      "epoch": 0.35753443589625467,
+      "grad_norm": 3.927533552407155,
+      "learning_rate": 0.0001488238778666986,
+      "loss": 1.1838,
+      "step": 7761
+    },
+    {
+      "epoch": 0.35758050398488966,
+      "grad_norm": 2.586509352402962,
+      "learning_rate": 0.0001488108557151053,
+      "loss": 1.3834,
+      "step": 7762
+    },
+    {
+      "epoch": 0.35762657207352466,
+      "grad_norm": 2.6882594020851536,
+      "learning_rate": 0.000148797832476822,
+      "loss": 1.4382,
+      "step": 7763
+    },
+    {
+      "epoch": 0.35767264016215966,
+      "grad_norm": 3.8823815333977163,
+      "learning_rate": 0.0001487848081521386,
+      "loss": 1.4056,
+      "step": 7764
+    },
+    {
+      "epoch": 0.35771870825079466,
+      "grad_norm": 1.5132700684889477,
+      "learning_rate": 0.00014877178274134514,
+      "loss": 1.5164,
+      "step": 7765
+    },
+    {
+      "epoch": 0.35776477633942966,
+      "grad_norm": 1.059592184967068,
+      "learning_rate": 0.00014875875624473155,
+      "loss": 1.7382,
+      "step": 7766
+    },
+    {
+      "epoch": 0.35781084442806466,
+      "grad_norm": 3.7932768951792575,
+      "learning_rate": 0.00014874572866258785,
+      "loss": 1.2315,
+      "step": 7767
+    },
+    {
+      "epoch": 0.35785691251669965,
+      "grad_norm": 2.559765655771637,
+      "learning_rate": 0.00014873269999520408,
+      "loss": 1.4314,
+      "step": 7768
+    },
+    {
+      "epoch": 0.3579029806053347,
+      "grad_norm": 3.456410089222315,
+      "learning_rate": 0.00014871967024287028,
+      "loss": 1.1159,
+      "step": 7769
+    },
+    {
+      "epoch": 0.3579490486939697,
+      "grad_norm": 5.778225099581439,
+      "learning_rate": 0.00014870663940587658,
+      "loss": 1.5255,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3579951167826047,
+      "grad_norm": 1.7888139037732402,
+      "learning_rate": 0.00014869360748451306,
+      "loss": 1.6581,
+      "step": 7771
+    },
+    {
+      "epoch": 0.3580411848712397,
+      "grad_norm": 1.8498018516268422,
+      "learning_rate": 0.00014868057447906986,
+      "loss": 1.7489,
+      "step": 7772
+    },
+    {
+      "epoch": 0.3580872529598747,
+      "grad_norm": 1.9785017458341345,
+      "learning_rate": 0.00014866754038983716,
+      "loss": 1.4486,
+      "step": 7773
+    },
+    {
+      "epoch": 0.3581333210485097,
+      "grad_norm": 2.4303633208534445,
+      "learning_rate": 0.00014865450521710512,
+      "loss": 1.2578,
+      "step": 7774
+    },
+    {
+      "epoch": 0.3581793891371447,
+      "grad_norm": 4.225361609964411,
+      "learning_rate": 0.0001486414689611639,
+      "loss": 1.7866,
+      "step": 7775
+    },
+    {
+      "epoch": 0.3582254572257797,
+      "grad_norm": 4.274559909289046,
+      "learning_rate": 0.00014862843162230383,
+      "loss": 1.3033,
+      "step": 7776
+    },
+    {
+      "epoch": 0.3582715253144147,
+      "grad_norm": 1.600835544384673,
+      "learning_rate": 0.0001486153932008151,
+      "loss": 1.6151,
+      "step": 7777
+    },
+    {
+      "epoch": 0.3583175934030497,
+      "grad_norm": 1.2502079856769979,
+      "learning_rate": 0.00014860235369698797,
+      "loss": 1.5097,
+      "step": 7778
+    },
+    {
+      "epoch": 0.3583636614916847,
+      "grad_norm": 2.970596292078744,
+      "learning_rate": 0.00014858931311111277,
+      "loss": 1.2991,
+      "step": 7779
+    },
+    {
+      "epoch": 0.3584097295803197,
+      "grad_norm": 4.749529708095376,
+      "learning_rate": 0.00014857627144347984,
+      "loss": 1.3237,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3584557976689547,
+      "grad_norm": 2.0679701543501396,
+      "learning_rate": 0.00014856322869437954,
+      "loss": 1.4199,
+      "step": 7781
+    },
+    {
+      "epoch": 0.35850186575758974,
+      "grad_norm": 10.355183359078042,
+      "learning_rate": 0.0001485501848641022,
+      "loss": 1.271,
+      "step": 7782
+    },
+    {
+      "epoch": 0.35854793384622474,
+      "grad_norm": 2.276050929052542,
+      "learning_rate": 0.00014853713995293823,
+      "loss": 1.4562,
+      "step": 7783
+    },
+    {
+      "epoch": 0.35859400193485974,
+      "grad_norm": 1.584577131505778,
+      "learning_rate": 0.0001485240939611781,
+      "loss": 1.3843,
+      "step": 7784
+    },
+    {
+      "epoch": 0.35864007002349474,
+      "grad_norm": 1.1641790255208486,
+      "learning_rate": 0.00014851104688911218,
+      "loss": 1.3336,
+      "step": 7785
+    },
+    {
+      "epoch": 0.35868613811212974,
+      "grad_norm": 1.9253245453701517,
+      "learning_rate": 0.000148497998737031,
+      "loss": 1.2657,
+      "step": 7786
+    },
+    {
+      "epoch": 0.35873220620076474,
+      "grad_norm": 1.8395643095654344,
+      "learning_rate": 0.000148484949505225,
+      "loss": 1.3468,
+      "step": 7787
+    },
+    {
+      "epoch": 0.35877827428939973,
+      "grad_norm": 1.8696800827823543,
+      "learning_rate": 0.00014847189919398477,
+      "loss": 1.5127,
+      "step": 7788
+    },
+    {
+      "epoch": 0.35882434237803473,
+      "grad_norm": 4.2973653759682575,
+      "learning_rate": 0.0001484588478036008,
+      "loss": 1.1639,
+      "step": 7789
+    },
+    {
+      "epoch": 0.35887041046666973,
+      "grad_norm": 3.50488584871783,
+      "learning_rate": 0.00014844579533436364,
+      "loss": 1.4632,
+      "step": 7790
+    },
+    {
+      "epoch": 0.35891647855530473,
+      "grad_norm": 2.063524428149067,
+      "learning_rate": 0.000148432741786564,
+      "loss": 1.5688,
+      "step": 7791
+    },
+    {
+      "epoch": 0.35896254664393973,
+      "grad_norm": 4.160474867482992,
+      "learning_rate": 0.00014841968716049233,
+      "loss": 1.2753,
+      "step": 7792
+    },
+    {
+      "epoch": 0.3590086147325747,
+      "grad_norm": 3.8772913723906504,
+      "learning_rate": 0.00014840663145643937,
+      "loss": 1.5948,
+      "step": 7793
+    },
+    {
+      "epoch": 0.3590546828212097,
+      "grad_norm": 1.9125868232016117,
+      "learning_rate": 0.0001483935746746958,
+      "loss": 1.3918,
+      "step": 7794
+    },
+    {
+      "epoch": 0.3591007509098447,
+      "grad_norm": 2.4502098155293868,
+      "learning_rate": 0.00014838051681555223,
+      "loss": 1.4003,
+      "step": 7795
+    },
+    {
+      "epoch": 0.3591468189984798,
+      "grad_norm": 2.6732507418881424,
+      "learning_rate": 0.0001483674578792994,
+      "loss": 1.466,
+      "step": 7796
+    },
+    {
+      "epoch": 0.3591928870871148,
+      "grad_norm": 1.7678169803802877,
+      "learning_rate": 0.00014835439786622806,
+      "loss": 1.491,
+      "step": 7797
+    },
+    {
+      "epoch": 0.3592389551757498,
+      "grad_norm": 39.32150507999535,
+      "learning_rate": 0.00014834133677662896,
+      "loss": 1.4349,
+      "step": 7798
+    },
+    {
+      "epoch": 0.3592850232643848,
+      "grad_norm": 1.1955985643743745,
+      "learning_rate": 0.0001483282746107929,
+      "loss": 1.605,
+      "step": 7799
+    },
+    {
+      "epoch": 0.35933109135301977,
+      "grad_norm": 5.361519445870314,
+      "learning_rate": 0.00014831521136901067,
+      "loss": 1.4413,
+      "step": 7800
+    },
+    {
+      "epoch": 0.35937715944165477,
+      "grad_norm": 2.166693128015606,
+      "learning_rate": 0.00014830214705157307,
+      "loss": 1.7383,
+      "step": 7801
+    },
+    {
+      "epoch": 0.35942322753028977,
+      "grad_norm": 9.559894461901525,
+      "learning_rate": 0.000148289081658771,
+      "loss": 1.7206,
+      "step": 7802
+    },
+    {
+      "epoch": 0.35946929561892477,
+      "grad_norm": 2.4805627944018123,
+      "learning_rate": 0.00014827601519089534,
+      "loss": 1.6096,
+      "step": 7803
+    },
+    {
+      "epoch": 0.35951536370755977,
+      "grad_norm": 2.8629348399986596,
+      "learning_rate": 0.00014826294764823698,
+      "loss": 1.6745,
+      "step": 7804
+    },
+    {
+      "epoch": 0.35956143179619476,
+      "grad_norm": 1.8015356722155664,
+      "learning_rate": 0.00014824987903108684,
+      "loss": 1.0164,
+      "step": 7805
+    },
+    {
+      "epoch": 0.35960749988482976,
+      "grad_norm": 4.36327597201529,
+      "learning_rate": 0.00014823680933973588,
+      "loss": 1.8077,
+      "step": 7806
+    },
+    {
+      "epoch": 0.35965356797346476,
+      "grad_norm": 1.820642813440044,
+      "learning_rate": 0.00014822373857447505,
+      "loss": 1.3146,
+      "step": 7807
+    },
+    {
+      "epoch": 0.35969963606209976,
+      "grad_norm": 2.7573438356015263,
+      "learning_rate": 0.00014821066673559536,
+      "loss": 1.5127,
+      "step": 7808
+    },
+    {
+      "epoch": 0.3597457041507348,
+      "grad_norm": 4.0271736616578675,
+      "learning_rate": 0.00014819759382338785,
+      "loss": 1.3929,
+      "step": 7809
+    },
+    {
+      "epoch": 0.3597917722393698,
+      "grad_norm": 3.086325401533296,
+      "learning_rate": 0.00014818451983814358,
+      "loss": 1.3028,
+      "step": 7810
+    },
+    {
+      "epoch": 0.3598378403280048,
+      "grad_norm": 2.368392273445397,
+      "learning_rate": 0.00014817144478015355,
+      "loss": 1.2971,
+      "step": 7811
+    },
+    {
+      "epoch": 0.3598839084166398,
+      "grad_norm": 2.6505931280907147,
+      "learning_rate": 0.00014815836864970894,
+      "loss": 1.2977,
+      "step": 7812
+    },
+    {
+      "epoch": 0.3599299765052748,
+      "grad_norm": 3.6418451657337045,
+      "learning_rate": 0.0001481452914471008,
+      "loss": 1.7765,
+      "step": 7813
+    },
+    {
+      "epoch": 0.3599760445939098,
+      "grad_norm": 2.1629671360824063,
+      "learning_rate": 0.0001481322131726203,
+      "loss": 1.3,
+      "step": 7814
+    },
+    {
+      "epoch": 0.3600221126825448,
+      "grad_norm": 1.8395972822643791,
+      "learning_rate": 0.0001481191338265586,
+      "loss": 1.1392,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3600681807711798,
+      "grad_norm": 2.4364718264301,
+      "learning_rate": 0.0001481060534092069,
+      "loss": 1.618,
+      "step": 7816
+    },
+    {
+      "epoch": 0.3601142488598148,
+      "grad_norm": 2.764263478902608,
+      "learning_rate": 0.0001480929719208564,
+      "loss": 1.2127,
+      "step": 7817
+    },
+    {
+      "epoch": 0.3601603169484498,
+      "grad_norm": 3.794340951566421,
+      "learning_rate": 0.00014807988936179833,
+      "loss": 1.487,
+      "step": 7818
+    },
+    {
+      "epoch": 0.3602063850370848,
+      "grad_norm": 3.593095187539386,
+      "learning_rate": 0.00014806680573232402,
+      "loss": 1.5258,
+      "step": 7819
+    },
+    {
+      "epoch": 0.3602524531257198,
+      "grad_norm": 1.7450234514934004,
+      "learning_rate": 0.00014805372103272464,
+      "loss": 1.4095,
+      "step": 7820
+    },
+    {
+      "epoch": 0.3602985212143548,
+      "grad_norm": 1.847028241466368,
+      "learning_rate": 0.00014804063526329157,
+      "loss": 1.6415,
+      "step": 7821
+    },
+    {
+      "epoch": 0.3603445893029898,
+      "grad_norm": 1.4812078508929052,
+      "learning_rate": 0.00014802754842431612,
+      "loss": 1.365,
+      "step": 7822
+    },
+    {
+      "epoch": 0.36039065739162485,
+      "grad_norm": 3.1690751425160113,
+      "learning_rate": 0.00014801446051608967,
+      "loss": 1.5126,
+      "step": 7823
+    },
+    {
+      "epoch": 0.36043672548025985,
+      "grad_norm": 1.7137936833092635,
+      "learning_rate": 0.00014800137153890357,
+      "loss": 1.1943,
+      "step": 7824
+    },
+    {
+      "epoch": 0.36048279356889484,
+      "grad_norm": 1.7002138291893718,
+      "learning_rate": 0.00014798828149304924,
+      "loss": 1.3391,
+      "step": 7825
+    },
+    {
+      "epoch": 0.36052886165752984,
+      "grad_norm": 2.1455503662443176,
+      "learning_rate": 0.0001479751903788181,
+      "loss": 1.4893,
+      "step": 7826
+    },
+    {
+      "epoch": 0.36057492974616484,
+      "grad_norm": 2.6731688024109594,
+      "learning_rate": 0.00014796209819650162,
+      "loss": 1.4317,
+      "step": 7827
+    },
+    {
+      "epoch": 0.36062099783479984,
+      "grad_norm": 2.2471108356138845,
+      "learning_rate": 0.00014794900494639125,
+      "loss": 1.5471,
+      "step": 7828
+    },
+    {
+      "epoch": 0.36066706592343484,
+      "grad_norm": 2.777473788703433,
+      "learning_rate": 0.0001479359106287785,
+      "loss": 1.5455,
+      "step": 7829
+    },
+    {
+      "epoch": 0.36071313401206984,
+      "grad_norm": 6.440548186690372,
+      "learning_rate": 0.0001479228152439549,
+      "loss": 1.6265,
+      "step": 7830
+    },
+    {
+      "epoch": 0.36075920210070483,
+      "grad_norm": 1.2313235928503203,
+      "learning_rate": 0.000147909718792212,
+      "loss": 1.4356,
+      "step": 7831
+    },
+    {
+      "epoch": 0.36080527018933983,
+      "grad_norm": 1.4652969856205278,
+      "learning_rate": 0.00014789662127384133,
+      "loss": 1.4761,
+      "step": 7832
+    },
+    {
+      "epoch": 0.36085133827797483,
+      "grad_norm": 2.008189110002522,
+      "learning_rate": 0.00014788352268913454,
+      "loss": 1.5254,
+      "step": 7833
+    },
+    {
+      "epoch": 0.36089740636660983,
+      "grad_norm": 2.083203101574905,
+      "learning_rate": 0.00014787042303838322,
+      "loss": 1.5207,
+      "step": 7834
+    },
+    {
+      "epoch": 0.36094347445524483,
+      "grad_norm": 3.4477609503746085,
+      "learning_rate": 0.00014785732232187898,
+      "loss": 1.7421,
+      "step": 7835
+    },
+    {
+      "epoch": 0.3609895425438799,
+      "grad_norm": 2.827388075398616,
+      "learning_rate": 0.00014784422053991354,
+      "loss": 1.5762,
+      "step": 7836
+    },
+    {
+      "epoch": 0.3610356106325149,
+      "grad_norm": 2.211251781841849,
+      "learning_rate": 0.00014783111769277855,
+      "loss": 1.0819,
+      "step": 7837
+    },
+    {
+      "epoch": 0.3610816787211499,
+      "grad_norm": 2.2827453674383142,
+      "learning_rate": 0.00014781801378076575,
+      "loss": 1.3343,
+      "step": 7838
+    },
+    {
+      "epoch": 0.3611277468097849,
+      "grad_norm": 1.6862376463178743,
+      "learning_rate": 0.00014780490880416687,
+      "loss": 1.4118,
+      "step": 7839
+    },
+    {
+      "epoch": 0.3611738148984199,
+      "grad_norm": 4.731088356814096,
+      "learning_rate": 0.00014779180276327367,
+      "loss": 1.3544,
+      "step": 7840
+    },
+    {
+      "epoch": 0.3612198829870549,
+      "grad_norm": 1.6863959127248678,
+      "learning_rate": 0.0001477786956583779,
+      "loss": 1.6921,
+      "step": 7841
+    },
+    {
+      "epoch": 0.3612659510756899,
+      "grad_norm": 4.006889269055493,
+      "learning_rate": 0.0001477655874897714,
+      "loss": 1.5485,
+      "step": 7842
+    },
+    {
+      "epoch": 0.36131201916432487,
+      "grad_norm": 1.3105330584515926,
+      "learning_rate": 0.000147752478257746,
+      "loss": 1.5266,
+      "step": 7843
+    },
+    {
+      "epoch": 0.36135808725295987,
+      "grad_norm": 1.6270844041298649,
+      "learning_rate": 0.00014773936796259356,
+      "loss": 1.4559,
+      "step": 7844
+    },
+    {
+      "epoch": 0.36140415534159487,
+      "grad_norm": 6.607021768774998,
+      "learning_rate": 0.0001477262566046059,
+      "loss": 1.4232,
+      "step": 7845
+    },
+    {
+      "epoch": 0.36145022343022987,
+      "grad_norm": 1.9123966930637666,
+      "learning_rate": 0.00014771314418407503,
+      "loss": 1.5808,
+      "step": 7846
+    },
+    {
+      "epoch": 0.36149629151886487,
+      "grad_norm": 2.4945531722757175,
+      "learning_rate": 0.00014770003070129278,
+      "loss": 1.6808,
+      "step": 7847
+    },
+    {
+      "epoch": 0.36154235960749986,
+      "grad_norm": 2.5141129075211186,
+      "learning_rate": 0.00014768691615655116,
+      "loss": 1.5731,
+      "step": 7848
+    },
+    {
+      "epoch": 0.36158842769613486,
+      "grad_norm": 1.9033859122637087,
+      "learning_rate": 0.0001476738005501421,
+      "loss": 1.5767,
+      "step": 7849
+    },
+    {
+      "epoch": 0.3616344957847699,
+      "grad_norm": 2.968371341928274,
+      "learning_rate": 0.00014766068388235762,
+      "loss": 1.6977,
+      "step": 7850
+    },
+    {
+      "epoch": 0.3616805638734049,
+      "grad_norm": 1.9871140143998696,
+      "learning_rate": 0.00014764756615348975,
+      "loss": 1.3354,
+      "step": 7851
+    },
+    {
+      "epoch": 0.3617266319620399,
+      "grad_norm": 4.017303364390912,
+      "learning_rate": 0.0001476344473638305,
+      "loss": 1.4052,
+      "step": 7852
+    },
+    {
+      "epoch": 0.3617727000506749,
+      "grad_norm": 3.8543062410005295,
+      "learning_rate": 0.00014762132751367193,
+      "loss": 1.2687,
+      "step": 7853
+    },
+    {
+      "epoch": 0.3618187681393099,
+      "grad_norm": 1.8412792503330087,
+      "learning_rate": 0.0001476082066033062,
+      "loss": 1.5021,
+      "step": 7854
+    },
+    {
+      "epoch": 0.3618648362279449,
+      "grad_norm": 1.0485683615700196,
+      "learning_rate": 0.00014759508463302535,
+      "loss": 1.6683,
+      "step": 7855
+    },
+    {
+      "epoch": 0.3619109043165799,
+      "grad_norm": 1.3734700388356775,
+      "learning_rate": 0.00014758196160312157,
+      "loss": 1.6172,
+      "step": 7856
+    },
+    {
+      "epoch": 0.3619569724052149,
+      "grad_norm": 1.6264164291156111,
+      "learning_rate": 0.00014756883751388696,
+      "loss": 1.5933,
+      "step": 7857
+    },
+    {
+      "epoch": 0.3620030404938499,
+      "grad_norm": 1.9702793164389536,
+      "learning_rate": 0.0001475557123656138,
+      "loss": 1.5096,
+      "step": 7858
+    },
+    {
+      "epoch": 0.3620491085824849,
+      "grad_norm": 2.7935868690194146,
+      "learning_rate": 0.00014754258615859423,
+      "loss": 1.6348,
+      "step": 7859
+    },
+    {
+      "epoch": 0.3620951766711199,
+      "grad_norm": 1.5485225301986298,
+      "learning_rate": 0.00014752945889312047,
+      "loss": 1.6208,
+      "step": 7860
+    },
+    {
+      "epoch": 0.3621412447597549,
+      "grad_norm": 2.7305061954296823,
+      "learning_rate": 0.00014751633056948483,
+      "loss": 1.4018,
+      "step": 7861
+    },
+    {
+      "epoch": 0.3621873128483899,
+      "grad_norm": 1.8396188720369673,
+      "learning_rate": 0.00014750320118797955,
+      "loss": 1.2912,
+      "step": 7862
+    },
+    {
+      "epoch": 0.3622333809370249,
+      "grad_norm": 1.7496336437117672,
+      "learning_rate": 0.00014749007074889696,
+      "loss": 1.6149,
+      "step": 7863
+    },
+    {
+      "epoch": 0.36227944902565995,
+      "grad_norm": 8.96611070135353,
+      "learning_rate": 0.00014747693925252933,
+      "loss": 1.5216,
+      "step": 7864
+    },
+    {
+      "epoch": 0.36232551711429495,
+      "grad_norm": 3.063215933457071,
+      "learning_rate": 0.0001474638066991691,
+      "loss": 1.6101,
+      "step": 7865
+    },
+    {
+      "epoch": 0.36237158520292995,
+      "grad_norm": 14.664052824249222,
+      "learning_rate": 0.00014745067308910857,
+      "loss": 1.1734,
+      "step": 7866
+    },
+    {
+      "epoch": 0.36241765329156495,
+      "grad_norm": 1.719555483937719,
+      "learning_rate": 0.00014743753842264018,
+      "loss": 1.2488,
+      "step": 7867
+    },
+    {
+      "epoch": 0.36246372138019994,
+      "grad_norm": 3.289323575333369,
+      "learning_rate": 0.00014742440270005636,
+      "loss": 1.3031,
+      "step": 7868
+    },
+    {
+      "epoch": 0.36250978946883494,
+      "grad_norm": 1.889862858023557,
+      "learning_rate": 0.0001474112659216495,
+      "loss": 1.4222,
+      "step": 7869
+    },
+    {
+      "epoch": 0.36255585755746994,
+      "grad_norm": 1.5668901967336664,
+      "learning_rate": 0.0001473981280877121,
+      "loss": 1.3143,
+      "step": 7870
+    },
+    {
+      "epoch": 0.36260192564610494,
+      "grad_norm": 3.63462863223449,
+      "learning_rate": 0.00014738498919853664,
+      "loss": 1.5746,
+      "step": 7871
+    },
+    {
+      "epoch": 0.36264799373473994,
+      "grad_norm": 18.598976515245617,
+      "learning_rate": 0.00014737184925441566,
+      "loss": 1.7171,
+      "step": 7872
+    },
+    {
+      "epoch": 0.36269406182337494,
+      "grad_norm": 2.405101373276524,
+      "learning_rate": 0.00014735870825564163,
+      "loss": 1.8493,
+      "step": 7873
+    },
+    {
+      "epoch": 0.36274012991200993,
+      "grad_norm": 7.871728581632738,
+      "learning_rate": 0.00014734556620250722,
+      "loss": 1.5432,
+      "step": 7874
+    },
+    {
+      "epoch": 0.36278619800064493,
+      "grad_norm": 2.18832630956583,
+      "learning_rate": 0.00014733242309530496,
+      "loss": 1.6415,
+      "step": 7875
+    },
+    {
+      "epoch": 0.36283226608927993,
+      "grad_norm": 5.5402739088345445,
+      "learning_rate": 0.00014731927893432743,
+      "loss": 1.7226,
+      "step": 7876
+    },
+    {
+      "epoch": 0.362878334177915,
+      "grad_norm": 6.130246358864138,
+      "learning_rate": 0.0001473061337198673,
+      "loss": 1.6915,
+      "step": 7877
+    },
+    {
+      "epoch": 0.36292440226655,
+      "grad_norm": 2.8164951173044375,
+      "learning_rate": 0.00014729298745221723,
+      "loss": 1.5916,
+      "step": 7878
+    },
+    {
+      "epoch": 0.362970470355185,
+      "grad_norm": 6.325898158150294,
+      "learning_rate": 0.00014727984013166988,
+      "loss": 1.5451,
+      "step": 7879
+    },
+    {
+      "epoch": 0.36301653844382,
+      "grad_norm": 8.94515224758415,
+      "learning_rate": 0.00014726669175851794,
+      "loss": 1.7109,
+      "step": 7880
+    },
+    {
+      "epoch": 0.363062606532455,
+      "grad_norm": 2.5324057395603323,
+      "learning_rate": 0.00014725354233305414,
+      "loss": 1.3868,
+      "step": 7881
+    },
+    {
+      "epoch": 0.36310867462109,
+      "grad_norm": 2.9000652876966684,
+      "learning_rate": 0.00014724039185557125,
+      "loss": 1.6244,
+      "step": 7882
+    },
+    {
+      "epoch": 0.363154742709725,
+      "grad_norm": 1.93027903352279,
+      "learning_rate": 0.00014722724032636205,
+      "loss": 1.7048,
+      "step": 7883
+    },
+    {
+      "epoch": 0.36320081079836,
+      "grad_norm": 3.7922475410463234,
+      "learning_rate": 0.00014721408774571933,
+      "loss": 1.6558,
+      "step": 7884
+    },
+    {
+      "epoch": 0.363246878886995,
+      "grad_norm": 2.5581008251662354,
+      "learning_rate": 0.0001472009341139359,
+      "loss": 1.8328,
+      "step": 7885
+    },
+    {
+      "epoch": 0.36329294697562997,
+      "grad_norm": 2.3750636939386984,
+      "learning_rate": 0.00014718777943130459,
+      "loss": 1.1892,
+      "step": 7886
+    },
+    {
+      "epoch": 0.36333901506426497,
+      "grad_norm": 3.2848464670405444,
+      "learning_rate": 0.00014717462369811832,
+      "loss": 1.2189,
+      "step": 7887
+    },
+    {
+      "epoch": 0.36338508315289997,
+      "grad_norm": 4.889727761068161,
+      "learning_rate": 0.0001471614669146699,
+      "loss": 1.4668,
+      "step": 7888
+    },
+    {
+      "epoch": 0.36343115124153497,
+      "grad_norm": 1.170147772834098,
+      "learning_rate": 0.00014714830908125226,
+      "loss": 1.4786,
+      "step": 7889
+    },
+    {
+      "epoch": 0.36347721933016996,
+      "grad_norm": 2.0934595506636278,
+      "learning_rate": 0.0001471351501981584,
+      "loss": 1.348,
+      "step": 7890
+    },
+    {
+      "epoch": 0.363523287418805,
+      "grad_norm": 4.882485747158777,
+      "learning_rate": 0.0001471219902656812,
+      "loss": 1.2367,
+      "step": 7891
+    },
+    {
+      "epoch": 0.36356935550744,
+      "grad_norm": 2.299211779699829,
+      "learning_rate": 0.00014710882928411373,
+      "loss": 1.6056,
+      "step": 7892
+    },
+    {
+      "epoch": 0.363615423596075,
+      "grad_norm": 4.46506584554107,
+      "learning_rate": 0.00014709566725374894,
+      "loss": 1.4841,
+      "step": 7893
+    },
+    {
+      "epoch": 0.36366149168471,
+      "grad_norm": 2.098100015710836,
+      "learning_rate": 0.00014708250417487986,
+      "loss": 1.43,
+      "step": 7894
+    },
+    {
+      "epoch": 0.363707559773345,
+      "grad_norm": 1.4861552684604542,
+      "learning_rate": 0.00014706934004779957,
+      "loss": 1.5739,
+      "step": 7895
+    },
+    {
+      "epoch": 0.36375362786198,
+      "grad_norm": 1.5605736537272554,
+      "learning_rate": 0.00014705617487280112,
+      "loss": 1.673,
+      "step": 7896
+    },
+    {
+      "epoch": 0.363799695950615,
+      "grad_norm": 3.3429975171251187,
+      "learning_rate": 0.00014704300865017763,
+      "loss": 1.612,
+      "step": 7897
+    },
+    {
+      "epoch": 0.36384576403925,
+      "grad_norm": 2.7107462958466666,
+      "learning_rate": 0.0001470298413802222,
+      "loss": 1.1527,
+      "step": 7898
+    },
+    {
+      "epoch": 0.363891832127885,
+      "grad_norm": 4.974942578097108,
+      "learning_rate": 0.00014701667306322798,
+      "loss": 1.2623,
+      "step": 7899
+    },
+    {
+      "epoch": 0.36393790021652,
+      "grad_norm": 3.4762800526300515,
+      "learning_rate": 0.0001470035036994882,
+      "loss": 1.6622,
+      "step": 7900
+    },
+    {
+      "epoch": 0.363983968305155,
+      "grad_norm": 1.000653985614761,
+      "learning_rate": 0.00014699033328929596,
+      "loss": 1.3596,
+      "step": 7901
+    },
+    {
+      "epoch": 0.36403003639379,
+      "grad_norm": 2.684618470798968,
+      "learning_rate": 0.0001469771618329445,
+      "loss": 1.6175,
+      "step": 7902
+    },
+    {
+      "epoch": 0.364076104482425,
+      "grad_norm": 1.8434683545772155,
+      "learning_rate": 0.00014696398933072713,
+      "loss": 1.3328,
+      "step": 7903
+    },
+    {
+      "epoch": 0.36412217257106005,
+      "grad_norm": 3.0736816967197558,
+      "learning_rate": 0.00014695081578293705,
+      "loss": 1.1991,
+      "step": 7904
+    },
+    {
+      "epoch": 0.36416824065969505,
+      "grad_norm": 3.423690811586396,
+      "learning_rate": 0.00014693764118986754,
+      "loss": 1.6242,
+      "step": 7905
+    },
+    {
+      "epoch": 0.36421430874833005,
+      "grad_norm": 6.296901184988064,
+      "learning_rate": 0.00014692446555181194,
+      "loss": 1.485,
+      "step": 7906
+    },
+    {
+      "epoch": 0.36426037683696505,
+      "grad_norm": 4.003466376390033,
+      "learning_rate": 0.00014691128886906357,
+      "loss": 1.1323,
+      "step": 7907
+    },
+    {
+      "epoch": 0.36430644492560005,
+      "grad_norm": 1.5675837663358814,
+      "learning_rate": 0.00014689811114191578,
+      "loss": 1.3396,
+      "step": 7908
+    },
+    {
+      "epoch": 0.36435251301423505,
+      "grad_norm": 4.376502991027423,
+      "learning_rate": 0.000146884932370662,
+      "loss": 1.4503,
+      "step": 7909
+    },
+    {
+      "epoch": 0.36439858110287005,
+      "grad_norm": 1.474424504628677,
+      "learning_rate": 0.00014687175255559556,
+      "loss": 1.2239,
+      "step": 7910
+    },
+    {
+      "epoch": 0.36444464919150504,
+      "grad_norm": 3.0863977213152616,
+      "learning_rate": 0.00014685857169700992,
+      "loss": 1.523,
+      "step": 7911
+    },
+    {
+      "epoch": 0.36449071728014004,
+      "grad_norm": 3.2432084596559236,
+      "learning_rate": 0.0001468453897951985,
+      "loss": 1.4071,
+      "step": 7912
+    },
+    {
+      "epoch": 0.36453678536877504,
+      "grad_norm": 5.065209414557431,
+      "learning_rate": 0.00014683220685045485,
+      "loss": 1.5678,
+      "step": 7913
+    },
+    {
+      "epoch": 0.36458285345741004,
+      "grad_norm": 2.5534970857250054,
+      "learning_rate": 0.0001468190228630724,
+      "loss": 1.3694,
+      "step": 7914
+    },
+    {
+      "epoch": 0.36462892154604504,
+      "grad_norm": 2.9418011725414255,
+      "learning_rate": 0.00014680583783334466,
+      "loss": 1.5584,
+      "step": 7915
+    },
+    {
+      "epoch": 0.36467498963468004,
+      "grad_norm": 2.0737515374521305,
+      "learning_rate": 0.00014679265176156524,
+      "loss": 1.3435,
+      "step": 7916
+    },
+    {
+      "epoch": 0.36472105772331503,
+      "grad_norm": 1.9378646178387062,
+      "learning_rate": 0.0001467794646480276,
+      "loss": 1.3916,
+      "step": 7917
+    },
+    {
+      "epoch": 0.3647671258119501,
+      "grad_norm": 3.039562385979564,
+      "learning_rate": 0.00014676627649302544,
+      "loss": 1.4671,
+      "step": 7918
+    },
+    {
+      "epoch": 0.3648131939005851,
+      "grad_norm": 1.5626766133880499,
+      "learning_rate": 0.0001467530872968523,
+      "loss": 1.6075,
+      "step": 7919
+    },
+    {
+      "epoch": 0.3648592619892201,
+      "grad_norm": 1.9956222505486594,
+      "learning_rate": 0.00014673989705980183,
+      "loss": 1.5658,
+      "step": 7920
+    },
+    {
+      "epoch": 0.3649053300778551,
+      "grad_norm": 4.427861401360613,
+      "learning_rate": 0.00014672670578216767,
+      "loss": 1.3385,
+      "step": 7921
+    },
+    {
+      "epoch": 0.3649513981664901,
+      "grad_norm": 3.530298977224219,
+      "learning_rate": 0.0001467135134642436,
+      "loss": 1.6467,
+      "step": 7922
+    },
+    {
+      "epoch": 0.3649974662551251,
+      "grad_norm": 3.3999319676845006,
+      "learning_rate": 0.0001467003201063232,
+      "loss": 1.4909,
+      "step": 7923
+    },
+    {
+      "epoch": 0.3650435343437601,
+      "grad_norm": 1.5210212378660573,
+      "learning_rate": 0.00014668712570870023,
+      "loss": 1.5933,
+      "step": 7924
+    },
+    {
+      "epoch": 0.3650896024323951,
+      "grad_norm": 3.3565070401592987,
+      "learning_rate": 0.00014667393027166848,
+      "loss": 1.2474,
+      "step": 7925
+    },
+    {
+      "epoch": 0.3651356705210301,
+      "grad_norm": 2.3601731792219622,
+      "learning_rate": 0.0001466607337955217,
+      "loss": 1.8347,
+      "step": 7926
+    },
+    {
+      "epoch": 0.3651817386096651,
+      "grad_norm": 2.6149084560695464,
+      "learning_rate": 0.00014664753628055368,
+      "loss": 1.3483,
+      "step": 7927
+    },
+    {
+      "epoch": 0.3652278066983001,
+      "grad_norm": 1.6490817466748313,
+      "learning_rate": 0.00014663433772705827,
+      "loss": 1.4169,
+      "step": 7928
+    },
+    {
+      "epoch": 0.36527387478693507,
+      "grad_norm": 7.182977712896731,
+      "learning_rate": 0.00014662113813532926,
+      "loss": 1.4128,
+      "step": 7929
+    },
+    {
+      "epoch": 0.36531994287557007,
+      "grad_norm": 1.8564271771208343,
+      "learning_rate": 0.00014660793750566055,
+      "loss": 1.5881,
+      "step": 7930
+    },
+    {
+      "epoch": 0.36536601096420507,
+      "grad_norm": 2.8881478859627276,
+      "learning_rate": 0.00014659473583834607,
+      "loss": 1.6723,
+      "step": 7931
+    },
+    {
+      "epoch": 0.3654120790528401,
+      "grad_norm": 3.4397211348478276,
+      "learning_rate": 0.00014658153313367966,
+      "loss": 1.6158,
+      "step": 7932
+    },
+    {
+      "epoch": 0.3654581471414751,
+      "grad_norm": 6.47092508621192,
+      "learning_rate": 0.00014656832939195525,
+      "loss": 1.6338,
+      "step": 7933
+    },
+    {
+      "epoch": 0.3655042152301101,
+      "grad_norm": 2.3290670683086767,
+      "learning_rate": 0.0001465551246134669,
+      "loss": 1.263,
+      "step": 7934
+    },
+    {
+      "epoch": 0.3655502833187451,
+      "grad_norm": 1.7191267599682378,
+      "learning_rate": 0.00014654191879850847,
+      "loss": 1.4302,
+      "step": 7935
+    },
+    {
+      "epoch": 0.3655963514073801,
+      "grad_norm": 2.638936275427064,
+      "learning_rate": 0.00014652871194737406,
+      "loss": 1.6357,
+      "step": 7936
+    },
+    {
+      "epoch": 0.3656424194960151,
+      "grad_norm": 4.797847361575053,
+      "learning_rate": 0.00014651550406035764,
+      "loss": 1.6601,
+      "step": 7937
+    },
+    {
+      "epoch": 0.3656884875846501,
+      "grad_norm": 1.8934543771404468,
+      "learning_rate": 0.00014650229513775328,
+      "loss": 1.3837,
+      "step": 7938
+    },
+    {
+      "epoch": 0.3657345556732851,
+      "grad_norm": 2.3772852830478115,
+      "learning_rate": 0.00014648908517985506,
+      "loss": 1.4275,
+      "step": 7939
+    },
+    {
+      "epoch": 0.3657806237619201,
+      "grad_norm": 2.4353966803761913,
+      "learning_rate": 0.00014647587418695706,
+      "loss": 1.3736,
+      "step": 7940
+    },
+    {
+      "epoch": 0.3658266918505551,
+      "grad_norm": 2.29146597455642,
+      "learning_rate": 0.00014646266215935343,
+      "loss": 1.4379,
+      "step": 7941
+    },
+    {
+      "epoch": 0.3658727599391901,
+      "grad_norm": 3.946204596744561,
+      "learning_rate": 0.00014644944909733828,
+      "loss": 1.389,
+      "step": 7942
+    },
+    {
+      "epoch": 0.3659188280278251,
+      "grad_norm": 2.8332838329213645,
+      "learning_rate": 0.00014643623500120578,
+      "loss": 1.4227,
+      "step": 7943
+    },
+    {
+      "epoch": 0.3659648961164601,
+      "grad_norm": 17.21066056001472,
+      "learning_rate": 0.00014642301987125014,
+      "loss": 1.3038,
+      "step": 7944
+    },
+    {
+      "epoch": 0.36601096420509516,
+      "grad_norm": 5.45140424753129,
+      "learning_rate": 0.00014640980370776556,
+      "loss": 1.3357,
+      "step": 7945
+    },
+    {
+      "epoch": 0.36605703229373016,
+      "grad_norm": 2.2643316871138355,
+      "learning_rate": 0.00014639658651104624,
+      "loss": 1.406,
+      "step": 7946
+    },
+    {
+      "epoch": 0.36610310038236515,
+      "grad_norm": 2.9979455939216297,
+      "learning_rate": 0.00014638336828138652,
+      "loss": 1.5177,
+      "step": 7947
+    },
+    {
+      "epoch": 0.36614916847100015,
+      "grad_norm": 1.748580149236944,
+      "learning_rate": 0.00014637014901908063,
+      "loss": 1.67,
+      "step": 7948
+    },
+    {
+      "epoch": 0.36619523655963515,
+      "grad_norm": 1.3969522054663357,
+      "learning_rate": 0.00014635692872442288,
+      "loss": 1.4573,
+      "step": 7949
+    },
+    {
+      "epoch": 0.36624130464827015,
+      "grad_norm": 2.4349526385238827,
+      "learning_rate": 0.0001463437073977076,
+      "loss": 1.641,
+      "step": 7950
+    },
+    {
+      "epoch": 0.36628737273690515,
+      "grad_norm": 2.3236280575220056,
+      "learning_rate": 0.00014633048503922912,
+      "loss": 1.3766,
+      "step": 7951
+    },
+    {
+      "epoch": 0.36633344082554015,
+      "grad_norm": 1.9243382188965417,
+      "learning_rate": 0.00014631726164928183,
+      "loss": 1.3763,
+      "step": 7952
+    },
+    {
+      "epoch": 0.36637950891417514,
+      "grad_norm": 11.416364080040209,
+      "learning_rate": 0.00014630403722816016,
+      "loss": 1.4753,
+      "step": 7953
+    },
+    {
+      "epoch": 0.36642557700281014,
+      "grad_norm": 1.3092549305072139,
+      "learning_rate": 0.00014629081177615846,
+      "loss": 1.7,
+      "step": 7954
+    },
+    {
+      "epoch": 0.36647164509144514,
+      "grad_norm": 1.0873558517453377,
+      "learning_rate": 0.0001462775852935712,
+      "loss": 1.586,
+      "step": 7955
+    },
+    {
+      "epoch": 0.36651771318008014,
+      "grad_norm": 2.5838025464920644,
+      "learning_rate": 0.00014626435778069288,
+      "loss": 1.4509,
+      "step": 7956
+    },
+    {
+      "epoch": 0.36656378126871514,
+      "grad_norm": 1.360886209219802,
+      "learning_rate": 0.00014625112923781795,
+      "loss": 1.5981,
+      "step": 7957
+    },
+    {
+      "epoch": 0.36660984935735014,
+      "grad_norm": 1.9246389525167762,
+      "learning_rate": 0.00014623789966524092,
+      "loss": 1.5523,
+      "step": 7958
+    },
+    {
+      "epoch": 0.3666559174459852,
+      "grad_norm": 2.216723599547953,
+      "learning_rate": 0.00014622466906325635,
+      "loss": 1.3301,
+      "step": 7959
+    },
+    {
+      "epoch": 0.3667019855346202,
+      "grad_norm": 1.5620257468689718,
+      "learning_rate": 0.0001462114374321588,
+      "loss": 1.5897,
+      "step": 7960
+    },
+    {
+      "epoch": 0.3667480536232552,
+      "grad_norm": 1.8677887526988712,
+      "learning_rate": 0.0001461982047722428,
+      "loss": 1.5625,
+      "step": 7961
+    },
+    {
+      "epoch": 0.3667941217118902,
+      "grad_norm": 1.234714718068745,
+      "learning_rate": 0.000146184971083803,
+      "loss": 1.6159,
+      "step": 7962
+    },
+    {
+      "epoch": 0.3668401898005252,
+      "grad_norm": 1.2148798297267367,
+      "learning_rate": 0.000146171736367134,
+      "loss": 1.4539,
+      "step": 7963
+    },
+    {
+      "epoch": 0.3668862578891602,
+      "grad_norm": 3.1381793152549995,
+      "learning_rate": 0.00014615850062253048,
+      "loss": 1.2243,
+      "step": 7964
+    },
+    {
+      "epoch": 0.3669323259777952,
+      "grad_norm": 1.1285967980982596,
+      "learning_rate": 0.00014614526385028703,
+      "loss": 1.3175,
+      "step": 7965
+    },
+    {
+      "epoch": 0.3669783940664302,
+      "grad_norm": 1.465514191634424,
+      "learning_rate": 0.00014613202605069848,
+      "loss": 1.5972,
+      "step": 7966
+    },
+    {
+      "epoch": 0.3670244621550652,
+      "grad_norm": 1.3366907094836964,
+      "learning_rate": 0.00014611878722405942,
+      "loss": 1.5961,
+      "step": 7967
+    },
+    {
+      "epoch": 0.3670705302437002,
+      "grad_norm": 4.067099706102899,
+      "learning_rate": 0.00014610554737066463,
+      "loss": 1.4467,
+      "step": 7968
+    },
+    {
+      "epoch": 0.3671165983323352,
+      "grad_norm": 1.8577676641545997,
+      "learning_rate": 0.00014609230649080893,
+      "loss": 1.5724,
+      "step": 7969
+    },
+    {
+      "epoch": 0.3671626664209702,
+      "grad_norm": 1.412480107860917,
+      "learning_rate": 0.00014607906458478703,
+      "loss": 1.6997,
+      "step": 7970
+    },
+    {
+      "epoch": 0.3672087345096052,
+      "grad_norm": 1.5394902522754168,
+      "learning_rate": 0.00014606582165289375,
+      "loss": 1.7132,
+      "step": 7971
+    },
+    {
+      "epoch": 0.3672548025982402,
+      "grad_norm": 1.8832140226408596,
+      "learning_rate": 0.00014605257769542392,
+      "loss": 1.5948,
+      "step": 7972
+    },
+    {
+      "epoch": 0.3673008706868752,
+      "grad_norm": 2.531681618584868,
+      "learning_rate": 0.00014603933271267248,
+      "loss": 1.2007,
+      "step": 7973
+    },
+    {
+      "epoch": 0.3673469387755102,
+      "grad_norm": 2.008674003536504,
+      "learning_rate": 0.0001460260867049342,
+      "loss": 1.319,
+      "step": 7974
+    },
+    {
+      "epoch": 0.3673930068641452,
+      "grad_norm": 1.7936743168477323,
+      "learning_rate": 0.000146012839672504,
+      "loss": 1.2233,
+      "step": 7975
+    },
+    {
+      "epoch": 0.3674390749527802,
+      "grad_norm": 2.168382936528419,
+      "learning_rate": 0.0001459995916156768,
+      "loss": 1.3873,
+      "step": 7976
+    },
+    {
+      "epoch": 0.3674851430414152,
+      "grad_norm": 2.0603350065193684,
+      "learning_rate": 0.00014598634253474763,
+      "loss": 1.5796,
+      "step": 7977
+    },
+    {
+      "epoch": 0.3675312111300502,
+      "grad_norm": 1.639549827323257,
+      "learning_rate": 0.00014597309243001138,
+      "loss": 1.3737,
+      "step": 7978
+    },
+    {
+      "epoch": 0.3675772792186852,
+      "grad_norm": 1.5864020304850965,
+      "learning_rate": 0.00014595984130176304,
+      "loss": 1.6271,
+      "step": 7979
+    },
+    {
+      "epoch": 0.3676233473073202,
+      "grad_norm": 3.354826735873579,
+      "learning_rate": 0.00014594658915029765,
+      "loss": 1.3738,
+      "step": 7980
+    },
+    {
+      "epoch": 0.3676694153959552,
+      "grad_norm": 2.788572575590543,
+      "learning_rate": 0.00014593333597591018,
+      "loss": 1.6148,
+      "step": 7981
+    },
+    {
+      "epoch": 0.3677154834845902,
+      "grad_norm": 1.6589146364044665,
+      "learning_rate": 0.0001459200817788958,
+      "loss": 1.537,
+      "step": 7982
+    },
+    {
+      "epoch": 0.3677615515732252,
+      "grad_norm": 2.614192834088506,
+      "learning_rate": 0.00014590682655954952,
+      "loss": 1.0954,
+      "step": 7983
+    },
+    {
+      "epoch": 0.3678076196618602,
+      "grad_norm": 3.991402389065211,
+      "learning_rate": 0.00014589357031816647,
+      "loss": 1.6091,
+      "step": 7984
+    },
+    {
+      "epoch": 0.3678536877504952,
+      "grad_norm": 1.67242202073557,
+      "learning_rate": 0.00014588031305504172,
+      "loss": 1.4986,
+      "step": 7985
+    },
+    {
+      "epoch": 0.36789975583913026,
+      "grad_norm": 2.4205710357321073,
+      "learning_rate": 0.00014586705477047054,
+      "loss": 1.5914,
+      "step": 7986
+    },
+    {
+      "epoch": 0.36794582392776526,
+      "grad_norm": 2.6146794280269248,
+      "learning_rate": 0.000145853795464748,
+      "loss": 1.6761,
+      "step": 7987
+    },
+    {
+      "epoch": 0.36799189201640026,
+      "grad_norm": 4.737184367724566,
+      "learning_rate": 0.00014584053513816932,
+      "loss": 1.5059,
+      "step": 7988
+    },
+    {
+      "epoch": 0.36803796010503526,
+      "grad_norm": 4.4194897526908505,
+      "learning_rate": 0.0001458272737910297,
+      "loss": 1.1557,
+      "step": 7989
+    },
+    {
+      "epoch": 0.36808402819367025,
+      "grad_norm": 25.66585119204001,
+      "learning_rate": 0.00014581401142362444,
+      "loss": 1.5701,
+      "step": 7990
+    },
+    {
+      "epoch": 0.36813009628230525,
+      "grad_norm": 2.9514645154383725,
+      "learning_rate": 0.00014580074803624877,
+      "loss": 1.5286,
+      "step": 7991
+    },
+    {
+      "epoch": 0.36817616437094025,
+      "grad_norm": 1.861501201668882,
+      "learning_rate": 0.00014578748362919797,
+      "loss": 1.5529,
+      "step": 7992
+    },
+    {
+      "epoch": 0.36822223245957525,
+      "grad_norm": 3.386325518182548,
+      "learning_rate": 0.00014577421820276734,
+      "loss": 1.4662,
+      "step": 7993
+    },
+    {
+      "epoch": 0.36826830054821025,
+      "grad_norm": 1.8896166337963776,
+      "learning_rate": 0.00014576095175725223,
+      "loss": 1.508,
+      "step": 7994
+    },
+    {
+      "epoch": 0.36831436863684525,
+      "grad_norm": 4.140892336885428,
+      "learning_rate": 0.00014574768429294802,
+      "loss": 1.737,
+      "step": 7995
+    },
+    {
+      "epoch": 0.36836043672548024,
+      "grad_norm": 3.3372397860564256,
+      "learning_rate": 0.00014573441581015005,
+      "loss": 1.4156,
+      "step": 7996
+    },
+    {
+      "epoch": 0.36840650481411524,
+      "grad_norm": 3.2050285837957664,
+      "learning_rate": 0.00014572114630915375,
+      "loss": 1.6234,
+      "step": 7997
+    },
+    {
+      "epoch": 0.36845257290275024,
+      "grad_norm": 4.174633850697286,
+      "learning_rate": 0.0001457078757902545,
+      "loss": 1.7154,
+      "step": 7998
+    },
+    {
+      "epoch": 0.36849864099138524,
+      "grad_norm": 1.9974848328581976,
+      "learning_rate": 0.00014569460425374774,
+      "loss": 1.4185,
+      "step": 7999
+    },
+    {
+      "epoch": 0.3685447090800203,
+      "grad_norm": 8.363771293735354,
+      "learning_rate": 0.000145681331699929,
+      "loss": 1.7732,
+      "step": 8000
+    },
+    {
+      "epoch": 0.3685907771686553,
+      "grad_norm": 3.6235184296812077,
+      "learning_rate": 0.00014566805812909372,
+      "loss": 1.6156,
+      "step": 8001
+    },
+    {
+      "epoch": 0.3686368452572903,
+      "grad_norm": 3.0375608497004984,
+      "learning_rate": 0.00014565478354153742,
+      "loss": 1.6922,
+      "step": 8002
+    },
+    {
+      "epoch": 0.3686829133459253,
+      "grad_norm": 1.6843506230318595,
+      "learning_rate": 0.00014564150793755566,
+      "loss": 1.4517,
+      "step": 8003
+    },
+    {
+      "epoch": 0.3687289814345603,
+      "grad_norm": 2.8433834223528724,
+      "learning_rate": 0.000145628231317444,
+      "loss": 1.4642,
+      "step": 8004
+    },
+    {
+      "epoch": 0.3687750495231953,
+      "grad_norm": 4.780956808363177,
+      "learning_rate": 0.000145614953681498,
+      "loss": 1.4702,
+      "step": 8005
+    },
+    {
+      "epoch": 0.3688211176118303,
+      "grad_norm": 1.6449264144092268,
+      "learning_rate": 0.00014560167503001328,
+      "loss": 1.7134,
+      "step": 8006
+    },
+    {
+      "epoch": 0.3688671857004653,
+      "grad_norm": 1.8915123790586048,
+      "learning_rate": 0.00014558839536328542,
+      "loss": 1.4472,
+      "step": 8007
+    },
+    {
+      "epoch": 0.3689132537891003,
+      "grad_norm": 4.213039655016178,
+      "learning_rate": 0.0001455751146816101,
+      "loss": 1.7597,
+      "step": 8008
+    },
+    {
+      "epoch": 0.3689593218777353,
+      "grad_norm": 1.7330090534574971,
+      "learning_rate": 0.00014556183298528304,
+      "loss": 1.4636,
+      "step": 8009
+    },
+    {
+      "epoch": 0.3690053899663703,
+      "grad_norm": 1.3708805875782832,
+      "learning_rate": 0.00014554855027459985,
+      "loss": 1.4347,
+      "step": 8010
+    },
+    {
+      "epoch": 0.3690514580550053,
+      "grad_norm": 4.7408477475623565,
+      "learning_rate": 0.0001455352665498563,
+      "loss": 1.7489,
+      "step": 8011
+    },
+    {
+      "epoch": 0.3690975261436403,
+      "grad_norm": 1.270286420838142,
+      "learning_rate": 0.00014552198181134814,
+      "loss": 1.5062,
+      "step": 8012
+    },
+    {
+      "epoch": 0.36914359423227533,
+      "grad_norm": 1.2816762454231152,
+      "learning_rate": 0.0001455086960593711,
+      "loss": 1.5912,
+      "step": 8013
+    },
+    {
+      "epoch": 0.36918966232091033,
+      "grad_norm": 1.1713576140176276,
+      "learning_rate": 0.00014549540929422095,
+      "loss": 1.1577,
+      "step": 8014
+    },
+    {
+      "epoch": 0.3692357304095453,
+      "grad_norm": 1.4121014893535575,
+      "learning_rate": 0.00014548212151619354,
+      "loss": 1.5721,
+      "step": 8015
+    },
+    {
+      "epoch": 0.3692817984981803,
+      "grad_norm": 1.85736646650291,
+      "learning_rate": 0.00014546883272558464,
+      "loss": 1.6627,
+      "step": 8016
+    },
+    {
+      "epoch": 0.3693278665868153,
+      "grad_norm": 4.9373737314897035,
+      "learning_rate": 0.00014545554292269019,
+      "loss": 1.55,
+      "step": 8017
+    },
+    {
+      "epoch": 0.3693739346754503,
+      "grad_norm": 2.54454860157983,
+      "learning_rate": 0.000145442252107806,
+      "loss": 1.5272,
+      "step": 8018
+    },
+    {
+      "epoch": 0.3694200027640853,
+      "grad_norm": 2.3880701164184277,
+      "learning_rate": 0.00014542896028122798,
+      "loss": 1.4342,
+      "step": 8019
+    },
+    {
+      "epoch": 0.3694660708527203,
+      "grad_norm": 6.198952461270359,
+      "learning_rate": 0.00014541566744325204,
+      "loss": 1.5792,
+      "step": 8020
+    },
+    {
+      "epoch": 0.3695121389413553,
+      "grad_norm": 10.451050891826378,
+      "learning_rate": 0.00014540237359417417,
+      "loss": 1.2808,
+      "step": 8021
+    },
+    {
+      "epoch": 0.3695582070299903,
+      "grad_norm": 4.5770883791476535,
+      "learning_rate": 0.00014538907873429027,
+      "loss": 1.279,
+      "step": 8022
+    },
+    {
+      "epoch": 0.3696042751186253,
+      "grad_norm": 2.8368007609910872,
+      "learning_rate": 0.00014537578286389637,
+      "loss": 1.4983,
+      "step": 8023
+    },
+    {
+      "epoch": 0.3696503432072603,
+      "grad_norm": 2.5063358298428122,
+      "learning_rate": 0.00014536248598328848,
+      "loss": 1.7274,
+      "step": 8024
+    },
+    {
+      "epoch": 0.3696964112958953,
+      "grad_norm": 1.996457433528675,
+      "learning_rate": 0.00014534918809276258,
+      "loss": 1.2861,
+      "step": 8025
+    },
+    {
+      "epoch": 0.3697424793845303,
+      "grad_norm": 8.195908576813203,
+      "learning_rate": 0.00014533588919261482,
+      "loss": 1.3809,
+      "step": 8026
+    },
+    {
+      "epoch": 0.36978854747316536,
+      "grad_norm": 3.8725515014140353,
+      "learning_rate": 0.00014532258928314117,
+      "loss": 1.3671,
+      "step": 8027
+    },
+    {
+      "epoch": 0.36983461556180036,
+      "grad_norm": 1.8463891622203557,
+      "learning_rate": 0.00014530928836463782,
+      "loss": 1.4612,
+      "step": 8028
+    },
+    {
+      "epoch": 0.36988068365043536,
+      "grad_norm": 7.976166292032256,
+      "learning_rate": 0.00014529598643740082,
+      "loss": 1.8839,
+      "step": 8029
+    },
+    {
+      "epoch": 0.36992675173907036,
+      "grad_norm": 1.5107615849478067,
+      "learning_rate": 0.00014528268350172638,
+      "loss": 1.6752,
+      "step": 8030
+    },
+    {
+      "epoch": 0.36997281982770536,
+      "grad_norm": 6.074301521501473,
+      "learning_rate": 0.0001452693795579106,
+      "loss": 1.3546,
+      "step": 8031
+    },
+    {
+      "epoch": 0.37001888791634036,
+      "grad_norm": 1.1409174119309486,
+      "learning_rate": 0.00014525607460624972,
+      "loss": 1.5964,
+      "step": 8032
+    },
+    {
+      "epoch": 0.37006495600497535,
+      "grad_norm": 1.5480037780798712,
+      "learning_rate": 0.00014524276864703997,
+      "loss": 1.4012,
+      "step": 8033
+    },
+    {
+      "epoch": 0.37011102409361035,
+      "grad_norm": 8.710541470892501,
+      "learning_rate": 0.0001452294616805775,
+      "loss": 1.3075,
+      "step": 8034
+    },
+    {
+      "epoch": 0.37015709218224535,
+      "grad_norm": 3.878313926632492,
+      "learning_rate": 0.00014521615370715867,
+      "loss": 1.4976,
+      "step": 8035
+    },
+    {
+      "epoch": 0.37020316027088035,
+      "grad_norm": 2.388947405123849,
+      "learning_rate": 0.00014520284472707966,
+      "loss": 1.5244,
+      "step": 8036
+    },
+    {
+      "epoch": 0.37024922835951535,
+      "grad_norm": 1.6775409226528502,
+      "learning_rate": 0.00014518953474063683,
+      "loss": 1.5826,
+      "step": 8037
+    },
+    {
+      "epoch": 0.37029529644815035,
+      "grad_norm": 3.4808815352864624,
+      "learning_rate": 0.0001451762237481265,
+      "loss": 1.4722,
+      "step": 8038
+    },
+    {
+      "epoch": 0.37034136453678534,
+      "grad_norm": 1.4654677230705566,
+      "learning_rate": 0.000145162911749845,
+      "loss": 1.6087,
+      "step": 8039
+    },
+    {
+      "epoch": 0.3703874326254204,
+      "grad_norm": 4.269924703577002,
+      "learning_rate": 0.00014514959874608872,
+      "loss": 1.3934,
+      "step": 8040
+    },
+    {
+      "epoch": 0.3704335007140554,
+      "grad_norm": 2.4747177354134036,
+      "learning_rate": 0.00014513628473715405,
+      "loss": 1.7049,
+      "step": 8041
+    },
+    {
+      "epoch": 0.3704795688026904,
+      "grad_norm": 1.2389575451761452,
+      "learning_rate": 0.00014512296972333735,
+      "loss": 1.2972,
+      "step": 8042
+    },
+    {
+      "epoch": 0.3705256368913254,
+      "grad_norm": 4.664558030074927,
+      "learning_rate": 0.0001451096537049351,
+      "loss": 1.5148,
+      "step": 8043
+    },
+    {
+      "epoch": 0.3705717049799604,
+      "grad_norm": 5.910762149790772,
+      "learning_rate": 0.0001450963366822438,
+      "loss": 1.5108,
+      "step": 8044
+    },
+    {
+      "epoch": 0.3706177730685954,
+      "grad_norm": 2.167791196586623,
+      "learning_rate": 0.00014508301865555988,
+      "loss": 1.4826,
+      "step": 8045
+    },
+    {
+      "epoch": 0.3706638411572304,
+      "grad_norm": 2.354335176184653,
+      "learning_rate": 0.00014506969962517981,
+      "loss": 1.4961,
+      "step": 8046
+    },
+    {
+      "epoch": 0.3707099092458654,
+      "grad_norm": 2.9047690791288403,
+      "learning_rate": 0.00014505637959140018,
+      "loss": 1.5967,
+      "step": 8047
+    },
+    {
+      "epoch": 0.3707559773345004,
+      "grad_norm": 2.974638033028197,
+      "learning_rate": 0.00014504305855451754,
+      "loss": 1.5026,
+      "step": 8048
+    },
+    {
+      "epoch": 0.3708020454231354,
+      "grad_norm": 1.6215913955086108,
+      "learning_rate": 0.0001450297365148284,
+      "loss": 1.4228,
+      "step": 8049
+    },
+    {
+      "epoch": 0.3708481135117704,
+      "grad_norm": 2.652089383534507,
+      "learning_rate": 0.0001450164134726294,
+      "loss": 1.488,
+      "step": 8050
+    },
+    {
+      "epoch": 0.3708941816004054,
+      "grad_norm": 2.5973998589652805,
+      "learning_rate": 0.00014500308942821713,
+      "loss": 1.7387,
+      "step": 8051
+    },
+    {
+      "epoch": 0.3709402496890404,
+      "grad_norm": 1.135258090755569,
+      "learning_rate": 0.00014498976438188826,
+      "loss": 1.7884,
+      "step": 8052
+    },
+    {
+      "epoch": 0.3709863177776754,
+      "grad_norm": 2.0499107603427,
+      "learning_rate": 0.0001449764383339394,
+      "loss": 1.3616,
+      "step": 8053
+    },
+    {
+      "epoch": 0.37103238586631043,
+      "grad_norm": 1.0157407108275698,
+      "learning_rate": 0.0001449631112846673,
+      "loss": 1.517,
+      "step": 8054
+    },
+    {
+      "epoch": 0.37107845395494543,
+      "grad_norm": 1.2168109515905279,
+      "learning_rate": 0.00014494978323436857,
+      "loss": 1.5609,
+      "step": 8055
+    },
+    {
+      "epoch": 0.37112452204358043,
+      "grad_norm": 1.5764661332627408,
+      "learning_rate": 0.00014493645418334002,
+      "loss": 1.6408,
+      "step": 8056
+    },
+    {
+      "epoch": 0.3711705901322154,
+      "grad_norm": 2.7819534494459854,
+      "learning_rate": 0.00014492312413187834,
+      "loss": 1.4445,
+      "step": 8057
+    },
+    {
+      "epoch": 0.3712166582208504,
+      "grad_norm": 1.4026831957840593,
+      "learning_rate": 0.00014490979308028035,
+      "loss": 1.514,
+      "step": 8058
+    },
+    {
+      "epoch": 0.3712627263094854,
+      "grad_norm": 1.213722651061903,
+      "learning_rate": 0.00014489646102884283,
+      "loss": 1.2907,
+      "step": 8059
+    },
+    {
+      "epoch": 0.3713087943981204,
+      "grad_norm": 2.5119016821564086,
+      "learning_rate": 0.00014488312797786258,
+      "loss": 1.6974,
+      "step": 8060
+    },
+    {
+      "epoch": 0.3713548624867554,
+      "grad_norm": 3.6323408732842855,
+      "learning_rate": 0.00014486979392763644,
+      "loss": 1.5363,
+      "step": 8061
+    },
+    {
+      "epoch": 0.3714009305753904,
+      "grad_norm": 2.144470780118938,
+      "learning_rate": 0.00014485645887846125,
+      "loss": 1.4787,
+      "step": 8062
+    },
+    {
+      "epoch": 0.3714469986640254,
+      "grad_norm": 5.230091493604424,
+      "learning_rate": 0.00014484312283063395,
+      "loss": 1.4296,
+      "step": 8063
+    },
+    {
+      "epoch": 0.3714930667526604,
+      "grad_norm": 1.5976037648574533,
+      "learning_rate": 0.0001448297857844514,
+      "loss": 1.4503,
+      "step": 8064
+    },
+    {
+      "epoch": 0.3715391348412954,
+      "grad_norm": 3.6586608155832043,
+      "learning_rate": 0.00014481644774021053,
+      "loss": 1.7591,
+      "step": 8065
+    },
+    {
+      "epoch": 0.3715852029299304,
+      "grad_norm": 1.8547935856567173,
+      "learning_rate": 0.00014480310869820827,
+      "loss": 1.7782,
+      "step": 8066
+    },
+    {
+      "epoch": 0.3716312710185654,
+      "grad_norm": 1.137814899449746,
+      "learning_rate": 0.00014478976865874165,
+      "loss": 1.7109,
+      "step": 8067
+    },
+    {
+      "epoch": 0.37167733910720047,
+      "grad_norm": 2.7065811409028098,
+      "learning_rate": 0.00014477642762210762,
+      "loss": 1.579,
+      "step": 8068
+    },
+    {
+      "epoch": 0.37172340719583546,
+      "grad_norm": 1.8117238113754774,
+      "learning_rate": 0.00014476308558860322,
+      "loss": 1.446,
+      "step": 8069
+    },
+    {
+      "epoch": 0.37176947528447046,
+      "grad_norm": 1.4892375099088144,
+      "learning_rate": 0.00014474974255852544,
+      "loss": 1.7295,
+      "step": 8070
+    },
+    {
+      "epoch": 0.37181554337310546,
+      "grad_norm": 3.021481155260866,
+      "learning_rate": 0.00014473639853217137,
+      "loss": 1.6935,
+      "step": 8071
+    },
+    {
+      "epoch": 0.37186161146174046,
+      "grad_norm": 1.50854482821164,
+      "learning_rate": 0.00014472305350983814,
+      "loss": 1.3676,
+      "step": 8072
+    },
+    {
+      "epoch": 0.37190767955037546,
+      "grad_norm": 2.40382883195836,
+      "learning_rate": 0.00014470970749182275,
+      "loss": 1.6233,
+      "step": 8073
+    },
+    {
+      "epoch": 0.37195374763901046,
+      "grad_norm": 1.915536586035009,
+      "learning_rate": 0.00014469636047842243,
+      "loss": 1.4082,
+      "step": 8074
+    },
+    {
+      "epoch": 0.37199981572764546,
+      "grad_norm": 1.7965207147968723,
+      "learning_rate": 0.00014468301246993426,
+      "loss": 1.3787,
+      "step": 8075
+    },
+    {
+      "epoch": 0.37204588381628045,
+      "grad_norm": 1.3997967254743582,
+      "learning_rate": 0.00014466966346665544,
+      "loss": 1.606,
+      "step": 8076
+    },
+    {
+      "epoch": 0.37209195190491545,
+      "grad_norm": 1.5521973156936018,
+      "learning_rate": 0.00014465631346888315,
+      "loss": 1.4249,
+      "step": 8077
+    },
+    {
+      "epoch": 0.37213801999355045,
+      "grad_norm": 1.281363974315087,
+      "learning_rate": 0.0001446429624769146,
+      "loss": 1.3399,
+      "step": 8078
+    },
+    {
+      "epoch": 0.37218408808218545,
+      "grad_norm": 1.0925320305934587,
+      "learning_rate": 0.00014462961049104705,
+      "loss": 1.6164,
+      "step": 8079
+    },
+    {
+      "epoch": 0.37223015617082045,
+      "grad_norm": 1.9441724665667157,
+      "learning_rate": 0.00014461625751157773,
+      "loss": 1.4182,
+      "step": 8080
+    },
+    {
+      "epoch": 0.3722762242594555,
+      "grad_norm": 1.7810904059894443,
+      "learning_rate": 0.00014460290353880397,
+      "loss": 1.1355,
+      "step": 8081
+    },
+    {
+      "epoch": 0.3723222923480905,
+      "grad_norm": 1.6652211253955331,
+      "learning_rate": 0.000144589548573023,
+      "loss": 1.4866,
+      "step": 8082
+    },
+    {
+      "epoch": 0.3723683604367255,
+      "grad_norm": 0.9444884052788686,
+      "learning_rate": 0.00014457619261453223,
+      "loss": 1.5126,
+      "step": 8083
+    },
+    {
+      "epoch": 0.3724144285253605,
+      "grad_norm": 1.1812968306682832,
+      "learning_rate": 0.00014456283566362891,
+      "loss": 1.3837,
+      "step": 8084
+    },
+    {
+      "epoch": 0.3724604966139955,
+      "grad_norm": 1.3316712869033116,
+      "learning_rate": 0.00014454947772061052,
+      "loss": 1.4711,
+      "step": 8085
+    },
+    {
+      "epoch": 0.3725065647026305,
+      "grad_norm": 2.2010356846896575,
+      "learning_rate": 0.00014453611878577436,
+      "loss": 1.5609,
+      "step": 8086
+    },
+    {
+      "epoch": 0.3725526327912655,
+      "grad_norm": 1.0229531378807875,
+      "learning_rate": 0.0001445227588594179,
+      "loss": 1.4278,
+      "step": 8087
+    },
+    {
+      "epoch": 0.3725987008799005,
+      "grad_norm": 1.9366794855992018,
+      "learning_rate": 0.00014450939794183852,
+      "loss": 1.3689,
+      "step": 8088
+    },
+    {
+      "epoch": 0.3726447689685355,
+      "grad_norm": 2.422752630232606,
+      "learning_rate": 0.0001444960360333337,
+      "loss": 1.5483,
+      "step": 8089
+    },
+    {
+      "epoch": 0.3726908370571705,
+      "grad_norm": 0.9455144721873809,
+      "learning_rate": 0.000144482673134201,
+      "loss": 1.6426,
+      "step": 8090
+    },
+    {
+      "epoch": 0.3727369051458055,
+      "grad_norm": 1.6674572437014328,
+      "learning_rate": 0.0001444693092447378,
+      "loss": 1.7542,
+      "step": 8091
+    },
+    {
+      "epoch": 0.3727829732344405,
+      "grad_norm": 1.4262231855528522,
+      "learning_rate": 0.0001444559443652417,
+      "loss": 1.0101,
+      "step": 8092
+    },
+    {
+      "epoch": 0.3728290413230755,
+      "grad_norm": 2.2796152436676285,
+      "learning_rate": 0.0001444425784960102,
+      "loss": 1.4993,
+      "step": 8093
+    },
+    {
+      "epoch": 0.3728751094117105,
+      "grad_norm": 1.2727307986136833,
+      "learning_rate": 0.00014442921163734096,
+      "loss": 1.495,
+      "step": 8094
+    },
+    {
+      "epoch": 0.37292117750034554,
+      "grad_norm": 1.0611382644029324,
+      "learning_rate": 0.00014441584378953142,
+      "loss": 1.3499,
+      "step": 8095
+    },
+    {
+      "epoch": 0.37296724558898053,
+      "grad_norm": 2.980277195122816,
+      "learning_rate": 0.00014440247495287934,
+      "loss": 1.4498,
+      "step": 8096
+    },
+    {
+      "epoch": 0.37301331367761553,
+      "grad_norm": 3.0597249233600645,
+      "learning_rate": 0.00014438910512768225,
+      "loss": 1.2605,
+      "step": 8097
+    },
+    {
+      "epoch": 0.37305938176625053,
+      "grad_norm": 1.2738086852182144,
+      "learning_rate": 0.0001443757343142379,
+      "loss": 1.5339,
+      "step": 8098
+    },
+    {
+      "epoch": 0.37310544985488553,
+      "grad_norm": 1.3635958657548986,
+      "learning_rate": 0.0001443623625128438,
+      "loss": 1.1405,
+      "step": 8099
+    },
+    {
+      "epoch": 0.3731515179435205,
+      "grad_norm": 1.9871980529168967,
+      "learning_rate": 0.00014434898972379785,
+      "loss": 1.1328,
+      "step": 8100
+    },
+    {
+      "epoch": 0.3731975860321555,
+      "grad_norm": 3.6889399736959994,
+      "learning_rate": 0.00014433561594739765,
+      "loss": 1.4893,
+      "step": 8101
+    },
+    {
+      "epoch": 0.3732436541207905,
+      "grad_norm": 1.5762170679891656,
+      "learning_rate": 0.000144322241183941,
+      "loss": 1.5411,
+      "step": 8102
+    },
+    {
+      "epoch": 0.3732897222094255,
+      "grad_norm": 2.0077565993828848,
+      "learning_rate": 0.00014430886543372566,
+      "loss": 1.4495,
+      "step": 8103
+    },
+    {
+      "epoch": 0.3733357902980605,
+      "grad_norm": 1.9244185252648847,
+      "learning_rate": 0.00014429548869704937,
+      "loss": 1.4296,
+      "step": 8104
+    },
+    {
+      "epoch": 0.3733818583866955,
+      "grad_norm": 2.462504218437136,
+      "learning_rate": 0.00014428211097421002,
+      "loss": 1.0806,
+      "step": 8105
+    },
+    {
+      "epoch": 0.3734279264753305,
+      "grad_norm": 1.6866504972815153,
+      "learning_rate": 0.00014426873226550536,
+      "loss": 1.6188,
+      "step": 8106
+    },
+    {
+      "epoch": 0.3734739945639655,
+      "grad_norm": 1.3530740030392328,
+      "learning_rate": 0.0001442553525712333,
+      "loss": 1.3032,
+      "step": 8107
+    },
+    {
+      "epoch": 0.37352006265260057,
+      "grad_norm": 1.1467500042257588,
+      "learning_rate": 0.0001442419718916917,
+      "loss": 1.1996,
+      "step": 8108
+    },
+    {
+      "epoch": 0.37356613074123557,
+      "grad_norm": 1.3128734660192223,
+      "learning_rate": 0.00014422859022717843,
+      "loss": 1.6826,
+      "step": 8109
+    },
+    {
+      "epoch": 0.37361219882987057,
+      "grad_norm": 1.6904883581054497,
+      "learning_rate": 0.00014421520757799142,
+      "loss": 1.183,
+      "step": 8110
+    },
+    {
+      "epoch": 0.37365826691850557,
+      "grad_norm": 1.8556988964363974,
+      "learning_rate": 0.00014420182394442866,
+      "loss": 1.3455,
+      "step": 8111
+    },
+    {
+      "epoch": 0.37370433500714056,
+      "grad_norm": 1.734242795092629,
+      "learning_rate": 0.00014418843932678807,
+      "loss": 1.564,
+      "step": 8112
+    },
+    {
+      "epoch": 0.37375040309577556,
+      "grad_norm": 1.2793925894248606,
+      "learning_rate": 0.00014417505372536763,
+      "loss": 1.6188,
+      "step": 8113
+    },
+    {
+      "epoch": 0.37379647118441056,
+      "grad_norm": 1.3737298131687363,
+      "learning_rate": 0.0001441616671404654,
+      "loss": 1.3238,
+      "step": 8114
+    },
+    {
+      "epoch": 0.37384253927304556,
+      "grad_norm": 2.2176091226090406,
+      "learning_rate": 0.00014414827957237936,
+      "loss": 1.3631,
+      "step": 8115
+    },
+    {
+      "epoch": 0.37388860736168056,
+      "grad_norm": 1.889710526410331,
+      "learning_rate": 0.0001441348910214076,
+      "loss": 1.2357,
+      "step": 8116
+    },
+    {
+      "epoch": 0.37393467545031556,
+      "grad_norm": 2.7220435824032876,
+      "learning_rate": 0.00014412150148784809,
+      "loss": 1.5362,
+      "step": 8117
+    },
+    {
+      "epoch": 0.37398074353895056,
+      "grad_norm": 2.2412547402126064,
+      "learning_rate": 0.00014410811097199906,
+      "loss": 1.4349,
+      "step": 8118
+    },
+    {
+      "epoch": 0.37402681162758555,
+      "grad_norm": 3.5648657429643698,
+      "learning_rate": 0.00014409471947415855,
+      "loss": 1.5781,
+      "step": 8119
+    },
+    {
+      "epoch": 0.37407287971622055,
+      "grad_norm": 2.296446559309773,
+      "learning_rate": 0.00014408132699462471,
+      "loss": 1.5255,
+      "step": 8120
+    },
+    {
+      "epoch": 0.37411894780485555,
+      "grad_norm": 1.7550203903380674,
+      "learning_rate": 0.00014406793353369571,
+      "loss": 1.2996,
+      "step": 8121
+    },
+    {
+      "epoch": 0.3741650158934906,
+      "grad_norm": 1.443414291699749,
+      "learning_rate": 0.00014405453909166976,
+      "loss": 1.1829,
+      "step": 8122
+    },
+    {
+      "epoch": 0.3742110839821256,
+      "grad_norm": 1.1818855187967707,
+      "learning_rate": 0.00014404114366884497,
+      "loss": 1.6554,
+      "step": 8123
+    },
+    {
+      "epoch": 0.3742571520707606,
+      "grad_norm": 1.5457473747243928,
+      "learning_rate": 0.0001440277472655197,
+      "loss": 1.3455,
+      "step": 8124
+    },
+    {
+      "epoch": 0.3743032201593956,
+      "grad_norm": 1.8337427390589949,
+      "learning_rate": 0.0001440143498819921,
+      "loss": 1.1298,
+      "step": 8125
+    },
+    {
+      "epoch": 0.3743492882480306,
+      "grad_norm": 2.14014720738217,
+      "learning_rate": 0.00014400095151856045,
+      "loss": 1.1656,
+      "step": 8126
+    },
+    {
+      "epoch": 0.3743953563366656,
+      "grad_norm": 1.5925015145577164,
+      "learning_rate": 0.0001439875521755231,
+      "loss": 1.6778,
+      "step": 8127
+    },
+    {
+      "epoch": 0.3744414244253006,
+      "grad_norm": 3.1865539807609107,
+      "learning_rate": 0.00014397415185317828,
+      "loss": 1.4624,
+      "step": 8128
+    },
+    {
+      "epoch": 0.3744874925139356,
+      "grad_norm": 1.9418475543059124,
+      "learning_rate": 0.0001439607505518244,
+      "loss": 1.3911,
+      "step": 8129
+    },
+    {
+      "epoch": 0.3745335606025706,
+      "grad_norm": 2.20233753159321,
+      "learning_rate": 0.00014394734827175978,
+      "loss": 1.3896,
+      "step": 8130
+    },
+    {
+      "epoch": 0.3745796286912056,
+      "grad_norm": 1.6155866409374984,
+      "learning_rate": 0.0001439339450132828,
+      "loss": 1.2249,
+      "step": 8131
+    },
+    {
+      "epoch": 0.3746256967798406,
+      "grad_norm": 1.4341380519293832,
+      "learning_rate": 0.00014392054077669186,
+      "loss": 1.2992,
+      "step": 8132
+    },
+    {
+      "epoch": 0.3746717648684756,
+      "grad_norm": 2.06225263543446,
+      "learning_rate": 0.00014390713556228539,
+      "loss": 1.2663,
+      "step": 8133
+    },
+    {
+      "epoch": 0.3747178329571106,
+      "grad_norm": 2.4722844030471753,
+      "learning_rate": 0.00014389372937036185,
+      "loss": 1.3729,
+      "step": 8134
+    },
+    {
+      "epoch": 0.3747639010457456,
+      "grad_norm": 1.2798484510553567,
+      "learning_rate": 0.00014388032220121966,
+      "loss": 1.4221,
+      "step": 8135
+    },
+    {
+      "epoch": 0.37480996913438064,
+      "grad_norm": 3.485714716640164,
+      "learning_rate": 0.00014386691405515736,
+      "loss": 1.3165,
+      "step": 8136
+    },
+    {
+      "epoch": 0.37485603722301564,
+      "grad_norm": 1.4246317888499775,
+      "learning_rate": 0.0001438535049324734,
+      "loss": 1.712,
+      "step": 8137
+    },
+    {
+      "epoch": 0.37490210531165064,
+      "grad_norm": 1.7931643446154113,
+      "learning_rate": 0.00014384009483346636,
+      "loss": 1.5141,
+      "step": 8138
+    },
+    {
+      "epoch": 0.37494817340028563,
+      "grad_norm": 1.3323738128086555,
+      "learning_rate": 0.00014382668375843477,
+      "loss": 1.4396,
+      "step": 8139
+    },
+    {
+      "epoch": 0.37499424148892063,
+      "grad_norm": 1.6131925325561267,
+      "learning_rate": 0.00014381327170767723,
+      "loss": 1.2591,
+      "step": 8140
+    },
+    {
+      "epoch": 0.37504030957755563,
+      "grad_norm": 1.4958236556288935,
+      "learning_rate": 0.00014379985868149232,
+      "loss": 1.7351,
+      "step": 8141
+    },
+    {
+      "epoch": 0.37508637766619063,
+      "grad_norm": 1.3229104155625317,
+      "learning_rate": 0.00014378644468017866,
+      "loss": 1.5739,
+      "step": 8142
+    },
+    {
+      "epoch": 0.3751324457548256,
+      "grad_norm": 9.708459456636922,
+      "learning_rate": 0.0001437730297040349,
+      "loss": 1.3166,
+      "step": 8143
+    },
+    {
+      "epoch": 0.3751785138434606,
+      "grad_norm": 1.2570666023681316,
+      "learning_rate": 0.00014375961375335964,
+      "loss": 1.2067,
+      "step": 8144
+    },
+    {
+      "epoch": 0.3752245819320956,
+      "grad_norm": 1.1865320733107456,
+      "learning_rate": 0.00014374619682845164,
+      "loss": 1.0852,
+      "step": 8145
+    },
+    {
+      "epoch": 0.3752706500207306,
+      "grad_norm": 2.4375915891250926,
+      "learning_rate": 0.0001437327789296096,
+      "loss": 1.4392,
+      "step": 8146
+    },
+    {
+      "epoch": 0.3753167181093656,
+      "grad_norm": 1.3681459237818234,
+      "learning_rate": 0.0001437193600571322,
+      "loss": 1.0523,
+      "step": 8147
+    },
+    {
+      "epoch": 0.3753627861980006,
+      "grad_norm": 1.379217328265326,
+      "learning_rate": 0.0001437059402113182,
+      "loss": 1.2812,
+      "step": 8148
+    },
+    {
+      "epoch": 0.3754088542866357,
+      "grad_norm": 2.088056351653029,
+      "learning_rate": 0.00014369251939246638,
+      "loss": 1.4063,
+      "step": 8149
+    },
+    {
+      "epoch": 0.37545492237527067,
+      "grad_norm": 1.106862728042769,
+      "learning_rate": 0.00014367909760087553,
+      "loss": 1.5763,
+      "step": 8150
+    },
+    {
+      "epoch": 0.37550099046390567,
+      "grad_norm": 2.5371750456406725,
+      "learning_rate": 0.00014366567483684451,
+      "loss": 1.2815,
+      "step": 8151
+    },
+    {
+      "epoch": 0.37554705855254067,
+      "grad_norm": 2.317872982719639,
+      "learning_rate": 0.00014365225110067207,
+      "loss": 1.5905,
+      "step": 8152
+    },
+    {
+      "epoch": 0.37559312664117567,
+      "grad_norm": 1.373714893530772,
+      "learning_rate": 0.0001436388263926571,
+      "loss": 1.2432,
+      "step": 8153
+    },
+    {
+      "epoch": 0.37563919472981067,
+      "grad_norm": 3.3183955505053375,
+      "learning_rate": 0.0001436254007130985,
+      "loss": 1.611,
+      "step": 8154
+    },
+    {
+      "epoch": 0.37568526281844566,
+      "grad_norm": 2.341063058095062,
+      "learning_rate": 0.00014361197406229513,
+      "loss": 1.4512,
+      "step": 8155
+    },
+    {
+      "epoch": 0.37573133090708066,
+      "grad_norm": 1.3125640438607995,
+      "learning_rate": 0.000143598546440546,
+      "loss": 1.4751,
+      "step": 8156
+    },
+    {
+      "epoch": 0.37577739899571566,
+      "grad_norm": 0.9736743540942873,
+      "learning_rate": 0.00014358511784814991,
+      "loss": 1.1549,
+      "step": 8157
+    },
+    {
+      "epoch": 0.37582346708435066,
+      "grad_norm": 1.3835059710170792,
+      "learning_rate": 0.00014357168828540596,
+      "loss": 1.3886,
+      "step": 8158
+    },
+    {
+      "epoch": 0.37586953517298566,
+      "grad_norm": 1.0850051919033956,
+      "learning_rate": 0.00014355825775261307,
+      "loss": 1.4094,
+      "step": 8159
+    },
+    {
+      "epoch": 0.37591560326162066,
+      "grad_norm": 2.2112402912637807,
+      "learning_rate": 0.00014354482625007027,
+      "loss": 1.4998,
+      "step": 8160
+    },
+    {
+      "epoch": 0.37596167135025566,
+      "grad_norm": 1.0670338603428469,
+      "learning_rate": 0.00014353139377807654,
+      "loss": 1.4765,
+      "step": 8161
+    },
+    {
+      "epoch": 0.37600773943889065,
+      "grad_norm": 1.1092779412433043,
+      "learning_rate": 0.00014351796033693098,
+      "loss": 1.3548,
+      "step": 8162
+    },
+    {
+      "epoch": 0.3760538075275257,
+      "grad_norm": 1.3942601943376387,
+      "learning_rate": 0.00014350452592693267,
+      "loss": 1.3344,
+      "step": 8163
+    },
+    {
+      "epoch": 0.3760998756161607,
+      "grad_norm": 1.330706652584019,
+      "learning_rate": 0.00014349109054838065,
+      "loss": 1.5079,
+      "step": 8164
+    },
+    {
+      "epoch": 0.3761459437047957,
+      "grad_norm": 4.492432078793375,
+      "learning_rate": 0.0001434776542015741,
+      "loss": 1.3231,
+      "step": 8165
+    },
+    {
+      "epoch": 0.3761920117934307,
+      "grad_norm": 1.5326342505268629,
+      "learning_rate": 0.0001434642168868121,
+      "loss": 1.6059,
+      "step": 8166
+    },
+    {
+      "epoch": 0.3762380798820657,
+      "grad_norm": 1.237242441604465,
+      "learning_rate": 0.00014345077860439387,
+      "loss": 1.5443,
+      "step": 8167
+    },
+    {
+      "epoch": 0.3762841479707007,
+      "grad_norm": 3.508686165455976,
+      "learning_rate": 0.00014343733935461852,
+      "loss": 1.3172,
+      "step": 8168
+    },
+    {
+      "epoch": 0.3763302160593357,
+      "grad_norm": 1.534614445440634,
+      "learning_rate": 0.0001434238991377853,
+      "loss": 1.2922,
+      "step": 8169
+    },
+    {
+      "epoch": 0.3763762841479707,
+      "grad_norm": 2.237612300519017,
+      "learning_rate": 0.00014341045795419346,
+      "loss": 1.4421,
+      "step": 8170
+    },
+    {
+      "epoch": 0.3764223522366057,
+      "grad_norm": 1.1117262761063822,
+      "learning_rate": 0.00014339701580414217,
+      "loss": 1.3958,
+      "step": 8171
+    },
+    {
+      "epoch": 0.3764684203252407,
+      "grad_norm": 1.1854531399666244,
+      "learning_rate": 0.00014338357268793073,
+      "loss": 1.3603,
+      "step": 8172
+    },
+    {
+      "epoch": 0.3765144884138757,
+      "grad_norm": 1.2433660376020563,
+      "learning_rate": 0.00014337012860585844,
+      "loss": 1.4803,
+      "step": 8173
+    },
+    {
+      "epoch": 0.3765605565025107,
+      "grad_norm": 1.5419174121909225,
+      "learning_rate": 0.0001433566835582246,
+      "loss": 1.1756,
+      "step": 8174
+    },
+    {
+      "epoch": 0.3766066245911457,
+      "grad_norm": 1.1162048312890736,
+      "learning_rate": 0.00014334323754532858,
+      "loss": 1.3186,
+      "step": 8175
+    },
+    {
+      "epoch": 0.37665269267978074,
+      "grad_norm": 1.571222156545218,
+      "learning_rate": 0.00014332979056746963,
+      "loss": 1.6548,
+      "step": 8176
+    },
+    {
+      "epoch": 0.37669876076841574,
+      "grad_norm": 1.7683518595648797,
+      "learning_rate": 0.0001433163426249472,
+      "loss": 1.3878,
+      "step": 8177
+    },
+    {
+      "epoch": 0.37674482885705074,
+      "grad_norm": 1.1905552561524,
+      "learning_rate": 0.0001433028937180607,
+      "loss": 1.4722,
+      "step": 8178
+    },
+    {
+      "epoch": 0.37679089694568574,
+      "grad_norm": 1.827576484077976,
+      "learning_rate": 0.00014328944384710951,
+      "loss": 1.3795,
+      "step": 8179
+    },
+    {
+      "epoch": 0.37683696503432074,
+      "grad_norm": 0.9562320591786848,
+      "learning_rate": 0.00014327599301239307,
+      "loss": 1.1783,
+      "step": 8180
+    },
+    {
+      "epoch": 0.37688303312295574,
+      "grad_norm": 1.4859023797236772,
+      "learning_rate": 0.00014326254121421086,
+      "loss": 1.5752,
+      "step": 8181
+    },
+    {
+      "epoch": 0.37692910121159073,
+      "grad_norm": 1.1632010504391566,
+      "learning_rate": 0.00014324908845286233,
+      "loss": 1.1548,
+      "step": 8182
+    },
+    {
+      "epoch": 0.37697516930022573,
+      "grad_norm": 1.7268366298627282,
+      "learning_rate": 0.00014323563472864702,
+      "loss": 1.3457,
+      "step": 8183
+    },
+    {
+      "epoch": 0.37702123738886073,
+      "grad_norm": 1.0756023515270785,
+      "learning_rate": 0.00014322218004186442,
+      "loss": 1.3094,
+      "step": 8184
+    },
+    {
+      "epoch": 0.37706730547749573,
+      "grad_norm": 1.3369429859802044,
+      "learning_rate": 0.0001432087243928141,
+      "loss": 1.3292,
+      "step": 8185
+    },
+    {
+      "epoch": 0.3771133735661307,
+      "grad_norm": 1.5934772926493568,
+      "learning_rate": 0.00014319526778179564,
+      "loss": 1.3151,
+      "step": 8186
+    },
+    {
+      "epoch": 0.3771594416547657,
+      "grad_norm": 1.0362516723834099,
+      "learning_rate": 0.00014318181020910858,
+      "loss": 1.4772,
+      "step": 8187
+    },
+    {
+      "epoch": 0.3772055097434007,
+      "grad_norm": 1.0489769005731173,
+      "learning_rate": 0.00014316835167505255,
+      "loss": 1.451,
+      "step": 8188
+    },
+    {
+      "epoch": 0.3772515778320357,
+      "grad_norm": 1.5476829934811782,
+      "learning_rate": 0.00014315489217992723,
+      "loss": 1.2386,
+      "step": 8189
+    },
+    {
+      "epoch": 0.3772976459206708,
+      "grad_norm": 1.477022276116668,
+      "learning_rate": 0.00014314143172403218,
+      "loss": 1.5428,
+      "step": 8190
+    },
+    {
+      "epoch": 0.3773437140093058,
+      "grad_norm": 1.9899007261861497,
+      "learning_rate": 0.00014312797030766717,
+      "loss": 1.4501,
+      "step": 8191
+    },
+    {
+      "epoch": 0.3773897820979408,
+      "grad_norm": 1.6232131065910518,
+      "learning_rate": 0.0001431145079311318,
+      "loss": 1.5769,
+      "step": 8192
+    },
+    {
+      "epoch": 0.37743585018657577,
+      "grad_norm": 1.210491136605227,
+      "learning_rate": 0.00014310104459472589,
+      "loss": 1.2629,
+      "step": 8193
+    },
+    {
+      "epoch": 0.37748191827521077,
+      "grad_norm": 1.189661174476252,
+      "learning_rate": 0.00014308758029874907,
+      "loss": 1.3688,
+      "step": 8194
+    },
+    {
+      "epoch": 0.37752798636384577,
+      "grad_norm": 1.421302562499319,
+      "learning_rate": 0.00014307411504350118,
+      "loss": 1.487,
+      "step": 8195
+    },
+    {
+      "epoch": 0.37757405445248077,
+      "grad_norm": 1.2623969462581726,
+      "learning_rate": 0.00014306064882928197,
+      "loss": 1.4812,
+      "step": 8196
+    },
+    {
+      "epoch": 0.37762012254111577,
+      "grad_norm": 1.228073852109991,
+      "learning_rate": 0.00014304718165639128,
+      "loss": 1.5321,
+      "step": 8197
+    },
+    {
+      "epoch": 0.37766619062975076,
+      "grad_norm": 1.1443560250080003,
+      "learning_rate": 0.00014303371352512886,
+      "loss": 1.3403,
+      "step": 8198
+    },
+    {
+      "epoch": 0.37771225871838576,
+      "grad_norm": 2.3869570183463913,
+      "learning_rate": 0.0001430202444357946,
+      "loss": 1.3352,
+      "step": 8199
+    },
+    {
+      "epoch": 0.37775832680702076,
+      "grad_norm": 1.9773541775441148,
+      "learning_rate": 0.0001430067743886884,
+      "loss": 1.5012,
+      "step": 8200
+    },
+    {
+      "epoch": 0.37780439489565576,
+      "grad_norm": 2.350900646940215,
+      "learning_rate": 0.00014299330338411005,
+      "loss": 1.2416,
+      "step": 8201
+    },
+    {
+      "epoch": 0.37785046298429076,
+      "grad_norm": 1.1186071513083977,
+      "learning_rate": 0.00014297983142235956,
+      "loss": 1.2486,
+      "step": 8202
+    },
+    {
+      "epoch": 0.3778965310729258,
+      "grad_norm": 1.0975635211929002,
+      "learning_rate": 0.00014296635850373677,
+      "loss": 1.2512,
+      "step": 8203
+    },
+    {
+      "epoch": 0.3779425991615608,
+      "grad_norm": 1.6796963406387115,
+      "learning_rate": 0.00014295288462854173,
+      "loss": 1.5052,
+      "step": 8204
+    },
+    {
+      "epoch": 0.3779886672501958,
+      "grad_norm": 1.0543969884119069,
+      "learning_rate": 0.00014293940979707433,
+      "loss": 1.142,
+      "step": 8205
+    },
+    {
+      "epoch": 0.3780347353388308,
+      "grad_norm": 1.6781549239268327,
+      "learning_rate": 0.00014292593400963465,
+      "loss": 1.3891,
+      "step": 8206
+    },
+    {
+      "epoch": 0.3780808034274658,
+      "grad_norm": 1.0824179463491763,
+      "learning_rate": 0.0001429124572665226,
+      "loss": 1.6134,
+      "step": 8207
+    },
+    {
+      "epoch": 0.3781268715161008,
+      "grad_norm": 1.5612459497499143,
+      "learning_rate": 0.0001428989795680383,
+      "loss": 1.3742,
+      "step": 8208
+    },
+    {
+      "epoch": 0.3781729396047358,
+      "grad_norm": 1.7149480360275509,
+      "learning_rate": 0.00014288550091448175,
+      "loss": 1.4818,
+      "step": 8209
+    },
+    {
+      "epoch": 0.3782190076933708,
+      "grad_norm": 1.6435216900223912,
+      "learning_rate": 0.00014287202130615307,
+      "loss": 1.1195,
+      "step": 8210
+    },
+    {
+      "epoch": 0.3782650757820058,
+      "grad_norm": 2.009019061187925,
+      "learning_rate": 0.00014285854074335232,
+      "loss": 1.3315,
+      "step": 8211
+    },
+    {
+      "epoch": 0.3783111438706408,
+      "grad_norm": 1.253537114191165,
+      "learning_rate": 0.00014284505922637964,
+      "loss": 1.1926,
+      "step": 8212
+    },
+    {
+      "epoch": 0.3783572119592758,
+      "grad_norm": 1.3725459223821843,
+      "learning_rate": 0.00014283157675553523,
+      "loss": 1.2909,
+      "step": 8213
+    },
+    {
+      "epoch": 0.3784032800479108,
+      "grad_norm": 1.5460904272399418,
+      "learning_rate": 0.0001428180933311192,
+      "loss": 1.4516,
+      "step": 8214
+    },
+    {
+      "epoch": 0.3784493481365458,
+      "grad_norm": 1.5525819456391534,
+      "learning_rate": 0.00014280460895343172,
+      "loss": 1.1354,
+      "step": 8215
+    },
+    {
+      "epoch": 0.3784954162251808,
+      "grad_norm": 1.3990346469155295,
+      "learning_rate": 0.000142791123622773,
+      "loss": 1.5351,
+      "step": 8216
+    },
+    {
+      "epoch": 0.37854148431381585,
+      "grad_norm": 0.8459361029441688,
+      "learning_rate": 0.0001427776373394433,
+      "loss": 1.4564,
+      "step": 8217
+    },
+    {
+      "epoch": 0.37858755240245084,
+      "grad_norm": 2.142075137970092,
+      "learning_rate": 0.00014276415010374283,
+      "loss": 1.325,
+      "step": 8218
+    },
+    {
+      "epoch": 0.37863362049108584,
+      "grad_norm": 1.2574353269907361,
+      "learning_rate": 0.00014275066191597193,
+      "loss": 1.2895,
+      "step": 8219
+    },
+    {
+      "epoch": 0.37867968857972084,
+      "grad_norm": 1.4211790807848692,
+      "learning_rate": 0.0001427371727764308,
+      "loss": 1.2243,
+      "step": 8220
+    },
+    {
+      "epoch": 0.37872575666835584,
+      "grad_norm": 1.3168280300882054,
+      "learning_rate": 0.00014272368268541981,
+      "loss": 1.467,
+      "step": 8221
+    },
+    {
+      "epoch": 0.37877182475699084,
+      "grad_norm": 1.1103191637020104,
+      "learning_rate": 0.00014271019164323928,
+      "loss": 1.4875,
+      "step": 8222
+    },
+    {
+      "epoch": 0.37881789284562584,
+      "grad_norm": 0.7656000326650007,
+      "learning_rate": 0.00014269669965018958,
+      "loss": 1.4358,
+      "step": 8223
+    },
+    {
+      "epoch": 0.37886396093426083,
+      "grad_norm": 1.1814305915702252,
+      "learning_rate": 0.00014268320670657108,
+      "loss": 1.3829,
+      "step": 8224
+    },
+    {
+      "epoch": 0.37891002902289583,
+      "grad_norm": 2.1758788551819648,
+      "learning_rate": 0.00014266971281268412,
+      "loss": 1.4078,
+      "step": 8225
+    },
+    {
+      "epoch": 0.37895609711153083,
+      "grad_norm": 1.7384261302570954,
+      "learning_rate": 0.00014265621796882922,
+      "loss": 1.4066,
+      "step": 8226
+    },
+    {
+      "epoch": 0.37900216520016583,
+      "grad_norm": 1.4222977789571596,
+      "learning_rate": 0.00014264272217530673,
+      "loss": 1.1678,
+      "step": 8227
+    },
+    {
+      "epoch": 0.37904823328880083,
+      "grad_norm": 1.1196203701054674,
+      "learning_rate": 0.00014262922543241717,
+      "loss": 1.3363,
+      "step": 8228
+    },
+    {
+      "epoch": 0.3790943013774358,
+      "grad_norm": 1.0076602667123042,
+      "learning_rate": 0.00014261572774046098,
+      "loss": 1.5529,
+      "step": 8229
+    },
+    {
+      "epoch": 0.3791403694660708,
+      "grad_norm": 1.59113036808754,
+      "learning_rate": 0.0001426022290997387,
+      "loss": 1.1716,
+      "step": 8230
+    },
+    {
+      "epoch": 0.3791864375547059,
+      "grad_norm": 1.1148700993268295,
+      "learning_rate": 0.00014258872951055081,
+      "loss": 1.3659,
+      "step": 8231
+    },
+    {
+      "epoch": 0.3792325056433409,
+      "grad_norm": 1.5202742427885365,
+      "learning_rate": 0.00014257522897319792,
+      "loss": 1.1529,
+      "step": 8232
+    },
+    {
+      "epoch": 0.3792785737319759,
+      "grad_norm": 1.5280505787723189,
+      "learning_rate": 0.0001425617274879805,
+      "loss": 1.4687,
+      "step": 8233
+    },
+    {
+      "epoch": 0.3793246418206109,
+      "grad_norm": 1.30695828531236,
+      "learning_rate": 0.00014254822505519927,
+      "loss": 1.5833,
+      "step": 8234
+    },
+    {
+      "epoch": 0.3793707099092459,
+      "grad_norm": 1.5753279010294483,
+      "learning_rate": 0.00014253472167515474,
+      "loss": 1.7119,
+      "step": 8235
+    },
+    {
+      "epoch": 0.37941677799788087,
+      "grad_norm": 1.0784096854290925,
+      "learning_rate": 0.0001425212173481475,
+      "loss": 1.5822,
+      "step": 8236
+    },
+    {
+      "epoch": 0.37946284608651587,
+      "grad_norm": 0.9995345032864558,
+      "learning_rate": 0.00014250771207447834,
+      "loss": 1.5639,
+      "step": 8237
+    },
+    {
+      "epoch": 0.37950891417515087,
+      "grad_norm": 1.076490441408451,
+      "learning_rate": 0.00014249420585444784,
+      "loss": 0.8529,
+      "step": 8238
+    },
+    {
+      "epoch": 0.37955498226378587,
+      "grad_norm": 1.1902858103521727,
+      "learning_rate": 0.0001424806986883567,
+      "loss": 1.4355,
+      "step": 8239
+    },
+    {
+      "epoch": 0.37960105035242087,
+      "grad_norm": 2.2742649766583014,
+      "learning_rate": 0.00014246719057650563,
+      "loss": 1.569,
+      "step": 8240
+    },
+    {
+      "epoch": 0.37964711844105586,
+      "grad_norm": 1.753710097376024,
+      "learning_rate": 0.0001424536815191954,
+      "loss": 1.6199,
+      "step": 8241
+    },
+    {
+      "epoch": 0.37969318652969086,
+      "grad_norm": 1.480649066792265,
+      "learning_rate": 0.00014244017151672674,
+      "loss": 1.2835,
+      "step": 8242
+    },
+    {
+      "epoch": 0.37973925461832586,
+      "grad_norm": 1.2204489099842457,
+      "learning_rate": 0.00014242666056940042,
+      "loss": 1.3707,
+      "step": 8243
+    },
+    {
+      "epoch": 0.3797853227069609,
+      "grad_norm": 1.1430099052160116,
+      "learning_rate": 0.00014241314867751724,
+      "loss": 1.5196,
+      "step": 8244
+    },
+    {
+      "epoch": 0.3798313907955959,
+      "grad_norm": 1.1562803265494461,
+      "learning_rate": 0.00014239963584137803,
+      "loss": 1.2823,
+      "step": 8245
+    },
+    {
+      "epoch": 0.3798774588842309,
+      "grad_norm": 1.4969628372545851,
+      "learning_rate": 0.00014238612206128365,
+      "loss": 1.2122,
+      "step": 8246
+    },
+    {
+      "epoch": 0.3799235269728659,
+      "grad_norm": 1.134801992345631,
+      "learning_rate": 0.00014237260733753493,
+      "loss": 1.253,
+      "step": 8247
+    },
+    {
+      "epoch": 0.3799695950615009,
+      "grad_norm": 1.2385906006842318,
+      "learning_rate": 0.00014235909167043274,
+      "loss": 1.4374,
+      "step": 8248
+    },
+    {
+      "epoch": 0.3800156631501359,
+      "grad_norm": 1.014801931865323,
+      "learning_rate": 0.000142345575060278,
+      "loss": 1.5002,
+      "step": 8249
+    },
+    {
+      "epoch": 0.3800617312387709,
+      "grad_norm": 1.7742997910802445,
+      "learning_rate": 0.00014233205750737167,
+      "loss": 1.4638,
+      "step": 8250
+    },
+    {
+      "epoch": 0.3801077993274059,
+      "grad_norm": 1.234376098667414,
+      "learning_rate": 0.00014231853901201466,
+      "loss": 1.5472,
+      "step": 8251
+    },
+    {
+      "epoch": 0.3801538674160409,
+      "grad_norm": 1.0595076658540774,
+      "learning_rate": 0.00014230501957450794,
+      "loss": 1.5037,
+      "step": 8252
+    },
+    {
+      "epoch": 0.3801999355046759,
+      "grad_norm": 1.2850145797194144,
+      "learning_rate": 0.0001422914991951525,
+      "loss": 1.4702,
+      "step": 8253
+    },
+    {
+      "epoch": 0.3802460035933109,
+      "grad_norm": 1.3632757513169251,
+      "learning_rate": 0.0001422779778742493,
+      "loss": 1.6033,
+      "step": 8254
+    },
+    {
+      "epoch": 0.3802920716819459,
+      "grad_norm": 0.8071259948747038,
+      "learning_rate": 0.00014226445561209947,
+      "loss": 1.5211,
+      "step": 8255
+    },
+    {
+      "epoch": 0.3803381397705809,
+      "grad_norm": 0.9388104073929391,
+      "learning_rate": 0.00014225093240900394,
+      "loss": 1.7267,
+      "step": 8256
+    },
+    {
+      "epoch": 0.3803842078592159,
+      "grad_norm": 1.7235285434928067,
+      "learning_rate": 0.00014223740826526394,
+      "loss": 1.0016,
+      "step": 8257
+    },
+    {
+      "epoch": 0.38043027594785095,
+      "grad_norm": 0.9625152879557065,
+      "learning_rate": 0.0001422238831811804,
+      "loss": 1.5429,
+      "step": 8258
+    },
+    {
+      "epoch": 0.38047634403648595,
+      "grad_norm": 1.2055673270575336,
+      "learning_rate": 0.0001422103571570545,
+      "loss": 1.6142,
+      "step": 8259
+    },
+    {
+      "epoch": 0.38052241212512095,
+      "grad_norm": 1.0486533683670054,
+      "learning_rate": 0.00014219683019318737,
+      "loss": 1.4809,
+      "step": 8260
+    },
+    {
+      "epoch": 0.38056848021375594,
+      "grad_norm": 1.2826431768596238,
+      "learning_rate": 0.0001421833022898802,
+      "loss": 1.2574,
+      "step": 8261
+    },
+    {
+      "epoch": 0.38061454830239094,
+      "grad_norm": 1.5678787999316268,
+      "learning_rate": 0.00014216977344743412,
+      "loss": 1.2463,
+      "step": 8262
+    },
+    {
+      "epoch": 0.38066061639102594,
+      "grad_norm": 1.2940893044202095,
+      "learning_rate": 0.00014215624366615035,
+      "loss": 1.4438,
+      "step": 8263
+    },
+    {
+      "epoch": 0.38070668447966094,
+      "grad_norm": 1.6673237706846584,
+      "learning_rate": 0.00014214271294633006,
+      "loss": 1.299,
+      "step": 8264
+    },
+    {
+      "epoch": 0.38075275256829594,
+      "grad_norm": 2.8115445045820806,
+      "learning_rate": 0.00014212918128827457,
+      "loss": 1.4532,
+      "step": 8265
+    },
+    {
+      "epoch": 0.38079882065693094,
+      "grad_norm": 2.013650601178024,
+      "learning_rate": 0.00014211564869228506,
+      "loss": 1.2873,
+      "step": 8266
+    },
+    {
+      "epoch": 0.38084488874556593,
+      "grad_norm": 0.7864420151285245,
+      "learning_rate": 0.00014210211515866284,
+      "loss": 1.2948,
+      "step": 8267
+    },
+    {
+      "epoch": 0.38089095683420093,
+      "grad_norm": 1.2936182661488653,
+      "learning_rate": 0.00014208858068770925,
+      "loss": 1.1894,
+      "step": 8268
+    },
+    {
+      "epoch": 0.38093702492283593,
+      "grad_norm": 1.1627619612113027,
+      "learning_rate": 0.00014207504527972555,
+      "loss": 1.5741,
+      "step": 8269
+    },
+    {
+      "epoch": 0.38098309301147093,
+      "grad_norm": 1.8207878417087464,
+      "learning_rate": 0.00014206150893501316,
+      "loss": 1.3428,
+      "step": 8270
+    },
+    {
+      "epoch": 0.381029161100106,
+      "grad_norm": 1.839893824145546,
+      "learning_rate": 0.00014204797165387334,
+      "loss": 1.5851,
+      "step": 8271
+    },
+    {
+      "epoch": 0.381075229188741,
+      "grad_norm": 1.3127719321608626,
+      "learning_rate": 0.00014203443343660754,
+      "loss": 1.5467,
+      "step": 8272
+    },
+    {
+      "epoch": 0.381121297277376,
+      "grad_norm": 1.0872872532723732,
+      "learning_rate": 0.00014202089428351714,
+      "loss": 1.3743,
+      "step": 8273
+    },
+    {
+      "epoch": 0.381167365366011,
+      "grad_norm": 1.6186941810264406,
+      "learning_rate": 0.00014200735419490362,
+      "loss": 1.4408,
+      "step": 8274
+    },
+    {
+      "epoch": 0.381213433454646,
+      "grad_norm": 1.2315363864643132,
+      "learning_rate": 0.00014199381317106836,
+      "loss": 1.2111,
+      "step": 8275
+    },
+    {
+      "epoch": 0.381259501543281,
+      "grad_norm": 1.9229382787116973,
+      "learning_rate": 0.00014198027121231283,
+      "loss": 1.2136,
+      "step": 8276
+    },
+    {
+      "epoch": 0.381305569631916,
+      "grad_norm": 1.1475526770486009,
+      "learning_rate": 0.00014196672831893857,
+      "loss": 1.4384,
+      "step": 8277
+    },
+    {
+      "epoch": 0.381351637720551,
+      "grad_norm": 3.059136689571875,
+      "learning_rate": 0.00014195318449124707,
+      "loss": 1.4132,
+      "step": 8278
+    },
+    {
+      "epoch": 0.38139770580918597,
+      "grad_norm": 1.1362481358013996,
+      "learning_rate": 0.00014193963972953985,
+      "loss": 1.4258,
+      "step": 8279
+    },
+    {
+      "epoch": 0.38144377389782097,
+      "grad_norm": 2.992544566360264,
+      "learning_rate": 0.00014192609403411842,
+      "loss": 1.507,
+      "step": 8280
+    },
+    {
+      "epoch": 0.38148984198645597,
+      "grad_norm": 2.919995169632167,
+      "learning_rate": 0.00014191254740528446,
+      "loss": 1.0758,
+      "step": 8281
+    },
+    {
+      "epoch": 0.38153591007509097,
+      "grad_norm": 1.4817814941625964,
+      "learning_rate": 0.00014189899984333944,
+      "loss": 1.4954,
+      "step": 8282
+    },
+    {
+      "epoch": 0.38158197816372597,
+      "grad_norm": 2.7494203711648573,
+      "learning_rate": 0.00014188545134858507,
+      "loss": 1.3808,
+      "step": 8283
+    },
+    {
+      "epoch": 0.38162804625236096,
+      "grad_norm": 1.15998733307976,
+      "learning_rate": 0.0001418719019213229,
+      "loss": 1.4345,
+      "step": 8284
+    },
+    {
+      "epoch": 0.381674114340996,
+      "grad_norm": 1.2241649856594992,
+      "learning_rate": 0.00014185835156185468,
+      "loss": 1.0515,
+      "step": 8285
+    },
+    {
+      "epoch": 0.381720182429631,
+      "grad_norm": 1.266992950419607,
+      "learning_rate": 0.00014184480027048198,
+      "loss": 1.3358,
+      "step": 8286
+    },
+    {
+      "epoch": 0.381766250518266,
+      "grad_norm": 1.5378858943304456,
+      "learning_rate": 0.0001418312480475066,
+      "loss": 1.3442,
+      "step": 8287
+    },
+    {
+      "epoch": 0.381812318606901,
+      "grad_norm": 1.1911302212891564,
+      "learning_rate": 0.0001418176948932302,
+      "loss": 1.3918,
+      "step": 8288
+    },
+    {
+      "epoch": 0.381858386695536,
+      "grad_norm": 1.572342444732778,
+      "learning_rate": 0.0001418041408079545,
+      "loss": 1.2278,
+      "step": 8289
+    },
+    {
+      "epoch": 0.381904454784171,
+      "grad_norm": 1.2432855669886318,
+      "learning_rate": 0.0001417905857919813,
+      "loss": 1.3765,
+      "step": 8290
+    },
+    {
+      "epoch": 0.381950522872806,
+      "grad_norm": 0.9109992281480165,
+      "learning_rate": 0.00014177702984561233,
+      "loss": 1.3887,
+      "step": 8291
+    },
+    {
+      "epoch": 0.381996590961441,
+      "grad_norm": 1.416290712328964,
+      "learning_rate": 0.00014176347296914949,
+      "loss": 1.1725,
+      "step": 8292
+    },
+    {
+      "epoch": 0.382042659050076,
+      "grad_norm": 6.903395752358738,
+      "learning_rate": 0.00014174991516289448,
+      "loss": 1.1053,
+      "step": 8293
+    },
+    {
+      "epoch": 0.382088727138711,
+      "grad_norm": 1.0548420475194729,
+      "learning_rate": 0.00014173635642714922,
+      "loss": 1.2297,
+      "step": 8294
+    },
+    {
+      "epoch": 0.382134795227346,
+      "grad_norm": 1.6083969830346583,
+      "learning_rate": 0.00014172279676221553,
+      "loss": 1.3034,
+      "step": 8295
+    },
+    {
+      "epoch": 0.382180863315981,
+      "grad_norm": 1.8813917044358435,
+      "learning_rate": 0.00014170923616839531,
+      "loss": 1.4417,
+      "step": 8296
+    },
+    {
+      "epoch": 0.382226931404616,
+      "grad_norm": 0.8795559550190925,
+      "learning_rate": 0.00014169567464599044,
+      "loss": 1.5224,
+      "step": 8297
+    },
+    {
+      "epoch": 0.382272999493251,
+      "grad_norm": 0.8863622695429466,
+      "learning_rate": 0.00014168211219530292,
+      "loss": 1.5594,
+      "step": 8298
+    },
+    {
+      "epoch": 0.38231906758188605,
+      "grad_norm": 1.3362733730229388,
+      "learning_rate": 0.00014166854881663462,
+      "loss": 1.3299,
+      "step": 8299
+    },
+    {
+      "epoch": 0.38236513567052105,
+      "grad_norm": 2.1941884950106494,
+      "learning_rate": 0.0001416549845102875,
+      "loss": 1.472,
+      "step": 8300
+    },
+    {
+      "epoch": 0.38241120375915605,
+      "grad_norm": 1.0589858520065891,
+      "learning_rate": 0.00014164141927656363,
+      "loss": 1.3786,
+      "step": 8301
+    },
+    {
+      "epoch": 0.38245727184779105,
+      "grad_norm": 1.3380256555233268,
+      "learning_rate": 0.00014162785311576493,
+      "loss": 1.3328,
+      "step": 8302
+    },
+    {
+      "epoch": 0.38250333993642605,
+      "grad_norm": 1.3040623227692607,
+      "learning_rate": 0.00014161428602819346,
+      "loss": 1.3324,
+      "step": 8303
+    },
+    {
+      "epoch": 0.38254940802506104,
+      "grad_norm": 1.5555610054291653,
+      "learning_rate": 0.00014160071801415125,
+      "loss": 1.2685,
+      "step": 8304
+    },
+    {
+      "epoch": 0.38259547611369604,
+      "grad_norm": 1.264729564015958,
+      "learning_rate": 0.0001415871490739404,
+      "loss": 1.4207,
+      "step": 8305
+    },
+    {
+      "epoch": 0.38264154420233104,
+      "grad_norm": 1.0840991145737082,
+      "learning_rate": 0.00014157357920786294,
+      "loss": 1.2619,
+      "step": 8306
+    },
+    {
+      "epoch": 0.38268761229096604,
+      "grad_norm": 1.990347236035144,
+      "learning_rate": 0.0001415600084162211,
+      "loss": 1.3847,
+      "step": 8307
+    },
+    {
+      "epoch": 0.38273368037960104,
+      "grad_norm": 2.6732914228764155,
+      "learning_rate": 0.00014154643669931686,
+      "loss": 1.2534,
+      "step": 8308
+    },
+    {
+      "epoch": 0.38277974846823604,
+      "grad_norm": 2.203282072981101,
+      "learning_rate": 0.0001415328640574525,
+      "loss": 1.1196,
+      "step": 8309
+    },
+    {
+      "epoch": 0.38282581655687103,
+      "grad_norm": 2.330780175734304,
+      "learning_rate": 0.00014151929049093008,
+      "loss": 1.5471,
+      "step": 8310
+    },
+    {
+      "epoch": 0.38287188464550603,
+      "grad_norm": 3.0206524898633207,
+      "learning_rate": 0.00014150571600005186,
+      "loss": 1.3421,
+      "step": 8311
+    },
+    {
+      "epoch": 0.3829179527341411,
+      "grad_norm": 1.0129583132892643,
+      "learning_rate": 0.00014149214058512008,
+      "loss": 1.437,
+      "step": 8312
+    },
+    {
+      "epoch": 0.3829640208227761,
+      "grad_norm": 1.0537588731834433,
+      "learning_rate": 0.00014147856424643688,
+      "loss": 1.3375,
+      "step": 8313
+    },
+    {
+      "epoch": 0.3830100889114111,
+      "grad_norm": 1.2679367410132825,
+      "learning_rate": 0.00014146498698430461,
+      "loss": 1.3072,
+      "step": 8314
+    },
+    {
+      "epoch": 0.3830561570000461,
+      "grad_norm": 2.2455488099926817,
+      "learning_rate": 0.00014145140879902549,
+      "loss": 1.2888,
+      "step": 8315
+    },
+    {
+      "epoch": 0.3831022250886811,
+      "grad_norm": 2.0582246556515678,
+      "learning_rate": 0.00014143782969090185,
+      "loss": 1.1778,
+      "step": 8316
+    },
+    {
+      "epoch": 0.3831482931773161,
+      "grad_norm": 1.1858600117932938,
+      "learning_rate": 0.00014142424966023595,
+      "loss": 1.5908,
+      "step": 8317
+    },
+    {
+      "epoch": 0.3831943612659511,
+      "grad_norm": 1.8402006945474685,
+      "learning_rate": 0.00014141066870733017,
+      "loss": 1.2191,
+      "step": 8318
+    },
+    {
+      "epoch": 0.3832404293545861,
+      "grad_norm": 3.635309515473574,
+      "learning_rate": 0.00014139708683248684,
+      "loss": 1.4203,
+      "step": 8319
+    },
+    {
+      "epoch": 0.3832864974432211,
+      "grad_norm": 2.106072000087631,
+      "learning_rate": 0.00014138350403600838,
+      "loss": 1.4174,
+      "step": 8320
+    },
+    {
+      "epoch": 0.3833325655318561,
+      "grad_norm": 1.8242704407264656,
+      "learning_rate": 0.00014136992031819713,
+      "loss": 1.1778,
+      "step": 8321
+    },
+    {
+      "epoch": 0.38337863362049107,
+      "grad_norm": 0.9472949358040232,
+      "learning_rate": 0.00014135633567935555,
+      "loss": 1.4038,
+      "step": 8322
+    },
+    {
+      "epoch": 0.38342470170912607,
+      "grad_norm": 1.5120738789251122,
+      "learning_rate": 0.00014134275011978606,
+      "loss": 1.3859,
+      "step": 8323
+    },
+    {
+      "epoch": 0.38347076979776107,
+      "grad_norm": 2.6472837077398865,
+      "learning_rate": 0.00014132916363979113,
+      "loss": 1.3226,
+      "step": 8324
+    },
+    {
+      "epoch": 0.38351683788639607,
+      "grad_norm": 1.424927744509352,
+      "learning_rate": 0.00014131557623967329,
+      "loss": 1.2245,
+      "step": 8325
+    },
+    {
+      "epoch": 0.3835629059750311,
+      "grad_norm": 1.0742438944925456,
+      "learning_rate": 0.00014130198791973493,
+      "loss": 1.501,
+      "step": 8326
+    },
+    {
+      "epoch": 0.3836089740636661,
+      "grad_norm": 1.7229923733784658,
+      "learning_rate": 0.00014128839868027864,
+      "loss": 1.3002,
+      "step": 8327
+    },
+    {
+      "epoch": 0.3836550421523011,
+      "grad_norm": 1.2534947052149779,
+      "learning_rate": 0.00014127480852160691,
+      "loss": 1.5168,
+      "step": 8328
+    },
+    {
+      "epoch": 0.3837011102409361,
+      "grad_norm": 1.3260812306468697,
+      "learning_rate": 0.0001412612174440224,
+      "loss": 1.283,
+      "step": 8329
+    },
+    {
+      "epoch": 0.3837471783295711,
+      "grad_norm": 1.0857215919840284,
+      "learning_rate": 0.0001412476254478276,
+      "loss": 1.4394,
+      "step": 8330
+    },
+    {
+      "epoch": 0.3837932464182061,
+      "grad_norm": 1.5241881473220629,
+      "learning_rate": 0.00014123403253332514,
+      "loss": 1.2006,
+      "step": 8331
+    },
+    {
+      "epoch": 0.3838393145068411,
+      "grad_norm": 1.5793275239029816,
+      "learning_rate": 0.00014122043870081765,
+      "loss": 1.5646,
+      "step": 8332
+    },
+    {
+      "epoch": 0.3838853825954761,
+      "grad_norm": 4.154413739530688,
+      "learning_rate": 0.00014120684395060778,
+      "loss": 1.3678,
+      "step": 8333
+    },
+    {
+      "epoch": 0.3839314506841111,
+      "grad_norm": 1.1033135796343891,
+      "learning_rate": 0.00014119324828299818,
+      "loss": 1.0837,
+      "step": 8334
+    },
+    {
+      "epoch": 0.3839775187727461,
+      "grad_norm": 2.070932020094797,
+      "learning_rate": 0.00014117965169829153,
+      "loss": 1.219,
+      "step": 8335
+    },
+    {
+      "epoch": 0.3840235868613811,
+      "grad_norm": 1.436658304054552,
+      "learning_rate": 0.00014116605419679055,
+      "loss": 1.3467,
+      "step": 8336
+    },
+    {
+      "epoch": 0.3840696549500161,
+      "grad_norm": 2.0971537758693706,
+      "learning_rate": 0.00014115245577879794,
+      "loss": 1.3307,
+      "step": 8337
+    },
+    {
+      "epoch": 0.3841157230386511,
+      "grad_norm": 2.1819825650787714,
+      "learning_rate": 0.0001411388564446165,
+      "loss": 1.2236,
+      "step": 8338
+    },
+    {
+      "epoch": 0.38416179112728616,
+      "grad_norm": 3.117599933029118,
+      "learning_rate": 0.00014112525619454895,
+      "loss": 1.0937,
+      "step": 8339
+    },
+    {
+      "epoch": 0.38420785921592115,
+      "grad_norm": 1.70961008919268,
+      "learning_rate": 0.00014111165502889806,
+      "loss": 1.4649,
+      "step": 8340
+    },
+    {
+      "epoch": 0.38425392730455615,
+      "grad_norm": 1.2432109729741596,
+      "learning_rate": 0.00014109805294796669,
+      "loss": 1.2249,
+      "step": 8341
+    },
+    {
+      "epoch": 0.38429999539319115,
+      "grad_norm": 1.692255566171062,
+      "learning_rate": 0.00014108444995205763,
+      "loss": 1.5716,
+      "step": 8342
+    },
+    {
+      "epoch": 0.38434606348182615,
+      "grad_norm": 0.998062633080742,
+      "learning_rate": 0.00014107084604147373,
+      "loss": 1.557,
+      "step": 8343
+    },
+    {
+      "epoch": 0.38439213157046115,
+      "grad_norm": 1.0585286430558882,
+      "learning_rate": 0.00014105724121651786,
+      "loss": 1.6162,
+      "step": 8344
+    },
+    {
+      "epoch": 0.38443819965909615,
+      "grad_norm": 1.6844476572537785,
+      "learning_rate": 0.00014104363547749293,
+      "loss": 1.3567,
+      "step": 8345
+    },
+    {
+      "epoch": 0.38448426774773115,
+      "grad_norm": 1.0861084211454237,
+      "learning_rate": 0.00014103002882470182,
+      "loss": 1.3685,
+      "step": 8346
+    },
+    {
+      "epoch": 0.38453033583636614,
+      "grad_norm": 1.717405470497314,
+      "learning_rate": 0.00014101642125844747,
+      "loss": 1.0952,
+      "step": 8347
+    },
+    {
+      "epoch": 0.38457640392500114,
+      "grad_norm": 1.4542960350775918,
+      "learning_rate": 0.00014100281277903281,
+      "loss": 1.3821,
+      "step": 8348
+    },
+    {
+      "epoch": 0.38462247201363614,
+      "grad_norm": 1.6493912805815312,
+      "learning_rate": 0.00014098920338676086,
+      "loss": 1.4804,
+      "step": 8349
+    },
+    {
+      "epoch": 0.38466854010227114,
+      "grad_norm": 1.787220571682267,
+      "learning_rate": 0.00014097559308193458,
+      "loss": 1.5167,
+      "step": 8350
+    },
+    {
+      "epoch": 0.38471460819090614,
+      "grad_norm": 1.904650240739698,
+      "learning_rate": 0.00014096198186485696,
+      "loss": 1.3649,
+      "step": 8351
+    },
+    {
+      "epoch": 0.38476067627954114,
+      "grad_norm": 1.29276336791618,
+      "learning_rate": 0.00014094836973583104,
+      "loss": 1.4489,
+      "step": 8352
+    },
+    {
+      "epoch": 0.3848067443681762,
+      "grad_norm": 1.149737736785709,
+      "learning_rate": 0.0001409347566951599,
+      "loss": 1.0849,
+      "step": 8353
+    },
+    {
+      "epoch": 0.3848528124568112,
+      "grad_norm": 2.6625174624647374,
+      "learning_rate": 0.00014092114274314658,
+      "loss": 1.3524,
+      "step": 8354
+    },
+    {
+      "epoch": 0.3848988805454462,
+      "grad_norm": 1.9261823044147386,
+      "learning_rate": 0.0001409075278800942,
+      "loss": 1.3776,
+      "step": 8355
+    },
+    {
+      "epoch": 0.3849449486340812,
+      "grad_norm": 2.0498051393320145,
+      "learning_rate": 0.00014089391210630583,
+      "loss": 1.4032,
+      "step": 8356
+    },
+    {
+      "epoch": 0.3849910167227162,
+      "grad_norm": 3.3308897552831622,
+      "learning_rate": 0.0001408802954220846,
+      "loss": 1.1503,
+      "step": 8357
+    },
+    {
+      "epoch": 0.3850370848113512,
+      "grad_norm": 1.0141191590055614,
+      "learning_rate": 0.00014086667782773376,
+      "loss": 1.2957,
+      "step": 8358
+    },
+    {
+      "epoch": 0.3850831528999862,
+      "grad_norm": 1.3440407120665117,
+      "learning_rate": 0.00014085305932355635,
+      "loss": 1.272,
+      "step": 8359
+    },
+    {
+      "epoch": 0.3851292209886212,
+      "grad_norm": 2.880153872003922,
+      "learning_rate": 0.00014083943990985564,
+      "loss": 1.591,
+      "step": 8360
+    },
+    {
+      "epoch": 0.3851752890772562,
+      "grad_norm": 1.1930316986650595,
+      "learning_rate": 0.00014082581958693483,
+      "loss": 1.2601,
+      "step": 8361
+    },
+    {
+      "epoch": 0.3852213571658912,
+      "grad_norm": 2.8425642930525066,
+      "learning_rate": 0.00014081219835509716,
+      "loss": 1.3339,
+      "step": 8362
+    },
+    {
+      "epoch": 0.3852674252545262,
+      "grad_norm": 1.6292831358711823,
+      "learning_rate": 0.00014079857621464585,
+      "loss": 1.3988,
+      "step": 8363
+    },
+    {
+      "epoch": 0.3853134933431612,
+      "grad_norm": 0.9907730836669165,
+      "learning_rate": 0.0001407849531658842,
+      "loss": 1.3582,
+      "step": 8364
+    },
+    {
+      "epoch": 0.38535956143179617,
+      "grad_norm": 2.509390011099579,
+      "learning_rate": 0.0001407713292091155,
+      "loss": 1.2889,
+      "step": 8365
+    },
+    {
+      "epoch": 0.38540562952043117,
+      "grad_norm": 1.5232721703845744,
+      "learning_rate": 0.00014075770434464306,
+      "loss": 1.2146,
+      "step": 8366
+    },
+    {
+      "epoch": 0.3854516976090662,
+      "grad_norm": 1.9348569933509752,
+      "learning_rate": 0.00014074407857277022,
+      "loss": 1.1598,
+      "step": 8367
+    },
+    {
+      "epoch": 0.3854977656977012,
+      "grad_norm": 2.6383848073451093,
+      "learning_rate": 0.0001407304518938003,
+      "loss": 1.1014,
+      "step": 8368
+    },
+    {
+      "epoch": 0.3855438337863362,
+      "grad_norm": 1.0754471779140446,
+      "learning_rate": 0.00014071682430803674,
+      "loss": 1.6397,
+      "step": 8369
+    },
+    {
+      "epoch": 0.3855899018749712,
+      "grad_norm": 3.7688025302560404,
+      "learning_rate": 0.0001407031958157829,
+      "loss": 1.1033,
+      "step": 8370
+    },
+    {
+      "epoch": 0.3856359699636062,
+      "grad_norm": 0.9700913412853929,
+      "learning_rate": 0.00014068956641734222,
+      "loss": 1.7221,
+      "step": 8371
+    },
+    {
+      "epoch": 0.3856820380522412,
+      "grad_norm": 1.5207676859386345,
+      "learning_rate": 0.0001406759361130181,
+      "loss": 1.5,
+      "step": 8372
+    },
+    {
+      "epoch": 0.3857281061408762,
+      "grad_norm": 2.033285084738011,
+      "learning_rate": 0.000140662304903114,
+      "loss": 1.257,
+      "step": 8373
+    },
+    {
+      "epoch": 0.3857741742295112,
+      "grad_norm": 11.159825836287935,
+      "learning_rate": 0.00014064867278793336,
+      "loss": 1.3557,
+      "step": 8374
+    },
+    {
+      "epoch": 0.3858202423181462,
+      "grad_norm": 2.667264581016671,
+      "learning_rate": 0.00014063503976777977,
+      "loss": 1.4805,
+      "step": 8375
+    },
+    {
+      "epoch": 0.3858663104067812,
+      "grad_norm": 2.055100797154295,
+      "learning_rate": 0.00014062140584295666,
+      "loss": 1.2786,
+      "step": 8376
+    },
+    {
+      "epoch": 0.3859123784954162,
+      "grad_norm": 1.7677770018002017,
+      "learning_rate": 0.00014060777101376762,
+      "loss": 1.3689,
+      "step": 8377
+    },
+    {
+      "epoch": 0.3859584465840512,
+      "grad_norm": 2.2810300047981635,
+      "learning_rate": 0.0001405941352805162,
+      "loss": 1.5538,
+      "step": 8378
+    },
+    {
+      "epoch": 0.3860045146726862,
+      "grad_norm": 1.184112905261385,
+      "learning_rate": 0.00014058049864350592,
+      "loss": 1.3415,
+      "step": 8379
+    },
+    {
+      "epoch": 0.38605058276132126,
+      "grad_norm": 5.428703666110903,
+      "learning_rate": 0.00014056686110304047,
+      "loss": 1.0476,
+      "step": 8380
+    },
+    {
+      "epoch": 0.38609665084995626,
+      "grad_norm": 1.6294619889506095,
+      "learning_rate": 0.00014055322265942336,
+      "loss": 1.395,
+      "step": 8381
+    },
+    {
+      "epoch": 0.38614271893859126,
+      "grad_norm": 1.2958100155518388,
+      "learning_rate": 0.00014053958331295834,
+      "loss": 1.4984,
+      "step": 8382
+    },
+    {
+      "epoch": 0.38618878702722625,
+      "grad_norm": 1.394332126757288,
+      "learning_rate": 0.00014052594306394894,
+      "loss": 1.3789,
+      "step": 8383
+    },
+    {
+      "epoch": 0.38623485511586125,
+      "grad_norm": 3.8013121773377194,
+      "learning_rate": 0.00014051230191269897,
+      "loss": 1.6131,
+      "step": 8384
+    },
+    {
+      "epoch": 0.38628092320449625,
+      "grad_norm": 2.574260172044585,
+      "learning_rate": 0.00014049865985951202,
+      "loss": 1.2593,
+      "step": 8385
+    },
+    {
+      "epoch": 0.38632699129313125,
+      "grad_norm": 1.190204995036209,
+      "learning_rate": 0.00014048501690469187,
+      "loss": 1.3577,
+      "step": 8386
+    },
+    {
+      "epoch": 0.38637305938176625,
+      "grad_norm": 1.806954585133878,
+      "learning_rate": 0.0001404713730485422,
+      "loss": 1.2683,
+      "step": 8387
+    },
+    {
+      "epoch": 0.38641912747040125,
+      "grad_norm": 5.715225267158835,
+      "learning_rate": 0.00014045772829136686,
+      "loss": 1.1313,
+      "step": 8388
+    },
+    {
+      "epoch": 0.38646519555903625,
+      "grad_norm": 1.605724289560926,
+      "learning_rate": 0.00014044408263346955,
+      "loss": 1.541,
+      "step": 8389
+    },
+    {
+      "epoch": 0.38651126364767124,
+      "grad_norm": 3.3928909424953964,
+      "learning_rate": 0.00014043043607515405,
+      "loss": 1.5098,
+      "step": 8390
+    },
+    {
+      "epoch": 0.38655733173630624,
+      "grad_norm": 2.1842737369340464,
+      "learning_rate": 0.00014041678861672426,
+      "loss": 1.4178,
+      "step": 8391
+    },
+    {
+      "epoch": 0.38660339982494124,
+      "grad_norm": 2.38971536275037,
+      "learning_rate": 0.00014040314025848394,
+      "loss": 1.3353,
+      "step": 8392
+    },
+    {
+      "epoch": 0.38664946791357624,
+      "grad_norm": 1.9309489453869035,
+      "learning_rate": 0.000140389491000737,
+      "loss": 1.2856,
+      "step": 8393
+    },
+    {
+      "epoch": 0.3866955360022113,
+      "grad_norm": 2.739161810248552,
+      "learning_rate": 0.00014037584084378727,
+      "loss": 1.3129,
+      "step": 8394
+    },
+    {
+      "epoch": 0.3867416040908463,
+      "grad_norm": 2.3014963597024525,
+      "learning_rate": 0.00014036218978793872,
+      "loss": 1.056,
+      "step": 8395
+    },
+    {
+      "epoch": 0.3867876721794813,
+      "grad_norm": 2.8897978950140906,
+      "learning_rate": 0.00014034853783349515,
+      "loss": 1.3317,
+      "step": 8396
+    },
+    {
+      "epoch": 0.3868337402681163,
+      "grad_norm": 0.9957395121361684,
+      "learning_rate": 0.00014033488498076065,
+      "loss": 1.5659,
+      "step": 8397
+    },
+    {
+      "epoch": 0.3868798083567513,
+      "grad_norm": 2.1199960481089866,
+      "learning_rate": 0.00014032123123003905,
+      "loss": 1.3279,
+      "step": 8398
+    },
+    {
+      "epoch": 0.3869258764453863,
+      "grad_norm": 2.3141180340002663,
+      "learning_rate": 0.00014030757658163435,
+      "loss": 1.6095,
+      "step": 8399
+    },
+    {
+      "epoch": 0.3869719445340213,
+      "grad_norm": 2.882176264497792,
+      "learning_rate": 0.0001402939210358506,
+      "loss": 1.193,
+      "step": 8400
+    },
+    {
+      "epoch": 0.3870180126226563,
+      "grad_norm": 1.1248682024887597,
+      "learning_rate": 0.00014028026459299178,
+      "loss": 1.3803,
+      "step": 8401
+    },
+    {
+      "epoch": 0.3870640807112913,
+      "grad_norm": 2.390565307228673,
+      "learning_rate": 0.00014026660725336196,
+      "loss": 1.2835,
+      "step": 8402
+    },
+    {
+      "epoch": 0.3871101487999263,
+      "grad_norm": 1.2606597098759493,
+      "learning_rate": 0.00014025294901726513,
+      "loss": 1.4847,
+      "step": 8403
+    },
+    {
+      "epoch": 0.3871562168885613,
+      "grad_norm": 2.609560287331329,
+      "learning_rate": 0.00014023928988500544,
+      "loss": 1.3705,
+      "step": 8404
+    },
+    {
+      "epoch": 0.3872022849771963,
+      "grad_norm": 1.177071505798566,
+      "learning_rate": 0.00014022562985688695,
+      "loss": 1.3874,
+      "step": 8405
+    },
+    {
+      "epoch": 0.3872483530658313,
+      "grad_norm": 1.7337286758507253,
+      "learning_rate": 0.00014021196893321378,
+      "loss": 1.603,
+      "step": 8406
+    },
+    {
+      "epoch": 0.38729442115446633,
+      "grad_norm": 2.539505156045141,
+      "learning_rate": 0.00014019830711429007,
+      "loss": 1.3734,
+      "step": 8407
+    },
+    {
+      "epoch": 0.3873404892431013,
+      "grad_norm": 2.9229195555957967,
+      "learning_rate": 0.00014018464440042,
+      "loss": 1.1048,
+      "step": 8408
+    },
+    {
+      "epoch": 0.3873865573317363,
+      "grad_norm": 1.0990935982112227,
+      "learning_rate": 0.00014017098079190769,
+      "loss": 1.2627,
+      "step": 8409
+    },
+    {
+      "epoch": 0.3874326254203713,
+      "grad_norm": 2.6103636105478367,
+      "learning_rate": 0.00014015731628905738,
+      "loss": 1.3418,
+      "step": 8410
+    },
+    {
+      "epoch": 0.3874786935090063,
+      "grad_norm": 5.445004900782183,
+      "learning_rate": 0.00014014365089217327,
+      "loss": 1.4407,
+      "step": 8411
+    },
+    {
+      "epoch": 0.3875247615976413,
+      "grad_norm": 5.541804921247628,
+      "learning_rate": 0.0001401299846015596,
+      "loss": 1.3391,
+      "step": 8412
+    },
+    {
+      "epoch": 0.3875708296862763,
+      "grad_norm": 1.847408940180611,
+      "learning_rate": 0.00014011631741752068,
+      "loss": 1.2697,
+      "step": 8413
+    },
+    {
+      "epoch": 0.3876168977749113,
+      "grad_norm": 1.708573315284923,
+      "learning_rate": 0.0001401026493403607,
+      "loss": 1.4176,
+      "step": 8414
+    },
+    {
+      "epoch": 0.3876629658635463,
+      "grad_norm": 1.933736571082727,
+      "learning_rate": 0.00014008898037038399,
+      "loss": 1.23,
+      "step": 8415
+    },
+    {
+      "epoch": 0.3877090339521813,
+      "grad_norm": 1.40235324841737,
+      "learning_rate": 0.00014007531050789487,
+      "loss": 1.7215,
+      "step": 8416
+    },
+    {
+      "epoch": 0.3877551020408163,
+      "grad_norm": 1.1954833238921863,
+      "learning_rate": 0.0001400616397531977,
+      "loss": 1.4586,
+      "step": 8417
+    },
+    {
+      "epoch": 0.3878011701294513,
+      "grad_norm": 1.7268222102743849,
+      "learning_rate": 0.00014004796810659678,
+      "loss": 1.5485,
+      "step": 8418
+    },
+    {
+      "epoch": 0.3878472382180863,
+      "grad_norm": 1.8426428589771855,
+      "learning_rate": 0.00014003429556839656,
+      "loss": 1.4741,
+      "step": 8419
+    },
+    {
+      "epoch": 0.3878933063067213,
+      "grad_norm": 1.3962740711628168,
+      "learning_rate": 0.00014002062213890135,
+      "loss": 1.3658,
+      "step": 8420
+    },
+    {
+      "epoch": 0.38793937439535636,
+      "grad_norm": 1.397375165753805,
+      "learning_rate": 0.00014000694781841563,
+      "loss": 1.5073,
+      "step": 8421
+    },
+    {
+      "epoch": 0.38798544248399136,
+      "grad_norm": 2.1327452084957508,
+      "learning_rate": 0.00013999327260724378,
+      "loss": 1.3711,
+      "step": 8422
+    },
+    {
+      "epoch": 0.38803151057262636,
+      "grad_norm": 1.9580889750364479,
+      "learning_rate": 0.00013997959650569034,
+      "loss": 1.1211,
+      "step": 8423
+    },
+    {
+      "epoch": 0.38807757866126136,
+      "grad_norm": 1.3331576430136878,
+      "learning_rate": 0.0001399659195140597,
+      "loss": 1.3828,
+      "step": 8424
+    },
+    {
+      "epoch": 0.38812364674989636,
+      "grad_norm": 1.3316231101420353,
+      "learning_rate": 0.0001399522416326564,
+      "loss": 1.2587,
+      "step": 8425
+    },
+    {
+      "epoch": 0.38816971483853135,
+      "grad_norm": 1.3286411787373937,
+      "learning_rate": 0.00013993856286178497,
+      "loss": 1.1567,
+      "step": 8426
+    },
+    {
+      "epoch": 0.38821578292716635,
+      "grad_norm": 0.9952131395084921,
+      "learning_rate": 0.0001399248832017499,
+      "loss": 1.3811,
+      "step": 8427
+    },
+    {
+      "epoch": 0.38826185101580135,
+      "grad_norm": 1.1815561812869868,
+      "learning_rate": 0.00013991120265285573,
+      "loss": 1.5589,
+      "step": 8428
+    },
+    {
+      "epoch": 0.38830791910443635,
+      "grad_norm": 1.912933224394773,
+      "learning_rate": 0.0001398975212154071,
+      "loss": 1.3634,
+      "step": 8429
+    },
+    {
+      "epoch": 0.38835398719307135,
+      "grad_norm": 1.2273912266387397,
+      "learning_rate": 0.00013988383888970858,
+      "loss": 1.3772,
+      "step": 8430
+    },
+    {
+      "epoch": 0.38840005528170635,
+      "grad_norm": 2.190709271218192,
+      "learning_rate": 0.0001398701556760647,
+      "loss": 1.5018,
+      "step": 8431
+    },
+    {
+      "epoch": 0.38844612337034135,
+      "grad_norm": 1.6949825013449349,
+      "learning_rate": 0.00013985647157478025,
+      "loss": 1.4189,
+      "step": 8432
+    },
+    {
+      "epoch": 0.38849219145897634,
+      "grad_norm": 8.092600144781775,
+      "learning_rate": 0.00013984278658615975,
+      "loss": 1.3495,
+      "step": 8433
+    },
+    {
+      "epoch": 0.38853825954761134,
+      "grad_norm": 2.814473977810638,
+      "learning_rate": 0.00013982910071050797,
+      "loss": 1.4537,
+      "step": 8434
+    },
+    {
+      "epoch": 0.3885843276362464,
+      "grad_norm": 1.515589443102626,
+      "learning_rate": 0.00013981541394812955,
+      "loss": 1.2184,
+      "step": 8435
+    },
+    {
+      "epoch": 0.3886303957248814,
+      "grad_norm": 1.3445780322393104,
+      "learning_rate": 0.00013980172629932916,
+      "loss": 1.4203,
+      "step": 8436
+    },
+    {
+      "epoch": 0.3886764638135164,
+      "grad_norm": 1.1776639752093303,
+      "learning_rate": 0.00013978803776441162,
+      "loss": 1.689,
+      "step": 8437
+    },
+    {
+      "epoch": 0.3887225319021514,
+      "grad_norm": 1.6719773725820706,
+      "learning_rate": 0.0001397743483436816,
+      "loss": 1.4179,
+      "step": 8438
+    },
+    {
+      "epoch": 0.3887685999907864,
+      "grad_norm": 2.844055555204671,
+      "learning_rate": 0.00013976065803744393,
+      "loss": 1.3821,
+      "step": 8439
+    },
+    {
+      "epoch": 0.3888146680794214,
+      "grad_norm": 1.3032842337083392,
+      "learning_rate": 0.00013974696684600338,
+      "loss": 1.3695,
+      "step": 8440
+    },
+    {
+      "epoch": 0.3888607361680564,
+      "grad_norm": 1.3886356306328114,
+      "learning_rate": 0.0001397332747696648,
+      "loss": 1.461,
+      "step": 8441
+    },
+    {
+      "epoch": 0.3889068042566914,
+      "grad_norm": 1.7027126925044849,
+      "learning_rate": 0.00013971958180873295,
+      "loss": 1.4823,
+      "step": 8442
+    },
+    {
+      "epoch": 0.3889528723453264,
+      "grad_norm": 1.1459747768826678,
+      "learning_rate": 0.00013970588796351273,
+      "loss": 1.5709,
+      "step": 8443
+    },
+    {
+      "epoch": 0.3889989404339614,
+      "grad_norm": 1.334171951490066,
+      "learning_rate": 0.000139692193234309,
+      "loss": 1.1654,
+      "step": 8444
+    },
+    {
+      "epoch": 0.3890450085225964,
+      "grad_norm": 1.2246907520695551,
+      "learning_rate": 0.0001396784976214266,
+      "loss": 1.7421,
+      "step": 8445
+    },
+    {
+      "epoch": 0.3890910766112314,
+      "grad_norm": 1.893546692934609,
+      "learning_rate": 0.00013966480112517055,
+      "loss": 1.2599,
+      "step": 8446
+    },
+    {
+      "epoch": 0.3891371446998664,
+      "grad_norm": 2.564967750102403,
+      "learning_rate": 0.00013965110374584568,
+      "loss": 1.2945,
+      "step": 8447
+    },
+    {
+      "epoch": 0.38918321278850143,
+      "grad_norm": 1.2890005446257962,
+      "learning_rate": 0.000139637405483757,
+      "loss": 1.4484,
+      "step": 8448
+    },
+    {
+      "epoch": 0.38922928087713643,
+      "grad_norm": 1.4215273602176737,
+      "learning_rate": 0.0001396237063392094,
+      "loss": 1.5951,
+      "step": 8449
+    },
+    {
+      "epoch": 0.38927534896577143,
+      "grad_norm": 1.4052204060707145,
+      "learning_rate": 0.00013961000631250794,
+      "loss": 1.3137,
+      "step": 8450
+    },
+    {
+      "epoch": 0.3893214170544064,
+      "grad_norm": 2.376055599312175,
+      "learning_rate": 0.0001395963054039576,
+      "loss": 1.0283,
+      "step": 8451
+    },
+    {
+      "epoch": 0.3893674851430414,
+      "grad_norm": 1.9835636539108752,
+      "learning_rate": 0.00013958260361386345,
+      "loss": 1.327,
+      "step": 8452
+    },
+    {
+      "epoch": 0.3894135532316764,
+      "grad_norm": 1.6615583369689233,
+      "learning_rate": 0.00013956890094253048,
+      "loss": 1.0569,
+      "step": 8453
+    },
+    {
+      "epoch": 0.3894596213203114,
+      "grad_norm": 1.7712386608752775,
+      "learning_rate": 0.00013955519739026378,
+      "loss": 1.2054,
+      "step": 8454
+    },
+    {
+      "epoch": 0.3895056894089464,
+      "grad_norm": 1.6814587963939074,
+      "learning_rate": 0.00013954149295736842,
+      "loss": 1.3659,
+      "step": 8455
+    },
+    {
+      "epoch": 0.3895517574975814,
+      "grad_norm": 1.3614363353168257,
+      "learning_rate": 0.00013952778764414952,
+      "loss": 1.4157,
+      "step": 8456
+    },
+    {
+      "epoch": 0.3895978255862164,
+      "grad_norm": 3.1352845757548797,
+      "learning_rate": 0.0001395140814509122,
+      "loss": 1.1064,
+      "step": 8457
+    },
+    {
+      "epoch": 0.3896438936748514,
+      "grad_norm": 1.277968009362131,
+      "learning_rate": 0.00013950037437796163,
+      "loss": 1.3973,
+      "step": 8458
+    },
+    {
+      "epoch": 0.3896899617634864,
+      "grad_norm": 3.500945579416599,
+      "learning_rate": 0.00013948666642560293,
+      "loss": 0.8922,
+      "step": 8459
+    },
+    {
+      "epoch": 0.3897360298521214,
+      "grad_norm": 1.4007879905338214,
+      "learning_rate": 0.0001394729575941413,
+      "loss": 1.6152,
+      "step": 8460
+    },
+    {
+      "epoch": 0.3897820979407564,
+      "grad_norm": 1.033846595837073,
+      "learning_rate": 0.000139459247883882,
+      "loss": 1.4511,
+      "step": 8461
+    },
+    {
+      "epoch": 0.38982816602939147,
+      "grad_norm": 1.2965255682562888,
+      "learning_rate": 0.00013944553729513016,
+      "loss": 1.6283,
+      "step": 8462
+    },
+    {
+      "epoch": 0.38987423411802646,
+      "grad_norm": 1.4444235354003623,
+      "learning_rate": 0.00013943182582819107,
+      "loss": 1.2676,
+      "step": 8463
+    },
+    {
+      "epoch": 0.38992030220666146,
+      "grad_norm": 1.412421445685047,
+      "learning_rate": 0.00013941811348336998,
+      "loss": 1.21,
+      "step": 8464
+    },
+    {
+      "epoch": 0.38996637029529646,
+      "grad_norm": 5.743333267624754,
+      "learning_rate": 0.00013940440026097224,
+      "loss": 1.154,
+      "step": 8465
+    },
+    {
+      "epoch": 0.39001243838393146,
+      "grad_norm": 1.0674609763215046,
+      "learning_rate": 0.00013939068616130302,
+      "loss": 1.3485,
+      "step": 8466
+    },
+    {
+      "epoch": 0.39005850647256646,
+      "grad_norm": 1.199455999275395,
+      "learning_rate": 0.00013937697118466774,
+      "loss": 1.0655,
+      "step": 8467
+    },
+    {
+      "epoch": 0.39010457456120146,
+      "grad_norm": 1.349178612303766,
+      "learning_rate": 0.0001393632553313717,
+      "loss": 1.1013,
+      "step": 8468
+    },
+    {
+      "epoch": 0.39015064264983645,
+      "grad_norm": 2.352789288794723,
+      "learning_rate": 0.00013934953860172028,
+      "loss": 1.4844,
+      "step": 8469
+    },
+    {
+      "epoch": 0.39019671073847145,
+      "grad_norm": 1.3233217699546194,
+      "learning_rate": 0.0001393358209960189,
+      "loss": 1.3213,
+      "step": 8470
+    },
+    {
+      "epoch": 0.39024277882710645,
+      "grad_norm": 1.478043338387027,
+      "learning_rate": 0.00013932210251457286,
+      "loss": 1.2437,
+      "step": 8471
+    },
+    {
+      "epoch": 0.39028884691574145,
+      "grad_norm": 1.4037793080459267,
+      "learning_rate": 0.00013930838315768767,
+      "loss": 1.3476,
+      "step": 8472
+    },
+    {
+      "epoch": 0.39033491500437645,
+      "grad_norm": 2.840438628537937,
+      "learning_rate": 0.0001392946629256687,
+      "loss": 1.3987,
+      "step": 8473
+    },
+    {
+      "epoch": 0.39038098309301145,
+      "grad_norm": 1.281280125200011,
+      "learning_rate": 0.00013928094181882143,
+      "loss": 1.364,
+      "step": 8474
+    },
+    {
+      "epoch": 0.3904270511816465,
+      "grad_norm": 0.9129271455102456,
+      "learning_rate": 0.00013926721983745135,
+      "loss": 1.5503,
+      "step": 8475
+    },
+    {
+      "epoch": 0.3904731192702815,
+      "grad_norm": 1.215346714244138,
+      "learning_rate": 0.00013925349698186396,
+      "loss": 1.6073,
+      "step": 8476
+    },
+    {
+      "epoch": 0.3905191873589165,
+      "grad_norm": 1.9027684137702416,
+      "learning_rate": 0.00013923977325236474,
+      "loss": 1.2629,
+      "step": 8477
+    },
+    {
+      "epoch": 0.3905652554475515,
+      "grad_norm": 1.4613708910065044,
+      "learning_rate": 0.0001392260486492593,
+      "loss": 1.1128,
+      "step": 8478
+    },
+    {
+      "epoch": 0.3906113235361865,
+      "grad_norm": 1.4327525024388859,
+      "learning_rate": 0.00013921232317285306,
+      "loss": 1.3176,
+      "step": 8479
+    },
+    {
+      "epoch": 0.3906573916248215,
+      "grad_norm": 3.1280861095711945,
+      "learning_rate": 0.00013919859682345173,
+      "loss": 1.3593,
+      "step": 8480
+    },
+    {
+      "epoch": 0.3907034597134565,
+      "grad_norm": 2.1948882914225454,
+      "learning_rate": 0.00013918486960136088,
+      "loss": 1.3165,
+      "step": 8481
+    },
+    {
+      "epoch": 0.3907495278020915,
+      "grad_norm": 1.6198405443731294,
+      "learning_rate": 0.00013917114150688607,
+      "loss": 1.3936,
+      "step": 8482
+    },
+    {
+      "epoch": 0.3907955958907265,
+      "grad_norm": 2.5846505520221332,
+      "learning_rate": 0.00013915741254033297,
+      "loss": 1.1442,
+      "step": 8483
+    },
+    {
+      "epoch": 0.3908416639793615,
+      "grad_norm": 1.2948744415264384,
+      "learning_rate": 0.0001391436827020072,
+      "loss": 1.5135,
+      "step": 8484
+    },
+    {
+      "epoch": 0.3908877320679965,
+      "grad_norm": 1.548831881837667,
+      "learning_rate": 0.00013912995199221446,
+      "loss": 1.3903,
+      "step": 8485
+    },
+    {
+      "epoch": 0.3909338001566315,
+      "grad_norm": 1.03710971943379,
+      "learning_rate": 0.00013911622041126042,
+      "loss": 1.5114,
+      "step": 8486
+    },
+    {
+      "epoch": 0.3909798682452665,
+      "grad_norm": 2.1344826455869046,
+      "learning_rate": 0.0001391024879594508,
+      "loss": 1.5764,
+      "step": 8487
+    },
+    {
+      "epoch": 0.3910259363339015,
+      "grad_norm": 1.1456578149279293,
+      "learning_rate": 0.00013908875463709136,
+      "loss": 1.1982,
+      "step": 8488
+    },
+    {
+      "epoch": 0.39107200442253653,
+      "grad_norm": 1.4110247063690873,
+      "learning_rate": 0.00013907502044448782,
+      "loss": 1.6268,
+      "step": 8489
+    },
+    {
+      "epoch": 0.39111807251117153,
+      "grad_norm": 1.3128344403636443,
+      "learning_rate": 0.00013906128538194594,
+      "loss": 1.4103,
+      "step": 8490
+    },
+    {
+      "epoch": 0.39116414059980653,
+      "grad_norm": 1.62070275245916,
+      "learning_rate": 0.0001390475494497715,
+      "loss": 1.1535,
+      "step": 8491
+    },
+    {
+      "epoch": 0.39121020868844153,
+      "grad_norm": 1.6253434119344559,
+      "learning_rate": 0.00013903381264827037,
+      "loss": 1.295,
+      "step": 8492
+    },
+    {
+      "epoch": 0.39125627677707653,
+      "grad_norm": 1.3296756138446442,
+      "learning_rate": 0.00013902007497774827,
+      "loss": 1.3517,
+      "step": 8493
+    },
+    {
+      "epoch": 0.3913023448657115,
+      "grad_norm": 1.406016574947095,
+      "learning_rate": 0.00013900633643851114,
+      "loss": 1.609,
+      "step": 8494
+    },
+    {
+      "epoch": 0.3913484129543465,
+      "grad_norm": 2.8118153468657474,
+      "learning_rate": 0.0001389925970308648,
+      "loss": 1.3689,
+      "step": 8495
+    },
+    {
+      "epoch": 0.3913944810429815,
+      "grad_norm": 1.242003088791439,
+      "learning_rate": 0.00013897885675511515,
+      "loss": 1.2896,
+      "step": 8496
+    },
+    {
+      "epoch": 0.3914405491316165,
+      "grad_norm": 1.745389575443586,
+      "learning_rate": 0.00013896511561156807,
+      "loss": 1.1767,
+      "step": 8497
+    },
+    {
+      "epoch": 0.3914866172202515,
+      "grad_norm": 1.3217820027531737,
+      "learning_rate": 0.00013895137360052953,
+      "loss": 1.1412,
+      "step": 8498
+    },
+    {
+      "epoch": 0.3915326853088865,
+      "grad_norm": 1.010808593372072,
+      "learning_rate": 0.00013893763072230547,
+      "loss": 1.3363,
+      "step": 8499
+    },
+    {
+      "epoch": 0.3915787533975215,
+      "grad_norm": 1.1387936198399955,
+      "learning_rate": 0.00013892388697720175,
+      "loss": 1.7089,
+      "step": 8500
+    },
+    {
+      "epoch": 0.3916248214861565,
+      "grad_norm": 1.398807361060656,
+      "learning_rate": 0.0001389101423655245,
+      "loss": 1.2402,
+      "step": 8501
+    },
+    {
+      "epoch": 0.3916708895747915,
+      "grad_norm": 3.160029121173083,
+      "learning_rate": 0.00013889639688757958,
+      "loss": 1.4811,
+      "step": 8502
+    },
+    {
+      "epoch": 0.39171695766342657,
+      "grad_norm": 1.2208075946516181,
+      "learning_rate": 0.0001388826505436731,
+      "loss": 1.5718,
+      "step": 8503
+    },
+    {
+      "epoch": 0.39176302575206157,
+      "grad_norm": 1.5756351320348028,
+      "learning_rate": 0.0001388689033341111,
+      "loss": 1.0309,
+      "step": 8504
+    },
+    {
+      "epoch": 0.39180909384069657,
+      "grad_norm": 2.8039380370511973,
+      "learning_rate": 0.00013885515525919958,
+      "loss": 1.561,
+      "step": 8505
+    },
+    {
+      "epoch": 0.39185516192933156,
+      "grad_norm": 1.7549039540859543,
+      "learning_rate": 0.00013884140631924467,
+      "loss": 1.721,
+      "step": 8506
+    },
+    {
+      "epoch": 0.39190123001796656,
+      "grad_norm": 1.6328527875982703,
+      "learning_rate": 0.00013882765651455244,
+      "loss": 1.1874,
+      "step": 8507
+    },
+    {
+      "epoch": 0.39194729810660156,
+      "grad_norm": 1.62877459622416,
+      "learning_rate": 0.00013881390584542905,
+      "loss": 1.4793,
+      "step": 8508
+    },
+    {
+      "epoch": 0.39199336619523656,
+      "grad_norm": 2.007082556914734,
+      "learning_rate": 0.00013880015431218053,
+      "loss": 1.2683,
+      "step": 8509
+    },
+    {
+      "epoch": 0.39203943428387156,
+      "grad_norm": 0.905618538707775,
+      "learning_rate": 0.00013878640191511315,
+      "loss": 1.4638,
+      "step": 8510
+    },
+    {
+      "epoch": 0.39208550237250656,
+      "grad_norm": 2.1044901836808614,
+      "learning_rate": 0.00013877264865453303,
+      "loss": 1.4184,
+      "step": 8511
+    },
+    {
+      "epoch": 0.39213157046114155,
+      "grad_norm": 1.060334504557824,
+      "learning_rate": 0.00013875889453074638,
+      "loss": 1.4025,
+      "step": 8512
+    },
+    {
+      "epoch": 0.39217763854977655,
+      "grad_norm": 1.8201159055273959,
+      "learning_rate": 0.00013874513954405937,
+      "loss": 1.2488,
+      "step": 8513
+    },
+    {
+      "epoch": 0.39222370663841155,
+      "grad_norm": 1.8133747814706824,
+      "learning_rate": 0.00013873138369477827,
+      "loss": 1.5283,
+      "step": 8514
+    },
+    {
+      "epoch": 0.39226977472704655,
+      "grad_norm": 3.1271757020189326,
+      "learning_rate": 0.0001387176269832093,
+      "loss": 1.4816,
+      "step": 8515
+    },
+    {
+      "epoch": 0.3923158428156816,
+      "grad_norm": 1.2364353586085872,
+      "learning_rate": 0.0001387038694096588,
+      "loss": 1.5355,
+      "step": 8516
+    },
+    {
+      "epoch": 0.3923619109043166,
+      "grad_norm": 2.6481049041366527,
+      "learning_rate": 0.000138690110974433,
+      "loss": 1.2785,
+      "step": 8517
+    },
+    {
+      "epoch": 0.3924079789929516,
+      "grad_norm": 2.806307041042768,
+      "learning_rate": 0.0001386763516778382,
+      "loss": 1.2318,
+      "step": 8518
+    },
+    {
+      "epoch": 0.3924540470815866,
+      "grad_norm": 2.613589692745652,
+      "learning_rate": 0.00013866259152018075,
+      "loss": 1.566,
+      "step": 8519
+    },
+    {
+      "epoch": 0.3925001151702216,
+      "grad_norm": 1.0996307447917786,
+      "learning_rate": 0.00013864883050176703,
+      "loss": 1.6247,
+      "step": 8520
+    },
+    {
+      "epoch": 0.3925461832588566,
+      "grad_norm": 0.9629449527546962,
+      "learning_rate": 0.0001386350686229033,
+      "loss": 1.3179,
+      "step": 8521
+    },
+    {
+      "epoch": 0.3925922513474916,
+      "grad_norm": 1.5285500081059453,
+      "learning_rate": 0.00013862130588389603,
+      "loss": 1.3953,
+      "step": 8522
+    },
+    {
+      "epoch": 0.3926383194361266,
+      "grad_norm": 1.6163852541709893,
+      "learning_rate": 0.00013860754228505164,
+      "loss": 1.3196,
+      "step": 8523
+    },
+    {
+      "epoch": 0.3926843875247616,
+      "grad_norm": 3.709923557424389,
+      "learning_rate": 0.00013859377782667647,
+      "loss": 1.2187,
+      "step": 8524
+    },
+    {
+      "epoch": 0.3927304556133966,
+      "grad_norm": 2.5828744142310556,
+      "learning_rate": 0.00013858001250907707,
+      "loss": 1.3447,
+      "step": 8525
+    },
+    {
+      "epoch": 0.3927765237020316,
+      "grad_norm": 1.885238171714547,
+      "learning_rate": 0.0001385662463325598,
+      "loss": 1.4352,
+      "step": 8526
+    },
+    {
+      "epoch": 0.3928225917906666,
+      "grad_norm": 2.7899638583869857,
+      "learning_rate": 0.00013855247929743124,
+      "loss": 1.3518,
+      "step": 8527
+    },
+    {
+      "epoch": 0.3928686598793016,
+      "grad_norm": 1.7229642575434643,
+      "learning_rate": 0.00013853871140399776,
+      "loss": 1.2948,
+      "step": 8528
+    },
+    {
+      "epoch": 0.3929147279679366,
+      "grad_norm": 1.3025170732649523,
+      "learning_rate": 0.00013852494265256598,
+      "loss": 1.0867,
+      "step": 8529
+    },
+    {
+      "epoch": 0.39296079605657164,
+      "grad_norm": 1.3447222672226051,
+      "learning_rate": 0.00013851117304344238,
+      "loss": 1.2592,
+      "step": 8530
+    },
+    {
+      "epoch": 0.39300686414520664,
+      "grad_norm": 2.1130807224718913,
+      "learning_rate": 0.00013849740257693358,
+      "loss": 1.4927,
+      "step": 8531
+    },
+    {
+      "epoch": 0.39305293223384163,
+      "grad_norm": 1.5584812507716148,
+      "learning_rate": 0.0001384836312533461,
+      "loss": 1.4763,
+      "step": 8532
+    },
+    {
+      "epoch": 0.39309900032247663,
+      "grad_norm": 4.436151564170309,
+      "learning_rate": 0.00013846985907298658,
+      "loss": 1.3402,
+      "step": 8533
+    },
+    {
+      "epoch": 0.39314506841111163,
+      "grad_norm": 0.9174663954296852,
+      "learning_rate": 0.00013845608603616158,
+      "loss": 1.3561,
+      "step": 8534
+    },
+    {
+      "epoch": 0.39319113649974663,
+      "grad_norm": 2.4330603003348164,
+      "learning_rate": 0.00013844231214317775,
+      "loss": 1.398,
+      "step": 8535
+    },
+    {
+      "epoch": 0.39323720458838163,
+      "grad_norm": 3.8442614917116216,
+      "learning_rate": 0.00013842853739434182,
+      "loss": 1.2232,
+      "step": 8536
+    },
+    {
+      "epoch": 0.3932832726770166,
+      "grad_norm": 3.400756087213631,
+      "learning_rate": 0.00013841476178996036,
+      "loss": 1.4141,
+      "step": 8537
+    },
+    {
+      "epoch": 0.3933293407656516,
+      "grad_norm": 1.2366618349305032,
+      "learning_rate": 0.00013840098533034008,
+      "loss": 1.4946,
+      "step": 8538
+    },
+    {
+      "epoch": 0.3933754088542866,
+      "grad_norm": 4.729639855171402,
+      "learning_rate": 0.00013838720801578772,
+      "loss": 1.266,
+      "step": 8539
+    },
+    {
+      "epoch": 0.3934214769429216,
+      "grad_norm": 3.0218368234177624,
+      "learning_rate": 0.00013837342984661,
+      "loss": 1.5167,
+      "step": 8540
+    },
+    {
+      "epoch": 0.3934675450315566,
+      "grad_norm": 2.122308109390923,
+      "learning_rate": 0.00013835965082311364,
+      "loss": 1.4141,
+      "step": 8541
+    },
+    {
+      "epoch": 0.3935136131201916,
+      "grad_norm": 4.29682005389261,
+      "learning_rate": 0.00013834587094560546,
+      "loss": 1.1859,
+      "step": 8542
+    },
+    {
+      "epoch": 0.3935596812088267,
+      "grad_norm": 2.0061632661140836,
+      "learning_rate": 0.00013833209021439218,
+      "loss": 1.2489,
+      "step": 8543
+    },
+    {
+      "epoch": 0.39360574929746167,
+      "grad_norm": 1.388324221465861,
+      "learning_rate": 0.00013831830862978063,
+      "loss": 1.3634,
+      "step": 8544
+    },
+    {
+      "epoch": 0.39365181738609667,
+      "grad_norm": 2.566508282419038,
+      "learning_rate": 0.00013830452619207769,
+      "loss": 1.2652,
+      "step": 8545
+    },
+    {
+      "epoch": 0.39369788547473167,
+      "grad_norm": 2.8235895099413937,
+      "learning_rate": 0.0001382907429015901,
+      "loss": 1.3747,
+      "step": 8546
+    },
+    {
+      "epoch": 0.39374395356336667,
+      "grad_norm": 1.1202301345405306,
+      "learning_rate": 0.00013827695875862482,
+      "loss": 1.4585,
+      "step": 8547
+    },
+    {
+      "epoch": 0.39379002165200166,
+      "grad_norm": 1.815141529965708,
+      "learning_rate": 0.00013826317376348864,
+      "loss": 1.6998,
+      "step": 8548
+    },
+    {
+      "epoch": 0.39383608974063666,
+      "grad_norm": 1.713171744039963,
+      "learning_rate": 0.00013824938791648852,
+      "loss": 1.2012,
+      "step": 8549
+    },
+    {
+      "epoch": 0.39388215782927166,
+      "grad_norm": 3.480360116093116,
+      "learning_rate": 0.00013823560121793135,
+      "loss": 1.3261,
+      "step": 8550
+    },
+    {
+      "epoch": 0.39392822591790666,
+      "grad_norm": 4.3136406106336755,
+      "learning_rate": 0.0001382218136681241,
+      "loss": 1.6516,
+      "step": 8551
+    },
+    {
+      "epoch": 0.39397429400654166,
+      "grad_norm": 0.9907533585461317,
+      "learning_rate": 0.00013820802526737369,
+      "loss": 1.2991,
+      "step": 8552
+    },
+    {
+      "epoch": 0.39402036209517666,
+      "grad_norm": 1.0648951206431194,
+      "learning_rate": 0.00013819423601598712,
+      "loss": 1.3628,
+      "step": 8553
+    },
+    {
+      "epoch": 0.39406643018381166,
+      "grad_norm": 7.640599316089681,
+      "learning_rate": 0.0001381804459142714,
+      "loss": 1.1454,
+      "step": 8554
+    },
+    {
+      "epoch": 0.39411249827244665,
+      "grad_norm": 1.3565344675141884,
+      "learning_rate": 0.00013816665496253346,
+      "loss": 1.2784,
+      "step": 8555
+    },
+    {
+      "epoch": 0.39415856636108165,
+      "grad_norm": 1.1674553278252275,
+      "learning_rate": 0.00013815286316108045,
+      "loss": 1.5162,
+      "step": 8556
+    },
+    {
+      "epoch": 0.3942046344497167,
+      "grad_norm": 2.6281814874801825,
+      "learning_rate": 0.0001381390705102193,
+      "loss": 1.6041,
+      "step": 8557
+    },
+    {
+      "epoch": 0.3942507025383517,
+      "grad_norm": 1.2667638210445145,
+      "learning_rate": 0.0001381252770102572,
+      "loss": 1.2894,
+      "step": 8558
+    },
+    {
+      "epoch": 0.3942967706269867,
+      "grad_norm": 4.3580588805307885,
+      "learning_rate": 0.0001381114826615011,
+      "loss": 1.1498,
+      "step": 8559
+    },
+    {
+      "epoch": 0.3943428387156217,
+      "grad_norm": 2.342416005495087,
+      "learning_rate": 0.00013809768746425827,
+      "loss": 1.3712,
+      "step": 8560
+    },
+    {
+      "epoch": 0.3943889068042567,
+      "grad_norm": 1.1423337271809826,
+      "learning_rate": 0.0001380838914188357,
+      "loss": 1.2983,
+      "step": 8561
+    },
+    {
+      "epoch": 0.3944349748928917,
+      "grad_norm": 3.4916468299309265,
+      "learning_rate": 0.00013807009452554062,
+      "loss": 1.4924,
+      "step": 8562
+    },
+    {
+      "epoch": 0.3944810429815267,
+      "grad_norm": 0.8731114658990798,
+      "learning_rate": 0.00013805629678468012,
+      "loss": 1.5055,
+      "step": 8563
+    },
+    {
+      "epoch": 0.3945271110701617,
+      "grad_norm": 1.7219856167563077,
+      "learning_rate": 0.00013804249819656148,
+      "loss": 1.5527,
+      "step": 8564
+    },
+    {
+      "epoch": 0.3945731791587967,
+      "grad_norm": 1.2200041606343794,
+      "learning_rate": 0.0001380286987614918,
+      "loss": 1.5241,
+      "step": 8565
+    },
+    {
+      "epoch": 0.3946192472474317,
+      "grad_norm": 2.5044086782159374,
+      "learning_rate": 0.00013801489847977837,
+      "loss": 1.2977,
+      "step": 8566
+    },
+    {
+      "epoch": 0.3946653153360667,
+      "grad_norm": 1.952574546365625,
+      "learning_rate": 0.00013800109735172836,
+      "loss": 1.4304,
+      "step": 8567
+    },
+    {
+      "epoch": 0.3947113834247017,
+      "grad_norm": 2.5573898637599433,
+      "learning_rate": 0.00013798729537764913,
+      "loss": 1.3076,
+      "step": 8568
+    },
+    {
+      "epoch": 0.3947574515133367,
+      "grad_norm": 2.0836145577806335,
+      "learning_rate": 0.0001379734925578479,
+      "loss": 1.6435,
+      "step": 8569
+    },
+    {
+      "epoch": 0.39480351960197174,
+      "grad_norm": 1.0263535991343191,
+      "learning_rate": 0.00013795968889263193,
+      "loss": 1.2523,
+      "step": 8570
+    },
+    {
+      "epoch": 0.39484958769060674,
+      "grad_norm": 0.9532100770367021,
+      "learning_rate": 0.00013794588438230862,
+      "loss": 1.6918,
+      "step": 8571
+    },
+    {
+      "epoch": 0.39489565577924174,
+      "grad_norm": 1.5084555533901483,
+      "learning_rate": 0.00013793207902718525,
+      "loss": 1.3139,
+      "step": 8572
+    },
+    {
+      "epoch": 0.39494172386787674,
+      "grad_norm": 2.5633192987230236,
+      "learning_rate": 0.00013791827282756916,
+      "loss": 1.4924,
+      "step": 8573
+    },
+    {
+      "epoch": 0.39498779195651174,
+      "grad_norm": 1.7078559323548947,
+      "learning_rate": 0.00013790446578376776,
+      "loss": 1.6668,
+      "step": 8574
+    },
+    {
+      "epoch": 0.39503386004514673,
+      "grad_norm": 2.401716194018826,
+      "learning_rate": 0.0001378906578960884,
+      "loss": 1.1755,
+      "step": 8575
+    },
+    {
+      "epoch": 0.39507992813378173,
+      "grad_norm": 1.0636901849076155,
+      "learning_rate": 0.0001378768491648385,
+      "loss": 1.5191,
+      "step": 8576
+    },
+    {
+      "epoch": 0.39512599622241673,
+      "grad_norm": 1.229953764632152,
+      "learning_rate": 0.00013786303959032554,
+      "loss": 1.4092,
+      "step": 8577
+    },
+    {
+      "epoch": 0.39517206431105173,
+      "grad_norm": 1.4903387556992123,
+      "learning_rate": 0.0001378492291728569,
+      "loss": 1.3898,
+      "step": 8578
+    },
+    {
+      "epoch": 0.39521813239968673,
+      "grad_norm": 1.081306730932269,
+      "learning_rate": 0.00013783541791274003,
+      "loss": 1.5701,
+      "step": 8579
+    },
+    {
+      "epoch": 0.3952642004883217,
+      "grad_norm": 0.9171837033759866,
+      "learning_rate": 0.00013782160581028255,
+      "loss": 1.1978,
+      "step": 8580
+    },
+    {
+      "epoch": 0.3953102685769567,
+      "grad_norm": 2.7266399163025006,
+      "learning_rate": 0.00013780779286579178,
+      "loss": 1.3628,
+      "step": 8581
+    },
+    {
+      "epoch": 0.3953563366655917,
+      "grad_norm": 1.202575415067881,
+      "learning_rate": 0.00013779397907957538,
+      "loss": 1.2361,
+      "step": 8582
+    },
+    {
+      "epoch": 0.3954024047542267,
+      "grad_norm": 1.4202573298598895,
+      "learning_rate": 0.0001377801644519408,
+      "loss": 1.3467,
+      "step": 8583
+    },
+    {
+      "epoch": 0.3954484728428618,
+      "grad_norm": 1.8465999253468488,
+      "learning_rate": 0.0001377663489831957,
+      "loss": 1.1827,
+      "step": 8584
+    },
+    {
+      "epoch": 0.3954945409314968,
+      "grad_norm": 5.249589521982165,
+      "learning_rate": 0.0001377525326736475,
+      "loss": 1.2643,
+      "step": 8585
+    },
+    {
+      "epoch": 0.3955406090201318,
+      "grad_norm": 2.0606545097122706,
+      "learning_rate": 0.00013773871552360398,
+      "loss": 1.2453,
+      "step": 8586
+    },
+    {
+      "epoch": 0.39558667710876677,
+      "grad_norm": 1.1377354182975934,
+      "learning_rate": 0.0001377248975333726,
+      "loss": 1.4565,
+      "step": 8587
+    },
+    {
+      "epoch": 0.39563274519740177,
+      "grad_norm": 1.2727938426728815,
+      "learning_rate": 0.0001377110787032611,
+      "loss": 1.4826,
+      "step": 8588
+    },
+    {
+      "epoch": 0.39567881328603677,
+      "grad_norm": 0.973256693136953,
+      "learning_rate": 0.00013769725903357708,
+      "loss": 1.6811,
+      "step": 8589
+    },
+    {
+      "epoch": 0.39572488137467177,
+      "grad_norm": 1.2365888981502127,
+      "learning_rate": 0.00013768343852462825,
+      "loss": 1.4959,
+      "step": 8590
+    },
+    {
+      "epoch": 0.39577094946330676,
+      "grad_norm": 0.9804596424506127,
+      "learning_rate": 0.00013766961717672226,
+      "loss": 1.5652,
+      "step": 8591
+    },
+    {
+      "epoch": 0.39581701755194176,
+      "grad_norm": 1.6864367204720478,
+      "learning_rate": 0.00013765579499016683,
+      "loss": 1.5368,
+      "step": 8592
+    },
+    {
+      "epoch": 0.39586308564057676,
+      "grad_norm": 1.7359063610002843,
+      "learning_rate": 0.0001376419719652697,
+      "loss": 1.2923,
+      "step": 8593
+    },
+    {
+      "epoch": 0.39590915372921176,
+      "grad_norm": 3.650335941207732,
+      "learning_rate": 0.00013762814810233857,
+      "loss": 1.4669,
+      "step": 8594
+    },
+    {
+      "epoch": 0.39595522181784676,
+      "grad_norm": 1.6339697581273327,
+      "learning_rate": 0.0001376143234016813,
+      "loss": 1.4847,
+      "step": 8595
+    },
+    {
+      "epoch": 0.39600128990648176,
+      "grad_norm": 0.8469058032254609,
+      "learning_rate": 0.00013760049786360555,
+      "loss": 1.621,
+      "step": 8596
+    },
+    {
+      "epoch": 0.39604735799511676,
+      "grad_norm": 1.0037374579864566,
+      "learning_rate": 0.00013758667148841924,
+      "loss": 1.3643,
+      "step": 8597
+    },
+    {
+      "epoch": 0.3960934260837518,
+      "grad_norm": 1.0848469869555735,
+      "learning_rate": 0.0001375728442764301,
+      "loss": 1.4662,
+      "step": 8598
+    },
+    {
+      "epoch": 0.3961394941723868,
+      "grad_norm": 1.5276061611706404,
+      "learning_rate": 0.00013755901622794603,
+      "loss": 1.3297,
+      "step": 8599
+    },
+    {
+      "epoch": 0.3961855622610218,
+      "grad_norm": 2.01017313912039,
+      "learning_rate": 0.00013754518734327483,
+      "loss": 1.1058,
+      "step": 8600
+    },
+    {
+      "epoch": 0.3962316303496568,
+      "grad_norm": 1.110008682565616,
+      "learning_rate": 0.00013753135762272444,
+      "loss": 1.4498,
+      "step": 8601
+    },
+    {
+      "epoch": 0.3962776984382918,
+      "grad_norm": 1.10593428406507,
+      "learning_rate": 0.00013751752706660272,
+      "loss": 1.43,
+      "step": 8602
+    },
+    {
+      "epoch": 0.3963237665269268,
+      "grad_norm": 1.1406898332509856,
+      "learning_rate": 0.0001375036956752176,
+      "loss": 1.3715,
+      "step": 8603
+    },
+    {
+      "epoch": 0.3963698346155618,
+      "grad_norm": 1.035641909775595,
+      "learning_rate": 0.00013748986344887698,
+      "loss": 1.6425,
+      "step": 8604
+    },
+    {
+      "epoch": 0.3964159027041968,
+      "grad_norm": 1.248252058060349,
+      "learning_rate": 0.00013747603038788883,
+      "loss": 1.3485,
+      "step": 8605
+    },
+    {
+      "epoch": 0.3964619707928318,
+      "grad_norm": 3.549842825677077,
+      "learning_rate": 0.00013746219649256113,
+      "loss": 1.3333,
+      "step": 8606
+    },
+    {
+      "epoch": 0.3965080388814668,
+      "grad_norm": 1.4700927028721595,
+      "learning_rate": 0.00013744836176320186,
+      "loss": 1.6066,
+      "step": 8607
+    },
+    {
+      "epoch": 0.3965541069701018,
+      "grad_norm": 0.9717479764679633,
+      "learning_rate": 0.00013743452620011904,
+      "loss": 1.3999,
+      "step": 8608
+    },
+    {
+      "epoch": 0.3966001750587368,
+      "grad_norm": 1.4128311705559402,
+      "learning_rate": 0.00013742068980362066,
+      "loss": 1.4881,
+      "step": 8609
+    },
+    {
+      "epoch": 0.3966462431473718,
+      "grad_norm": 13.895733930309559,
+      "learning_rate": 0.00013740685257401477,
+      "loss": 1.3312,
+      "step": 8610
+    },
+    {
+      "epoch": 0.39669231123600684,
+      "grad_norm": 1.6267231857204334,
+      "learning_rate": 0.00013739301451160947,
+      "loss": 1.3624,
+      "step": 8611
+    },
+    {
+      "epoch": 0.39673837932464184,
+      "grad_norm": 1.11387348817458,
+      "learning_rate": 0.0001373791756167128,
+      "loss": 1.554,
+      "step": 8612
+    },
+    {
+      "epoch": 0.39678444741327684,
+      "grad_norm": 1.7538566084603704,
+      "learning_rate": 0.00013736533588963292,
+      "loss": 1.2859,
+      "step": 8613
+    },
+    {
+      "epoch": 0.39683051550191184,
+      "grad_norm": 5.7185599239584874,
+      "learning_rate": 0.00013735149533067786,
+      "loss": 1.5553,
+      "step": 8614
+    },
+    {
+      "epoch": 0.39687658359054684,
+      "grad_norm": 3.097146311108555,
+      "learning_rate": 0.0001373376539401558,
+      "loss": 1.5323,
+      "step": 8615
+    },
+    {
+      "epoch": 0.39692265167918184,
+      "grad_norm": 2.8103348342096104,
+      "learning_rate": 0.00013732381171837493,
+      "loss": 0.9728,
+      "step": 8616
+    },
+    {
+      "epoch": 0.39696871976781684,
+      "grad_norm": 3.052412720533225,
+      "learning_rate": 0.00013730996866564337,
+      "loss": 1.3517,
+      "step": 8617
+    },
+    {
+      "epoch": 0.39701478785645183,
+      "grad_norm": 1.1051578304594385,
+      "learning_rate": 0.0001372961247822693,
+      "loss": 1.4096,
+      "step": 8618
+    },
+    {
+      "epoch": 0.39706085594508683,
+      "grad_norm": 2.6891243616365457,
+      "learning_rate": 0.00013728228006856103,
+      "loss": 1.2876,
+      "step": 8619
+    },
+    {
+      "epoch": 0.39710692403372183,
+      "grad_norm": 1.0105527624813584,
+      "learning_rate": 0.00013726843452482664,
+      "loss": 1.5286,
+      "step": 8620
+    },
+    {
+      "epoch": 0.39715299212235683,
+      "grad_norm": 1.0550127377583451,
+      "learning_rate": 0.0001372545881513745,
+      "loss": 1.4072,
+      "step": 8621
+    },
+    {
+      "epoch": 0.39719906021099183,
+      "grad_norm": 1.6437506114861415,
+      "learning_rate": 0.00013724074094851284,
+      "loss": 1.2006,
+      "step": 8622
+    },
+    {
+      "epoch": 0.3972451282996268,
+      "grad_norm": 2.124207738295244,
+      "learning_rate": 0.00013722689291654992,
+      "loss": 1.5638,
+      "step": 8623
+    },
+    {
+      "epoch": 0.3972911963882618,
+      "grad_norm": 1.2169984441600505,
+      "learning_rate": 0.00013721304405579408,
+      "loss": 1.5681,
+      "step": 8624
+    },
+    {
+      "epoch": 0.3973372644768969,
+      "grad_norm": 2.559373643470672,
+      "learning_rate": 0.0001371991943665536,
+      "loss": 1.3801,
+      "step": 8625
+    },
+    {
+      "epoch": 0.3973833325655319,
+      "grad_norm": 2.1395843719478838,
+      "learning_rate": 0.00013718534384913684,
+      "loss": 1.5556,
+      "step": 8626
+    },
+    {
+      "epoch": 0.3974294006541669,
+      "grad_norm": 5.319741789839764,
+      "learning_rate": 0.00013717149250385216,
+      "loss": 1.4931,
+      "step": 8627
+    },
+    {
+      "epoch": 0.3974754687428019,
+      "grad_norm": 2.0126542360551927,
+      "learning_rate": 0.00013715764033100797,
+      "loss": 1.5007,
+      "step": 8628
+    },
+    {
+      "epoch": 0.3975215368314369,
+      "grad_norm": 1.910975935718435,
+      "learning_rate": 0.00013714378733091257,
+      "loss": 1.6374,
+      "step": 8629
+    },
+    {
+      "epoch": 0.39756760492007187,
+      "grad_norm": 1.4333383680996346,
+      "learning_rate": 0.00013712993350387448,
+      "loss": 1.8255,
+      "step": 8630
+    },
+    {
+      "epoch": 0.39761367300870687,
+      "grad_norm": 1.5056263186135959,
+      "learning_rate": 0.00013711607885020206,
+      "loss": 1.4523,
+      "step": 8631
+    },
+    {
+      "epoch": 0.39765974109734187,
+      "grad_norm": 1.242275558588456,
+      "learning_rate": 0.0001371022233702038,
+      "loss": 1.159,
+      "step": 8632
+    },
+    {
+      "epoch": 0.39770580918597687,
+      "grad_norm": 1.9952558052150355,
+      "learning_rate": 0.00013708836706418813,
+      "loss": 1.3117,
+      "step": 8633
+    },
+    {
+      "epoch": 0.39775187727461186,
+      "grad_norm": 2.319786054364668,
+      "learning_rate": 0.00013707450993246355,
+      "loss": 1.3299,
+      "step": 8634
+    },
+    {
+      "epoch": 0.39779794536324686,
+      "grad_norm": 3.087222500551155,
+      "learning_rate": 0.00013706065197533862,
+      "loss": 1.2463,
+      "step": 8635
+    },
+    {
+      "epoch": 0.39784401345188186,
+      "grad_norm": 0.8843431130231586,
+      "learning_rate": 0.00013704679319312182,
+      "loss": 1.3017,
+      "step": 8636
+    },
+    {
+      "epoch": 0.39789008154051686,
+      "grad_norm": 2.6196500361662927,
+      "learning_rate": 0.00013703293358612167,
+      "loss": 1.1579,
+      "step": 8637
+    },
+    {
+      "epoch": 0.3979361496291519,
+      "grad_norm": 1.6457036899097415,
+      "learning_rate": 0.00013701907315464675,
+      "loss": 1.2509,
+      "step": 8638
+    },
+    {
+      "epoch": 0.3979822177177869,
+      "grad_norm": 1.432928573123534,
+      "learning_rate": 0.00013700521189900563,
+      "loss": 1.3964,
+      "step": 8639
+    },
+    {
+      "epoch": 0.3980282858064219,
+      "grad_norm": 1.6854476806988508,
+      "learning_rate": 0.00013699134981950693,
+      "loss": 0.9665,
+      "step": 8640
+    },
+    {
+      "epoch": 0.3980743538950569,
+      "grad_norm": 1.6393158407525206,
+      "learning_rate": 0.00013697748691645926,
+      "loss": 1.3578,
+      "step": 8641
+    },
+    {
+      "epoch": 0.3981204219836919,
+      "grad_norm": 1.500106438304118,
+      "learning_rate": 0.00013696362319017124,
+      "loss": 1.3073,
+      "step": 8642
+    },
+    {
+      "epoch": 0.3981664900723269,
+      "grad_norm": 1.203360680540753,
+      "learning_rate": 0.00013694975864095155,
+      "loss": 1.3227,
+      "step": 8643
+    },
+    {
+      "epoch": 0.3982125581609619,
+      "grad_norm": 1.0313703998978339,
+      "learning_rate": 0.00013693589326910878,
+      "loss": 1.6724,
+      "step": 8644
+    },
+    {
+      "epoch": 0.3982586262495969,
+      "grad_norm": 1.1277356738024562,
+      "learning_rate": 0.00013692202707495175,
+      "loss": 1.5289,
+      "step": 8645
+    },
+    {
+      "epoch": 0.3983046943382319,
+      "grad_norm": 4.1565586191262565,
+      "learning_rate": 0.00013690816005878907,
+      "loss": 1.3884,
+      "step": 8646
+    },
+    {
+      "epoch": 0.3983507624268669,
+      "grad_norm": 2.165240367424902,
+      "learning_rate": 0.00013689429222092946,
+      "loss": 1.3528,
+      "step": 8647
+    },
+    {
+      "epoch": 0.3983968305155019,
+      "grad_norm": 3.2083643445906147,
+      "learning_rate": 0.00013688042356168172,
+      "loss": 1.3907,
+      "step": 8648
+    },
+    {
+      "epoch": 0.3984428986041369,
+      "grad_norm": 1.6017135864656085,
+      "learning_rate": 0.00013686655408135456,
+      "loss": 1.2134,
+      "step": 8649
+    },
+    {
+      "epoch": 0.3984889666927719,
+      "grad_norm": 1.6171170640240984,
+      "learning_rate": 0.0001368526837802568,
+      "loss": 1.2724,
+      "step": 8650
+    },
+    {
+      "epoch": 0.3985350347814069,
+      "grad_norm": 2.127542495506973,
+      "learning_rate": 0.00013683881265869725,
+      "loss": 1.3258,
+      "step": 8651
+    },
+    {
+      "epoch": 0.39858110287004195,
+      "grad_norm": 1.8045877949526008,
+      "learning_rate": 0.0001368249407169847,
+      "loss": 1.3597,
+      "step": 8652
+    },
+    {
+      "epoch": 0.39862717095867695,
+      "grad_norm": 2.2891798119430273,
+      "learning_rate": 0.00013681106795542793,
+      "loss": 1.3707,
+      "step": 8653
+    },
+    {
+      "epoch": 0.39867323904731194,
+      "grad_norm": 1.2320554272702942,
+      "learning_rate": 0.0001367971943743359,
+      "loss": 1.1999,
+      "step": 8654
+    },
+    {
+      "epoch": 0.39871930713594694,
+      "grad_norm": 1.802083259462525,
+      "learning_rate": 0.00013678331997401744,
+      "loss": 1.3387,
+      "step": 8655
+    },
+    {
+      "epoch": 0.39876537522458194,
+      "grad_norm": 1.1633322910350787,
+      "learning_rate": 0.0001367694447547814,
+      "loss": 1.3417,
+      "step": 8656
+    },
+    {
+      "epoch": 0.39881144331321694,
+      "grad_norm": 2.7316678719648047,
+      "learning_rate": 0.00013675556871693673,
+      "loss": 1.1886,
+      "step": 8657
+    },
+    {
+      "epoch": 0.39885751140185194,
+      "grad_norm": 1.6999771643835246,
+      "learning_rate": 0.00013674169186079233,
+      "loss": 1.4642,
+      "step": 8658
+    },
+    {
+      "epoch": 0.39890357949048694,
+      "grad_norm": 1.852515070211721,
+      "learning_rate": 0.00013672781418665717,
+      "loss": 1.3762,
+      "step": 8659
+    },
+    {
+      "epoch": 0.39894964757912194,
+      "grad_norm": 3.466158376340824,
+      "learning_rate": 0.0001367139356948402,
+      "loss": 1.6153,
+      "step": 8660
+    },
+    {
+      "epoch": 0.39899571566775693,
+      "grad_norm": 3.530019834136805,
+      "learning_rate": 0.00013670005638565042,
+      "loss": 1.2351,
+      "step": 8661
+    },
+    {
+      "epoch": 0.39904178375639193,
+      "grad_norm": 1.447669382447936,
+      "learning_rate": 0.00013668617625939677,
+      "loss": 1.319,
+      "step": 8662
+    },
+    {
+      "epoch": 0.39908785184502693,
+      "grad_norm": 1.0958266645131802,
+      "learning_rate": 0.00013667229531638838,
+      "loss": 1.4762,
+      "step": 8663
+    },
+    {
+      "epoch": 0.39913391993366193,
+      "grad_norm": 2.5213887082315116,
+      "learning_rate": 0.00013665841355693417,
+      "loss": 0.991,
+      "step": 8664
+    },
+    {
+      "epoch": 0.3991799880222969,
+      "grad_norm": 0.9303697465514601,
+      "learning_rate": 0.00013664453098134323,
+      "loss": 1.5953,
+      "step": 8665
+    },
+    {
+      "epoch": 0.399226056110932,
+      "grad_norm": 1.5962428919701488,
+      "learning_rate": 0.00013663064758992466,
+      "loss": 1.3073,
+      "step": 8666
+    },
+    {
+      "epoch": 0.399272124199567,
+      "grad_norm": 1.8633913631421437,
+      "learning_rate": 0.0001366167633829875,
+      "loss": 1.4964,
+      "step": 8667
+    },
+    {
+      "epoch": 0.399318192288202,
+      "grad_norm": 3.3085626686994183,
+      "learning_rate": 0.0001366028783608409,
+      "loss": 1.1195,
+      "step": 8668
+    },
+    {
+      "epoch": 0.399364260376837,
+      "grad_norm": 1.8742513672500212,
+      "learning_rate": 0.000136588992523794,
+      "loss": 1.491,
+      "step": 8669
+    },
+    {
+      "epoch": 0.399410328465472,
+      "grad_norm": 1.1041921981635647,
+      "learning_rate": 0.00013657510587215588,
+      "loss": 1.389,
+      "step": 8670
+    },
+    {
+      "epoch": 0.399456396554107,
+      "grad_norm": 1.8656306980111073,
+      "learning_rate": 0.00013656121840623575,
+      "loss": 1.2505,
+      "step": 8671
+    },
+    {
+      "epoch": 0.399502464642742,
+      "grad_norm": 2.061695141495733,
+      "learning_rate": 0.00013654733012634279,
+      "loss": 1.3282,
+      "step": 8672
+    },
+    {
+      "epoch": 0.39954853273137697,
+      "grad_norm": 1.6917425447627128,
+      "learning_rate": 0.0001365334410327862,
+      "loss": 1.5787,
+      "step": 8673
+    },
+    {
+      "epoch": 0.39959460082001197,
+      "grad_norm": 1.073632930825056,
+      "learning_rate": 0.00013651955112587517,
+      "loss": 1.3699,
+      "step": 8674
+    },
+    {
+      "epoch": 0.39964066890864697,
+      "grad_norm": 0.9832576702455672,
+      "learning_rate": 0.00013650566040591898,
+      "loss": 1.2935,
+      "step": 8675
+    },
+    {
+      "epoch": 0.39968673699728197,
+      "grad_norm": 1.9183808482754998,
+      "learning_rate": 0.00013649176887322684,
+      "loss": 1.2286,
+      "step": 8676
+    },
+    {
+      "epoch": 0.39973280508591696,
+      "grad_norm": 1.219935208148789,
+      "learning_rate": 0.000136477876528108,
+      "loss": 1.4914,
+      "step": 8677
+    },
+    {
+      "epoch": 0.39977887317455196,
+      "grad_norm": 1.364891055732989,
+      "learning_rate": 0.00013646398337087182,
+      "loss": 1.1518,
+      "step": 8678
+    },
+    {
+      "epoch": 0.399824941263187,
+      "grad_norm": 2.0041771292087134,
+      "learning_rate": 0.00013645008940182756,
+      "loss": 1.4504,
+      "step": 8679
+    },
+    {
+      "epoch": 0.399871009351822,
+      "grad_norm": 2.070639555530608,
+      "learning_rate": 0.00013643619462128456,
+      "loss": 1.3453,
+      "step": 8680
+    },
+    {
+      "epoch": 0.399917077440457,
+      "grad_norm": 1.111687638385308,
+      "learning_rate": 0.00013642229902955218,
+      "loss": 1.5003,
+      "step": 8681
+    },
+    {
+      "epoch": 0.399963145529092,
+      "grad_norm": 1.4603862860357786,
+      "learning_rate": 0.00013640840262693974,
+      "loss": 1.2016,
+      "step": 8682
+    },
+    {
+      "epoch": 0.400009213617727,
+      "grad_norm": 1.795679409930521,
+      "learning_rate": 0.00013639450541375666,
+      "loss": 1.5634,
+      "step": 8683
+    },
+    {
+      "epoch": 0.400055281706362,
+      "grad_norm": 1.5056556997044754,
+      "learning_rate": 0.0001363806073903123,
+      "loss": 1.3055,
+      "step": 8684
+    },
+    {
+      "epoch": 0.400101349794997,
+      "grad_norm": 1.5046221434754938,
+      "learning_rate": 0.00013636670855691612,
+      "loss": 1.3625,
+      "step": 8685
+    },
+    {
+      "epoch": 0.400147417883632,
+      "grad_norm": 4.4756334885022415,
+      "learning_rate": 0.00013635280891387752,
+      "loss": 1.3665,
+      "step": 8686
+    },
+    {
+      "epoch": 0.400193485972267,
+      "grad_norm": 2.1386781010719518,
+      "learning_rate": 0.00013633890846150595,
+      "loss": 1.0953,
+      "step": 8687
+    },
+    {
+      "epoch": 0.400239554060902,
+      "grad_norm": 1.626610369340214,
+      "learning_rate": 0.0001363250072001109,
+      "loss": 1.264,
+      "step": 8688
+    },
+    {
+      "epoch": 0.400285622149537,
+      "grad_norm": 1.264887040087227,
+      "learning_rate": 0.00013631110513000186,
+      "loss": 1.2873,
+      "step": 8689
+    },
+    {
+      "epoch": 0.400331690238172,
+      "grad_norm": 1.9226498109514545,
+      "learning_rate": 0.0001362972022514883,
+      "loss": 1.2397,
+      "step": 8690
+    },
+    {
+      "epoch": 0.400377758326807,
+      "grad_norm": 2.036985768880877,
+      "learning_rate": 0.00013628329856487978,
+      "loss": 1.4923,
+      "step": 8691
+    },
+    {
+      "epoch": 0.400423826415442,
+      "grad_norm": 2.518385566517694,
+      "learning_rate": 0.00013626939407048586,
+      "loss": 1.2102,
+      "step": 8692
+    },
+    {
+      "epoch": 0.40046989450407705,
+      "grad_norm": 1.2163370943371865,
+      "learning_rate": 0.000136255488768616,
+      "loss": 1.2982,
+      "step": 8693
+    },
+    {
+      "epoch": 0.40051596259271205,
+      "grad_norm": 1.5117678948048683,
+      "learning_rate": 0.0001362415826595799,
+      "loss": 1.3852,
+      "step": 8694
+    },
+    {
+      "epoch": 0.40056203068134705,
+      "grad_norm": 1.0556725383044334,
+      "learning_rate": 0.00013622767574368708,
+      "loss": 1.7452,
+      "step": 8695
+    },
+    {
+      "epoch": 0.40060809876998205,
+      "grad_norm": 0.938087354159488,
+      "learning_rate": 0.00013621376802124718,
+      "loss": 1.4169,
+      "step": 8696
+    },
+    {
+      "epoch": 0.40065416685861704,
+      "grad_norm": 1.643304870765272,
+      "learning_rate": 0.00013619985949256985,
+      "loss": 1.1686,
+      "step": 8697
+    },
+    {
+      "epoch": 0.40070023494725204,
+      "grad_norm": 2.953931487751063,
+      "learning_rate": 0.00013618595015796472,
+      "loss": 1.5145,
+      "step": 8698
+    },
+    {
+      "epoch": 0.40074630303588704,
+      "grad_norm": 2.9389416923608134,
+      "learning_rate": 0.00013617204001774142,
+      "loss": 1.215,
+      "step": 8699
+    },
+    {
+      "epoch": 0.40079237112452204,
+      "grad_norm": 1.3065260037459798,
+      "learning_rate": 0.00013615812907220967,
+      "loss": 1.2643,
+      "step": 8700
+    },
+    {
+      "epoch": 0.40083843921315704,
+      "grad_norm": 1.504799056847038,
+      "learning_rate": 0.00013614421732167922,
+      "loss": 1.3892,
+      "step": 8701
+    },
+    {
+      "epoch": 0.40088450730179204,
+      "grad_norm": 1.230549024440678,
+      "learning_rate": 0.00013613030476645972,
+      "loss": 1.1988,
+      "step": 8702
+    },
+    {
+      "epoch": 0.40093057539042704,
+      "grad_norm": 2.3641282593150885,
+      "learning_rate": 0.00013611639140686095,
+      "loss": 1.1928,
+      "step": 8703
+    },
+    {
+      "epoch": 0.40097664347906203,
+      "grad_norm": 1.5372820494840906,
+      "learning_rate": 0.00013610247724319262,
+      "loss": 1.3123,
+      "step": 8704
+    },
+    {
+      "epoch": 0.40102271156769703,
+      "grad_norm": 1.2775925652228723,
+      "learning_rate": 0.00013608856227576455,
+      "loss": 1.2348,
+      "step": 8705
+    },
+    {
+      "epoch": 0.4010687796563321,
+      "grad_norm": 1.2519744911190662,
+      "learning_rate": 0.00013607464650488652,
+      "loss": 1.6203,
+      "step": 8706
+    },
+    {
+      "epoch": 0.4011148477449671,
+      "grad_norm": 2.1328750796858875,
+      "learning_rate": 0.00013606072993086835,
+      "loss": 1.4285,
+      "step": 8707
+    },
+    {
+      "epoch": 0.4011609158336021,
+      "grad_norm": 2.029484474454642,
+      "learning_rate": 0.00013604681255401983,
+      "loss": 1.3845,
+      "step": 8708
+    },
+    {
+      "epoch": 0.4012069839222371,
+      "grad_norm": 5.076229376680115,
+      "learning_rate": 0.00013603289437465085,
+      "loss": 1.2844,
+      "step": 8709
+    },
+    {
+      "epoch": 0.4012530520108721,
+      "grad_norm": 4.254111492391304,
+      "learning_rate": 0.00013601897539307128,
+      "loss": 1.7155,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4012991200995071,
+      "grad_norm": 1.5002900965221548,
+      "learning_rate": 0.00013600505560959096,
+      "loss": 1.1018,
+      "step": 8711
+    },
+    {
+      "epoch": 0.4013451881881421,
+      "grad_norm": 1.4450071060251184,
+      "learning_rate": 0.00013599113502451982,
+      "loss": 1.5819,
+      "step": 8712
+    },
+    {
+      "epoch": 0.4013912562767771,
+      "grad_norm": 1.996444613373762,
+      "learning_rate": 0.00013597721363816777,
+      "loss": 1.382,
+      "step": 8713
+    },
+    {
+      "epoch": 0.4014373243654121,
+      "grad_norm": 1.1360488467468859,
+      "learning_rate": 0.00013596329145084475,
+      "loss": 1.324,
+      "step": 8714
+    },
+    {
+      "epoch": 0.40148339245404707,
+      "grad_norm": 1.1387138949710476,
+      "learning_rate": 0.0001359493684628607,
+      "loss": 1.092,
+      "step": 8715
+    },
+    {
+      "epoch": 0.40152946054268207,
+      "grad_norm": 1.8063113230494692,
+      "learning_rate": 0.00013593544467452561,
+      "loss": 1.1784,
+      "step": 8716
+    },
+    {
+      "epoch": 0.40157552863131707,
+      "grad_norm": 1.8137822200064588,
+      "learning_rate": 0.00013592152008614947,
+      "loss": 1.2612,
+      "step": 8717
+    },
+    {
+      "epoch": 0.40162159671995207,
+      "grad_norm": 1.365449819087104,
+      "learning_rate": 0.00013590759469804227,
+      "loss": 1.3928,
+      "step": 8718
+    },
+    {
+      "epoch": 0.40166766480858707,
+      "grad_norm": 1.2700839031855433,
+      "learning_rate": 0.00013589366851051403,
+      "loss": 1.3142,
+      "step": 8719
+    },
+    {
+      "epoch": 0.4017137328972221,
+      "grad_norm": 0.9790880943858317,
+      "learning_rate": 0.00013587974152387482,
+      "loss": 1.3046,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4017598009858571,
+      "grad_norm": 0.8521422207760011,
+      "learning_rate": 0.00013586581373843467,
+      "loss": 1.2769,
+      "step": 8721
+    },
+    {
+      "epoch": 0.4018058690744921,
+      "grad_norm": 1.3836369843428786,
+      "learning_rate": 0.0001358518851545037,
+      "loss": 1.3887,
+      "step": 8722
+    },
+    {
+      "epoch": 0.4018519371631271,
+      "grad_norm": 1.317238156211136,
+      "learning_rate": 0.00013583795577239196,
+      "loss": 1.2648,
+      "step": 8723
+    },
+    {
+      "epoch": 0.4018980052517621,
+      "grad_norm": 1.2360041191191726,
+      "learning_rate": 0.00013582402559240956,
+      "loss": 1.4694,
+      "step": 8724
+    },
+    {
+      "epoch": 0.4019440733403971,
+      "grad_norm": 1.8465324697936563,
+      "learning_rate": 0.00013581009461486668,
+      "loss": 1.5151,
+      "step": 8725
+    },
+    {
+      "epoch": 0.4019901414290321,
+      "grad_norm": 2.0470425830954864,
+      "learning_rate": 0.00013579616284007347,
+      "loss": 0.8812,
+      "step": 8726
+    },
+    {
+      "epoch": 0.4020362095176671,
+      "grad_norm": 2.6682167582364653,
+      "learning_rate": 0.00013578223026834002,
+      "loss": 1.1046,
+      "step": 8727
+    },
+    {
+      "epoch": 0.4020822776063021,
+      "grad_norm": 1.4411195481701147,
+      "learning_rate": 0.00013576829689997664,
+      "loss": 1.6649,
+      "step": 8728
+    },
+    {
+      "epoch": 0.4021283456949371,
+      "grad_norm": 3.1466904133520583,
+      "learning_rate": 0.0001357543627352934,
+      "loss": 1.1854,
+      "step": 8729
+    },
+    {
+      "epoch": 0.4021744137835721,
+      "grad_norm": 1.2513737271777214,
+      "learning_rate": 0.0001357404277746006,
+      "loss": 1.3956,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4022204818722071,
+      "grad_norm": 1.0361471079647013,
+      "learning_rate": 0.00013572649201820845,
+      "loss": 1.6754,
+      "step": 8731
+    },
+    {
+      "epoch": 0.4022665499608421,
+      "grad_norm": 1.3391942356981863,
+      "learning_rate": 0.00013571255546642719,
+      "loss": 1.3937,
+      "step": 8732
+    },
+    {
+      "epoch": 0.4023126180494771,
+      "grad_norm": 1.5327193045681227,
+      "learning_rate": 0.00013569861811956716,
+      "loss": 1.2744,
+      "step": 8733
+    },
+    {
+      "epoch": 0.40235868613811215,
+      "grad_norm": 2.9428005516780744,
+      "learning_rate": 0.00013568467997793858,
+      "loss": 1.2559,
+      "step": 8734
+    },
+    {
+      "epoch": 0.40240475422674715,
+      "grad_norm": 0.9395057104590732,
+      "learning_rate": 0.00013567074104185175,
+      "loss": 1.2882,
+      "step": 8735
+    },
+    {
+      "epoch": 0.40245082231538215,
+      "grad_norm": 1.0469175261522532,
+      "learning_rate": 0.0001356568013116171,
+      "loss": 1.486,
+      "step": 8736
+    },
+    {
+      "epoch": 0.40249689040401715,
+      "grad_norm": 2.865082432738962,
+      "learning_rate": 0.0001356428607875449,
+      "loss": 1.2699,
+      "step": 8737
+    },
+    {
+      "epoch": 0.40254295849265215,
+      "grad_norm": 1.4435678046057354,
+      "learning_rate": 0.00013562891946994552,
+      "loss": 1.359,
+      "step": 8738
+    },
+    {
+      "epoch": 0.40258902658128715,
+      "grad_norm": 1.1783250032549362,
+      "learning_rate": 0.0001356149773591293,
+      "loss": 1.142,
+      "step": 8739
+    },
+    {
+      "epoch": 0.40263509466992214,
+      "grad_norm": 2.314382241890599,
+      "learning_rate": 0.0001356010344554067,
+      "loss": 1.1431,
+      "step": 8740
+    },
+    {
+      "epoch": 0.40268116275855714,
+      "grad_norm": 1.8632959124107118,
+      "learning_rate": 0.0001355870907590881,
+      "loss": 1.5146,
+      "step": 8741
+    },
+    {
+      "epoch": 0.40272723084719214,
+      "grad_norm": 1.4033906329738473,
+      "learning_rate": 0.00013557314627048395,
+      "loss": 1.6592,
+      "step": 8742
+    },
+    {
+      "epoch": 0.40277329893582714,
+      "grad_norm": 1.3765989745159146,
+      "learning_rate": 0.0001355592009899047,
+      "loss": 1.6168,
+      "step": 8743
+    },
+    {
+      "epoch": 0.40281936702446214,
+      "grad_norm": 1.0949116175698186,
+      "learning_rate": 0.00013554525491766078,
+      "loss": 1.2141,
+      "step": 8744
+    },
+    {
+      "epoch": 0.40286543511309714,
+      "grad_norm": 4.300035286004678,
+      "learning_rate": 0.00013553130805406274,
+      "loss": 0.9589,
+      "step": 8745
+    },
+    {
+      "epoch": 0.40291150320173214,
+      "grad_norm": 1.4920886211533502,
+      "learning_rate": 0.00013551736039942104,
+      "loss": 1.4769,
+      "step": 8746
+    },
+    {
+      "epoch": 0.4029575712903672,
+      "grad_norm": 1.7778549845928224,
+      "learning_rate": 0.00013550341195404624,
+      "loss": 1.26,
+      "step": 8747
+    },
+    {
+      "epoch": 0.4030036393790022,
+      "grad_norm": 1.0170562191971768,
+      "learning_rate": 0.00013548946271824878,
+      "loss": 1.6132,
+      "step": 8748
+    },
+    {
+      "epoch": 0.4030497074676372,
+      "grad_norm": 1.3146826322454228,
+      "learning_rate": 0.00013547551269233935,
+      "loss": 1.4688,
+      "step": 8749
+    },
+    {
+      "epoch": 0.4030957755562722,
+      "grad_norm": 1.3263138101924166,
+      "learning_rate": 0.0001354615618766284,
+      "loss": 1.3228,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4031418436449072,
+      "grad_norm": 1.2251409128329436,
+      "learning_rate": 0.00013544761027142662,
+      "loss": 1.1511,
+      "step": 8751
+    },
+    {
+      "epoch": 0.4031879117335422,
+      "grad_norm": 1.1320343249128189,
+      "learning_rate": 0.00013543365787704455,
+      "loss": 1.1165,
+      "step": 8752
+    },
+    {
+      "epoch": 0.4032339798221772,
+      "grad_norm": 2.3173613235017565,
+      "learning_rate": 0.0001354197046937929,
+      "loss": 1.2589,
+      "step": 8753
+    },
+    {
+      "epoch": 0.4032800479108122,
+      "grad_norm": 2.080095752094254,
+      "learning_rate": 0.00013540575072198217,
+      "loss": 1.5256,
+      "step": 8754
+    },
+    {
+      "epoch": 0.4033261159994472,
+      "grad_norm": 2.1529194744948925,
+      "learning_rate": 0.00013539179596192317,
+      "loss": 0.9541,
+      "step": 8755
+    },
+    {
+      "epoch": 0.4033721840880822,
+      "grad_norm": 1.640685826593117,
+      "learning_rate": 0.0001353778404139265,
+      "loss": 1.553,
+      "step": 8756
+    },
+    {
+      "epoch": 0.4034182521767172,
+      "grad_norm": 1.0382585561803113,
+      "learning_rate": 0.00013536388407830286,
+      "loss": 1.3867,
+      "step": 8757
+    },
+    {
+      "epoch": 0.40346432026535217,
+      "grad_norm": 3.7331262922897928,
+      "learning_rate": 0.000135349926955363,
+      "loss": 0.9774,
+      "step": 8758
+    },
+    {
+      "epoch": 0.40351038835398717,
+      "grad_norm": 1.2513658333799431,
+      "learning_rate": 0.0001353359690454176,
+      "loss": 1.6463,
+      "step": 8759
+    },
+    {
+      "epoch": 0.40355645644262217,
+      "grad_norm": 1.1805893824848412,
+      "learning_rate": 0.00013532201034877747,
+      "loss": 1.2308,
+      "step": 8760
+    },
+    {
+      "epoch": 0.4036025245312572,
+      "grad_norm": 3.185905201620781,
+      "learning_rate": 0.0001353080508657533,
+      "loss": 1.1686,
+      "step": 8761
+    },
+    {
+      "epoch": 0.4036485926198922,
+      "grad_norm": 1.1928532186498946,
+      "learning_rate": 0.00013529409059665596,
+      "loss": 1.4429,
+      "step": 8762
+    },
+    {
+      "epoch": 0.4036946607085272,
+      "grad_norm": 0.7268862936906236,
+      "learning_rate": 0.00013528012954179617,
+      "loss": 1.5949,
+      "step": 8763
+    },
+    {
+      "epoch": 0.4037407287971622,
+      "grad_norm": 1.5160533706941393,
+      "learning_rate": 0.00013526616770148485,
+      "loss": 1.7745,
+      "step": 8764
+    },
+    {
+      "epoch": 0.4037867968857972,
+      "grad_norm": 1.3699679191353682,
+      "learning_rate": 0.00013525220507603275,
+      "loss": 1.5106,
+      "step": 8765
+    },
+    {
+      "epoch": 0.4038328649744322,
+      "grad_norm": 0.8729392408743857,
+      "learning_rate": 0.00013523824166575073,
+      "loss": 1.7608,
+      "step": 8766
+    },
+    {
+      "epoch": 0.4038789330630672,
+      "grad_norm": 0.8887868133124872,
+      "learning_rate": 0.00013522427747094968,
+      "loss": 1.3683,
+      "step": 8767
+    },
+    {
+      "epoch": 0.4039250011517022,
+      "grad_norm": 1.6484731820543803,
+      "learning_rate": 0.0001352103124919405,
+      "loss": 1.4782,
+      "step": 8768
+    },
+    {
+      "epoch": 0.4039710692403372,
+      "grad_norm": 5.558182480044083,
+      "learning_rate": 0.00013519634672903407,
+      "loss": 1.2371,
+      "step": 8769
+    },
+    {
+      "epoch": 0.4040171373289722,
+      "grad_norm": 1.7580749348537439,
+      "learning_rate": 0.0001351823801825413,
+      "loss": 1.4157,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4040632054176072,
+      "grad_norm": 1.4523747015893937,
+      "learning_rate": 0.00013516841285277325,
+      "loss": 1.3484,
+      "step": 8771
+    },
+    {
+      "epoch": 0.4041092735062422,
+      "grad_norm": 1.5689214351184895,
+      "learning_rate": 0.0001351544447400407,
+      "loss": 1.369,
+      "step": 8772
+    },
+    {
+      "epoch": 0.4041553415948772,
+      "grad_norm": 1.1853790443477334,
+      "learning_rate": 0.00013514047584465481,
+      "loss": 1.3788,
+      "step": 8773
+    },
+    {
+      "epoch": 0.40420140968351226,
+      "grad_norm": 1.3601460582035427,
+      "learning_rate": 0.00013512650616692641,
+      "loss": 1.1634,
+      "step": 8774
+    },
+    {
+      "epoch": 0.40424747777214726,
+      "grad_norm": 1.652130924092803,
+      "learning_rate": 0.0001351125357071666,
+      "loss": 1.2933,
+      "step": 8775
+    },
+    {
+      "epoch": 0.40429354586078226,
+      "grad_norm": 1.456050226145577,
+      "learning_rate": 0.00013509856446568635,
+      "loss": 1.4297,
+      "step": 8776
+    },
+    {
+      "epoch": 0.40433961394941725,
+      "grad_norm": 0.7603776251552163,
+      "learning_rate": 0.00013508459244279678,
+      "loss": 1.3575,
+      "step": 8777
+    },
+    {
+      "epoch": 0.40438568203805225,
+      "grad_norm": 1.8147211063626296,
+      "learning_rate": 0.00013507061963880892,
+      "loss": 1.4017,
+      "step": 8778
+    },
+    {
+      "epoch": 0.40443175012668725,
+      "grad_norm": 0.9735913608416445,
+      "learning_rate": 0.0001350566460540338,
+      "loss": 1.2648,
+      "step": 8779
+    },
+    {
+      "epoch": 0.40447781821532225,
+      "grad_norm": 2.024719760081601,
+      "learning_rate": 0.00013504267168878262,
+      "loss": 1.2971,
+      "step": 8780
+    },
+    {
+      "epoch": 0.40452388630395725,
+      "grad_norm": 0.8653429988305066,
+      "learning_rate": 0.0001350286965433664,
+      "loss": 1.7448,
+      "step": 8781
+    },
+    {
+      "epoch": 0.40456995439259225,
+      "grad_norm": 0.911905845770762,
+      "learning_rate": 0.00013501472061809634,
+      "loss": 1.3835,
+      "step": 8782
+    },
+    {
+      "epoch": 0.40461602248122724,
+      "grad_norm": 1.7062401055910268,
+      "learning_rate": 0.00013500074391328357,
+      "loss": 1.3276,
+      "step": 8783
+    },
+    {
+      "epoch": 0.40466209056986224,
+      "grad_norm": 2.662184169163377,
+      "learning_rate": 0.00013498676642923925,
+      "loss": 1.5196,
+      "step": 8784
+    },
+    {
+      "epoch": 0.40470815865849724,
+      "grad_norm": 1.0683733832113393,
+      "learning_rate": 0.00013497278816627454,
+      "loss": 0.967,
+      "step": 8785
+    },
+    {
+      "epoch": 0.40475422674713224,
+      "grad_norm": 1.345657613910586,
+      "learning_rate": 0.0001349588091247007,
+      "loss": 1.1103,
+      "step": 8786
+    },
+    {
+      "epoch": 0.40480029483576724,
+      "grad_norm": 1.3400933747303996,
+      "learning_rate": 0.00013494482930482884,
+      "loss": 1.4626,
+      "step": 8787
+    },
+    {
+      "epoch": 0.4048463629244023,
+      "grad_norm": 1.4532465053567423,
+      "learning_rate": 0.00013493084870697038,
+      "loss": 1.0947,
+      "step": 8788
+    },
+    {
+      "epoch": 0.4048924310130373,
+      "grad_norm": 3.0980575270571227,
+      "learning_rate": 0.0001349168673314364,
+      "loss": 1.3573,
+      "step": 8789
+    },
+    {
+      "epoch": 0.4049384991016723,
+      "grad_norm": 1.0488931542057114,
+      "learning_rate": 0.00013490288517853825,
+      "loss": 1.2,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4049845671903073,
+      "grad_norm": 0.9794306510951182,
+      "learning_rate": 0.0001348889022485872,
+      "loss": 1.4145,
+      "step": 8791
+    },
+    {
+      "epoch": 0.4050306352789423,
+      "grad_norm": 1.0033238953776775,
+      "learning_rate": 0.0001348749185418946,
+      "loss": 1.3654,
+      "step": 8792
+    },
+    {
+      "epoch": 0.4050767033675773,
+      "grad_norm": 1.8455183017651025,
+      "learning_rate": 0.0001348609340587717,
+      "loss": 1.1809,
+      "step": 8793
+    },
+    {
+      "epoch": 0.4051227714562123,
+      "grad_norm": 1.3915982653603562,
+      "learning_rate": 0.00013484694879952987,
+      "loss": 1.2678,
+      "step": 8794
+    },
+    {
+      "epoch": 0.4051688395448473,
+      "grad_norm": 2.229111214156957,
+      "learning_rate": 0.0001348329627644805,
+      "loss": 1.3388,
+      "step": 8795
+    },
+    {
+      "epoch": 0.4052149076334823,
+      "grad_norm": 2.8747194299354994,
+      "learning_rate": 0.00013481897595393495,
+      "loss": 1.4006,
+      "step": 8796
+    },
+    {
+      "epoch": 0.4052609757221173,
+      "grad_norm": 1.3549485920184878,
+      "learning_rate": 0.00013480498836820456,
+      "loss": 1.3336,
+      "step": 8797
+    },
+    {
+      "epoch": 0.4053070438107523,
+      "grad_norm": 1.371009183514079,
+      "learning_rate": 0.0001347910000076008,
+      "loss": 1.4131,
+      "step": 8798
+    },
+    {
+      "epoch": 0.4053531118993873,
+      "grad_norm": 17.91491663266644,
+      "learning_rate": 0.00013477701087243507,
+      "loss": 1.3427,
+      "step": 8799
+    },
+    {
+      "epoch": 0.4053991799880223,
+      "grad_norm": 1.0859571082710813,
+      "learning_rate": 0.00013476302096301885,
+      "loss": 1.1392,
+      "step": 8800
+    },
+    {
+      "epoch": 0.40544524807665727,
+      "grad_norm": 2.047996978256827,
+      "learning_rate": 0.00013474903027966358,
+      "loss": 1.3467,
+      "step": 8801
+    },
+    {
+      "epoch": 0.4054913161652923,
+      "grad_norm": 1.7162806169903895,
+      "learning_rate": 0.0001347350388226807,
+      "loss": 1.1231,
+      "step": 8802
+    },
+    {
+      "epoch": 0.4055373842539273,
+      "grad_norm": 1.122392543904598,
+      "learning_rate": 0.00013472104659238173,
+      "loss": 1.5356,
+      "step": 8803
+    },
+    {
+      "epoch": 0.4055834523425623,
+      "grad_norm": 1.7201371629394888,
+      "learning_rate": 0.0001347070535890782,
+      "loss": 1.5603,
+      "step": 8804
+    },
+    {
+      "epoch": 0.4056295204311973,
+      "grad_norm": 1.0792244656274594,
+      "learning_rate": 0.00013469305981308167,
+      "loss": 1.0539,
+      "step": 8805
+    },
+    {
+      "epoch": 0.4056755885198323,
+      "grad_norm": 1.4493261954398249,
+      "learning_rate": 0.00013467906526470362,
+      "loss": 1.3576,
+      "step": 8806
+    },
+    {
+      "epoch": 0.4057216566084673,
+      "grad_norm": 1.5281148844098604,
+      "learning_rate": 0.00013466506994425564,
+      "loss": 1.3292,
+      "step": 8807
+    },
+    {
+      "epoch": 0.4057677246971023,
+      "grad_norm": 1.3354903194500873,
+      "learning_rate": 0.00013465107385204935,
+      "loss": 1.1808,
+      "step": 8808
+    },
+    {
+      "epoch": 0.4058137927857373,
+      "grad_norm": 1.0952944872232482,
+      "learning_rate": 0.00013463707698839627,
+      "loss": 1.7073,
+      "step": 8809
+    },
+    {
+      "epoch": 0.4058598608743723,
+      "grad_norm": 1.5611346795698484,
+      "learning_rate": 0.00013462307935360808,
+      "loss": 1.3104,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4059059289630073,
+      "grad_norm": 0.7781826974345573,
+      "learning_rate": 0.0001346090809479964,
+      "loss": 1.6267,
+      "step": 8811
+    },
+    {
+      "epoch": 0.4059519970516423,
+      "grad_norm": 2.04476724961724,
+      "learning_rate": 0.00013459508177187285,
+      "loss": 1.3798,
+      "step": 8812
+    },
+    {
+      "epoch": 0.4059980651402773,
+      "grad_norm": 1.148779797150713,
+      "learning_rate": 0.00013458108182554915,
+      "loss": 1.5851,
+      "step": 8813
+    },
+    {
+      "epoch": 0.4060441332289123,
+      "grad_norm": 1.1617619169166133,
+      "learning_rate": 0.00013456708110933693,
+      "loss": 1.2165,
+      "step": 8814
+    },
+    {
+      "epoch": 0.40609020131754736,
+      "grad_norm": 1.0366068775707045,
+      "learning_rate": 0.00013455307962354795,
+      "loss": 1.2928,
+      "step": 8815
+    },
+    {
+      "epoch": 0.40613626940618236,
+      "grad_norm": 1.1889242117786114,
+      "learning_rate": 0.00013453907736849386,
+      "loss": 1.3491,
+      "step": 8816
+    },
+    {
+      "epoch": 0.40618233749481736,
+      "grad_norm": 1.342707681298014,
+      "learning_rate": 0.00013452507434448646,
+      "loss": 1.4685,
+      "step": 8817
+    },
+    {
+      "epoch": 0.40622840558345236,
+      "grad_norm": 5.074047889228625,
+      "learning_rate": 0.00013451107055183746,
+      "loss": 1.0352,
+      "step": 8818
+    },
+    {
+      "epoch": 0.40627447367208736,
+      "grad_norm": 0.8508419427643835,
+      "learning_rate": 0.0001344970659908587,
+      "loss": 1.47,
+      "step": 8819
+    },
+    {
+      "epoch": 0.40632054176072235,
+      "grad_norm": 2.4282843844668585,
+      "learning_rate": 0.00013448306066186187,
+      "loss": 1.2942,
+      "step": 8820
+    },
+    {
+      "epoch": 0.40636660984935735,
+      "grad_norm": 1.0227166199812563,
+      "learning_rate": 0.0001344690545651588,
+      "loss": 1.3625,
+      "step": 8821
+    },
+    {
+      "epoch": 0.40641267793799235,
+      "grad_norm": 1.7984481686706237,
+      "learning_rate": 0.00013445504770106138,
+      "loss": 1.3193,
+      "step": 8822
+    },
+    {
+      "epoch": 0.40645874602662735,
+      "grad_norm": 1.8182825968376626,
+      "learning_rate": 0.00013444104006988137,
+      "loss": 1.2871,
+      "step": 8823
+    },
+    {
+      "epoch": 0.40650481411526235,
+      "grad_norm": 2.29193087761114,
+      "learning_rate": 0.00013442703167193068,
+      "loss": 1.84,
+      "step": 8824
+    },
+    {
+      "epoch": 0.40655088220389735,
+      "grad_norm": 1.027998773773041,
+      "learning_rate": 0.00013441302250752113,
+      "loss": 1.1573,
+      "step": 8825
+    },
+    {
+      "epoch": 0.40659695029253234,
+      "grad_norm": 3.9604887740098924,
+      "learning_rate": 0.00013439901257696467,
+      "loss": 1.2611,
+      "step": 8826
+    },
+    {
+      "epoch": 0.40664301838116734,
+      "grad_norm": 1.176285883500121,
+      "learning_rate": 0.00013438500188057314,
+      "loss": 1.4276,
+      "step": 8827
+    },
+    {
+      "epoch": 0.40668908646980234,
+      "grad_norm": 0.8916950213384552,
+      "learning_rate": 0.00013437099041865857,
+      "loss": 1.1561,
+      "step": 8828
+    },
+    {
+      "epoch": 0.4067351545584374,
+      "grad_norm": 1.1325567198375637,
+      "learning_rate": 0.00013435697819153276,
+      "loss": 1.3518,
+      "step": 8829
+    },
+    {
+      "epoch": 0.4067812226470724,
+      "grad_norm": 1.2770565981134063,
+      "learning_rate": 0.00013434296519950777,
+      "loss": 1.3458,
+      "step": 8830
+    },
+    {
+      "epoch": 0.4068272907357074,
+      "grad_norm": 1.0594899088882483,
+      "learning_rate": 0.00013432895144289554,
+      "loss": 1.361,
+      "step": 8831
+    },
+    {
+      "epoch": 0.4068733588243424,
+      "grad_norm": 1.1410178041028414,
+      "learning_rate": 0.00013431493692200804,
+      "loss": 1.3315,
+      "step": 8832
+    },
+    {
+      "epoch": 0.4069194269129774,
+      "grad_norm": 1.0093932377534756,
+      "learning_rate": 0.00013430092163715736,
+      "loss": 1.4373,
+      "step": 8833
+    },
+    {
+      "epoch": 0.4069654950016124,
+      "grad_norm": 1.2984540093334895,
+      "learning_rate": 0.00013428690558865545,
+      "loss": 1.2073,
+      "step": 8834
+    },
+    {
+      "epoch": 0.4070115630902474,
+      "grad_norm": 0.9500342491027316,
+      "learning_rate": 0.00013427288877681436,
+      "loss": 1.2928,
+      "step": 8835
+    },
+    {
+      "epoch": 0.4070576311788824,
+      "grad_norm": 0.8771553900537433,
+      "learning_rate": 0.00013425887120194617,
+      "loss": 1.5398,
+      "step": 8836
+    },
+    {
+      "epoch": 0.4071036992675174,
+      "grad_norm": 0.8645612797988258,
+      "learning_rate": 0.00013424485286436298,
+      "loss": 1.4485,
+      "step": 8837
+    },
+    {
+      "epoch": 0.4071497673561524,
+      "grad_norm": 1.7483103827303816,
+      "learning_rate": 0.00013423083376437687,
+      "loss": 1.509,
+      "step": 8838
+    },
+    {
+      "epoch": 0.4071958354447874,
+      "grad_norm": 1.0942707271348546,
+      "learning_rate": 0.00013421681390229993,
+      "loss": 1.1156,
+      "step": 8839
+    },
+    {
+      "epoch": 0.4072419035334224,
+      "grad_norm": 0.9371677675769393,
+      "learning_rate": 0.00013420279327844425,
+      "loss": 1.3195,
+      "step": 8840
+    },
+    {
+      "epoch": 0.4072879716220574,
+      "grad_norm": 1.234087548535588,
+      "learning_rate": 0.00013418877189312208,
+      "loss": 1.1581,
+      "step": 8841
+    },
+    {
+      "epoch": 0.40733403971069243,
+      "grad_norm": 1.6095697261533242,
+      "learning_rate": 0.00013417474974664547,
+      "loss": 1.4141,
+      "step": 8842
+    },
+    {
+      "epoch": 0.40738010779932743,
+      "grad_norm": 0.822214923310346,
+      "learning_rate": 0.0001341607268393267,
+      "loss": 1.5645,
+      "step": 8843
+    },
+    {
+      "epoch": 0.4074261758879624,
+      "grad_norm": 3.002518320866853,
+      "learning_rate": 0.00013414670317147789,
+      "loss": 1.2968,
+      "step": 8844
+    },
+    {
+      "epoch": 0.4074722439765974,
+      "grad_norm": 1.478539707255822,
+      "learning_rate": 0.0001341326787434113,
+      "loss": 1.4899,
+      "step": 8845
+    },
+    {
+      "epoch": 0.4075183120652324,
+      "grad_norm": 1.385512491578341,
+      "learning_rate": 0.00013411865355543914,
+      "loss": 1.197,
+      "step": 8846
+    },
+    {
+      "epoch": 0.4075643801538674,
+      "grad_norm": 1.6078892887758485,
+      "learning_rate": 0.00013410462760787365,
+      "loss": 1.2587,
+      "step": 8847
+    },
+    {
+      "epoch": 0.4076104482425024,
+      "grad_norm": 0.7638214282025144,
+      "learning_rate": 0.0001340906009010271,
+      "loss": 1.1835,
+      "step": 8848
+    },
+    {
+      "epoch": 0.4076565163311374,
+      "grad_norm": 1.29491570929376,
+      "learning_rate": 0.0001340765734352118,
+      "loss": 1.2703,
+      "step": 8849
+    },
+    {
+      "epoch": 0.4077025844197724,
+      "grad_norm": 0.9075201303765505,
+      "learning_rate": 0.00013406254521074003,
+      "loss": 1.1796,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4077486525084074,
+      "grad_norm": 1.150954595360353,
+      "learning_rate": 0.00013404851622792404,
+      "loss": 1.1987,
+      "step": 8851
+    },
+    {
+      "epoch": 0.4077947205970424,
+      "grad_norm": 0.8397657506251401,
+      "learning_rate": 0.00013403448648707626,
+      "loss": 1.0051,
+      "step": 8852
+    },
+    {
+      "epoch": 0.4078407886856774,
+      "grad_norm": 3.268334771333029,
+      "learning_rate": 0.000134020455988509,
+      "loss": 1.3965,
+      "step": 8853
+    },
+    {
+      "epoch": 0.4078868567743124,
+      "grad_norm": 1.110145844554267,
+      "learning_rate": 0.0001340064247325346,
+      "loss": 1.3966,
+      "step": 8854
+    },
+    {
+      "epoch": 0.4079329248629474,
+      "grad_norm": 1.14610833924757,
+      "learning_rate": 0.00013399239271946544,
+      "loss": 1.2861,
+      "step": 8855
+    },
+    {
+      "epoch": 0.40797899295158246,
+      "grad_norm": 0.8487815500458055,
+      "learning_rate": 0.000133978359949614,
+      "loss": 1.3644,
+      "step": 8856
+    },
+    {
+      "epoch": 0.40802506104021746,
+      "grad_norm": 1.107411545973041,
+      "learning_rate": 0.0001339643264232926,
+      "loss": 1.3145,
+      "step": 8857
+    },
+    {
+      "epoch": 0.40807112912885246,
+      "grad_norm": 1.1948476913268238,
+      "learning_rate": 0.0001339502921408137,
+      "loss": 1.4063,
+      "step": 8858
+    },
+    {
+      "epoch": 0.40811719721748746,
+      "grad_norm": 1.3973228194415444,
+      "learning_rate": 0.00013393625710248977,
+      "loss": 1.4725,
+      "step": 8859
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 2.9116141594558456,
+      "learning_rate": 0.00013392222130863325,
+      "loss": 1.6165,
+      "step": 8860
+    },
+    {
+      "epoch": 0.40820933339475746,
+      "grad_norm": 1.3619615681448294,
+      "learning_rate": 0.00013390818475955666,
+      "loss": 1.3248,
+      "step": 8861
+    },
+    {
+      "epoch": 0.40825540148339245,
+      "grad_norm": 1.4087568478092263,
+      "learning_rate": 0.00013389414745557248,
+      "loss": 1.2291,
+      "step": 8862
+    },
+    {
+      "epoch": 0.40830146957202745,
+      "grad_norm": 0.7992804490614646,
+      "learning_rate": 0.0001338801093969932,
+      "loss": 1.5738,
+      "step": 8863
+    },
+    {
+      "epoch": 0.40834753766066245,
+      "grad_norm": 1.0183264371176153,
+      "learning_rate": 0.00013386607058413138,
+      "loss": 1.1535,
+      "step": 8864
+    },
+    {
+      "epoch": 0.40839360574929745,
+      "grad_norm": 2.387871291237416,
+      "learning_rate": 0.00013385203101729957,
+      "loss": 1.2729,
+      "step": 8865
+    },
+    {
+      "epoch": 0.40843967383793245,
+      "grad_norm": 0.9265655982949055,
+      "learning_rate": 0.00013383799069681035,
+      "loss": 1.3322,
+      "step": 8866
+    },
+    {
+      "epoch": 0.40848574192656745,
+      "grad_norm": 0.961510969614737,
+      "learning_rate": 0.00013382394962297624,
+      "loss": 1.4837,
+      "step": 8867
+    },
+    {
+      "epoch": 0.40853181001520245,
+      "grad_norm": 1.2433984089910644,
+      "learning_rate": 0.00013380990779610994,
+      "loss": 1.5861,
+      "step": 8868
+    },
+    {
+      "epoch": 0.40857787810383744,
+      "grad_norm": 1.1755816156331196,
+      "learning_rate": 0.000133795865216524,
+      "loss": 1.3392,
+      "step": 8869
+    },
+    {
+      "epoch": 0.4086239461924725,
+      "grad_norm": 0.9251922323135331,
+      "learning_rate": 0.00013378182188453105,
+      "loss": 1.4973,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4086700142811075,
+      "grad_norm": 2.7832650971761907,
+      "learning_rate": 0.00013376777780044376,
+      "loss": 1.1416,
+      "step": 8871
+    },
+    {
+      "epoch": 0.4087160823697425,
+      "grad_norm": 0.9555013713377866,
+      "learning_rate": 0.00013375373296457482,
+      "loss": 1.5914,
+      "step": 8872
+    },
+    {
+      "epoch": 0.4087621504583775,
+      "grad_norm": 0.8538345957172868,
+      "learning_rate": 0.00013373968737723688,
+      "loss": 1.0542,
+      "step": 8873
+    },
+    {
+      "epoch": 0.4088082185470125,
+      "grad_norm": 1.745720019341563,
+      "learning_rate": 0.0001337256410387427,
+      "loss": 1.352,
+      "step": 8874
+    },
+    {
+      "epoch": 0.4088542866356475,
+      "grad_norm": 1.005768772770039,
+      "learning_rate": 0.00013371159394940486,
+      "loss": 1.3747,
+      "step": 8875
+    },
+    {
+      "epoch": 0.4089003547242825,
+      "grad_norm": 1.0443680226944694,
+      "learning_rate": 0.0001336975461095362,
+      "loss": 1.4331,
+      "step": 8876
+    },
+    {
+      "epoch": 0.4089464228129175,
+      "grad_norm": 0.9281955930744453,
+      "learning_rate": 0.0001336834975194495,
+      "loss": 1.1812,
+      "step": 8877
+    },
+    {
+      "epoch": 0.4089924909015525,
+      "grad_norm": 0.8513892891177028,
+      "learning_rate": 0.00013366944817945744,
+      "loss": 1.253,
+      "step": 8878
+    },
+    {
+      "epoch": 0.4090385589901875,
+      "grad_norm": 2.296744676718522,
+      "learning_rate": 0.00013365539808987288,
+      "loss": 1.4903,
+      "step": 8879
+    },
+    {
+      "epoch": 0.4090846270788225,
+      "grad_norm": 1.6289333909019514,
+      "learning_rate": 0.00013364134725100856,
+      "loss": 1.2954,
+      "step": 8880
+    },
+    {
+      "epoch": 0.4091306951674575,
+      "grad_norm": 1.208095248734138,
+      "learning_rate": 0.00013362729566317736,
+      "loss": 1.3082,
+      "step": 8881
+    },
+    {
+      "epoch": 0.4091767632560925,
+      "grad_norm": 1.1670510499043072,
+      "learning_rate": 0.00013361324332669203,
+      "loss": 1.4551,
+      "step": 8882
+    },
+    {
+      "epoch": 0.40922283134472753,
+      "grad_norm": 3.171254960437044,
+      "learning_rate": 0.00013359919024186553,
+      "loss": 1.2619,
+      "step": 8883
+    },
+    {
+      "epoch": 0.40926889943336253,
+      "grad_norm": 2.0380228185037845,
+      "learning_rate": 0.00013358513640901064,
+      "loss": 1.2634,
+      "step": 8884
+    },
+    {
+      "epoch": 0.40931496752199753,
+      "grad_norm": 1.9280913528054786,
+      "learning_rate": 0.0001335710818284403,
+      "loss": 1.3104,
+      "step": 8885
+    },
+    {
+      "epoch": 0.40936103561063253,
+      "grad_norm": 2.0215452462437122,
+      "learning_rate": 0.00013355702650046734,
+      "loss": 1.2901,
+      "step": 8886
+    },
+    {
+      "epoch": 0.4094071036992675,
+      "grad_norm": 1.0467030502012749,
+      "learning_rate": 0.00013354297042540475,
+      "loss": 1.3194,
+      "step": 8887
+    },
+    {
+      "epoch": 0.4094531717879025,
+      "grad_norm": 1.1319732456410105,
+      "learning_rate": 0.00013352891360356544,
+      "loss": 1.1457,
+      "step": 8888
+    },
+    {
+      "epoch": 0.4094992398765375,
+      "grad_norm": 1.1995557389147433,
+      "learning_rate": 0.00013351485603526234,
+      "loss": 1.19,
+      "step": 8889
+    },
+    {
+      "epoch": 0.4095453079651725,
+      "grad_norm": 1.442256093681951,
+      "learning_rate": 0.00013350079772080843,
+      "loss": 1.6592,
+      "step": 8890
+    },
+    {
+      "epoch": 0.4095913760538075,
+      "grad_norm": 0.9332702174534206,
+      "learning_rate": 0.00013348673866051672,
+      "loss": 1.0766,
+      "step": 8891
+    },
+    {
+      "epoch": 0.4096374441424425,
+      "grad_norm": 1.0314485545515868,
+      "learning_rate": 0.00013347267885470022,
+      "loss": 1.3095,
+      "step": 8892
+    },
+    {
+      "epoch": 0.4096835122310775,
+      "grad_norm": 1.0370803077888218,
+      "learning_rate": 0.0001334586183036719,
+      "loss": 1.4894,
+      "step": 8893
+    },
+    {
+      "epoch": 0.4097295803197125,
+      "grad_norm": 2.5423292288769583,
+      "learning_rate": 0.00013344455700774478,
+      "loss": 1.0772,
+      "step": 8894
+    },
+    {
+      "epoch": 0.4097756484083475,
+      "grad_norm": 1.266843745257189,
+      "learning_rate": 0.00013343049496723194,
+      "loss": 1.2443,
+      "step": 8895
+    },
+    {
+      "epoch": 0.4098217164969825,
+      "grad_norm": 1.188729994191268,
+      "learning_rate": 0.00013341643218244649,
+      "loss": 1.3541,
+      "step": 8896
+    },
+    {
+      "epoch": 0.40986778458561757,
+      "grad_norm": 1.0738433833517873,
+      "learning_rate": 0.00013340236865370147,
+      "loss": 1.3828,
+      "step": 8897
+    },
+    {
+      "epoch": 0.40991385267425257,
+      "grad_norm": 0.9243162489147803,
+      "learning_rate": 0.00013338830438130997,
+      "loss": 1.2736,
+      "step": 8898
+    },
+    {
+      "epoch": 0.40995992076288756,
+      "grad_norm": 1.5654690568798904,
+      "learning_rate": 0.00013337423936558512,
+      "loss": 1.0482,
+      "step": 8899
+    },
+    {
+      "epoch": 0.41000598885152256,
+      "grad_norm": 3.6706825757766417,
+      "learning_rate": 0.00013336017360684004,
+      "loss": 1.6483,
+      "step": 8900
+    },
+    {
+      "epoch": 0.41005205694015756,
+      "grad_norm": 0.950844696346734,
+      "learning_rate": 0.0001333461071053879,
+      "loss": 1.3264,
+      "step": 8901
+    },
+    {
+      "epoch": 0.41009812502879256,
+      "grad_norm": 1.6368125758355372,
+      "learning_rate": 0.00013333203986154191,
+      "loss": 1.4747,
+      "step": 8902
+    },
+    {
+      "epoch": 0.41014419311742756,
+      "grad_norm": 1.1509297410770487,
+      "learning_rate": 0.00013331797187561515,
+      "loss": 1.4453,
+      "step": 8903
+    },
+    {
+      "epoch": 0.41019026120606256,
+      "grad_norm": 1.7822960157894228,
+      "learning_rate": 0.0001333039031479209,
+      "loss": 1.215,
+      "step": 8904
+    },
+    {
+      "epoch": 0.41023632929469755,
+      "grad_norm": 1.436068854324963,
+      "learning_rate": 0.00013328983367877237,
+      "loss": 1.5338,
+      "step": 8905
+    },
+    {
+      "epoch": 0.41028239738333255,
+      "grad_norm": 1.321075442901914,
+      "learning_rate": 0.00013327576346848272,
+      "loss": 1.3484,
+      "step": 8906
+    },
+    {
+      "epoch": 0.41032846547196755,
+      "grad_norm": 1.2213583972259447,
+      "learning_rate": 0.00013326169251736528,
+      "loss": 1.4234,
+      "step": 8907
+    },
+    {
+      "epoch": 0.41037453356060255,
+      "grad_norm": 1.0242427522630466,
+      "learning_rate": 0.0001332476208257333,
+      "loss": 1.3938,
+      "step": 8908
+    },
+    {
+      "epoch": 0.41042060164923755,
+      "grad_norm": 1.1696189883022696,
+      "learning_rate": 0.00013323354839390002,
+      "loss": 1.3646,
+      "step": 8909
+    },
+    {
+      "epoch": 0.4104666697378726,
+      "grad_norm": 1.379899745507512,
+      "learning_rate": 0.0001332194752221788,
+      "loss": 1.2011,
+      "step": 8910
+    },
+    {
+      "epoch": 0.4105127378265076,
+      "grad_norm": 0.9302108458708497,
+      "learning_rate": 0.00013320540131088292,
+      "loss": 1.2396,
+      "step": 8911
+    },
+    {
+      "epoch": 0.4105588059151426,
+      "grad_norm": 1.3500530800745816,
+      "learning_rate": 0.0001331913266603257,
+      "loss": 1.4033,
+      "step": 8912
+    },
+    {
+      "epoch": 0.4106048740037776,
+      "grad_norm": 1.4674280149003474,
+      "learning_rate": 0.0001331772512708205,
+      "loss": 1.3143,
+      "step": 8913
+    },
+    {
+      "epoch": 0.4106509420924126,
+      "grad_norm": 1.148725007733727,
+      "learning_rate": 0.0001331631751426807,
+      "loss": 1.3447,
+      "step": 8914
+    },
+    {
+      "epoch": 0.4106970101810476,
+      "grad_norm": 1.6115482603500533,
+      "learning_rate": 0.00013314909827621968,
+      "loss": 1.4621,
+      "step": 8915
+    },
+    {
+      "epoch": 0.4107430782696826,
+      "grad_norm": 1.2393564379601014,
+      "learning_rate": 0.0001331350206717508,
+      "loss": 1.1684,
+      "step": 8916
+    },
+    {
+      "epoch": 0.4107891463583176,
+      "grad_norm": 1.4704031932878008,
+      "learning_rate": 0.0001331209423295875,
+      "loss": 1.371,
+      "step": 8917
+    },
+    {
+      "epoch": 0.4108352144469526,
+      "grad_norm": 1.003825577455045,
+      "learning_rate": 0.00013310686325004322,
+      "loss": 1.1004,
+      "step": 8918
+    },
+    {
+      "epoch": 0.4108812825355876,
+      "grad_norm": 1.7646795697503967,
+      "learning_rate": 0.00013309278343343137,
+      "loss": 1.3327,
+      "step": 8919
+    },
+    {
+      "epoch": 0.4109273506242226,
+      "grad_norm": 1.1667481553550492,
+      "learning_rate": 0.00013307870288006547,
+      "loss": 1.1936,
+      "step": 8920
+    },
+    {
+      "epoch": 0.4109734187128576,
+      "grad_norm": 1.0965311995995943,
+      "learning_rate": 0.00013306462159025896,
+      "loss": 1.4457,
+      "step": 8921
+    },
+    {
+      "epoch": 0.4110194868014926,
+      "grad_norm": 1.3484422234515336,
+      "learning_rate": 0.00013305053956432533,
+      "loss": 1.434,
+      "step": 8922
+    },
+    {
+      "epoch": 0.4110655548901276,
+      "grad_norm": 2.293965143251389,
+      "learning_rate": 0.00013303645680257814,
+      "loss": 1.3984,
+      "step": 8923
+    },
+    {
+      "epoch": 0.41111162297876264,
+      "grad_norm": 0.9200052297049092,
+      "learning_rate": 0.00013302237330533085,
+      "loss": 1.4154,
+      "step": 8924
+    },
+    {
+      "epoch": 0.41115769106739763,
+      "grad_norm": 0.9433400728266118,
+      "learning_rate": 0.00013300828907289707,
+      "loss": 1.4123,
+      "step": 8925
+    },
+    {
+      "epoch": 0.41120375915603263,
+      "grad_norm": 1.1853847926943248,
+      "learning_rate": 0.00013299420410559028,
+      "loss": 1.3232,
+      "step": 8926
+    },
+    {
+      "epoch": 0.41124982724466763,
+      "grad_norm": 1.1420917578014929,
+      "learning_rate": 0.00013298011840372418,
+      "loss": 1.2277,
+      "step": 8927
+    },
+    {
+      "epoch": 0.41129589533330263,
+      "grad_norm": 1.4177588966909653,
+      "learning_rate": 0.00013296603196761225,
+      "loss": 1.273,
+      "step": 8928
+    },
+    {
+      "epoch": 0.41134196342193763,
+      "grad_norm": 1.5723965163601654,
+      "learning_rate": 0.00013295194479756817,
+      "loss": 1.332,
+      "step": 8929
+    },
+    {
+      "epoch": 0.4113880315105726,
+      "grad_norm": 1.2442413426208674,
+      "learning_rate": 0.0001329378568939055,
+      "loss": 1.2678,
+      "step": 8930
+    },
+    {
+      "epoch": 0.4114340995992076,
+      "grad_norm": 2.2123396020908332,
+      "learning_rate": 0.00013292376825693793,
+      "loss": 1.2042,
+      "step": 8931
+    },
+    {
+      "epoch": 0.4114801676878426,
+      "grad_norm": 1.2396288363775008,
+      "learning_rate": 0.00013290967888697911,
+      "loss": 1.5147,
+      "step": 8932
+    },
+    {
+      "epoch": 0.4115262357764776,
+      "grad_norm": 0.9719240412149261,
+      "learning_rate": 0.00013289558878434274,
+      "loss": 1.3285,
+      "step": 8933
+    },
+    {
+      "epoch": 0.4115723038651126,
+      "grad_norm": 1.4852125633526274,
+      "learning_rate": 0.0001328814979493425,
+      "loss": 1.6165,
+      "step": 8934
+    },
+    {
+      "epoch": 0.4116183719537476,
+      "grad_norm": 1.0798015896031088,
+      "learning_rate": 0.00013286740638229205,
+      "loss": 1.4843,
+      "step": 8935
+    },
+    {
+      "epoch": 0.4116644400423826,
+      "grad_norm": 1.114860024250705,
+      "learning_rate": 0.00013285331408350516,
+      "loss": 1.1063,
+      "step": 8936
+    },
+    {
+      "epoch": 0.41171050813101767,
+      "grad_norm": 1.3717628868159362,
+      "learning_rate": 0.00013283922105329556,
+      "loss": 1.5374,
+      "step": 8937
+    },
+    {
+      "epoch": 0.41175657621965267,
+      "grad_norm": 0.9822882399208942,
+      "learning_rate": 0.00013282512729197708,
+      "loss": 1.5365,
+      "step": 8938
+    },
+    {
+      "epoch": 0.41180264430828767,
+      "grad_norm": 1.4633427865358428,
+      "learning_rate": 0.00013281103279986334,
+      "loss": 1.1875,
+      "step": 8939
+    },
+    {
+      "epoch": 0.41184871239692267,
+      "grad_norm": 1.6641502793980412,
+      "learning_rate": 0.00013279693757726826,
+      "loss": 1.3342,
+      "step": 8940
+    },
+    {
+      "epoch": 0.41189478048555767,
+      "grad_norm": 1.1782783523307487,
+      "learning_rate": 0.00013278284162450556,
+      "loss": 1.3194,
+      "step": 8941
+    },
+    {
+      "epoch": 0.41194084857419266,
+      "grad_norm": 1.2482692459815907,
+      "learning_rate": 0.00013276874494188914,
+      "loss": 1.529,
+      "step": 8942
+    },
+    {
+      "epoch": 0.41198691666282766,
+      "grad_norm": 1.7910660679067467,
+      "learning_rate": 0.00013275464752973278,
+      "loss": 1.0889,
+      "step": 8943
+    },
+    {
+      "epoch": 0.41203298475146266,
+      "grad_norm": 1.421230472470317,
+      "learning_rate": 0.00013274054938835033,
+      "loss": 1.551,
+      "step": 8944
+    },
+    {
+      "epoch": 0.41207905284009766,
+      "grad_norm": 1.6916148304929948,
+      "learning_rate": 0.00013272645051805573,
+      "loss": 1.182,
+      "step": 8945
+    },
+    {
+      "epoch": 0.41212512092873266,
+      "grad_norm": 3.56904562674323,
+      "learning_rate": 0.0001327123509191628,
+      "loss": 1.3752,
+      "step": 8946
+    },
+    {
+      "epoch": 0.41217118901736766,
+      "grad_norm": 1.8308106298471876,
+      "learning_rate": 0.00013269825059198547,
+      "loss": 1.1907,
+      "step": 8947
+    },
+    {
+      "epoch": 0.41221725710600265,
+      "grad_norm": 1.5061219932622953,
+      "learning_rate": 0.00013268414953683766,
+      "loss": 1.2137,
+      "step": 8948
+    },
+    {
+      "epoch": 0.41226332519463765,
+      "grad_norm": 1.317216726536338,
+      "learning_rate": 0.0001326700477540333,
+      "loss": 1.2109,
+      "step": 8949
+    },
+    {
+      "epoch": 0.41230939328327265,
+      "grad_norm": 1.4810228924453899,
+      "learning_rate": 0.00013265594524388633,
+      "loss": 1.4475,
+      "step": 8950
+    },
+    {
+      "epoch": 0.4123554613719077,
+      "grad_norm": 2.703059965173799,
+      "learning_rate": 0.00013264184200671077,
+      "loss": 1.275,
+      "step": 8951
+    },
+    {
+      "epoch": 0.4124015294605427,
+      "grad_norm": 1.5396952504157306,
+      "learning_rate": 0.00013262773804282053,
+      "loss": 1.1909,
+      "step": 8952
+    },
+    {
+      "epoch": 0.4124475975491777,
+      "grad_norm": 2.1462350850084433,
+      "learning_rate": 0.00013261363335252967,
+      "loss": 1.1077,
+      "step": 8953
+    },
+    {
+      "epoch": 0.4124936656378127,
+      "grad_norm": 1.0698149907188366,
+      "learning_rate": 0.0001325995279361522,
+      "loss": 1.3436,
+      "step": 8954
+    },
+    {
+      "epoch": 0.4125397337264477,
+      "grad_norm": 0.8839615152921196,
+      "learning_rate": 0.00013258542179400207,
+      "loss": 1.4688,
+      "step": 8955
+    },
+    {
+      "epoch": 0.4125858018150827,
+      "grad_norm": 3.7756498902592113,
+      "learning_rate": 0.00013257131492639345,
+      "loss": 1.4641,
+      "step": 8956
+    },
+    {
+      "epoch": 0.4126318699037177,
+      "grad_norm": 1.074870684112147,
+      "learning_rate": 0.00013255720733364035,
+      "loss": 1.3245,
+      "step": 8957
+    },
+    {
+      "epoch": 0.4126779379923527,
+      "grad_norm": 1.3702138976043494,
+      "learning_rate": 0.00013254309901605685,
+      "loss": 1.1795,
+      "step": 8958
+    },
+    {
+      "epoch": 0.4127240060809877,
+      "grad_norm": 1.1551475224041075,
+      "learning_rate": 0.000132528989973957,
+      "loss": 1.4682,
+      "step": 8959
+    },
+    {
+      "epoch": 0.4127700741696227,
+      "grad_norm": 1.2290799498416596,
+      "learning_rate": 0.00013251488020765503,
+      "loss": 1.4928,
+      "step": 8960
+    },
+    {
+      "epoch": 0.4128161422582577,
+      "grad_norm": 4.343172786365784,
+      "learning_rate": 0.00013250076971746497,
+      "loss": 1.1401,
+      "step": 8961
+    },
+    {
+      "epoch": 0.4128622103468927,
+      "grad_norm": 2.5392609993244477,
+      "learning_rate": 0.00013248665850370102,
+      "loss": 1.4516,
+      "step": 8962
+    },
+    {
+      "epoch": 0.4129082784355277,
+      "grad_norm": 1.5192611521669706,
+      "learning_rate": 0.00013247254656667727,
+      "loss": 1.3731,
+      "step": 8963
+    },
+    {
+      "epoch": 0.4129543465241627,
+      "grad_norm": 1.072769949934057,
+      "learning_rate": 0.000132458433906708,
+      "loss": 1.3302,
+      "step": 8964
+    },
+    {
+      "epoch": 0.41300041461279774,
+      "grad_norm": 1.004929271142936,
+      "learning_rate": 0.00013244432052410734,
+      "loss": 1.3228,
+      "step": 8965
+    },
+    {
+      "epoch": 0.41304648270143274,
+      "grad_norm": 2.1288680150660397,
+      "learning_rate": 0.0001324302064191895,
+      "loss": 0.9803,
+      "step": 8966
+    },
+    {
+      "epoch": 0.41309255079006774,
+      "grad_norm": 1.4648022164618386,
+      "learning_rate": 0.00013241609159226876,
+      "loss": 1.1328,
+      "step": 8967
+    },
+    {
+      "epoch": 0.41313861887870273,
+      "grad_norm": 1.3813089781563959,
+      "learning_rate": 0.00013240197604365926,
+      "loss": 1.4074,
+      "step": 8968
+    },
+    {
+      "epoch": 0.41318468696733773,
+      "grad_norm": 1.6021167611059774,
+      "learning_rate": 0.00013238785977367538,
+      "loss": 1.5007,
+      "step": 8969
+    },
+    {
+      "epoch": 0.41323075505597273,
+      "grad_norm": 2.7710946310838422,
+      "learning_rate": 0.0001323737427826313,
+      "loss": 1.3243,
+      "step": 8970
+    },
+    {
+      "epoch": 0.41327682314460773,
+      "grad_norm": 4.722673723470433,
+      "learning_rate": 0.00013235962507084136,
+      "loss": 1.1273,
+      "step": 8971
+    },
+    {
+      "epoch": 0.41332289123324273,
+      "grad_norm": 2.7347034957404586,
+      "learning_rate": 0.00013234550663861983,
+      "loss": 1.2704,
+      "step": 8972
+    },
+    {
+      "epoch": 0.4133689593218777,
+      "grad_norm": 1.1326267570690653,
+      "learning_rate": 0.0001323313874862811,
+      "loss": 1.1521,
+      "step": 8973
+    },
+    {
+      "epoch": 0.4134150274105127,
+      "grad_norm": 1.2636572480463262,
+      "learning_rate": 0.00013231726761413944,
+      "loss": 1.6793,
+      "step": 8974
+    },
+    {
+      "epoch": 0.4134610954991477,
+      "grad_norm": 2.3611034623038076,
+      "learning_rate": 0.00013230314702250922,
+      "loss": 1.0356,
+      "step": 8975
+    },
+    {
+      "epoch": 0.4135071635877827,
+      "grad_norm": 2.648456105410163,
+      "learning_rate": 0.00013228902571170485,
+      "loss": 1.2143,
+      "step": 8976
+    },
+    {
+      "epoch": 0.4135532316764177,
+      "grad_norm": 2.6827199906730996,
+      "learning_rate": 0.00013227490368204064,
+      "loss": 1.4803,
+      "step": 8977
+    },
+    {
+      "epoch": 0.4135992997650528,
+      "grad_norm": 2.24549939732952,
+      "learning_rate": 0.00013226078093383107,
+      "loss": 1.258,
+      "step": 8978
+    },
+    {
+      "epoch": 0.4136453678536878,
+      "grad_norm": 2.3699168312233985,
+      "learning_rate": 0.00013224665746739051,
+      "loss": 1.5602,
+      "step": 8979
+    },
+    {
+      "epoch": 0.41369143594232277,
+      "grad_norm": 3.5818038436963717,
+      "learning_rate": 0.00013223253328303343,
+      "loss": 1.4581,
+      "step": 8980
+    },
+    {
+      "epoch": 0.41373750403095777,
+      "grad_norm": 1.5562438853305356,
+      "learning_rate": 0.00013221840838107424,
+      "loss": 1.2877,
+      "step": 8981
+    },
+    {
+      "epoch": 0.41378357211959277,
+      "grad_norm": 10.120783052298904,
+      "learning_rate": 0.00013220428276182746,
+      "loss": 1.4457,
+      "step": 8982
+    },
+    {
+      "epoch": 0.41382964020822777,
+      "grad_norm": 1.6323706955056854,
+      "learning_rate": 0.0001321901564256075,
+      "loss": 1.3152,
+      "step": 8983
+    },
+    {
+      "epoch": 0.41387570829686277,
+      "grad_norm": 1.6412790090592682,
+      "learning_rate": 0.00013217602937272898,
+      "loss": 1.2884,
+      "step": 8984
+    },
+    {
+      "epoch": 0.41392177638549776,
+      "grad_norm": 2.838434406411495,
+      "learning_rate": 0.0001321619016035063,
+      "loss": 1.4697,
+      "step": 8985
+    },
+    {
+      "epoch": 0.41396784447413276,
+      "grad_norm": 1.6887385579526901,
+      "learning_rate": 0.000132147773118254,
+      "loss": 1.3501,
+      "step": 8986
+    },
+    {
+      "epoch": 0.41401391256276776,
+      "grad_norm": 1.7280441501746828,
+      "learning_rate": 0.00013213364391728667,
+      "loss": 1.2873,
+      "step": 8987
+    },
+    {
+      "epoch": 0.41405998065140276,
+      "grad_norm": 8.62693041200915,
+      "learning_rate": 0.00013211951400091885,
+      "loss": 1.2583,
+      "step": 8988
+    },
+    {
+      "epoch": 0.41410604874003776,
+      "grad_norm": 2.259329516678038,
+      "learning_rate": 0.00013210538336946512,
+      "loss": 1.4859,
+      "step": 8989
+    },
+    {
+      "epoch": 0.41415211682867276,
+      "grad_norm": 5.63163614783156,
+      "learning_rate": 0.00013209125202324007,
+      "loss": 1.4726,
+      "step": 8990
+    },
+    {
+      "epoch": 0.41419818491730775,
+      "grad_norm": 1.7605153790495212,
+      "learning_rate": 0.00013207711996255834,
+      "loss": 1.5303,
+      "step": 8991
+    },
+    {
+      "epoch": 0.4142442530059428,
+      "grad_norm": 1.582627115077177,
+      "learning_rate": 0.00013206298718773454,
+      "loss": 1.2869,
+      "step": 8992
+    },
+    {
+      "epoch": 0.4142903210945778,
+      "grad_norm": 1.9722332282987365,
+      "learning_rate": 0.0001320488536990833,
+      "loss": 1.3381,
+      "step": 8993
+    },
+    {
+      "epoch": 0.4143363891832128,
+      "grad_norm": 0.9303799300944425,
+      "learning_rate": 0.00013203471949691927,
+      "loss": 1.7284,
+      "step": 8994
+    },
+    {
+      "epoch": 0.4143824572718478,
+      "grad_norm": 5.865827197957227,
+      "learning_rate": 0.00013202058458155715,
+      "loss": 1.3861,
+      "step": 8995
+    },
+    {
+      "epoch": 0.4144285253604828,
+      "grad_norm": 1.6897655611223787,
+      "learning_rate": 0.00013200644895331162,
+      "loss": 1.3569,
+      "step": 8996
+    },
+    {
+      "epoch": 0.4144745934491178,
+      "grad_norm": 4.0381387875762496,
+      "learning_rate": 0.00013199231261249738,
+      "loss": 1.3608,
+      "step": 8997
+    },
+    {
+      "epoch": 0.4145206615377528,
+      "grad_norm": 2.5345764589998048,
+      "learning_rate": 0.00013197817555942915,
+      "loss": 1.4212,
+      "step": 8998
+    },
+    {
+      "epoch": 0.4145667296263878,
+      "grad_norm": 2.8090375213131558,
+      "learning_rate": 0.00013196403779442168,
+      "loss": 1.2354,
+      "step": 8999
+    },
+    {
+      "epoch": 0.4146127977150228,
+      "grad_norm": 3.3178601522648834,
+      "learning_rate": 0.0001319498993177897,
+      "loss": 1.3804,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4146588658036578,
+      "grad_norm": 1.7423122440388557,
+      "learning_rate": 0.000131935760129848,
+      "loss": 1.3761,
+      "step": 9001
+    },
+    {
+      "epoch": 0.4147049338922928,
+      "grad_norm": 5.3297592959370865,
+      "learning_rate": 0.0001319216202309114,
+      "loss": 1.3175,
+      "step": 9002
+    },
+    {
+      "epoch": 0.4147510019809278,
+      "grad_norm": 1.9586034906411873,
+      "learning_rate": 0.0001319074796212946,
+      "loss": 1.6569,
+      "step": 9003
+    },
+    {
+      "epoch": 0.4147970700695628,
+      "grad_norm": 2.5852083936104844,
+      "learning_rate": 0.0001318933383013125,
+      "loss": 1.1874,
+      "step": 9004
+    },
+    {
+      "epoch": 0.41484313815819784,
+      "grad_norm": 1.324866300917451,
+      "learning_rate": 0.00013187919627127988,
+      "loss": 1.5334,
+      "step": 9005
+    },
+    {
+      "epoch": 0.41488920624683284,
+      "grad_norm": 1.8733052851205634,
+      "learning_rate": 0.00013186505353151166,
+      "loss": 1.2581,
+      "step": 9006
+    },
+    {
+      "epoch": 0.41493527433546784,
+      "grad_norm": 5.151798286517676,
+      "learning_rate": 0.00013185091008232263,
+      "loss": 1.6156,
+      "step": 9007
+    },
+    {
+      "epoch": 0.41498134242410284,
+      "grad_norm": 2.1476922126317235,
+      "learning_rate": 0.00013183676592402774,
+      "loss": 1.1221,
+      "step": 9008
+    },
+    {
+      "epoch": 0.41502741051273784,
+      "grad_norm": 1.6777474832207488,
+      "learning_rate": 0.0001318226210569418,
+      "loss": 1.5563,
+      "step": 9009
+    },
+    {
+      "epoch": 0.41507347860137284,
+      "grad_norm": 1.9982210471695592,
+      "learning_rate": 0.00013180847548137977,
+      "loss": 1.3461,
+      "step": 9010
+    },
+    {
+      "epoch": 0.41511954669000783,
+      "grad_norm": 1.5958893467618678,
+      "learning_rate": 0.00013179432919765662,
+      "loss": 1.2971,
+      "step": 9011
+    },
+    {
+      "epoch": 0.41516561477864283,
+      "grad_norm": 2.118672966426837,
+      "learning_rate": 0.00013178018220608717,
+      "loss": 1.1797,
+      "step": 9012
+    },
+    {
+      "epoch": 0.41521168286727783,
+      "grad_norm": 5.8490565420387695,
+      "learning_rate": 0.00013176603450698653,
+      "loss": 1.5476,
+      "step": 9013
+    },
+    {
+      "epoch": 0.41525775095591283,
+      "grad_norm": 1.9179081544777368,
+      "learning_rate": 0.00013175188610066956,
+      "loss": 1.6721,
+      "step": 9014
+    },
+    {
+      "epoch": 0.41530381904454783,
+      "grad_norm": 3.0798618846354113,
+      "learning_rate": 0.00013173773698745128,
+      "loss": 1.5069,
+      "step": 9015
+    },
+    {
+      "epoch": 0.4153498871331828,
+      "grad_norm": 4.868860665824462,
+      "learning_rate": 0.0001317235871676467,
+      "loss": 1.2645,
+      "step": 9016
+    },
+    {
+      "epoch": 0.4153959552218178,
+      "grad_norm": 0.9698374434251916,
+      "learning_rate": 0.0001317094366415709,
+      "loss": 1.4199,
+      "step": 9017
+    },
+    {
+      "epoch": 0.4154420233104528,
+      "grad_norm": 2.768706586686723,
+      "learning_rate": 0.0001316952854095388,
+      "loss": 1.5555,
+      "step": 9018
+    },
+    {
+      "epoch": 0.4154880913990879,
+      "grad_norm": 2.7359582143504237,
+      "learning_rate": 0.00013168113347186556,
+      "loss": 1.6441,
+      "step": 9019
+    },
+    {
+      "epoch": 0.4155341594877229,
+      "grad_norm": 1.6058924962931616,
+      "learning_rate": 0.0001316669808288662,
+      "loss": 1.1476,
+      "step": 9020
+    },
+    {
+      "epoch": 0.4155802275763579,
+      "grad_norm": 1.2064291451766946,
+      "learning_rate": 0.0001316528274808558,
+      "loss": 1.4536,
+      "step": 9021
+    },
+    {
+      "epoch": 0.4156262956649929,
+      "grad_norm": 4.230031096095758,
+      "learning_rate": 0.00013163867342814946,
+      "loss": 1.3522,
+      "step": 9022
+    },
+    {
+      "epoch": 0.41567236375362787,
+      "grad_norm": 1.1756824807808932,
+      "learning_rate": 0.00013162451867106232,
+      "loss": 1.5831,
+      "step": 9023
+    },
+    {
+      "epoch": 0.41571843184226287,
+      "grad_norm": 1.391783989989877,
+      "learning_rate": 0.0001316103632099095,
+      "loss": 1.337,
+      "step": 9024
+    },
+    {
+      "epoch": 0.41576449993089787,
+      "grad_norm": 1.5835666478173798,
+      "learning_rate": 0.0001315962070450061,
+      "loss": 1.2606,
+      "step": 9025
+    },
+    {
+      "epoch": 0.41581056801953287,
+      "grad_norm": 2.5760629652656286,
+      "learning_rate": 0.00013158205017666736,
+      "loss": 1.693,
+      "step": 9026
+    },
+    {
+      "epoch": 0.41585663610816787,
+      "grad_norm": 1.0800593395074267,
+      "learning_rate": 0.0001315678926052084,
+      "loss": 1.5743,
+      "step": 9027
+    },
+    {
+      "epoch": 0.41590270419680286,
+      "grad_norm": 4.975823854722844,
+      "learning_rate": 0.0001315537343309445,
+      "loss": 1.0952,
+      "step": 9028
+    },
+    {
+      "epoch": 0.41594877228543786,
+      "grad_norm": 1.6768221386090993,
+      "learning_rate": 0.00013153957535419074,
+      "loss": 1.7053,
+      "step": 9029
+    },
+    {
+      "epoch": 0.41599484037407286,
+      "grad_norm": 1.1579776140183802,
+      "learning_rate": 0.00013152541567526245,
+      "loss": 1.5637,
+      "step": 9030
+    },
+    {
+      "epoch": 0.41604090846270786,
+      "grad_norm": 1.577111233563021,
+      "learning_rate": 0.00013151125529447483,
+      "loss": 1.1653,
+      "step": 9031
+    },
+    {
+      "epoch": 0.41608697655134286,
+      "grad_norm": 1.104901418408314,
+      "learning_rate": 0.00013149709421214314,
+      "loss": 1.4331,
+      "step": 9032
+    },
+    {
+      "epoch": 0.4161330446399779,
+      "grad_norm": 1.3390467451655332,
+      "learning_rate": 0.00013148293242858265,
+      "loss": 1.3825,
+      "step": 9033
+    },
+    {
+      "epoch": 0.4161791127286129,
+      "grad_norm": 1.3256239693174665,
+      "learning_rate": 0.00013146876994410867,
+      "loss": 1.5558,
+      "step": 9034
+    },
+    {
+      "epoch": 0.4162251808172479,
+      "grad_norm": 1.0110611400081402,
+      "learning_rate": 0.00013145460675903644,
+      "loss": 1.2876,
+      "step": 9035
+    },
+    {
+      "epoch": 0.4162712489058829,
+      "grad_norm": 9.60592084071775,
+      "learning_rate": 0.00013144044287368136,
+      "loss": 1.3977,
+      "step": 9036
+    },
+    {
+      "epoch": 0.4163173169945179,
+      "grad_norm": 1.7540049354546414,
+      "learning_rate": 0.00013142627828835872,
+      "loss": 1.2996,
+      "step": 9037
+    },
+    {
+      "epoch": 0.4163633850831529,
+      "grad_norm": 0.9075327740605813,
+      "learning_rate": 0.00013141211300338385,
+      "loss": 1.7113,
+      "step": 9038
+    },
+    {
+      "epoch": 0.4164094531717879,
+      "grad_norm": 2.452390441270735,
+      "learning_rate": 0.00013139794701907221,
+      "loss": 1.3441,
+      "step": 9039
+    },
+    {
+      "epoch": 0.4164555212604229,
+      "grad_norm": 1.3968063307577256,
+      "learning_rate": 0.00013138378033573909,
+      "loss": 1.2997,
+      "step": 9040
+    },
+    {
+      "epoch": 0.4165015893490579,
+      "grad_norm": 2.886994686747676,
+      "learning_rate": 0.0001313696129536999,
+      "loss": 1.1722,
+      "step": 9041
+    },
+    {
+      "epoch": 0.4165476574376929,
+      "grad_norm": 2.2828404630686876,
+      "learning_rate": 0.00013135544487327005,
+      "loss": 1.3903,
+      "step": 9042
+    },
+    {
+      "epoch": 0.4165937255263279,
+      "grad_norm": 6.584736554557089,
+      "learning_rate": 0.00013134127609476499,
+      "loss": 1.4727,
+      "step": 9043
+    },
+    {
+      "epoch": 0.4166397936149629,
+      "grad_norm": 2.554281644556027,
+      "learning_rate": 0.00013132710661850017,
+      "loss": 1.2952,
+      "step": 9044
+    },
+    {
+      "epoch": 0.4166858617035979,
+      "grad_norm": 0.771622778216391,
+      "learning_rate": 0.000131312936444791,
+      "loss": 1.3481,
+      "step": 9045
+    },
+    {
+      "epoch": 0.41673192979223295,
+      "grad_norm": 1.2860165089759537,
+      "learning_rate": 0.00013129876557395304,
+      "loss": 1.4831,
+      "step": 9046
+    },
+    {
+      "epoch": 0.41677799788086795,
+      "grad_norm": 17.925385454189943,
+      "learning_rate": 0.00013128459400630168,
+      "loss": 1.2196,
+      "step": 9047
+    },
+    {
+      "epoch": 0.41682406596950294,
+      "grad_norm": 14.566141935154938,
+      "learning_rate": 0.00013127042174215255,
+      "loss": 1.3524,
+      "step": 9048
+    },
+    {
+      "epoch": 0.41687013405813794,
+      "grad_norm": 1.2103971257106805,
+      "learning_rate": 0.00013125624878182102,
+      "loss": 1.4674,
+      "step": 9049
+    },
+    {
+      "epoch": 0.41691620214677294,
+      "grad_norm": 2.251603186245073,
+      "learning_rate": 0.00013124207512562277,
+      "loss": 1.5884,
+      "step": 9050
+    },
+    {
+      "epoch": 0.41696227023540794,
+      "grad_norm": 2.4974911021026895,
+      "learning_rate": 0.00013122790077387323,
+      "loss": 1.2276,
+      "step": 9051
+    },
+    {
+      "epoch": 0.41700833832404294,
+      "grad_norm": 1.1401480593638014,
+      "learning_rate": 0.00013121372572688807,
+      "loss": 1.6268,
+      "step": 9052
+    },
+    {
+      "epoch": 0.41705440641267794,
+      "grad_norm": 1.7906214586867426,
+      "learning_rate": 0.0001311995499849828,
+      "loss": 1.033,
+      "step": 9053
+    },
+    {
+      "epoch": 0.41710047450131293,
+      "grad_norm": 1.5722795109579206,
+      "learning_rate": 0.00013118537354847305,
+      "loss": 1.5266,
+      "step": 9054
+    },
+    {
+      "epoch": 0.41714654258994793,
+      "grad_norm": 3.7211331176825575,
+      "learning_rate": 0.0001311711964176744,
+      "loss": 1.4591,
+      "step": 9055
+    },
+    {
+      "epoch": 0.41719261067858293,
+      "grad_norm": 1.4500000474916335,
+      "learning_rate": 0.00013115701859290255,
+      "loss": 1.4022,
+      "step": 9056
+    },
+    {
+      "epoch": 0.41723867876721793,
+      "grad_norm": 1.8300169747001402,
+      "learning_rate": 0.00013114284007447312,
+      "loss": 1.4683,
+      "step": 9057
+    },
+    {
+      "epoch": 0.41728474685585293,
+      "grad_norm": 2.358021462876334,
+      "learning_rate": 0.0001311286608627017,
+      "loss": 1.3659,
+      "step": 9058
+    },
+    {
+      "epoch": 0.4173308149444879,
+      "grad_norm": 1.6197158426426284,
+      "learning_rate": 0.00013111448095790403,
+      "loss": 1.2979,
+      "step": 9059
+    },
+    {
+      "epoch": 0.417376883033123,
+      "grad_norm": 1.7285331101876262,
+      "learning_rate": 0.00013110030036039583,
+      "loss": 1.3036,
+      "step": 9060
+    },
+    {
+      "epoch": 0.417422951121758,
+      "grad_norm": 2.130523327541515,
+      "learning_rate": 0.00013108611907049276,
+      "loss": 1.2647,
+      "step": 9061
+    },
+    {
+      "epoch": 0.417469019210393,
+      "grad_norm": 3.618890566784322,
+      "learning_rate": 0.00013107193708851051,
+      "loss": 1.2876,
+      "step": 9062
+    },
+    {
+      "epoch": 0.417515087299028,
+      "grad_norm": 1.5948114754893314,
+      "learning_rate": 0.0001310577544147649,
+      "loss": 1.3847,
+      "step": 9063
+    },
+    {
+      "epoch": 0.417561155387663,
+      "grad_norm": 1.6554903102928935,
+      "learning_rate": 0.0001310435710495716,
+      "loss": 1.3468,
+      "step": 9064
+    },
+    {
+      "epoch": 0.417607223476298,
+      "grad_norm": 2.4134743946991466,
+      "learning_rate": 0.00013102938699324646,
+      "loss": 1.1914,
+      "step": 9065
+    },
+    {
+      "epoch": 0.41765329156493297,
+      "grad_norm": 2.2566986255721746,
+      "learning_rate": 0.0001310152022461052,
+      "loss": 1.5351,
+      "step": 9066
+    },
+    {
+      "epoch": 0.41769935965356797,
+      "grad_norm": 1.946073179137308,
+      "learning_rate": 0.00013100101680846364,
+      "loss": 1.4658,
+      "step": 9067
+    },
+    {
+      "epoch": 0.41774542774220297,
+      "grad_norm": 1.8470784465456787,
+      "learning_rate": 0.00013098683068063762,
+      "loss": 1.4403,
+      "step": 9068
+    },
+    {
+      "epoch": 0.41779149583083797,
+      "grad_norm": 2.589157940589137,
+      "learning_rate": 0.0001309726438629429,
+      "loss": 1.6241,
+      "step": 9069
+    },
+    {
+      "epoch": 0.41783756391947297,
+      "grad_norm": 2.1786701859095134,
+      "learning_rate": 0.00013095845635569542,
+      "loss": 1.5466,
+      "step": 9070
+    },
+    {
+      "epoch": 0.41788363200810796,
+      "grad_norm": 2.3290814561527213,
+      "learning_rate": 0.00013094426815921095,
+      "loss": 1.297,
+      "step": 9071
+    },
+    {
+      "epoch": 0.41792970009674296,
+      "grad_norm": 3.3497786499629645,
+      "learning_rate": 0.00013093007927380542,
+      "loss": 1.1787,
+      "step": 9072
+    },
+    {
+      "epoch": 0.417975768185378,
+      "grad_norm": 2.291273281044222,
+      "learning_rate": 0.00013091588969979472,
+      "loss": 1.5737,
+      "step": 9073
+    },
+    {
+      "epoch": 0.418021836274013,
+      "grad_norm": 1.3845632850834142,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 1.442,
+      "step": 9074
+    },
+    {
+      "epoch": 0.418067904362648,
+      "grad_norm": 1.2408893891286146,
+      "learning_rate": 0.00013088750848722143,
+      "loss": 1.1706,
+      "step": 9075
+    },
+    {
+      "epoch": 0.418113972451283,
+      "grad_norm": 1.291385817640082,
+      "learning_rate": 0.00013087331684929067,
+      "loss": 1.5243,
+      "step": 9076
+    },
+    {
+      "epoch": 0.418160040539918,
+      "grad_norm": 3.353918164534614,
+      "learning_rate": 0.00013085912452401847,
+      "loss": 1.5305,
+      "step": 9077
+    },
+    {
+      "epoch": 0.418206108628553,
+      "grad_norm": 1.4863463329499593,
+      "learning_rate": 0.00013084493151172076,
+      "loss": 1.3431,
+      "step": 9078
+    },
+    {
+      "epoch": 0.418252176717188,
+      "grad_norm": 3.5739952688311725,
+      "learning_rate": 0.00013083073781271356,
+      "loss": 1.35,
+      "step": 9079
+    },
+    {
+      "epoch": 0.418298244805823,
+      "grad_norm": 1.2490307647956558,
+      "learning_rate": 0.0001308165434273128,
+      "loss": 1.4371,
+      "step": 9080
+    },
+    {
+      "epoch": 0.418344312894458,
+      "grad_norm": 2.5260811996718777,
+      "learning_rate": 0.0001308023483558346,
+      "loss": 1.4544,
+      "step": 9081
+    },
+    {
+      "epoch": 0.418390380983093,
+      "grad_norm": 2.743978741398916,
+      "learning_rate": 0.00013078815259859487,
+      "loss": 1.1732,
+      "step": 9082
+    },
+    {
+      "epoch": 0.418436449071728,
+      "grad_norm": 2.4342941436184673,
+      "learning_rate": 0.00013077395615590975,
+      "loss": 1.1193,
+      "step": 9083
+    },
+    {
+      "epoch": 0.418482517160363,
+      "grad_norm": 2.5778581058431254,
+      "learning_rate": 0.00013075975902809522,
+      "loss": 1.1016,
+      "step": 9084
+    },
+    {
+      "epoch": 0.418528585248998,
+      "grad_norm": 1.5576959432630049,
+      "learning_rate": 0.00013074556121546747,
+      "loss": 1.4157,
+      "step": 9085
+    },
+    {
+      "epoch": 0.418574653337633,
+      "grad_norm": 1.1402918223865932,
+      "learning_rate": 0.00013073136271834246,
+      "loss": 1.6734,
+      "step": 9086
+    },
+    {
+      "epoch": 0.41862072142626805,
+      "grad_norm": 1.138478503585696,
+      "learning_rate": 0.00013071716353703635,
+      "loss": 1.3976,
+      "step": 9087
+    },
+    {
+      "epoch": 0.41866678951490305,
+      "grad_norm": 5.0697719003710855,
+      "learning_rate": 0.0001307029636718653,
+      "loss": 1.4067,
+      "step": 9088
+    },
+    {
+      "epoch": 0.41871285760353805,
+      "grad_norm": 1.4419047950226331,
+      "learning_rate": 0.00013068876312314538,
+      "loss": 1.5375,
+      "step": 9089
+    },
+    {
+      "epoch": 0.41875892569217305,
+      "grad_norm": 4.321849872182242,
+      "learning_rate": 0.00013067456189119278,
+      "loss": 1.4518,
+      "step": 9090
+    },
+    {
+      "epoch": 0.41880499378080804,
+      "grad_norm": 1.7995208980795578,
+      "learning_rate": 0.00013066035997632365,
+      "loss": 1.4037,
+      "step": 9091
+    },
+    {
+      "epoch": 0.41885106186944304,
+      "grad_norm": 1.3164990212782859,
+      "learning_rate": 0.00013064615737885417,
+      "loss": 1.3673,
+      "step": 9092
+    },
+    {
+      "epoch": 0.41889712995807804,
+      "grad_norm": 13.703247534461061,
+      "learning_rate": 0.00013063195409910056,
+      "loss": 1.5724,
+      "step": 9093
+    },
+    {
+      "epoch": 0.41894319804671304,
+      "grad_norm": 4.714007114478558,
+      "learning_rate": 0.00013061775013737902,
+      "loss": 1.179,
+      "step": 9094
+    },
+    {
+      "epoch": 0.41898926613534804,
+      "grad_norm": 2.222773278246749,
+      "learning_rate": 0.00013060354549400576,
+      "loss": 1.4625,
+      "step": 9095
+    },
+    {
+      "epoch": 0.41903533422398304,
+      "grad_norm": 1.4853616340181908,
+      "learning_rate": 0.00013058934016929702,
+      "loss": 1.434,
+      "step": 9096
+    },
+    {
+      "epoch": 0.41908140231261803,
+      "grad_norm": 1.3944889148075077,
+      "learning_rate": 0.00013057513416356908,
+      "loss": 1.4369,
+      "step": 9097
+    },
+    {
+      "epoch": 0.41912747040125303,
+      "grad_norm": 1.7195251004966972,
+      "learning_rate": 0.00013056092747713822,
+      "loss": 1.2109,
+      "step": 9098
+    },
+    {
+      "epoch": 0.41917353848988803,
+      "grad_norm": 1.2964885705507783,
+      "learning_rate": 0.00013054672011032068,
+      "loss": 1.4592,
+      "step": 9099
+    },
+    {
+      "epoch": 0.41921960657852303,
+      "grad_norm": 2.0950055766384064,
+      "learning_rate": 0.0001305325120634328,
+      "loss": 1.3434,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4192656746671581,
+      "grad_norm": 2.3726451000325,
+      "learning_rate": 0.0001305183033367909,
+      "loss": 1.416,
+      "step": 9101
+    },
+    {
+      "epoch": 0.4193117427557931,
+      "grad_norm": 2.0296240867520714,
+      "learning_rate": 0.0001305040939307113,
+      "loss": 1.655,
+      "step": 9102
+    },
+    {
+      "epoch": 0.4193578108444281,
+      "grad_norm": 8.353701440473792,
+      "learning_rate": 0.0001304898838455104,
+      "loss": 1.4278,
+      "step": 9103
+    },
+    {
+      "epoch": 0.4194038789330631,
+      "grad_norm": 1.363170252400761,
+      "learning_rate": 0.00013047567308150445,
+      "loss": 1.2517,
+      "step": 9104
+    },
+    {
+      "epoch": 0.4194499470216981,
+      "grad_norm": 2.065618891939056,
+      "learning_rate": 0.0001304614616390099,
+      "loss": 1.2425,
+      "step": 9105
+    },
+    {
+      "epoch": 0.4194960151103331,
+      "grad_norm": 2.2818287919043767,
+      "learning_rate": 0.00013044724951834314,
+      "loss": 1.3558,
+      "step": 9106
+    },
+    {
+      "epoch": 0.4195420831989681,
+      "grad_norm": 172.50090984156358,
+      "learning_rate": 0.0001304330367198206,
+      "loss": 1.4628,
+      "step": 9107
+    },
+    {
+      "epoch": 0.4195881512876031,
+      "grad_norm": 1.287039288586875,
+      "learning_rate": 0.00013041882324375866,
+      "loss": 1.2511,
+      "step": 9108
+    },
+    {
+      "epoch": 0.41963421937623807,
+      "grad_norm": 1.5866933813811608,
+      "learning_rate": 0.00013040460909047378,
+      "loss": 1.2405,
+      "step": 9109
+    },
+    {
+      "epoch": 0.41968028746487307,
+      "grad_norm": 2.0429567396184924,
+      "learning_rate": 0.00013039039426028238,
+      "loss": 1.4528,
+      "step": 9110
+    },
+    {
+      "epoch": 0.41972635555350807,
+      "grad_norm": 3.12678464571349,
+      "learning_rate": 0.00013037617875350104,
+      "loss": 1.8172,
+      "step": 9111
+    },
+    {
+      "epoch": 0.41977242364214307,
+      "grad_norm": 3.6956980200156044,
+      "learning_rate": 0.0001303619625704461,
+      "loss": 1.4283,
+      "step": 9112
+    },
+    {
+      "epoch": 0.41981849173077807,
+      "grad_norm": 3.455349317636957,
+      "learning_rate": 0.00013034774571143412,
+      "loss": 1.9116,
+      "step": 9113
+    },
+    {
+      "epoch": 0.4198645598194131,
+      "grad_norm": 1.4367807558799672,
+      "learning_rate": 0.00013033352817678166,
+      "loss": 1.3421,
+      "step": 9114
+    },
+    {
+      "epoch": 0.4199106279080481,
+      "grad_norm": 1.8759625923649512,
+      "learning_rate": 0.00013031930996680518,
+      "loss": 1.7994,
+      "step": 9115
+    },
+    {
+      "epoch": 0.4199566959966831,
+      "grad_norm": 2.5368720813992995,
+      "learning_rate": 0.00013030509108182125,
+      "loss": 1.2819,
+      "step": 9116
+    },
+    {
+      "epoch": 0.4200027640853181,
+      "grad_norm": 1.4370293182668603,
+      "learning_rate": 0.00013029087152214642,
+      "loss": 1.6174,
+      "step": 9117
+    },
+    {
+      "epoch": 0.4200488321739531,
+      "grad_norm": 2.1694312070775914,
+      "learning_rate": 0.0001302766512880973,
+      "loss": 1.493,
+      "step": 9118
+    },
+    {
+      "epoch": 0.4200949002625881,
+      "grad_norm": 3.9384934650238375,
+      "learning_rate": 0.00013026243037999043,
+      "loss": 1.7384,
+      "step": 9119
+    },
+    {
+      "epoch": 0.4201409683512231,
+      "grad_norm": 1.5271369384594449,
+      "learning_rate": 0.00013024820879814245,
+      "loss": 1.3248,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4201870364398581,
+      "grad_norm": 2.2822233682430864,
+      "learning_rate": 0.00013023398654287,
+      "loss": 1.4094,
+      "step": 9121
+    },
+    {
+      "epoch": 0.4202331045284931,
+      "grad_norm": 6.735520770087131,
+      "learning_rate": 0.0001302197636144896,
+      "loss": 1.4689,
+      "step": 9122
+    },
+    {
+      "epoch": 0.4202791726171281,
+      "grad_norm": 2.9061017356372347,
+      "learning_rate": 0.00013020554001331802,
+      "loss": 1.4689,
+      "step": 9123
+    },
+    {
+      "epoch": 0.4203252407057631,
+      "grad_norm": 5.196710996883663,
+      "learning_rate": 0.00013019131573967186,
+      "loss": 1.5702,
+      "step": 9124
+    },
+    {
+      "epoch": 0.4203713087943981,
+      "grad_norm": 1.5635353129655296,
+      "learning_rate": 0.00013017709079386788,
+      "loss": 1.4107,
+      "step": 9125
+    },
+    {
+      "epoch": 0.4204173768830331,
+      "grad_norm": 2.0780115609527883,
+      "learning_rate": 0.00013016286517622265,
+      "loss": 1.5519,
+      "step": 9126
+    },
+    {
+      "epoch": 0.4204634449716681,
+      "grad_norm": 1.5692127325081466,
+      "learning_rate": 0.000130148638887053,
+      "loss": 1.5539,
+      "step": 9127
+    },
+    {
+      "epoch": 0.42050951306030315,
+      "grad_norm": 2.9227160328523993,
+      "learning_rate": 0.00013013441192667556,
+      "loss": 1.5827,
+      "step": 9128
+    },
+    {
+      "epoch": 0.42055558114893815,
+      "grad_norm": 1.7605024794912056,
+      "learning_rate": 0.00013012018429540712,
+      "loss": 1.4648,
+      "step": 9129
+    },
+    {
+      "epoch": 0.42060164923757315,
+      "grad_norm": 1.2728640986339754,
+      "learning_rate": 0.00013010595599356446,
+      "loss": 1.5302,
+      "step": 9130
+    },
+    {
+      "epoch": 0.42064771732620815,
+      "grad_norm": 1.3102026076872808,
+      "learning_rate": 0.00013009172702146423,
+      "loss": 1.5876,
+      "step": 9131
+    },
+    {
+      "epoch": 0.42069378541484315,
+      "grad_norm": 4.203204626640872,
+      "learning_rate": 0.00013007749737942335,
+      "loss": 1.2158,
+      "step": 9132
+    },
+    {
+      "epoch": 0.42073985350347815,
+      "grad_norm": 8.729533667684057,
+      "learning_rate": 0.00013006326706775857,
+      "loss": 1.4795,
+      "step": 9133
+    },
+    {
+      "epoch": 0.42078592159211314,
+      "grad_norm": 1.1100766776358884,
+      "learning_rate": 0.00013004903608678666,
+      "loss": 1.4394,
+      "step": 9134
+    },
+    {
+      "epoch": 0.42083198968074814,
+      "grad_norm": 2.955478531737165,
+      "learning_rate": 0.0001300348044368245,
+      "loss": 1.4486,
+      "step": 9135
+    },
+    {
+      "epoch": 0.42087805776938314,
+      "grad_norm": 3.2386103704021227,
+      "learning_rate": 0.00013002057211818893,
+      "loss": 1.6203,
+      "step": 9136
+    },
+    {
+      "epoch": 0.42092412585801814,
+      "grad_norm": 2.9363300836413444,
+      "learning_rate": 0.0001300063391311968,
+      "loss": 1.2482,
+      "step": 9137
+    },
+    {
+      "epoch": 0.42097019394665314,
+      "grad_norm": 1.7372693144938136,
+      "learning_rate": 0.00012999210547616494,
+      "loss": 1.5737,
+      "step": 9138
+    },
+    {
+      "epoch": 0.42101626203528814,
+      "grad_norm": 1.633824975945175,
+      "learning_rate": 0.00012997787115341032,
+      "loss": 1.5979,
+      "step": 9139
+    },
+    {
+      "epoch": 0.42106233012392313,
+      "grad_norm": 1.8997424917097647,
+      "learning_rate": 0.00012996363616324977,
+      "loss": 1.5198,
+      "step": 9140
+    },
+    {
+      "epoch": 0.4211083982125582,
+      "grad_norm": 2.0703144268943015,
+      "learning_rate": 0.00012994940050600024,
+      "loss": 1.5205,
+      "step": 9141
+    },
+    {
+      "epoch": 0.4211544663011932,
+      "grad_norm": 4.843390270829138,
+      "learning_rate": 0.0001299351641819787,
+      "loss": 1.2747,
+      "step": 9142
+    },
+    {
+      "epoch": 0.4212005343898282,
+      "grad_norm": 5.587926502565565,
+      "learning_rate": 0.000129920927191502,
+      "loss": 1.7887,
+      "step": 9143
+    },
+    {
+      "epoch": 0.4212466024784632,
+      "grad_norm": 3.013998569090412,
+      "learning_rate": 0.0001299066895348872,
+      "loss": 1.6302,
+      "step": 9144
+    },
+    {
+      "epoch": 0.4212926705670982,
+      "grad_norm": 1.112274843492997,
+      "learning_rate": 0.00012989245121245121,
+      "loss": 1.7254,
+      "step": 9145
+    },
+    {
+      "epoch": 0.4213387386557332,
+      "grad_norm": 1.333896187954236,
+      "learning_rate": 0.00012987821222451104,
+      "loss": 1.3398,
+      "step": 9146
+    },
+    {
+      "epoch": 0.4213848067443682,
+      "grad_norm": 1.5783065291622016,
+      "learning_rate": 0.00012986397257138375,
+      "loss": 1.318,
+      "step": 9147
+    },
+    {
+      "epoch": 0.4214308748330032,
+      "grad_norm": 2.807947645057014,
+      "learning_rate": 0.00012984973225338626,
+      "loss": 1.4402,
+      "step": 9148
+    },
+    {
+      "epoch": 0.4214769429216382,
+      "grad_norm": 2.370283306537643,
+      "learning_rate": 0.00012983549127083575,
+      "loss": 1.3444,
+      "step": 9149
+    },
+    {
+      "epoch": 0.4215230110102732,
+      "grad_norm": 1.1036563072780254,
+      "learning_rate": 0.0001298212496240491,
+      "loss": 1.7498,
+      "step": 9150
+    },
+    {
+      "epoch": 0.4215690790989082,
+      "grad_norm": 2.344981325705418,
+      "learning_rate": 0.0001298070073133435,
+      "loss": 1.3106,
+      "step": 9151
+    },
+    {
+      "epoch": 0.42161514718754317,
+      "grad_norm": 1.4278564508309253,
+      "learning_rate": 0.00012979276433903599,
+      "loss": 1.2214,
+      "step": 9152
+    },
+    {
+      "epoch": 0.42166121527617817,
+      "grad_norm": 2.103686550430998,
+      "learning_rate": 0.00012977852070144365,
+      "loss": 1.4624,
+      "step": 9153
+    },
+    {
+      "epoch": 0.42170728336481317,
+      "grad_norm": 4.650765979308418,
+      "learning_rate": 0.0001297642764008836,
+      "loss": 1.3747,
+      "step": 9154
+    },
+    {
+      "epoch": 0.4217533514534482,
+      "grad_norm": 2.8289996137221123,
+      "learning_rate": 0.00012975003143767298,
+      "loss": 1.5822,
+      "step": 9155
+    },
+    {
+      "epoch": 0.4217994195420832,
+      "grad_norm": 1.7019558811716489,
+      "learning_rate": 0.0001297357858121289,
+      "loss": 1.5487,
+      "step": 9156
+    },
+    {
+      "epoch": 0.4218454876307182,
+      "grad_norm": 1.4576918201762692,
+      "learning_rate": 0.00012972153952456856,
+      "loss": 1.4399,
+      "step": 9157
+    },
+    {
+      "epoch": 0.4218915557193532,
+      "grad_norm": 1.0717824770985616,
+      "learning_rate": 0.00012970729257530912,
+      "loss": 1.3984,
+      "step": 9158
+    },
+    {
+      "epoch": 0.4219376238079882,
+      "grad_norm": 2.2307256086689393,
+      "learning_rate": 0.0001296930449646677,
+      "loss": 1.281,
+      "step": 9159
+    },
+    {
+      "epoch": 0.4219836918966232,
+      "grad_norm": 1.2229707109144328,
+      "learning_rate": 0.0001296787966929616,
+      "loss": 1.7676,
+      "step": 9160
+    },
+    {
+      "epoch": 0.4220297599852582,
+      "grad_norm": 1.0683854127233172,
+      "learning_rate": 0.00012966454776050794,
+      "loss": 1.4797,
+      "step": 9161
+    },
+    {
+      "epoch": 0.4220758280738932,
+      "grad_norm": 1.5363022537740425,
+      "learning_rate": 0.000129650298167624,
+      "loss": 1.5581,
+      "step": 9162
+    },
+    {
+      "epoch": 0.4221218961625282,
+      "grad_norm": 2.464856426523245,
+      "learning_rate": 0.00012963604791462702,
+      "loss": 1.379,
+      "step": 9163
+    },
+    {
+      "epoch": 0.4221679642511632,
+      "grad_norm": 3.0779399979643993,
+      "learning_rate": 0.00012962179700183422,
+      "loss": 1.4391,
+      "step": 9164
+    },
+    {
+      "epoch": 0.4222140323397982,
+      "grad_norm": 4.275613272349249,
+      "learning_rate": 0.00012960754542956293,
+      "loss": 1.4066,
+      "step": 9165
+    },
+    {
+      "epoch": 0.4222601004284332,
+      "grad_norm": 2.1902004774245567,
+      "learning_rate": 0.0001295932931981304,
+      "loss": 1.8077,
+      "step": 9166
+    },
+    {
+      "epoch": 0.4223061685170682,
+      "grad_norm": 1.6428464210777658,
+      "learning_rate": 0.00012957904030785395,
+      "loss": 1.7016,
+      "step": 9167
+    },
+    {
+      "epoch": 0.4223522366057032,
+      "grad_norm": 2.1938545031935663,
+      "learning_rate": 0.00012956478675905085,
+      "loss": 1.4756,
+      "step": 9168
+    },
+    {
+      "epoch": 0.42239830469433826,
+      "grad_norm": 1.1513406773032415,
+      "learning_rate": 0.0001295505325520385,
+      "loss": 1.7204,
+      "step": 9169
+    },
+    {
+      "epoch": 0.42244437278297325,
+      "grad_norm": 1.2470660313438302,
+      "learning_rate": 0.0001295362776871342,
+      "loss": 1.3603,
+      "step": 9170
+    },
+    {
+      "epoch": 0.42249044087160825,
+      "grad_norm": 2.1566907151337253,
+      "learning_rate": 0.0001295220221646553,
+      "loss": 1.3422,
+      "step": 9171
+    },
+    {
+      "epoch": 0.42253650896024325,
+      "grad_norm": 1.731434561807486,
+      "learning_rate": 0.00012950776598491922,
+      "loss": 1.6408,
+      "step": 9172
+    },
+    {
+      "epoch": 0.42258257704887825,
+      "grad_norm": 1.7462196956254294,
+      "learning_rate": 0.00012949350914824334,
+      "loss": 1.4368,
+      "step": 9173
+    },
+    {
+      "epoch": 0.42262864513751325,
+      "grad_norm": 1.4773050261107008,
+      "learning_rate": 0.000129479251654945,
+      "loss": 1.6023,
+      "step": 9174
+    },
+    {
+      "epoch": 0.42267471322614825,
+      "grad_norm": 2.284140299108633,
+      "learning_rate": 0.00012946499350534172,
+      "loss": 1.6049,
+      "step": 9175
+    },
+    {
+      "epoch": 0.42272078131478324,
+      "grad_norm": 1.3734246215071255,
+      "learning_rate": 0.00012945073469975086,
+      "loss": 1.4527,
+      "step": 9176
+    },
+    {
+      "epoch": 0.42276684940341824,
+      "grad_norm": 2.0832908712999796,
+      "learning_rate": 0.00012943647523848986,
+      "loss": 1.4813,
+      "step": 9177
+    },
+    {
+      "epoch": 0.42281291749205324,
+      "grad_norm": 1.664031668735344,
+      "learning_rate": 0.00012942221512187625,
+      "loss": 1.6311,
+      "step": 9178
+    },
+    {
+      "epoch": 0.42285898558068824,
+      "grad_norm": 1.0977071148337831,
+      "learning_rate": 0.00012940795435022746,
+      "loss": 1.6675,
+      "step": 9179
+    },
+    {
+      "epoch": 0.42290505366932324,
+      "grad_norm": 1.5051465803108492,
+      "learning_rate": 0.000129393692923861,
+      "loss": 1.3921,
+      "step": 9180
+    },
+    {
+      "epoch": 0.42295112175795824,
+      "grad_norm": 1.0572621270141351,
+      "learning_rate": 0.00012937943084309435,
+      "loss": 1.2764,
+      "step": 9181
+    },
+    {
+      "epoch": 0.4229971898465933,
+      "grad_norm": 2.03015137209657,
+      "learning_rate": 0.00012936516810824506,
+      "loss": 1.3924,
+      "step": 9182
+    },
+    {
+      "epoch": 0.4230432579352283,
+      "grad_norm": 1.378451503177681,
+      "learning_rate": 0.00012935090471963062,
+      "loss": 1.4832,
+      "step": 9183
+    },
+    {
+      "epoch": 0.4230893260238633,
+      "grad_norm": 1.4630706970625098,
+      "learning_rate": 0.00012933664067756867,
+      "loss": 1.565,
+      "step": 9184
+    },
+    {
+      "epoch": 0.4231353941124983,
+      "grad_norm": 1.3168761103509896,
+      "learning_rate": 0.00012932237598237668,
+      "loss": 1.4655,
+      "step": 9185
+    },
+    {
+      "epoch": 0.4231814622011333,
+      "grad_norm": 1.0299197830622304,
+      "learning_rate": 0.00012930811063437226,
+      "loss": 1.3659,
+      "step": 9186
+    },
+    {
+      "epoch": 0.4232275302897683,
+      "grad_norm": 1.2274954235262259,
+      "learning_rate": 0.000129293844633873,
+      "loss": 1.5626,
+      "step": 9187
+    },
+    {
+      "epoch": 0.4232735983784033,
+      "grad_norm": 1.4101512873368167,
+      "learning_rate": 0.00012927957798119653,
+      "loss": 1.3317,
+      "step": 9188
+    },
+    {
+      "epoch": 0.4233196664670383,
+      "grad_norm": 2.990477182406341,
+      "learning_rate": 0.0001292653106766605,
+      "loss": 1.3428,
+      "step": 9189
+    },
+    {
+      "epoch": 0.4233657345556733,
+      "grad_norm": 1.2867956656271762,
+      "learning_rate": 0.00012925104272058246,
+      "loss": 1.2481,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4234118026443083,
+      "grad_norm": 1.7084356740017497,
+      "learning_rate": 0.00012923677411328012,
+      "loss": 1.4709,
+      "step": 9191
+    },
+    {
+      "epoch": 0.4234578707329433,
+      "grad_norm": 2.491750277558529,
+      "learning_rate": 0.00012922250485507112,
+      "loss": 1.3175,
+      "step": 9192
+    },
+    {
+      "epoch": 0.4235039388215783,
+      "grad_norm": 1.4129299499903303,
+      "learning_rate": 0.0001292082349462732,
+      "loss": 1.3266,
+      "step": 9193
+    },
+    {
+      "epoch": 0.4235500069102133,
+      "grad_norm": 1.3261804869095333,
+      "learning_rate": 0.00012919396438720399,
+      "loss": 1.3665,
+      "step": 9194
+    },
+    {
+      "epoch": 0.42359607499884827,
+      "grad_norm": 1.9773427048129126,
+      "learning_rate": 0.00012917969317818121,
+      "loss": 1.2916,
+      "step": 9195
+    },
+    {
+      "epoch": 0.4236421430874833,
+      "grad_norm": 1.2497414279806334,
+      "learning_rate": 0.00012916542131952258,
+      "loss": 1.4761,
+      "step": 9196
+    },
+    {
+      "epoch": 0.4236882111761183,
+      "grad_norm": 1.4319762052152916,
+      "learning_rate": 0.0001291511488115459,
+      "loss": 1.5275,
+      "step": 9197
+    },
+    {
+      "epoch": 0.4237342792647533,
+      "grad_norm": 1.2946849448428643,
+      "learning_rate": 0.0001291368756545688,
+      "loss": 1.4383,
+      "step": 9198
+    },
+    {
+      "epoch": 0.4237803473533883,
+      "grad_norm": 0.8231146069018231,
+      "learning_rate": 0.00012912260184890916,
+      "loss": 1.6331,
+      "step": 9199
+    },
+    {
+      "epoch": 0.4238264154420233,
+      "grad_norm": 4.055474551367202,
+      "learning_rate": 0.00012910832739488473,
+      "loss": 1.3253,
+      "step": 9200
+    },
+    {
+      "epoch": 0.4238724835306583,
+      "grad_norm": 3.5024857636762636,
+      "learning_rate": 0.00012909405229281332,
+      "loss": 1.3841,
+      "step": 9201
+    },
+    {
+      "epoch": 0.4239185516192933,
+      "grad_norm": 1.7397061403453402,
+      "learning_rate": 0.0001290797765430127,
+      "loss": 1.4828,
+      "step": 9202
+    },
+    {
+      "epoch": 0.4239646197079283,
+      "grad_norm": 1.1930334312862942,
+      "learning_rate": 0.0001290655001458007,
+      "loss": 1.6735,
+      "step": 9203
+    },
+    {
+      "epoch": 0.4240106877965633,
+      "grad_norm": 1.6341395696280574,
+      "learning_rate": 0.0001290512231014952,
+      "loss": 1.3186,
+      "step": 9204
+    },
+    {
+      "epoch": 0.4240567558851983,
+      "grad_norm": 1.2330780459083537,
+      "learning_rate": 0.00012903694541041403,
+      "loss": 1.3963,
+      "step": 9205
+    },
+    {
+      "epoch": 0.4241028239738333,
+      "grad_norm": 1.5884202455399212,
+      "learning_rate": 0.00012902266707287506,
+      "loss": 1.4749,
+      "step": 9206
+    },
+    {
+      "epoch": 0.4241488920624683,
+      "grad_norm": 1.639790082516019,
+      "learning_rate": 0.00012900838808919615,
+      "loss": 1.5985,
+      "step": 9207
+    },
+    {
+      "epoch": 0.4241949601511033,
+      "grad_norm": 1.0708870542302564,
+      "learning_rate": 0.00012899410845969525,
+      "loss": 1.216,
+      "step": 9208
+    },
+    {
+      "epoch": 0.42424102823973836,
+      "grad_norm": 2.2723891369556517,
+      "learning_rate": 0.00012897982818469022,
+      "loss": 1.1819,
+      "step": 9209
+    },
+    {
+      "epoch": 0.42428709632837336,
+      "grad_norm": 1.2755684131947662,
+      "learning_rate": 0.000128965547264499,
+      "loss": 1.3651,
+      "step": 9210
+    },
+    {
+      "epoch": 0.42433316441700836,
+      "grad_norm": 1.0299006105424382,
+      "learning_rate": 0.00012895126569943953,
+      "loss": 1.7041,
+      "step": 9211
+    },
+    {
+      "epoch": 0.42437923250564336,
+      "grad_norm": 1.0472963066539847,
+      "learning_rate": 0.00012893698348982976,
+      "loss": 1.4839,
+      "step": 9212
+    },
+    {
+      "epoch": 0.42442530059427835,
+      "grad_norm": 1.6192267904232134,
+      "learning_rate": 0.00012892270063598771,
+      "loss": 1.4662,
+      "step": 9213
+    },
+    {
+      "epoch": 0.42447136868291335,
+      "grad_norm": 1.0128175883618586,
+      "learning_rate": 0.00012890841713823133,
+      "loss": 1.5197,
+      "step": 9214
+    },
+    {
+      "epoch": 0.42451743677154835,
+      "grad_norm": 2.9971757980171674,
+      "learning_rate": 0.00012889413299687857,
+      "loss": 1.38,
+      "step": 9215
+    },
+    {
+      "epoch": 0.42456350486018335,
+      "grad_norm": 1.581565021455275,
+      "learning_rate": 0.00012887984821224747,
+      "loss": 1.5677,
+      "step": 9216
+    },
+    {
+      "epoch": 0.42460957294881835,
+      "grad_norm": 4.66661144505456,
+      "learning_rate": 0.00012886556278465612,
+      "loss": 1.316,
+      "step": 9217
+    },
+    {
+      "epoch": 0.42465564103745335,
+      "grad_norm": 3.0482078033530913,
+      "learning_rate": 0.00012885127671442246,
+      "loss": 1.2911,
+      "step": 9218
+    },
+    {
+      "epoch": 0.42470170912608834,
+      "grad_norm": 5.022110716326002,
+      "learning_rate": 0.00012883699000186462,
+      "loss": 1.3431,
+      "step": 9219
+    },
+    {
+      "epoch": 0.42474777721472334,
+      "grad_norm": 2.3356354929878056,
+      "learning_rate": 0.00012882270264730062,
+      "loss": 1.3948,
+      "step": 9220
+    },
+    {
+      "epoch": 0.42479384530335834,
+      "grad_norm": 2.07629755503154,
+      "learning_rate": 0.0001288084146510486,
+      "loss": 1.2871,
+      "step": 9221
+    },
+    {
+      "epoch": 0.42483991339199334,
+      "grad_norm": 2.438492113204229,
+      "learning_rate": 0.0001287941260134266,
+      "loss": 1.5357,
+      "step": 9222
+    },
+    {
+      "epoch": 0.4248859814806284,
+      "grad_norm": 1.9283070384497967,
+      "learning_rate": 0.00012877983673475276,
+      "loss": 1.4222,
+      "step": 9223
+    },
+    {
+      "epoch": 0.4249320495692634,
+      "grad_norm": 1.4114255197396262,
+      "learning_rate": 0.0001287655468153452,
+      "loss": 1.354,
+      "step": 9224
+    },
+    {
+      "epoch": 0.4249781176578984,
+      "grad_norm": 2.1903943163383417,
+      "learning_rate": 0.00012875125625552207,
+      "loss": 1.0339,
+      "step": 9225
+    },
+    {
+      "epoch": 0.4250241857465334,
+      "grad_norm": 2.9116261548179554,
+      "learning_rate": 0.00012873696505560152,
+      "loss": 1.5766,
+      "step": 9226
+    },
+    {
+      "epoch": 0.4250702538351684,
+      "grad_norm": 2.068847641727024,
+      "learning_rate": 0.0001287226732159017,
+      "loss": 1.2276,
+      "step": 9227
+    },
+    {
+      "epoch": 0.4251163219238034,
+      "grad_norm": 1.7443188762325776,
+      "learning_rate": 0.00012870838073674083,
+      "loss": 1.3686,
+      "step": 9228
+    },
+    {
+      "epoch": 0.4251623900124384,
+      "grad_norm": 1.288807215576158,
+      "learning_rate": 0.00012869408761843705,
+      "loss": 1.3158,
+      "step": 9229
+    },
+    {
+      "epoch": 0.4252084581010734,
+      "grad_norm": 1.2804392003024943,
+      "learning_rate": 0.00012867979386130865,
+      "loss": 1.5938,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4252545261897084,
+      "grad_norm": 1.4325533612399466,
+      "learning_rate": 0.00012866549946567382,
+      "loss": 1.4891,
+      "step": 9231
+    },
+    {
+      "epoch": 0.4253005942783434,
+      "grad_norm": 1.5631586084242646,
+      "learning_rate": 0.00012865120443185076,
+      "loss": 1.5417,
+      "step": 9232
+    },
+    {
+      "epoch": 0.4253466623669784,
+      "grad_norm": 2.4179262579585994,
+      "learning_rate": 0.00012863690876015776,
+      "loss": 1.4414,
+      "step": 9233
+    },
+    {
+      "epoch": 0.4253927304556134,
+      "grad_norm": 1.1678994043670723,
+      "learning_rate": 0.00012862261245091312,
+      "loss": 1.4656,
+      "step": 9234
+    },
+    {
+      "epoch": 0.4254387985442484,
+      "grad_norm": 1.7482962154686004,
+      "learning_rate": 0.00012860831550443505,
+      "loss": 1.5208,
+      "step": 9235
+    },
+    {
+      "epoch": 0.4254848666328834,
+      "grad_norm": 2.4780487790890806,
+      "learning_rate": 0.0001285940179210419,
+      "loss": 1.2954,
+      "step": 9236
+    },
+    {
+      "epoch": 0.42553093472151843,
+      "grad_norm": 1.2331209336415478,
+      "learning_rate": 0.000128579719701052,
+      "loss": 1.4227,
+      "step": 9237
+    },
+    {
+      "epoch": 0.4255770028101534,
+      "grad_norm": 3.7289858986382765,
+      "learning_rate": 0.00012856542084478364,
+      "loss": 1.2497,
+      "step": 9238
+    },
+    {
+      "epoch": 0.4256230708987884,
+      "grad_norm": 1.9036896091219715,
+      "learning_rate": 0.00012855112135255514,
+      "loss": 1.1293,
+      "step": 9239
+    },
+    {
+      "epoch": 0.4256691389874234,
+      "grad_norm": 3.2576701057312274,
+      "learning_rate": 0.0001285368212246849,
+      "loss": 1.3877,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4257152070760584,
+      "grad_norm": 1.3792183512260978,
+      "learning_rate": 0.00012852252046149126,
+      "loss": 1.7256,
+      "step": 9241
+    },
+    {
+      "epoch": 0.4257612751646934,
+      "grad_norm": 0.9635493415679552,
+      "learning_rate": 0.0001285082190632926,
+      "loss": 1.6404,
+      "step": 9242
+    },
+    {
+      "epoch": 0.4258073432533284,
+      "grad_norm": 5.452246583382888,
+      "learning_rate": 0.0001284939170304073,
+      "loss": 1.1809,
+      "step": 9243
+    },
+    {
+      "epoch": 0.4258534113419634,
+      "grad_norm": 1.1689465477680119,
+      "learning_rate": 0.00012847961436315383,
+      "loss": 1.5313,
+      "step": 9244
+    },
+    {
+      "epoch": 0.4258994794305984,
+      "grad_norm": 1.9051568671429528,
+      "learning_rate": 0.00012846531106185056,
+      "loss": 1.3125,
+      "step": 9245
+    },
+    {
+      "epoch": 0.4259455475192334,
+      "grad_norm": 4.171295438074941,
+      "learning_rate": 0.00012845100712681595,
+      "loss": 1.6207,
+      "step": 9246
+    },
+    {
+      "epoch": 0.4259916156078684,
+      "grad_norm": 4.201973142910413,
+      "learning_rate": 0.00012843670255836844,
+      "loss": 1.1073,
+      "step": 9247
+    },
+    {
+      "epoch": 0.4260376836965034,
+      "grad_norm": 1.6923141236970565,
+      "learning_rate": 0.00012842239735682653,
+      "loss": 1.3945,
+      "step": 9248
+    },
+    {
+      "epoch": 0.4260837517851384,
+      "grad_norm": 1.3952461403328178,
+      "learning_rate": 0.00012840809152250867,
+      "loss": 1.3967,
+      "step": 9249
+    },
+    {
+      "epoch": 0.42612981987377346,
+      "grad_norm": 2.1583691684119484,
+      "learning_rate": 0.0001283937850557334,
+      "loss": 1.3787,
+      "step": 9250
+    },
+    {
+      "epoch": 0.42617588796240846,
+      "grad_norm": 2.9489247635878653,
+      "learning_rate": 0.00012837947795681911,
+      "loss": 1.3687,
+      "step": 9251
+    },
+    {
+      "epoch": 0.42622195605104346,
+      "grad_norm": 1.895558768111015,
+      "learning_rate": 0.00012836517022608444,
+      "loss": 1.4139,
+      "step": 9252
+    },
+    {
+      "epoch": 0.42626802413967846,
+      "grad_norm": 1.1088606561907741,
+      "learning_rate": 0.00012835086186384787,
+      "loss": 1.3754,
+      "step": 9253
+    },
+    {
+      "epoch": 0.42631409222831346,
+      "grad_norm": 2.1972211185916266,
+      "learning_rate": 0.00012833655287042804,
+      "loss": 1.6755,
+      "step": 9254
+    },
+    {
+      "epoch": 0.42636016031694846,
+      "grad_norm": 8.987945684313123,
+      "learning_rate": 0.00012832224324614336,
+      "loss": 1.2724,
+      "step": 9255
+    },
+    {
+      "epoch": 0.42640622840558345,
+      "grad_norm": 2.382710249552418,
+      "learning_rate": 0.00012830793299131255,
+      "loss": 1.5052,
+      "step": 9256
+    },
+    {
+      "epoch": 0.42645229649421845,
+      "grad_norm": 1.6802227884985717,
+      "learning_rate": 0.00012829362210625412,
+      "loss": 1.2402,
+      "step": 9257
+    },
+    {
+      "epoch": 0.42649836458285345,
+      "grad_norm": 1.2693172652739764,
+      "learning_rate": 0.00012827931059128676,
+      "loss": 1.5771,
+      "step": 9258
+    },
+    {
+      "epoch": 0.42654443267148845,
+      "grad_norm": 2.7933335094013603,
+      "learning_rate": 0.000128264998446729,
+      "loss": 1.4898,
+      "step": 9259
+    },
+    {
+      "epoch": 0.42659050076012345,
+      "grad_norm": 1.081559280101749,
+      "learning_rate": 0.0001282506856728995,
+      "loss": 1.614,
+      "step": 9260
+    },
+    {
+      "epoch": 0.42663656884875845,
+      "grad_norm": 7.014034331404385,
+      "learning_rate": 0.00012823637227011696,
+      "loss": 1.5618,
+      "step": 9261
+    },
+    {
+      "epoch": 0.42668263693739344,
+      "grad_norm": 0.9789855909250532,
+      "learning_rate": 0.00012822205823869996,
+      "loss": 1.3512,
+      "step": 9262
+    },
+    {
+      "epoch": 0.42672870502602844,
+      "grad_norm": 1.226901756388501,
+      "learning_rate": 0.00012820774357896725,
+      "loss": 1.4185,
+      "step": 9263
+    },
+    {
+      "epoch": 0.4267747731146635,
+      "grad_norm": 2.2626325792084816,
+      "learning_rate": 0.00012819342829123751,
+      "loss": 1.3516,
+      "step": 9264
+    },
+    {
+      "epoch": 0.4268208412032985,
+      "grad_norm": 1.408156458923982,
+      "learning_rate": 0.0001281791123758294,
+      "loss": 1.6897,
+      "step": 9265
+    },
+    {
+      "epoch": 0.4268669092919335,
+      "grad_norm": 3.064667747778725,
+      "learning_rate": 0.00012816479583306167,
+      "loss": 1.3852,
+      "step": 9266
+    },
+    {
+      "epoch": 0.4269129773805685,
+      "grad_norm": 1.8774984649418012,
+      "learning_rate": 0.0001281504786632531,
+      "loss": 1.7117,
+      "step": 9267
+    },
+    {
+      "epoch": 0.4269590454692035,
+      "grad_norm": 2.5502953270941,
+      "learning_rate": 0.00012813616086672236,
+      "loss": 1.4729,
+      "step": 9268
+    },
+    {
+      "epoch": 0.4270051135578385,
+      "grad_norm": 1.5396272419880876,
+      "learning_rate": 0.0001281218424437882,
+      "loss": 1.4398,
+      "step": 9269
+    },
+    {
+      "epoch": 0.4270511816464735,
+      "grad_norm": 1.569835390150231,
+      "learning_rate": 0.0001281075233947695,
+      "loss": 1.6586,
+      "step": 9270
+    },
+    {
+      "epoch": 0.4270972497351085,
+      "grad_norm": 3.63988672257139,
+      "learning_rate": 0.00012809320371998494,
+      "loss": 1.3236,
+      "step": 9271
+    },
+    {
+      "epoch": 0.4271433178237435,
+      "grad_norm": 2.981565799949222,
+      "learning_rate": 0.0001280788834197534,
+      "loss": 1.6064,
+      "step": 9272
+    },
+    {
+      "epoch": 0.4271893859123785,
+      "grad_norm": 4.634627218238301,
+      "learning_rate": 0.00012806456249439365,
+      "loss": 1.358,
+      "step": 9273
+    },
+    {
+      "epoch": 0.4272354540010135,
+      "grad_norm": 1.5038005806492978,
+      "learning_rate": 0.00012805024094422456,
+      "loss": 1.4417,
+      "step": 9274
+    },
+    {
+      "epoch": 0.4272815220896485,
+      "grad_norm": 3.8550220680870986,
+      "learning_rate": 0.0001280359187695649,
+      "loss": 1.5733,
+      "step": 9275
+    },
+    {
+      "epoch": 0.4273275901782835,
+      "grad_norm": 1.8400963258587777,
+      "learning_rate": 0.00012802159597073365,
+      "loss": 1.3699,
+      "step": 9276
+    },
+    {
+      "epoch": 0.42737365826691853,
+      "grad_norm": 1.7869623579010676,
+      "learning_rate": 0.00012800727254804958,
+      "loss": 1.3831,
+      "step": 9277
+    },
+    {
+      "epoch": 0.42741972635555353,
+      "grad_norm": 1.385852186491599,
+      "learning_rate": 0.00012799294850183158,
+      "loss": 1.2957,
+      "step": 9278
+    },
+    {
+      "epoch": 0.42746579444418853,
+      "grad_norm": 1.6383082337827621,
+      "learning_rate": 0.00012797862383239863,
+      "loss": 1.6079,
+      "step": 9279
+    },
+    {
+      "epoch": 0.42751186253282353,
+      "grad_norm": 1.5090580248785888,
+      "learning_rate": 0.00012796429854006956,
+      "loss": 1.4796,
+      "step": 9280
+    },
+    {
+      "epoch": 0.4275579306214585,
+      "grad_norm": 5.750330440413282,
+      "learning_rate": 0.00012794997262516336,
+      "loss": 1.4084,
+      "step": 9281
+    },
+    {
+      "epoch": 0.4276039987100935,
+      "grad_norm": 2.070048592597527,
+      "learning_rate": 0.00012793564608799888,
+      "loss": 1.2945,
+      "step": 9282
+    },
+    {
+      "epoch": 0.4276500667987285,
+      "grad_norm": 2.6950688980070936,
+      "learning_rate": 0.0001279213189288952,
+      "loss": 1.3936,
+      "step": 9283
+    },
+    {
+      "epoch": 0.4276961348873635,
+      "grad_norm": 2.8228773686062816,
+      "learning_rate": 0.0001279069911481712,
+      "loss": 1.3667,
+      "step": 9284
+    },
+    {
+      "epoch": 0.4277422029759985,
+      "grad_norm": 1.924597815984349,
+      "learning_rate": 0.00012789266274614591,
+      "loss": 1.6477,
+      "step": 9285
+    },
+    {
+      "epoch": 0.4277882710646335,
+      "grad_norm": 6.639528605506925,
+      "learning_rate": 0.00012787833372313831,
+      "loss": 1.4004,
+      "step": 9286
+    },
+    {
+      "epoch": 0.4278343391532685,
+      "grad_norm": 1.6950237226214817,
+      "learning_rate": 0.0001278640040794674,
+      "loss": 1.7169,
+      "step": 9287
+    },
+    {
+      "epoch": 0.4278804072419035,
+      "grad_norm": 2.4341813047443246,
+      "learning_rate": 0.00012784967381545222,
+      "loss": 1.506,
+      "step": 9288
+    },
+    {
+      "epoch": 0.4279264753305385,
+      "grad_norm": 2.4498067763828675,
+      "learning_rate": 0.00012783534293141178,
+      "loss": 1.6009,
+      "step": 9289
+    },
+    {
+      "epoch": 0.4279725434191735,
+      "grad_norm": 2.499812822863923,
+      "learning_rate": 0.0001278210114276652,
+      "loss": 1.6228,
+      "step": 9290
+    },
+    {
+      "epoch": 0.42801861150780857,
+      "grad_norm": 4.326740490784885,
+      "learning_rate": 0.00012780667930453147,
+      "loss": 1.5995,
+      "step": 9291
+    },
+    {
+      "epoch": 0.42806467959644356,
+      "grad_norm": 2.6829062649228623,
+      "learning_rate": 0.00012779234656232972,
+      "loss": 1.4529,
+      "step": 9292
+    },
+    {
+      "epoch": 0.42811074768507856,
+      "grad_norm": 2.620285028953929,
+      "learning_rate": 0.000127778013201379,
+      "loss": 1.6564,
+      "step": 9293
+    },
+    {
+      "epoch": 0.42815681577371356,
+      "grad_norm": 2.0531396758937284,
+      "learning_rate": 0.0001277636792219985,
+      "loss": 1.5465,
+      "step": 9294
+    },
+    {
+      "epoch": 0.42820288386234856,
+      "grad_norm": 2.109857593532122,
+      "learning_rate": 0.00012774934462450726,
+      "loss": 1.3643,
+      "step": 9295
+    },
+    {
+      "epoch": 0.42824895195098356,
+      "grad_norm": 3.0761676397669904,
+      "learning_rate": 0.00012773500940922445,
+      "loss": 1.2609,
+      "step": 9296
+    },
+    {
+      "epoch": 0.42829502003961856,
+      "grad_norm": 4.541764483705353,
+      "learning_rate": 0.0001277206735764692,
+      "loss": 1.4195,
+      "step": 9297
+    },
+    {
+      "epoch": 0.42834108812825356,
+      "grad_norm": 4.497949511369293,
+      "learning_rate": 0.00012770633712656066,
+      "loss": 1.4471,
+      "step": 9298
+    },
+    {
+      "epoch": 0.42838715621688855,
+      "grad_norm": 3.149068864737677,
+      "learning_rate": 0.0001276920000598181,
+      "loss": 1.477,
+      "step": 9299
+    },
+    {
+      "epoch": 0.42843322430552355,
+      "grad_norm": 2.5695752599644908,
+      "learning_rate": 0.00012767766237656053,
+      "loss": 1.4211,
+      "step": 9300
+    },
+    {
+      "epoch": 0.42847929239415855,
+      "grad_norm": 2.0012535217103506,
+      "learning_rate": 0.00012766332407710735,
+      "loss": 1.427,
+      "step": 9301
+    },
+    {
+      "epoch": 0.42852536048279355,
+      "grad_norm": 1.6293906930932798,
+      "learning_rate": 0.00012764898516177766,
+      "loss": 1.4936,
+      "step": 9302
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.5139523647820272,
+      "learning_rate": 0.00012763464563089074,
+      "loss": 1.6621,
+      "step": 9303
+    },
+    {
+      "epoch": 0.4286174966600636,
+      "grad_norm": 1.6979743941767118,
+      "learning_rate": 0.00012762030548476582,
+      "loss": 1.4805,
+      "step": 9304
+    },
+    {
+      "epoch": 0.4286635647486986,
+      "grad_norm": 10.429655348862747,
+      "learning_rate": 0.00012760596472372214,
+      "loss": 1.5271,
+      "step": 9305
+    },
+    {
+      "epoch": 0.4287096328373336,
+      "grad_norm": 1.8789958876502904,
+      "learning_rate": 0.000127591623348079,
+      "loss": 1.3132,
+      "step": 9306
+    },
+    {
+      "epoch": 0.4287557009259686,
+      "grad_norm": 1.7113990184630767,
+      "learning_rate": 0.0001275772813581557,
+      "loss": 1.5166,
+      "step": 9307
+    },
+    {
+      "epoch": 0.4288017690146036,
+      "grad_norm": 2.72883786439901,
+      "learning_rate": 0.00012756293875427146,
+      "loss": 1.6273,
+      "step": 9308
+    },
+    {
+      "epoch": 0.4288478371032386,
+      "grad_norm": 2.8646797364835845,
+      "learning_rate": 0.00012754859553674568,
+      "loss": 1.4432,
+      "step": 9309
+    },
+    {
+      "epoch": 0.4288939051918736,
+      "grad_norm": 1.6021596530737006,
+      "learning_rate": 0.00012753425170589764,
+      "loss": 1.4036,
+      "step": 9310
+    },
+    {
+      "epoch": 0.4289399732805086,
+      "grad_norm": 1.846786613846402,
+      "learning_rate": 0.0001275199072620467,
+      "loss": 1.4406,
+      "step": 9311
+    },
+    {
+      "epoch": 0.4289860413691436,
+      "grad_norm": 2.4125556922500584,
+      "learning_rate": 0.00012750556220551224,
+      "loss": 1.5784,
+      "step": 9312
+    },
+    {
+      "epoch": 0.4290321094577786,
+      "grad_norm": 1.9725866638897602,
+      "learning_rate": 0.00012749121653661357,
+      "loss": 1.3,
+      "step": 9313
+    },
+    {
+      "epoch": 0.4290781775464136,
+      "grad_norm": 2.943423827024301,
+      "learning_rate": 0.00012747687025567014,
+      "loss": 1.4961,
+      "step": 9314
+    },
+    {
+      "epoch": 0.4291242456350486,
+      "grad_norm": 1.4173644527051452,
+      "learning_rate": 0.00012746252336300127,
+      "loss": 1.2479,
+      "step": 9315
+    },
+    {
+      "epoch": 0.4291703137236836,
+      "grad_norm": 2.297986673610897,
+      "learning_rate": 0.0001274481758589264,
+      "loss": 1.7644,
+      "step": 9316
+    },
+    {
+      "epoch": 0.4292163818123186,
+      "grad_norm": 6.9886295814245685,
+      "learning_rate": 0.00012743382774376498,
+      "loss": 1.502,
+      "step": 9317
+    },
+    {
+      "epoch": 0.42926244990095364,
+      "grad_norm": 2.7486759252038815,
+      "learning_rate": 0.00012741947901783642,
+      "loss": 1.2558,
+      "step": 9318
+    },
+    {
+      "epoch": 0.42930851798958863,
+      "grad_norm": 4.951397424946718,
+      "learning_rate": 0.00012740512968146015,
+      "loss": 1.3574,
+      "step": 9319
+    },
+    {
+      "epoch": 0.42935458607822363,
+      "grad_norm": 2.252971794135259,
+      "learning_rate": 0.0001273907797349557,
+      "loss": 1.4948,
+      "step": 9320
+    },
+    {
+      "epoch": 0.42940065416685863,
+      "grad_norm": 2.8847481716614203,
+      "learning_rate": 0.00012737642917864248,
+      "loss": 1.426,
+      "step": 9321
+    },
+    {
+      "epoch": 0.42944672225549363,
+      "grad_norm": 1.4337514759965302,
+      "learning_rate": 0.00012736207801283998,
+      "loss": 1.456,
+      "step": 9322
+    },
+    {
+      "epoch": 0.4294927903441286,
+      "grad_norm": 1.2173685628454889,
+      "learning_rate": 0.0001273477262378678,
+      "loss": 1.7884,
+      "step": 9323
+    },
+    {
+      "epoch": 0.4295388584327636,
+      "grad_norm": 2.4760916927875396,
+      "learning_rate": 0.0001273333738540453,
+      "loss": 1.0508,
+      "step": 9324
+    },
+    {
+      "epoch": 0.4295849265213986,
+      "grad_norm": 1.0801156845752151,
+      "learning_rate": 0.00012731902086169215,
+      "loss": 1.5235,
+      "step": 9325
+    },
+    {
+      "epoch": 0.4296309946100336,
+      "grad_norm": 3.87313111126457,
+      "learning_rate": 0.00012730466726112785,
+      "loss": 1.1269,
+      "step": 9326
+    },
+    {
+      "epoch": 0.4296770626986686,
+      "grad_norm": 2.080806962597685,
+      "learning_rate": 0.00012729031305267193,
+      "loss": 1.4987,
+      "step": 9327
+    },
+    {
+      "epoch": 0.4297231307873036,
+      "grad_norm": 2.0787432095633007,
+      "learning_rate": 0.00012727595823664397,
+      "loss": 1.4888,
+      "step": 9328
+    },
+    {
+      "epoch": 0.4297691988759386,
+      "grad_norm": 1.821571610217875,
+      "learning_rate": 0.0001272616028133636,
+      "loss": 1.4306,
+      "step": 9329
+    },
+    {
+      "epoch": 0.4298152669645736,
+      "grad_norm": 2.5525353577154224,
+      "learning_rate": 0.00012724724678315038,
+      "loss": 1.3535,
+      "step": 9330
+    },
+    {
+      "epoch": 0.4298613350532086,
+      "grad_norm": 2.047334880775977,
+      "learning_rate": 0.0001272328901463239,
+      "loss": 1.1644,
+      "step": 9331
+    },
+    {
+      "epoch": 0.42990740314184367,
+      "grad_norm": 1.378054031418294,
+      "learning_rate": 0.00012721853290320384,
+      "loss": 1.5267,
+      "step": 9332
+    },
+    {
+      "epoch": 0.42995347123047867,
+      "grad_norm": 2.6910377669167755,
+      "learning_rate": 0.0001272041750541098,
+      "loss": 1.4809,
+      "step": 9333
+    },
+    {
+      "epoch": 0.42999953931911367,
+      "grad_norm": 6.115226020585796,
+      "learning_rate": 0.00012718981659936147,
+      "loss": 1.3833,
+      "step": 9334
+    },
+    {
+      "epoch": 0.43004560740774866,
+      "grad_norm": 1.4729104374529827,
+      "learning_rate": 0.00012717545753927848,
+      "loss": 1.2861,
+      "step": 9335
+    },
+    {
+      "epoch": 0.43009167549638366,
+      "grad_norm": 2.1510543656578136,
+      "learning_rate": 0.00012716109787418053,
+      "loss": 1.4454,
+      "step": 9336
+    },
+    {
+      "epoch": 0.43013774358501866,
+      "grad_norm": 3.8485701668808803,
+      "learning_rate": 0.00012714673760438728,
+      "loss": 1.3153,
+      "step": 9337
+    },
+    {
+      "epoch": 0.43018381167365366,
+      "grad_norm": 1.377017442265225,
+      "learning_rate": 0.00012713237673021852,
+      "loss": 1.4106,
+      "step": 9338
+    },
+    {
+      "epoch": 0.43022987976228866,
+      "grad_norm": 2.791684133549563,
+      "learning_rate": 0.00012711801525199388,
+      "loss": 1.2456,
+      "step": 9339
+    },
+    {
+      "epoch": 0.43027594785092366,
+      "grad_norm": 1.9971945682814087,
+      "learning_rate": 0.0001271036531700331,
+      "loss": 1.4582,
+      "step": 9340
+    },
+    {
+      "epoch": 0.43032201593955866,
+      "grad_norm": 4.584505074544851,
+      "learning_rate": 0.000127089290484656,
+      "loss": 1.4686,
+      "step": 9341
+    },
+    {
+      "epoch": 0.43036808402819365,
+      "grad_norm": 2.6643777894165033,
+      "learning_rate": 0.00012707492719618227,
+      "loss": 1.1867,
+      "step": 9342
+    },
+    {
+      "epoch": 0.43041415211682865,
+      "grad_norm": 2.7750852068304885,
+      "learning_rate": 0.00012706056330493175,
+      "loss": 1.3286,
+      "step": 9343
+    },
+    {
+      "epoch": 0.43046022020546365,
+      "grad_norm": 2.374547757386886,
+      "learning_rate": 0.00012704619881122412,
+      "loss": 1.5118,
+      "step": 9344
+    },
+    {
+      "epoch": 0.4305062882940987,
+      "grad_norm": 1.0325850292435443,
+      "learning_rate": 0.0001270318337153793,
+      "loss": 1.5195,
+      "step": 9345
+    },
+    {
+      "epoch": 0.4305523563827337,
+      "grad_norm": 2.150903980090625,
+      "learning_rate": 0.00012701746801771702,
+      "loss": 1.6111,
+      "step": 9346
+    },
+    {
+      "epoch": 0.4305984244713687,
+      "grad_norm": 1.8113914423968822,
+      "learning_rate": 0.00012700310171855718,
+      "loss": 1.3517,
+      "step": 9347
+    },
+    {
+      "epoch": 0.4306444925600037,
+      "grad_norm": 1.2477771253202918,
+      "learning_rate": 0.00012698873481821953,
+      "loss": 1.6117,
+      "step": 9348
+    },
+    {
+      "epoch": 0.4306905606486387,
+      "grad_norm": 1.2875047754711206,
+      "learning_rate": 0.000126974367317024,
+      "loss": 1.0755,
+      "step": 9349
+    },
+    {
+      "epoch": 0.4307366287372737,
+      "grad_norm": 2.7383220203193446,
+      "learning_rate": 0.00012695999921529048,
+      "loss": 1.1638,
+      "step": 9350
+    },
+    {
+      "epoch": 0.4307826968259087,
+      "grad_norm": 1.3093086351300776,
+      "learning_rate": 0.00012694563051333875,
+      "loss": 1.3065,
+      "step": 9351
+    },
+    {
+      "epoch": 0.4308287649145437,
+      "grad_norm": 2.12103122510533,
+      "learning_rate": 0.00012693126121148874,
+      "loss": 1.1484,
+      "step": 9352
+    },
+    {
+      "epoch": 0.4308748330031787,
+      "grad_norm": 1.4279840119017997,
+      "learning_rate": 0.00012691689131006043,
+      "loss": 1.3848,
+      "step": 9353
+    },
+    {
+      "epoch": 0.4309209010918137,
+      "grad_norm": 2.083439502531259,
+      "learning_rate": 0.00012690252080937363,
+      "loss": 1.1948,
+      "step": 9354
+    },
+    {
+      "epoch": 0.4309669691804487,
+      "grad_norm": 3.0742562359366246,
+      "learning_rate": 0.00012688814970974836,
+      "loss": 1.3406,
+      "step": 9355
+    },
+    {
+      "epoch": 0.4310130372690837,
+      "grad_norm": 1.632949130342209,
+      "learning_rate": 0.00012687377801150456,
+      "loss": 1.5369,
+      "step": 9356
+    },
+    {
+      "epoch": 0.4310591053577187,
+      "grad_norm": 4.712336136768866,
+      "learning_rate": 0.00012685940571496214,
+      "loss": 0.9654,
+      "step": 9357
+    },
+    {
+      "epoch": 0.4311051734463537,
+      "grad_norm": 1.4469088183427643,
+      "learning_rate": 0.00012684503282044113,
+      "loss": 1.2366,
+      "step": 9358
+    },
+    {
+      "epoch": 0.43115124153498874,
+      "grad_norm": 2.0758484270240287,
+      "learning_rate": 0.00012683065932826148,
+      "loss": 1.5714,
+      "step": 9359
+    },
+    {
+      "epoch": 0.43119730962362374,
+      "grad_norm": 1.7854314304999956,
+      "learning_rate": 0.00012681628523874322,
+      "loss": 1.3897,
+      "step": 9360
+    },
+    {
+      "epoch": 0.43124337771225874,
+      "grad_norm": 2.7649678216371605,
+      "learning_rate": 0.00012680191055220634,
+      "loss": 1.1423,
+      "step": 9361
+    },
+    {
+      "epoch": 0.43128944580089373,
+      "grad_norm": 2.2235957062414213,
+      "learning_rate": 0.00012678753526897086,
+      "loss": 1.5248,
+      "step": 9362
+    },
+    {
+      "epoch": 0.43133551388952873,
+      "grad_norm": 1.4960928076214228,
+      "learning_rate": 0.00012677315938935686,
+      "loss": 1.518,
+      "step": 9363
+    },
+    {
+      "epoch": 0.43138158197816373,
+      "grad_norm": 1.9907493373512524,
+      "learning_rate": 0.00012675878291368436,
+      "loss": 1.6491,
+      "step": 9364
+    },
+    {
+      "epoch": 0.43142765006679873,
+      "grad_norm": 1.371746848332013,
+      "learning_rate": 0.00012674440584227345,
+      "loss": 1.4041,
+      "step": 9365
+    },
+    {
+      "epoch": 0.4314737181554337,
+      "grad_norm": 1.6545719275122344,
+      "learning_rate": 0.00012673002817544418,
+      "loss": 1.3519,
+      "step": 9366
+    },
+    {
+      "epoch": 0.4315197862440687,
+      "grad_norm": 2.1739306248687793,
+      "learning_rate": 0.00012671564991351668,
+      "loss": 1.3435,
+      "step": 9367
+    },
+    {
+      "epoch": 0.4315658543327037,
+      "grad_norm": 3.137094668009164,
+      "learning_rate": 0.00012670127105681104,
+      "loss": 1.3935,
+      "step": 9368
+    },
+    {
+      "epoch": 0.4316119224213387,
+      "grad_norm": 1.7321221438044758,
+      "learning_rate": 0.0001266868916056474,
+      "loss": 1.4387,
+      "step": 9369
+    },
+    {
+      "epoch": 0.4316579905099737,
+      "grad_norm": 1.3970743761852495,
+      "learning_rate": 0.00012667251156034585,
+      "loss": 1.2917,
+      "step": 9370
+    },
+    {
+      "epoch": 0.4317040585986087,
+      "grad_norm": 3.0996846864591183,
+      "learning_rate": 0.0001266581309212266,
+      "loss": 1.2322,
+      "step": 9371
+    },
+    {
+      "epoch": 0.4317501266872438,
+      "grad_norm": 1.341022568796218,
+      "learning_rate": 0.00012664374968860973,
+      "loss": 1.2086,
+      "step": 9372
+    },
+    {
+      "epoch": 0.43179619477587877,
+      "grad_norm": 1.4356118567057568,
+      "learning_rate": 0.00012662936786281547,
+      "loss": 1.319,
+      "step": 9373
+    },
+    {
+      "epoch": 0.43184226286451377,
+      "grad_norm": 4.719043474612854,
+      "learning_rate": 0.000126614985444164,
+      "loss": 1.1854,
+      "step": 9374
+    },
+    {
+      "epoch": 0.43188833095314877,
+      "grad_norm": 1.8512267053308884,
+      "learning_rate": 0.0001266006024329755,
+      "loss": 1.516,
+      "step": 9375
+    },
+    {
+      "epoch": 0.43193439904178377,
+      "grad_norm": 1.0757796397668813,
+      "learning_rate": 0.00012658621882957022,
+      "loss": 1.4036,
+      "step": 9376
+    },
+    {
+      "epoch": 0.43198046713041877,
+      "grad_norm": 3.1280841372422548,
+      "learning_rate": 0.00012657183463426835,
+      "loss": 1.0529,
+      "step": 9377
+    },
+    {
+      "epoch": 0.43202653521905376,
+      "grad_norm": 1.6203113464940786,
+      "learning_rate": 0.00012655744984739017,
+      "loss": 1.2405,
+      "step": 9378
+    },
+    {
+      "epoch": 0.43207260330768876,
+      "grad_norm": 3.4246356695252604,
+      "learning_rate": 0.00012654306446925587,
+      "loss": 1.5751,
+      "step": 9379
+    },
+    {
+      "epoch": 0.43211867139632376,
+      "grad_norm": 2.811556573073817,
+      "learning_rate": 0.00012652867850018577,
+      "loss": 1.3844,
+      "step": 9380
+    },
+    {
+      "epoch": 0.43216473948495876,
+      "grad_norm": 1.129546470648232,
+      "learning_rate": 0.00012651429194050014,
+      "loss": 1.3339,
+      "step": 9381
+    },
+    {
+      "epoch": 0.43221080757359376,
+      "grad_norm": 1.3568979254881781,
+      "learning_rate": 0.00012649990479051922,
+      "loss": 1.483,
+      "step": 9382
+    },
+    {
+      "epoch": 0.43225687566222876,
+      "grad_norm": 1.948467165283276,
+      "learning_rate": 0.0001264855170505634,
+      "loss": 1.3275,
+      "step": 9383
+    },
+    {
+      "epoch": 0.43230294375086376,
+      "grad_norm": 1.4661233758754173,
+      "learning_rate": 0.0001264711287209529,
+      "loss": 1.407,
+      "step": 9384
+    },
+    {
+      "epoch": 0.43234901183949875,
+      "grad_norm": 1.974939692934869,
+      "learning_rate": 0.00012645673980200817,
+      "loss": 1.2119,
+      "step": 9385
+    },
+    {
+      "epoch": 0.4323950799281338,
+      "grad_norm": 1.4411664803235547,
+      "learning_rate": 0.00012644235029404948,
+      "loss": 1.5277,
+      "step": 9386
+    },
+    {
+      "epoch": 0.4324411480167688,
+      "grad_norm": 1.1123889480550846,
+      "learning_rate": 0.0001264279601973972,
+      "loss": 1.3275,
+      "step": 9387
+    },
+    {
+      "epoch": 0.4324872161054038,
+      "grad_norm": 1.568122091102267,
+      "learning_rate": 0.0001264135695123717,
+      "loss": 1.3428,
+      "step": 9388
+    },
+    {
+      "epoch": 0.4325332841940388,
+      "grad_norm": 1.4223441763814082,
+      "learning_rate": 0.00012639917823929335,
+      "loss": 1.3333,
+      "step": 9389
+    },
+    {
+      "epoch": 0.4325793522826738,
+      "grad_norm": 5.713455525046865,
+      "learning_rate": 0.00012638478637848254,
+      "loss": 1.3223,
+      "step": 9390
+    },
+    {
+      "epoch": 0.4326254203713088,
+      "grad_norm": 1.374488047148518,
+      "learning_rate": 0.00012637039393025972,
+      "loss": 1.4141,
+      "step": 9391
+    },
+    {
+      "epoch": 0.4326714884599438,
+      "grad_norm": 0.9748155479925771,
+      "learning_rate": 0.0001263560008949453,
+      "loss": 1.5999,
+      "step": 9392
+    },
+    {
+      "epoch": 0.4327175565485788,
+      "grad_norm": 1.3482029326390876,
+      "learning_rate": 0.00012634160727285974,
+      "loss": 0.9845,
+      "step": 9393
+    },
+    {
+      "epoch": 0.4327636246372138,
+      "grad_norm": 0.8926088035363245,
+      "learning_rate": 0.0001263272130643234,
+      "loss": 1.2321,
+      "step": 9394
+    },
+    {
+      "epoch": 0.4328096927258488,
+      "grad_norm": 0.7060836195793424,
+      "learning_rate": 0.00012631281826965687,
+      "loss": 1.6544,
+      "step": 9395
+    },
+    {
+      "epoch": 0.4328557608144838,
+      "grad_norm": 1.5678617019644785,
+      "learning_rate": 0.0001262984228891805,
+      "loss": 1.5776,
+      "step": 9396
+    },
+    {
+      "epoch": 0.4329018289031188,
+      "grad_norm": 1.8025963519484327,
+      "learning_rate": 0.00012628402692321485,
+      "loss": 1.6102,
+      "step": 9397
+    },
+    {
+      "epoch": 0.4329478969917538,
+      "grad_norm": 2.262881321062213,
+      "learning_rate": 0.00012626963037208044,
+      "loss": 1.3587,
+      "step": 9398
+    },
+    {
+      "epoch": 0.4329939650803888,
+      "grad_norm": 1.0630821836653301,
+      "learning_rate": 0.0001262552332360977,
+      "loss": 1.5614,
+      "step": 9399
+    },
+    {
+      "epoch": 0.43304003316902384,
+      "grad_norm": 1.402142954388183,
+      "learning_rate": 0.00012624083551558726,
+      "loss": 1.6465,
+      "step": 9400
+    },
+    {
+      "epoch": 0.43308610125765884,
+      "grad_norm": 1.206047076119773,
+      "learning_rate": 0.00012622643721086957,
+      "loss": 1.3379,
+      "step": 9401
+    },
+    {
+      "epoch": 0.43313216934629384,
+      "grad_norm": 1.3944038996860928,
+      "learning_rate": 0.00012621203832226526,
+      "loss": 1.1852,
+      "step": 9402
+    },
+    {
+      "epoch": 0.43317823743492884,
+      "grad_norm": 1.7669282553957166,
+      "learning_rate": 0.00012619763885009483,
+      "loss": 1.3892,
+      "step": 9403
+    },
+    {
+      "epoch": 0.43322430552356384,
+      "grad_norm": 1.8595845741213224,
+      "learning_rate": 0.00012618323879467892,
+      "loss": 1.2507,
+      "step": 9404
+    },
+    {
+      "epoch": 0.43327037361219883,
+      "grad_norm": 0.8886621087297676,
+      "learning_rate": 0.00012616883815633808,
+      "loss": 1.2874,
+      "step": 9405
+    },
+    {
+      "epoch": 0.43331644170083383,
+      "grad_norm": 1.1729918595659872,
+      "learning_rate": 0.00012615443693539296,
+      "loss": 1.2485,
+      "step": 9406
+    },
+    {
+      "epoch": 0.43336250978946883,
+      "grad_norm": 1.0723477411897906,
+      "learning_rate": 0.0001261400351321641,
+      "loss": 1.2501,
+      "step": 9407
+    },
+    {
+      "epoch": 0.43340857787810383,
+      "grad_norm": 1.5261084700206942,
+      "learning_rate": 0.0001261256327469722,
+      "loss": 1.4296,
+      "step": 9408
+    },
+    {
+      "epoch": 0.4334546459667388,
+      "grad_norm": 0.9506791355663267,
+      "learning_rate": 0.00012611122978013786,
+      "loss": 1.6825,
+      "step": 9409
+    },
+    {
+      "epoch": 0.4335007140553738,
+      "grad_norm": 3.0513256478366095,
+      "learning_rate": 0.00012609682623198175,
+      "loss": 1.4779,
+      "step": 9410
+    },
+    {
+      "epoch": 0.4335467821440088,
+      "grad_norm": 1.4633937639758798,
+      "learning_rate": 0.00012608242210282458,
+      "loss": 1.2951,
+      "step": 9411
+    },
+    {
+      "epoch": 0.4335928502326438,
+      "grad_norm": 0.9218150323621518,
+      "learning_rate": 0.000126068017392987,
+      "loss": 1.4889,
+      "step": 9412
+    },
+    {
+      "epoch": 0.4336389183212789,
+      "grad_norm": 2.117376260813873,
+      "learning_rate": 0.00012605361210278973,
+      "loss": 1.1634,
+      "step": 9413
+    },
+    {
+      "epoch": 0.4336849864099139,
+      "grad_norm": 1.5915992174917832,
+      "learning_rate": 0.00012603920623255343,
+      "loss": 1.4047,
+      "step": 9414
+    },
+    {
+      "epoch": 0.4337310544985489,
+      "grad_norm": 1.1285840471208077,
+      "learning_rate": 0.00012602479978259887,
+      "loss": 1.3051,
+      "step": 9415
+    },
+    {
+      "epoch": 0.43377712258718387,
+      "grad_norm": 1.2497070669728876,
+      "learning_rate": 0.00012601039275324675,
+      "loss": 1.2937,
+      "step": 9416
+    },
+    {
+      "epoch": 0.43382319067581887,
+      "grad_norm": 1.2769903477391162,
+      "learning_rate": 0.00012599598514481786,
+      "loss": 1.3946,
+      "step": 9417
+    },
+    {
+      "epoch": 0.43386925876445387,
+      "grad_norm": 1.6046957375037556,
+      "learning_rate": 0.00012598157695763287,
+      "loss": 1.203,
+      "step": 9418
+    },
+    {
+      "epoch": 0.43391532685308887,
+      "grad_norm": 1.2074314590058395,
+      "learning_rate": 0.0001259671681920127,
+      "loss": 1.2072,
+      "step": 9419
+    },
+    {
+      "epoch": 0.43396139494172387,
+      "grad_norm": 1.2851507369620463,
+      "learning_rate": 0.00012595275884827797,
+      "loss": 1.5617,
+      "step": 9420
+    },
+    {
+      "epoch": 0.43400746303035886,
+      "grad_norm": 1.9707396138936437,
+      "learning_rate": 0.00012593834892674964,
+      "loss": 1.4143,
+      "step": 9421
+    },
+    {
+      "epoch": 0.43405353111899386,
+      "grad_norm": 0.9797739298741623,
+      "learning_rate": 0.00012592393842774839,
+      "loss": 1.3874,
+      "step": 9422
+    },
+    {
+      "epoch": 0.43409959920762886,
+      "grad_norm": 2.19580255808039,
+      "learning_rate": 0.00012590952735159515,
+      "loss": 1.3872,
+      "step": 9423
+    },
+    {
+      "epoch": 0.43414566729626386,
+      "grad_norm": 1.5620107653537778,
+      "learning_rate": 0.00012589511569861069,
+      "loss": 1.3548,
+      "step": 9424
+    },
+    {
+      "epoch": 0.43419173538489886,
+      "grad_norm": 3.316693249770102,
+      "learning_rate": 0.00012588070346911584,
+      "loss": 1.2786,
+      "step": 9425
+    },
+    {
+      "epoch": 0.43423780347353386,
+      "grad_norm": 1.4094228385896743,
+      "learning_rate": 0.00012586629066343156,
+      "loss": 1.1747,
+      "step": 9426
+    },
+    {
+      "epoch": 0.4342838715621689,
+      "grad_norm": 10.846787193090195,
+      "learning_rate": 0.00012585187728187864,
+      "loss": 1.4934,
+      "step": 9427
+    },
+    {
+      "epoch": 0.4343299396508039,
+      "grad_norm": 3.83699548192251,
+      "learning_rate": 0.00012583746332477802,
+      "loss": 1.2783,
+      "step": 9428
+    },
+    {
+      "epoch": 0.4343760077394389,
+      "grad_norm": 1.4007168122193447,
+      "learning_rate": 0.00012582304879245055,
+      "loss": 1.4153,
+      "step": 9429
+    },
+    {
+      "epoch": 0.4344220758280739,
+      "grad_norm": 1.852476778323208,
+      "learning_rate": 0.0001258086336852172,
+      "loss": 1.3195,
+      "step": 9430
+    },
+    {
+      "epoch": 0.4344681439167089,
+      "grad_norm": 3.1986071132957914,
+      "learning_rate": 0.00012579421800339884,
+      "loss": 1.2577,
+      "step": 9431
+    },
+    {
+      "epoch": 0.4345142120053439,
+      "grad_norm": 1.7454267411456375,
+      "learning_rate": 0.0001257798017473165,
+      "loss": 1.3925,
+      "step": 9432
+    },
+    {
+      "epoch": 0.4345602800939789,
+      "grad_norm": 1.577357994547793,
+      "learning_rate": 0.00012576538491729103,
+      "loss": 1.6169,
+      "step": 9433
+    },
+    {
+      "epoch": 0.4346063481826139,
+      "grad_norm": 1.1427734777423648,
+      "learning_rate": 0.00012575096751364347,
+      "loss": 1.4127,
+      "step": 9434
+    },
+    {
+      "epoch": 0.4346524162712489,
+      "grad_norm": 1.852196370430504,
+      "learning_rate": 0.00012573654953669476,
+      "loss": 1.7661,
+      "step": 9435
+    },
+    {
+      "epoch": 0.4346984843598839,
+      "grad_norm": 1.615201645277666,
+      "learning_rate": 0.00012572213098676588,
+      "loss": 1.3876,
+      "step": 9436
+    },
+    {
+      "epoch": 0.4347445524485189,
+      "grad_norm": 1.5114916172195514,
+      "learning_rate": 0.0001257077118641779,
+      "loss": 1.1194,
+      "step": 9437
+    },
+    {
+      "epoch": 0.4347906205371539,
+      "grad_norm": 2.3767882805070215,
+      "learning_rate": 0.00012569329216925175,
+      "loss": 1.4521,
+      "step": 9438
+    },
+    {
+      "epoch": 0.4348366886257889,
+      "grad_norm": 2.183808038312182,
+      "learning_rate": 0.00012567887190230857,
+      "loss": 1.4853,
+      "step": 9439
+    },
+    {
+      "epoch": 0.43488275671442395,
+      "grad_norm": 1.8458093834123606,
+      "learning_rate": 0.00012566445106366927,
+      "loss": 1.259,
+      "step": 9440
+    },
+    {
+      "epoch": 0.43492882480305894,
+      "grad_norm": 1.192966098835879,
+      "learning_rate": 0.000125650029653655,
+      "loss": 1.7225,
+      "step": 9441
+    },
+    {
+      "epoch": 0.43497489289169394,
+      "grad_norm": 1.1773754209663434,
+      "learning_rate": 0.0001256356076725868,
+      "loss": 1.5618,
+      "step": 9442
+    },
+    {
+      "epoch": 0.43502096098032894,
+      "grad_norm": 2.1322147951347676,
+      "learning_rate": 0.00012562118512078575,
+      "loss": 1.7152,
+      "step": 9443
+    },
+    {
+      "epoch": 0.43506702906896394,
+      "grad_norm": 1.8319063717094082,
+      "learning_rate": 0.00012560676199857294,
+      "loss": 1.4564,
+      "step": 9444
+    },
+    {
+      "epoch": 0.43511309715759894,
+      "grad_norm": 2.8187294296117247,
+      "learning_rate": 0.00012559233830626947,
+      "loss": 1.5236,
+      "step": 9445
+    },
+    {
+      "epoch": 0.43515916524623394,
+      "grad_norm": 1.3359039879101022,
+      "learning_rate": 0.0001255779140441965,
+      "loss": 1.3416,
+      "step": 9446
+    },
+    {
+      "epoch": 0.43520523333486894,
+      "grad_norm": 3.3313289239356854,
+      "learning_rate": 0.00012556348921267512,
+      "loss": 1.5384,
+      "step": 9447
+    },
+    {
+      "epoch": 0.43525130142350393,
+      "grad_norm": 1.562510790113455,
+      "learning_rate": 0.0001255490638120265,
+      "loss": 1.5312,
+      "step": 9448
+    },
+    {
+      "epoch": 0.43529736951213893,
+      "grad_norm": 2.8274378098520168,
+      "learning_rate": 0.00012553463784257174,
+      "loss": 1.1,
+      "step": 9449
+    },
+    {
+      "epoch": 0.43534343760077393,
+      "grad_norm": 1.761756157909338,
+      "learning_rate": 0.00012552021130463208,
+      "loss": 1.6547,
+      "step": 9450
+    },
+    {
+      "epoch": 0.43538950568940893,
+      "grad_norm": 1.183585477600852,
+      "learning_rate": 0.00012550578419852866,
+      "loss": 1.2137,
+      "step": 9451
+    },
+    {
+      "epoch": 0.4354355737780439,
+      "grad_norm": 1.6302845136163295,
+      "learning_rate": 0.0001254913565245827,
+      "loss": 1.2664,
+      "step": 9452
+    },
+    {
+      "epoch": 0.4354816418666789,
+      "grad_norm": 3.3286151888646422,
+      "learning_rate": 0.00012547692828311538,
+      "loss": 1.1468,
+      "step": 9453
+    },
+    {
+      "epoch": 0.435527709955314,
+      "grad_norm": 1.6123886200385507,
+      "learning_rate": 0.00012546249947444793,
+      "loss": 1.3338,
+      "step": 9454
+    },
+    {
+      "epoch": 0.435573778043949,
+      "grad_norm": 1.1723682834926568,
+      "learning_rate": 0.00012544807009890163,
+      "loss": 1.5325,
+      "step": 9455
+    },
+    {
+      "epoch": 0.435619846132584,
+      "grad_norm": 1.1575444878297252,
+      "learning_rate": 0.00012543364015679762,
+      "loss": 1.3205,
+      "step": 9456
+    },
+    {
+      "epoch": 0.435665914221219,
+      "grad_norm": 1.162559589031715,
+      "learning_rate": 0.00012541920964845727,
+      "loss": 1.4916,
+      "step": 9457
+    },
+    {
+      "epoch": 0.435711982309854,
+      "grad_norm": 1.5341374216972177,
+      "learning_rate": 0.00012540477857420177,
+      "loss": 1.3266,
+      "step": 9458
+    },
+    {
+      "epoch": 0.43575805039848897,
+      "grad_norm": 1.388100807030221,
+      "learning_rate": 0.00012539034693435247,
+      "loss": 1.4564,
+      "step": 9459
+    },
+    {
+      "epoch": 0.43580411848712397,
+      "grad_norm": 1.5946410324011562,
+      "learning_rate": 0.00012537591472923062,
+      "loss": 1.3446,
+      "step": 9460
+    },
+    {
+      "epoch": 0.43585018657575897,
+      "grad_norm": 1.578543296232885,
+      "learning_rate": 0.00012536148195915753,
+      "loss": 1.3755,
+      "step": 9461
+    },
+    {
+      "epoch": 0.43589625466439397,
+      "grad_norm": 1.5736755316033382,
+      "learning_rate": 0.00012534704862445452,
+      "loss": 1.6577,
+      "step": 9462
+    },
+    {
+      "epoch": 0.43594232275302897,
+      "grad_norm": 1.6905920794192426,
+      "learning_rate": 0.00012533261472544297,
+      "loss": 0.8667,
+      "step": 9463
+    },
+    {
+      "epoch": 0.43598839084166396,
+      "grad_norm": 2.068205590868379,
+      "learning_rate": 0.00012531818026244415,
+      "loss": 1.081,
+      "step": 9464
+    },
+    {
+      "epoch": 0.43603445893029896,
+      "grad_norm": 1.4899771662751091,
+      "learning_rate": 0.00012530374523577948,
+      "loss": 1.3425,
+      "step": 9465
+    },
+    {
+      "epoch": 0.43608052701893396,
+      "grad_norm": 1.756596343572258,
+      "learning_rate": 0.0001252893096457703,
+      "loss": 1.2634,
+      "step": 9466
+    },
+    {
+      "epoch": 0.43612659510756896,
+      "grad_norm": 2.367160214341238,
+      "learning_rate": 0.000125274873492738,
+      "loss": 1.4458,
+      "step": 9467
+    },
+    {
+      "epoch": 0.436172663196204,
+      "grad_norm": 3.4838827948187623,
+      "learning_rate": 0.00012526043677700397,
+      "loss": 1.3289,
+      "step": 9468
+    },
+    {
+      "epoch": 0.436218731284839,
+      "grad_norm": 1.4080245531190603,
+      "learning_rate": 0.00012524599949888964,
+      "loss": 1.2391,
+      "step": 9469
+    },
+    {
+      "epoch": 0.436264799373474,
+      "grad_norm": 1.4639488192500147,
+      "learning_rate": 0.00012523156165871642,
+      "loss": 1.5011,
+      "step": 9470
+    },
+    {
+      "epoch": 0.436310867462109,
+      "grad_norm": 1.115278751981585,
+      "learning_rate": 0.00012521712325680575,
+      "loss": 1.572,
+      "step": 9471
+    },
+    {
+      "epoch": 0.436356935550744,
+      "grad_norm": 1.6453278182657993,
+      "learning_rate": 0.00012520268429347907,
+      "loss": 1.6153,
+      "step": 9472
+    },
+    {
+      "epoch": 0.436403003639379,
+      "grad_norm": 1.173149413740306,
+      "learning_rate": 0.00012518824476905781,
+      "loss": 1.2754,
+      "step": 9473
+    },
+    {
+      "epoch": 0.436449071728014,
+      "grad_norm": 1.2713549426550559,
+      "learning_rate": 0.00012517380468386348,
+      "loss": 1.2383,
+      "step": 9474
+    },
+    {
+      "epoch": 0.436495139816649,
+      "grad_norm": 1.718638917497141,
+      "learning_rate": 0.00012515936403821755,
+      "loss": 1.3586,
+      "step": 9475
+    },
+    {
+      "epoch": 0.436541207905284,
+      "grad_norm": 1.994984902254957,
+      "learning_rate": 0.00012514492283244156,
+      "loss": 1.1355,
+      "step": 9476
+    },
+    {
+      "epoch": 0.436587275993919,
+      "grad_norm": 1.1312341466073517,
+      "learning_rate": 0.0001251304810668569,
+      "loss": 1.4924,
+      "step": 9477
+    },
+    {
+      "epoch": 0.436633344082554,
+      "grad_norm": 1.2561954446042338,
+      "learning_rate": 0.00012511603874178526,
+      "loss": 1.2178,
+      "step": 9478
+    },
+    {
+      "epoch": 0.436679412171189,
+      "grad_norm": 1.6655852859029758,
+      "learning_rate": 0.00012510159585754803,
+      "loss": 1.3068,
+      "step": 9479
+    },
+    {
+      "epoch": 0.436725480259824,
+      "grad_norm": 2.0380383165093066,
+      "learning_rate": 0.0001250871524144668,
+      "loss": 1.4589,
+      "step": 9480
+    },
+    {
+      "epoch": 0.43677154834845905,
+      "grad_norm": 1.2109230953744248,
+      "learning_rate": 0.00012507270841286317,
+      "loss": 1.515,
+      "step": 9481
+    },
+    {
+      "epoch": 0.43681761643709405,
+      "grad_norm": 2.230100204897014,
+      "learning_rate": 0.00012505826385305865,
+      "loss": 1.6273,
+      "step": 9482
+    },
+    {
+      "epoch": 0.43686368452572905,
+      "grad_norm": 1.0102634895739893,
+      "learning_rate": 0.00012504381873537485,
+      "loss": 1.3986,
+      "step": 9483
+    },
+    {
+      "epoch": 0.43690975261436404,
+      "grad_norm": 2.4849305734589815,
+      "learning_rate": 0.00012502937306013337,
+      "loss": 0.948,
+      "step": 9484
+    },
+    {
+      "epoch": 0.43695582070299904,
+      "grad_norm": 3.219715046675327,
+      "learning_rate": 0.00012501492682765583,
+      "loss": 1.2687,
+      "step": 9485
+    },
+    {
+      "epoch": 0.43700188879163404,
+      "grad_norm": 1.8452623283192955,
+      "learning_rate": 0.0001250004800382638,
+      "loss": 1.253,
+      "step": 9486
+    },
+    {
+      "epoch": 0.43704795688026904,
+      "grad_norm": 1.3436013206999902,
+      "learning_rate": 0.000124986032692279,
+      "loss": 1.3524,
+      "step": 9487
+    },
+    {
+      "epoch": 0.43709402496890404,
+      "grad_norm": 1.3454567102546995,
+      "learning_rate": 0.000124971584790023,
+      "loss": 1.4619,
+      "step": 9488
+    },
+    {
+      "epoch": 0.43714009305753904,
+      "grad_norm": 1.6748353355795285,
+      "learning_rate": 0.00012495713633181747,
+      "loss": 1.3316,
+      "step": 9489
+    },
+    {
+      "epoch": 0.43718616114617403,
+      "grad_norm": 2.425877219031016,
+      "learning_rate": 0.00012494268731798411,
+      "loss": 1.3712,
+      "step": 9490
+    },
+    {
+      "epoch": 0.43723222923480903,
+      "grad_norm": 1.148450150164431,
+      "learning_rate": 0.00012492823774884457,
+      "loss": 1.4242,
+      "step": 9491
+    },
+    {
+      "epoch": 0.43727829732344403,
+      "grad_norm": 1.2154987074702752,
+      "learning_rate": 0.00012491378762472056,
+      "loss": 1.446,
+      "step": 9492
+    },
+    {
+      "epoch": 0.43732436541207903,
+      "grad_norm": 1.4815685376325032,
+      "learning_rate": 0.0001248993369459338,
+      "loss": 1.3138,
+      "step": 9493
+    },
+    {
+      "epoch": 0.43737043350071403,
+      "grad_norm": 1.197664610093745,
+      "learning_rate": 0.000124884885712806,
+      "loss": 1.6298,
+      "step": 9494
+    },
+    {
+      "epoch": 0.4374165015893491,
+      "grad_norm": 2.1136545267217075,
+      "learning_rate": 0.00012487043392565888,
+      "loss": 1.2808,
+      "step": 9495
+    },
+    {
+      "epoch": 0.4374625696779841,
+      "grad_norm": 1.8750150855573349,
+      "learning_rate": 0.00012485598158481419,
+      "loss": 1.5712,
+      "step": 9496
+    },
+    {
+      "epoch": 0.4375086377666191,
+      "grad_norm": 1.2393545419865009,
+      "learning_rate": 0.00012484152869059368,
+      "loss": 1.1978,
+      "step": 9497
+    },
+    {
+      "epoch": 0.4375547058552541,
+      "grad_norm": 2.452111598862184,
+      "learning_rate": 0.00012482707524331913,
+      "loss": 1.1878,
+      "step": 9498
+    },
+    {
+      "epoch": 0.4376007739438891,
+      "grad_norm": 1.9688981902328027,
+      "learning_rate": 0.00012481262124331235,
+      "loss": 1.1459,
+      "step": 9499
+    },
+    {
+      "epoch": 0.4376468420325241,
+      "grad_norm": 2.0427491032595175,
+      "learning_rate": 0.00012479816669089505,
+      "loss": 1.1656,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4376929101211591,
+      "grad_norm": 0.9641612032754184,
+      "learning_rate": 0.00012478371158638912,
+      "loss": 1.4309,
+      "step": 9501
+    },
+    {
+      "epoch": 0.43773897820979407,
+      "grad_norm": 1.5016950244477008,
+      "learning_rate": 0.00012476925593011634,
+      "loss": 1.0271,
+      "step": 9502
+    },
+    {
+      "epoch": 0.43778504629842907,
+      "grad_norm": 1.4600201860820623,
+      "learning_rate": 0.00012475479972239855,
+      "loss": 1.3953,
+      "step": 9503
+    },
+    {
+      "epoch": 0.43783111438706407,
+      "grad_norm": 2.6433526990539162,
+      "learning_rate": 0.0001247403429635576,
+      "loss": 1.416,
+      "step": 9504
+    },
+    {
+      "epoch": 0.43787718247569907,
+      "grad_norm": 1.3777077967359608,
+      "learning_rate": 0.00012472588565391534,
+      "loss": 1.2245,
+      "step": 9505
+    },
+    {
+      "epoch": 0.43792325056433407,
+      "grad_norm": 1.5428686361210988,
+      "learning_rate": 0.0001247114277937936,
+      "loss": 1.3961,
+      "step": 9506
+    },
+    {
+      "epoch": 0.43796931865296906,
+      "grad_norm": 1.524339779471337,
+      "learning_rate": 0.0001246969693835143,
+      "loss": 1.1867,
+      "step": 9507
+    },
+    {
+      "epoch": 0.4380153867416041,
+      "grad_norm": 3.8686001855063914,
+      "learning_rate": 0.00012468251042339933,
+      "loss": 0.9872,
+      "step": 9508
+    },
+    {
+      "epoch": 0.4380614548302391,
+      "grad_norm": 4.959820978363578,
+      "learning_rate": 0.00012466805091377057,
+      "loss": 1.3143,
+      "step": 9509
+    },
+    {
+      "epoch": 0.4381075229188741,
+      "grad_norm": 4.680200856693485,
+      "learning_rate": 0.00012465359085495,
+      "loss": 1.245,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4381535910075091,
+      "grad_norm": 1.4198372719505172,
+      "learning_rate": 0.00012463913024725947,
+      "loss": 1.2486,
+      "step": 9511
+    },
+    {
+      "epoch": 0.4381996590961441,
+      "grad_norm": 1.3344631771193232,
+      "learning_rate": 0.00012462466909102094,
+      "loss": 1.0578,
+      "step": 9512
+    },
+    {
+      "epoch": 0.4382457271847791,
+      "grad_norm": 1.1253443845214954,
+      "learning_rate": 0.00012461020738655637,
+      "loss": 1.0185,
+      "step": 9513
+    },
+    {
+      "epoch": 0.4382917952734141,
+      "grad_norm": 2.3527740110641218,
+      "learning_rate": 0.00012459574513418777,
+      "loss": 1.4033,
+      "step": 9514
+    },
+    {
+      "epoch": 0.4383378633620491,
+      "grad_norm": 1.3826465978723204,
+      "learning_rate": 0.00012458128233423706,
+      "loss": 1.4145,
+      "step": 9515
+    },
+    {
+      "epoch": 0.4383839314506841,
+      "grad_norm": 1.6845972803888598,
+      "learning_rate": 0.00012456681898702625,
+      "loss": 1.4061,
+      "step": 9516
+    },
+    {
+      "epoch": 0.4384299995393191,
+      "grad_norm": 1.3281413785681413,
+      "learning_rate": 0.00012455235509287733,
+      "loss": 1.04,
+      "step": 9517
+    },
+    {
+      "epoch": 0.4384760676279541,
+      "grad_norm": 1.1144857296687292,
+      "learning_rate": 0.00012453789065211231,
+      "loss": 1.4488,
+      "step": 9518
+    },
+    {
+      "epoch": 0.4385221357165891,
+      "grad_norm": 1.915978912857706,
+      "learning_rate": 0.00012452342566505323,
+      "loss": 1.3086,
+      "step": 9519
+    },
+    {
+      "epoch": 0.4385682038052241,
+      "grad_norm": 1.0526317309450948,
+      "learning_rate": 0.00012450896013202215,
+      "loss": 1.397,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4386142718938591,
+      "grad_norm": 1.4151546333820055,
+      "learning_rate": 0.00012449449405334108,
+      "loss": 1.3214,
+      "step": 9521
+    },
+    {
+      "epoch": 0.43866033998249415,
+      "grad_norm": 1.3743690837380154,
+      "learning_rate": 0.00012448002742933208,
+      "loss": 1.5191,
+      "step": 9522
+    },
+    {
+      "epoch": 0.43870640807112915,
+      "grad_norm": 1.4089848906803577,
+      "learning_rate": 0.0001244655602603173,
+      "loss": 1.467,
+      "step": 9523
+    },
+    {
+      "epoch": 0.43875247615976415,
+      "grad_norm": 1.338262991477982,
+      "learning_rate": 0.00012445109254661873,
+      "loss": 1.5426,
+      "step": 9524
+    },
+    {
+      "epoch": 0.43879854424839915,
+      "grad_norm": 1.0428169932588185,
+      "learning_rate": 0.00012443662428855852,
+      "loss": 1.5245,
+      "step": 9525
+    },
+    {
+      "epoch": 0.43884461233703415,
+      "grad_norm": 1.689492341345552,
+      "learning_rate": 0.00012442215548645876,
+      "loss": 1.278,
+      "step": 9526
+    },
+    {
+      "epoch": 0.43889068042566914,
+      "grad_norm": 1.2765598952666977,
+      "learning_rate": 0.00012440768614064158,
+      "loss": 1.3767,
+      "step": 9527
+    },
+    {
+      "epoch": 0.43893674851430414,
+      "grad_norm": 1.1854908680975478,
+      "learning_rate": 0.00012439321625142913,
+      "loss": 1.4433,
+      "step": 9528
+    },
+    {
+      "epoch": 0.43898281660293914,
+      "grad_norm": 1.9008790941244733,
+      "learning_rate": 0.00012437874581914355,
+      "loss": 1.4507,
+      "step": 9529
+    },
+    {
+      "epoch": 0.43902888469157414,
+      "grad_norm": 1.958725378988545,
+      "learning_rate": 0.00012436427484410696,
+      "loss": 1.3675,
+      "step": 9530
+    },
+    {
+      "epoch": 0.43907495278020914,
+      "grad_norm": 1.091010013079928,
+      "learning_rate": 0.00012434980332664162,
+      "loss": 1.6952,
+      "step": 9531
+    },
+    {
+      "epoch": 0.43912102086884414,
+      "grad_norm": 0.9873486162082578,
+      "learning_rate": 0.0001243353312670696,
+      "loss": 1.6101,
+      "step": 9532
+    },
+    {
+      "epoch": 0.43916708895747913,
+      "grad_norm": 3.5322290753281336,
+      "learning_rate": 0.00012432085866571321,
+      "loss": 1.5025,
+      "step": 9533
+    },
+    {
+      "epoch": 0.43921315704611413,
+      "grad_norm": 1.1394789326331716,
+      "learning_rate": 0.00012430638552289459,
+      "loss": 1.4479,
+      "step": 9534
+    },
+    {
+      "epoch": 0.43925922513474913,
+      "grad_norm": 9.518064964999908,
+      "learning_rate": 0.00012429191183893596,
+      "loss": 1.3819,
+      "step": 9535
+    },
+    {
+      "epoch": 0.4393052932233842,
+      "grad_norm": 1.7337755534796375,
+      "learning_rate": 0.00012427743761415957,
+      "loss": 1.0158,
+      "step": 9536
+    },
+    {
+      "epoch": 0.4393513613120192,
+      "grad_norm": 2.1536804266517056,
+      "learning_rate": 0.00012426296284888765,
+      "loss": 1.3889,
+      "step": 9537
+    },
+    {
+      "epoch": 0.4393974294006542,
+      "grad_norm": 1.4371040008508082,
+      "learning_rate": 0.00012424848754344247,
+      "loss": 1.5103,
+      "step": 9538
+    },
+    {
+      "epoch": 0.4394434974892892,
+      "grad_norm": 1.2299758407717425,
+      "learning_rate": 0.00012423401169814628,
+      "loss": 1.3419,
+      "step": 9539
+    },
+    {
+      "epoch": 0.4394895655779242,
+      "grad_norm": 1.1490538584818863,
+      "learning_rate": 0.0001242195353133214,
+      "loss": 1.4947,
+      "step": 9540
+    },
+    {
+      "epoch": 0.4395356336665592,
+      "grad_norm": 1.6469259277918469,
+      "learning_rate": 0.00012420505838929007,
+      "loss": 1.5743,
+      "step": 9541
+    },
+    {
+      "epoch": 0.4395817017551942,
+      "grad_norm": 1.8982453378222777,
+      "learning_rate": 0.00012419058092637462,
+      "loss": 1.3049,
+      "step": 9542
+    },
+    {
+      "epoch": 0.4396277698438292,
+      "grad_norm": 1.1363584578976929,
+      "learning_rate": 0.00012417610292489738,
+      "loss": 1.6364,
+      "step": 9543
+    },
+    {
+      "epoch": 0.4396738379324642,
+      "grad_norm": 1.184136776048587,
+      "learning_rate": 0.00012416162438518068,
+      "loss": 1.6609,
+      "step": 9544
+    },
+    {
+      "epoch": 0.43971990602109917,
+      "grad_norm": 2.721512527153145,
+      "learning_rate": 0.0001241471453075468,
+      "loss": 1.2539,
+      "step": 9545
+    },
+    {
+      "epoch": 0.43976597410973417,
+      "grad_norm": 1.9134407726617346,
+      "learning_rate": 0.00012413266569231813,
+      "loss": 1.3536,
+      "step": 9546
+    },
+    {
+      "epoch": 0.43981204219836917,
+      "grad_norm": 1.5065855767264038,
+      "learning_rate": 0.00012411818553981706,
+      "loss": 1.414,
+      "step": 9547
+    },
+    {
+      "epoch": 0.43985811028700417,
+      "grad_norm": 3.0651107480937902,
+      "learning_rate": 0.00012410370485036593,
+      "loss": 1.6243,
+      "step": 9548
+    },
+    {
+      "epoch": 0.4399041783756392,
+      "grad_norm": 1.391790535065336,
+      "learning_rate": 0.00012408922362428716,
+      "loss": 1.7122,
+      "step": 9549
+    },
+    {
+      "epoch": 0.4399502464642742,
+      "grad_norm": 1.9886346717051073,
+      "learning_rate": 0.0001240747418619031,
+      "loss": 1.3766,
+      "step": 9550
+    },
+    {
+      "epoch": 0.4399963145529092,
+      "grad_norm": 1.4400651682465233,
+      "learning_rate": 0.00012406025956353625,
+      "loss": 1.6346,
+      "step": 9551
+    },
+    {
+      "epoch": 0.4400423826415442,
+      "grad_norm": 1.8868852076698956,
+      "learning_rate": 0.00012404577672950893,
+      "loss": 1.4333,
+      "step": 9552
+    },
+    {
+      "epoch": 0.4400884507301792,
+      "grad_norm": 1.9117557958831473,
+      "learning_rate": 0.0001240312933601436,
+      "loss": 1.588,
+      "step": 9553
+    },
+    {
+      "epoch": 0.4401345188188142,
+      "grad_norm": 2.1195125770431225,
+      "learning_rate": 0.00012401680945576277,
+      "loss": 1.277,
+      "step": 9554
+    },
+    {
+      "epoch": 0.4401805869074492,
+      "grad_norm": 1.3496619001750034,
+      "learning_rate": 0.00012400232501668884,
+      "loss": 1.2367,
+      "step": 9555
+    },
+    {
+      "epoch": 0.4402266549960842,
+      "grad_norm": 6.322263819748935,
+      "learning_rate": 0.0001239878400432443,
+      "loss": 1.1517,
+      "step": 9556
+    },
+    {
+      "epoch": 0.4402727230847192,
+      "grad_norm": 1.842433403351581,
+      "learning_rate": 0.0001239733545357516,
+      "loss": 1.1118,
+      "step": 9557
+    },
+    {
+      "epoch": 0.4403187911733542,
+      "grad_norm": 1.2167661019743647,
+      "learning_rate": 0.0001239588684945333,
+      "loss": 1.4262,
+      "step": 9558
+    },
+    {
+      "epoch": 0.4403648592619892,
+      "grad_norm": 1.6511881372649633,
+      "learning_rate": 0.0001239443819199119,
+      "loss": 1.3804,
+      "step": 9559
+    },
+    {
+      "epoch": 0.4404109273506242,
+      "grad_norm": 1.53296802262181,
+      "learning_rate": 0.00012392989481220987,
+      "loss": 1.5365,
+      "step": 9560
+    },
+    {
+      "epoch": 0.4404569954392592,
+      "grad_norm": 1.8037799652847428,
+      "learning_rate": 0.00012391540717174974,
+      "loss": 1.2132,
+      "step": 9561
+    },
+    {
+      "epoch": 0.4405030635278942,
+      "grad_norm": 2.3986113194956262,
+      "learning_rate": 0.00012390091899885412,
+      "loss": 1.5643,
+      "step": 9562
+    },
+    {
+      "epoch": 0.44054913161652925,
+      "grad_norm": 2.329369929603619,
+      "learning_rate": 0.00012388643029384548,
+      "loss": 1.44,
+      "step": 9563
+    },
+    {
+      "epoch": 0.44059519970516425,
+      "grad_norm": 1.0230475975288646,
+      "learning_rate": 0.00012387194105704645,
+      "loss": 1.4633,
+      "step": 9564
+    },
+    {
+      "epoch": 0.44064126779379925,
+      "grad_norm": 2.169485520613914,
+      "learning_rate": 0.00012385745128877954,
+      "loss": 1.1678,
+      "step": 9565
+    },
+    {
+      "epoch": 0.44068733588243425,
+      "grad_norm": 1.379982267531996,
+      "learning_rate": 0.0001238429609893674,
+      "loss": 1.3335,
+      "step": 9566
+    },
+    {
+      "epoch": 0.44073340397106925,
+      "grad_norm": 3.8347934396470005,
+      "learning_rate": 0.00012382847015913266,
+      "loss": 1.3325,
+      "step": 9567
+    },
+    {
+      "epoch": 0.44077947205970425,
+      "grad_norm": 1.5835078151525903,
+      "learning_rate": 0.00012381397879839784,
+      "loss": 1.4296,
+      "step": 9568
+    },
+    {
+      "epoch": 0.44082554014833925,
+      "grad_norm": 1.5397139763388452,
+      "learning_rate": 0.00012379948690748564,
+      "loss": 1.4447,
+      "step": 9569
+    },
+    {
+      "epoch": 0.44087160823697424,
+      "grad_norm": 2.9740953982857468,
+      "learning_rate": 0.00012378499448671866,
+      "loss": 1.618,
+      "step": 9570
+    },
+    {
+      "epoch": 0.44091767632560924,
+      "grad_norm": 4.220231258286569,
+      "learning_rate": 0.00012377050153641962,
+      "loss": 1.0356,
+      "step": 9571
+    },
+    {
+      "epoch": 0.44096374441424424,
+      "grad_norm": 1.4545342047796128,
+      "learning_rate": 0.00012375600805691103,
+      "loss": 1.5857,
+      "step": 9572
+    },
+    {
+      "epoch": 0.44100981250287924,
+      "grad_norm": 1.653241135614489,
+      "learning_rate": 0.00012374151404851572,
+      "loss": 1.2853,
+      "step": 9573
+    },
+    {
+      "epoch": 0.44105588059151424,
+      "grad_norm": 10.751503648251136,
+      "learning_rate": 0.00012372701951155626,
+      "loss": 1.1347,
+      "step": 9574
+    },
+    {
+      "epoch": 0.44110194868014924,
+      "grad_norm": 0.8677460923672122,
+      "learning_rate": 0.00012371252444635543,
+      "loss": 1.3093,
+      "step": 9575
+    },
+    {
+      "epoch": 0.4411480167687843,
+      "grad_norm": 1.8864451669201923,
+      "learning_rate": 0.00012369802885323585,
+      "loss": 1.3373,
+      "step": 9576
+    },
+    {
+      "epoch": 0.4411940848574193,
+      "grad_norm": 2.2440790400002637,
+      "learning_rate": 0.00012368353273252034,
+      "loss": 1.2097,
+      "step": 9577
+    },
+    {
+      "epoch": 0.4412401529460543,
+      "grad_norm": 1.66232852101788,
+      "learning_rate": 0.00012366903608453155,
+      "loss": 1.5438,
+      "step": 9578
+    },
+    {
+      "epoch": 0.4412862210346893,
+      "grad_norm": 4.888836580149807,
+      "learning_rate": 0.00012365453890959228,
+      "loss": 1.2125,
+      "step": 9579
+    },
+    {
+      "epoch": 0.4413322891233243,
+      "grad_norm": 1.5148354088036857,
+      "learning_rate": 0.00012364004120802525,
+      "loss": 1.4703,
+      "step": 9580
+    },
+    {
+      "epoch": 0.4413783572119593,
+      "grad_norm": 3.0171502001490365,
+      "learning_rate": 0.00012362554298015323,
+      "loss": 1.1949,
+      "step": 9581
+    },
+    {
+      "epoch": 0.4414244253005943,
+      "grad_norm": 1.9065876295141964,
+      "learning_rate": 0.00012361104422629903,
+      "loss": 1.5538,
+      "step": 9582
+    },
+    {
+      "epoch": 0.4414704933892293,
+      "grad_norm": 1.9217403135251663,
+      "learning_rate": 0.00012359654494678538,
+      "loss": 1.7071,
+      "step": 9583
+    },
+    {
+      "epoch": 0.4415165614778643,
+      "grad_norm": 3.2309757879452516,
+      "learning_rate": 0.00012358204514193515,
+      "loss": 1.2811,
+      "step": 9584
+    },
+    {
+      "epoch": 0.4415626295664993,
+      "grad_norm": 1.5407359999809525,
+      "learning_rate": 0.0001235675448120711,
+      "loss": 1.5288,
+      "step": 9585
+    },
+    {
+      "epoch": 0.4416086976551343,
+      "grad_norm": 1.5638048090749164,
+      "learning_rate": 0.00012355304395751608,
+      "loss": 1.3717,
+      "step": 9586
+    },
+    {
+      "epoch": 0.4416547657437693,
+      "grad_norm": 2.5717569983110984,
+      "learning_rate": 0.00012353854257859292,
+      "loss": 1.3765,
+      "step": 9587
+    },
+    {
+      "epoch": 0.44170083383240427,
+      "grad_norm": 1.1854843290107533,
+      "learning_rate": 0.00012352404067562445,
+      "loss": 1.6045,
+      "step": 9588
+    },
+    {
+      "epoch": 0.44174690192103927,
+      "grad_norm": 1.260923599327443,
+      "learning_rate": 0.0001235095382489336,
+      "loss": 1.401,
+      "step": 9589
+    },
+    {
+      "epoch": 0.4417929700096743,
+      "grad_norm": 1.5099734774713398,
+      "learning_rate": 0.00012349503529884314,
+      "loss": 1.2837,
+      "step": 9590
+    },
+    {
+      "epoch": 0.4418390380983093,
+      "grad_norm": 5.133625955148212,
+      "learning_rate": 0.00012348053182567603,
+      "loss": 1.3623,
+      "step": 9591
+    },
+    {
+      "epoch": 0.4418851061869443,
+      "grad_norm": 2.9233634601699103,
+      "learning_rate": 0.0001234660278297551,
+      "loss": 1.4125,
+      "step": 9592
+    },
+    {
+      "epoch": 0.4419311742755793,
+      "grad_norm": 1.4102877989143454,
+      "learning_rate": 0.00012345152331140336,
+      "loss": 1.2195,
+      "step": 9593
+    },
+    {
+      "epoch": 0.4419772423642143,
+      "grad_norm": 1.655704519429623,
+      "learning_rate": 0.00012343701827094363,
+      "loss": 1.6922,
+      "step": 9594
+    },
+    {
+      "epoch": 0.4420233104528493,
+      "grad_norm": 2.823085687568915,
+      "learning_rate": 0.00012342251270869889,
+      "loss": 1.3531,
+      "step": 9595
+    },
+    {
+      "epoch": 0.4420693785414843,
+      "grad_norm": 2.9309571635168306,
+      "learning_rate": 0.00012340800662499204,
+      "loss": 1.4613,
+      "step": 9596
+    },
+    {
+      "epoch": 0.4421154466301193,
+      "grad_norm": 2.511906623090244,
+      "learning_rate": 0.00012339350002014607,
+      "loss": 1.4349,
+      "step": 9597
+    },
+    {
+      "epoch": 0.4421615147187543,
+      "grad_norm": 4.619758400828155,
+      "learning_rate": 0.00012337899289448394,
+      "loss": 1.2714,
+      "step": 9598
+    },
+    {
+      "epoch": 0.4422075828073893,
+      "grad_norm": 2.682033259260995,
+      "learning_rate": 0.0001233644852483286,
+      "loss": 1.3801,
+      "step": 9599
+    },
+    {
+      "epoch": 0.4422536508960243,
+      "grad_norm": 1.3378512343112203,
+      "learning_rate": 0.0001233499770820031,
+      "loss": 1.5218,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4422997189846593,
+      "grad_norm": 1.570714729653691,
+      "learning_rate": 0.00012333546839583034,
+      "loss": 1.6125,
+      "step": 9601
+    },
+    {
+      "epoch": 0.4423457870732943,
+      "grad_norm": 2.575078510722149,
+      "learning_rate": 0.0001233209591901334,
+      "loss": 1.2095,
+      "step": 9602
+    },
+    {
+      "epoch": 0.4423918551619293,
+      "grad_norm": 2.78985561301156,
+      "learning_rate": 0.00012330644946523532,
+      "loss": 1.256,
+      "step": 9603
+    },
+    {
+      "epoch": 0.44243792325056436,
+      "grad_norm": 1.9976190872543518,
+      "learning_rate": 0.0001232919392214591,
+      "loss": 1.5967,
+      "step": 9604
+    },
+    {
+      "epoch": 0.44248399133919936,
+      "grad_norm": 2.264410335092586,
+      "learning_rate": 0.00012327742845912777,
+      "loss": 1.4803,
+      "step": 9605
+    },
+    {
+      "epoch": 0.44253005942783435,
+      "grad_norm": 3.2250200559492086,
+      "learning_rate": 0.00012326291717856447,
+      "loss": 1.2108,
+      "step": 9606
+    },
+    {
+      "epoch": 0.44257612751646935,
+      "grad_norm": 1.6937259458772616,
+      "learning_rate": 0.00012324840538009213,
+      "loss": 1.2451,
+      "step": 9607
+    },
+    {
+      "epoch": 0.44262219560510435,
+      "grad_norm": 1.733732043244852,
+      "learning_rate": 0.00012323389306403393,
+      "loss": 1.1325,
+      "step": 9608
+    },
+    {
+      "epoch": 0.44266826369373935,
+      "grad_norm": 1.4895045290548636,
+      "learning_rate": 0.00012321938023071292,
+      "loss": 1.3127,
+      "step": 9609
+    },
+    {
+      "epoch": 0.44271433178237435,
+      "grad_norm": 2.7905047615812792,
+      "learning_rate": 0.00012320486688045226,
+      "loss": 1.4181,
+      "step": 9610
+    },
+    {
+      "epoch": 0.44276039987100935,
+      "grad_norm": 1.7473269277192096,
+      "learning_rate": 0.000123190353013575,
+      "loss": 1.416,
+      "step": 9611
+    },
+    {
+      "epoch": 0.44280646795964435,
+      "grad_norm": 1.448569100740025,
+      "learning_rate": 0.0001231758386304043,
+      "loss": 1.439,
+      "step": 9612
+    },
+    {
+      "epoch": 0.44285253604827934,
+      "grad_norm": 1.8502832774344296,
+      "learning_rate": 0.0001231613237312633,
+      "loss": 1.3544,
+      "step": 9613
+    },
+    {
+      "epoch": 0.44289860413691434,
+      "grad_norm": 1.7062570485851474,
+      "learning_rate": 0.00012314680831647512,
+      "loss": 1.1347,
+      "step": 9614
+    },
+    {
+      "epoch": 0.44294467222554934,
+      "grad_norm": 1.6392833873635637,
+      "learning_rate": 0.00012313229238636298,
+      "loss": 1.6226,
+      "step": 9615
+    },
+    {
+      "epoch": 0.44299074031418434,
+      "grad_norm": 2.4438417246313144,
+      "learning_rate": 0.00012311777594124998,
+      "loss": 1.4248,
+      "step": 9616
+    },
+    {
+      "epoch": 0.4430368084028194,
+      "grad_norm": 1.596080731554679,
+      "learning_rate": 0.00012310325898145936,
+      "loss": 1.5441,
+      "step": 9617
+    },
+    {
+      "epoch": 0.4430828764914544,
+      "grad_norm": 2.230318766311692,
+      "learning_rate": 0.00012308874150731427,
+      "loss": 1.1623,
+      "step": 9618
+    },
+    {
+      "epoch": 0.4431289445800894,
+      "grad_norm": 1.654487183785855,
+      "learning_rate": 0.00012307422351913797,
+      "loss": 1.3423,
+      "step": 9619
+    },
+    {
+      "epoch": 0.4431750126687244,
+      "grad_norm": 1.8048480839271852,
+      "learning_rate": 0.00012305970501725358,
+      "loss": 1.3188,
+      "step": 9620
+    },
+    {
+      "epoch": 0.4432210807573594,
+      "grad_norm": 1.076644364337536,
+      "learning_rate": 0.00012304518600198444,
+      "loss": 1.3503,
+      "step": 9621
+    },
+    {
+      "epoch": 0.4432671488459944,
+      "grad_norm": 3.879483565803413,
+      "learning_rate": 0.00012303066647365376,
+      "loss": 1.593,
+      "step": 9622
+    },
+    {
+      "epoch": 0.4433132169346294,
+      "grad_norm": 1.3381030366281441,
+      "learning_rate": 0.00012301614643258477,
+      "loss": 1.3057,
+      "step": 9623
+    },
+    {
+      "epoch": 0.4433592850232644,
+      "grad_norm": 2.7108556775779657,
+      "learning_rate": 0.00012300162587910077,
+      "loss": 1.6013,
+      "step": 9624
+    },
+    {
+      "epoch": 0.4434053531118994,
+      "grad_norm": 4.1855845273226855,
+      "learning_rate": 0.00012298710481352496,
+      "loss": 1.3707,
+      "step": 9625
+    },
+    {
+      "epoch": 0.4434514212005344,
+      "grad_norm": 2.0712922059766545,
+      "learning_rate": 0.0001229725832361807,
+      "loss": 1.1867,
+      "step": 9626
+    },
+    {
+      "epoch": 0.4434974892891694,
+      "grad_norm": 1.7841778332077212,
+      "learning_rate": 0.00012295806114739127,
+      "loss": 1.8202,
+      "step": 9627
+    },
+    {
+      "epoch": 0.4435435573778044,
+      "grad_norm": 2.975499386398513,
+      "learning_rate": 0.00012294353854747995,
+      "loss": 1.122,
+      "step": 9628
+    },
+    {
+      "epoch": 0.4435896254664394,
+      "grad_norm": 2.1718046514143166,
+      "learning_rate": 0.00012292901543677008,
+      "loss": 1.0249,
+      "step": 9629
+    },
+    {
+      "epoch": 0.4436356935550744,
+      "grad_norm": 1.0426950318292059,
+      "learning_rate": 0.00012291449181558504,
+      "loss": 1.6906,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4436817616437094,
+      "grad_norm": 3.47839605369651,
+      "learning_rate": 0.0001228999676842481,
+      "loss": 1.4549,
+      "step": 9631
+    },
+    {
+      "epoch": 0.4437278297323444,
+      "grad_norm": 6.722956260135303,
+      "learning_rate": 0.00012288544304308265,
+      "loss": 1.1864,
+      "step": 9632
+    },
+    {
+      "epoch": 0.4437738978209794,
+      "grad_norm": 1.8808996639976927,
+      "learning_rate": 0.00012287091789241203,
+      "loss": 1.5504,
+      "step": 9633
+    },
+    {
+      "epoch": 0.4438199659096144,
+      "grad_norm": 1.4038140724010733,
+      "learning_rate": 0.00012285639223255968,
+      "loss": 1.2965,
+      "step": 9634
+    },
+    {
+      "epoch": 0.4438660339982494,
+      "grad_norm": 2.0462973046552984,
+      "learning_rate": 0.00012284186606384894,
+      "loss": 1.2227,
+      "step": 9635
+    },
+    {
+      "epoch": 0.4439121020868844,
+      "grad_norm": 1.6521780998790105,
+      "learning_rate": 0.00012282733938660318,
+      "loss": 1.4403,
+      "step": 9636
+    },
+    {
+      "epoch": 0.4439581701755194,
+      "grad_norm": 7.5600329658090155,
+      "learning_rate": 0.0001228128122011459,
+      "loss": 1.3544,
+      "step": 9637
+    },
+    {
+      "epoch": 0.4440042382641544,
+      "grad_norm": 1.7979250981641408,
+      "learning_rate": 0.00012279828450780043,
+      "loss": 1.3084,
+      "step": 9638
+    },
+    {
+      "epoch": 0.4440503063527894,
+      "grad_norm": 1.9924048185174679,
+      "learning_rate": 0.00012278375630689025,
+      "loss": 1.333,
+      "step": 9639
+    },
+    {
+      "epoch": 0.4440963744414244,
+      "grad_norm": 1.7994933746754742,
+      "learning_rate": 0.0001227692275987388,
+      "loss": 1.2735,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4441424425300594,
+      "grad_norm": 1.3677394766949929,
+      "learning_rate": 0.00012275469838366956,
+      "loss": 1.092,
+      "step": 9641
+    },
+    {
+      "epoch": 0.4441885106186944,
+      "grad_norm": 1.8508079566944695,
+      "learning_rate": 0.00012274016866200595,
+      "loss": 1.1749,
+      "step": 9642
+    },
+    {
+      "epoch": 0.4442345787073294,
+      "grad_norm": 2.035137922693421,
+      "learning_rate": 0.0001227256384340715,
+      "loss": 1.2937,
+      "step": 9643
+    },
+    {
+      "epoch": 0.44428064679596446,
+      "grad_norm": 1.119555553913424,
+      "learning_rate": 0.00012271110770018968,
+      "loss": 1.8275,
+      "step": 9644
+    },
+    {
+      "epoch": 0.44432671488459946,
+      "grad_norm": 2.3498560434266387,
+      "learning_rate": 0.00012269657646068395,
+      "loss": 1.3775,
+      "step": 9645
+    },
+    {
+      "epoch": 0.44437278297323446,
+      "grad_norm": 2.0299992011344496,
+      "learning_rate": 0.0001226820447158779,
+      "loss": 1.2985,
+      "step": 9646
+    },
+    {
+      "epoch": 0.44441885106186946,
+      "grad_norm": 1.5347884419330764,
+      "learning_rate": 0.000122667512466095,
+      "loss": 1.399,
+      "step": 9647
+    },
+    {
+      "epoch": 0.44446491915050446,
+      "grad_norm": 2.1527802977661334,
+      "learning_rate": 0.0001226529797116588,
+      "loss": 1.3014,
+      "step": 9648
+    },
+    {
+      "epoch": 0.44451098723913945,
+      "grad_norm": 1.2413089090302452,
+      "learning_rate": 0.00012263844645289286,
+      "loss": 1.6792,
+      "step": 9649
+    },
+    {
+      "epoch": 0.44455705532777445,
+      "grad_norm": 1.2307654440514215,
+      "learning_rate": 0.00012262391269012073,
+      "loss": 1.5673,
+      "step": 9650
+    },
+    {
+      "epoch": 0.44460312341640945,
+      "grad_norm": 1.9496703288845418,
+      "learning_rate": 0.00012260937842366594,
+      "loss": 1.2565,
+      "step": 9651
+    },
+    {
+      "epoch": 0.44464919150504445,
+      "grad_norm": 1.8357651613595356,
+      "learning_rate": 0.00012259484365385214,
+      "loss": 1.3456,
+      "step": 9652
+    },
+    {
+      "epoch": 0.44469525959367945,
+      "grad_norm": 2.1545589465253916,
+      "learning_rate": 0.0001225803083810029,
+      "loss": 1.4569,
+      "step": 9653
+    },
+    {
+      "epoch": 0.44474132768231445,
+      "grad_norm": 1.574508224879516,
+      "learning_rate": 0.00012256577260544177,
+      "loss": 1.4421,
+      "step": 9654
+    },
+    {
+      "epoch": 0.44478739577094945,
+      "grad_norm": 1.5334325408418201,
+      "learning_rate": 0.00012255123632749245,
+      "loss": 1.3798,
+      "step": 9655
+    },
+    {
+      "epoch": 0.44483346385958444,
+      "grad_norm": 1.32599188746642,
+      "learning_rate": 0.00012253669954747848,
+      "loss": 1.5344,
+      "step": 9656
+    },
+    {
+      "epoch": 0.44487953194821944,
+      "grad_norm": 2.2173887284479554,
+      "learning_rate": 0.00012252216226572355,
+      "loss": 1.3093,
+      "step": 9657
+    },
+    {
+      "epoch": 0.4449256000368545,
+      "grad_norm": 1.2961879353114603,
+      "learning_rate": 0.00012250762448255128,
+      "loss": 1.3204,
+      "step": 9658
+    },
+    {
+      "epoch": 0.4449716681254895,
+      "grad_norm": 1.2559278012265906,
+      "learning_rate": 0.0001224930861982854,
+      "loss": 1.0526,
+      "step": 9659
+    },
+    {
+      "epoch": 0.4450177362141245,
+      "grad_norm": 1.2114754968627754,
+      "learning_rate": 0.00012247854741324944,
+      "loss": 1.3334,
+      "step": 9660
+    },
+    {
+      "epoch": 0.4450638043027595,
+      "grad_norm": 1.8939603889508794,
+      "learning_rate": 0.00012246400812776726,
+      "loss": 1.192,
+      "step": 9661
+    },
+    {
+      "epoch": 0.4451098723913945,
+      "grad_norm": 2.088628791109306,
+      "learning_rate": 0.00012244946834216237,
+      "loss": 1.0188,
+      "step": 9662
+    },
+    {
+      "epoch": 0.4451559404800295,
+      "grad_norm": 25.54643278993968,
+      "learning_rate": 0.00012243492805675863,
+      "loss": 1.0641,
+      "step": 9663
+    },
+    {
+      "epoch": 0.4452020085686645,
+      "grad_norm": 1.5252056767961994,
+      "learning_rate": 0.00012242038727187961,
+      "loss": 1.5367,
+      "step": 9664
+    },
+    {
+      "epoch": 0.4452480766572995,
+      "grad_norm": 5.563916242561843,
+      "learning_rate": 0.00012240584598784916,
+      "loss": 1.464,
+      "step": 9665
+    },
+    {
+      "epoch": 0.4452941447459345,
+      "grad_norm": 1.818207392422256,
+      "learning_rate": 0.00012239130420499094,
+      "loss": 1.2234,
+      "step": 9666
+    },
+    {
+      "epoch": 0.4453402128345695,
+      "grad_norm": 2.5495307963691367,
+      "learning_rate": 0.00012237676192362873,
+      "loss": 1.5434,
+      "step": 9667
+    },
+    {
+      "epoch": 0.4453862809232045,
+      "grad_norm": 1.177637225763678,
+      "learning_rate": 0.0001223622191440863,
+      "loss": 1.5739,
+      "step": 9668
+    },
+    {
+      "epoch": 0.4454323490118395,
+      "grad_norm": 3.4324606034369847,
+      "learning_rate": 0.0001223476758666874,
+      "loss": 1.6018,
+      "step": 9669
+    },
+    {
+      "epoch": 0.4454784171004745,
+      "grad_norm": 3.8152217422333785,
+      "learning_rate": 0.00012233313209175582,
+      "loss": 1.5722,
+      "step": 9670
+    },
+    {
+      "epoch": 0.44552448518910953,
+      "grad_norm": 2.022663590023681,
+      "learning_rate": 0.00012231858781961533,
+      "loss": 1.3851,
+      "step": 9671
+    },
+    {
+      "epoch": 0.44557055327774453,
+      "grad_norm": 2.3990431130176266,
+      "learning_rate": 0.00012230404305058979,
+      "loss": 1.4781,
+      "step": 9672
+    },
+    {
+      "epoch": 0.44561662136637953,
+      "grad_norm": 1.5581484247783157,
+      "learning_rate": 0.00012228949778500292,
+      "loss": 1.4406,
+      "step": 9673
+    },
+    {
+      "epoch": 0.4456626894550145,
+      "grad_norm": 1.1032123440248867,
+      "learning_rate": 0.00012227495202317865,
+      "loss": 1.6775,
+      "step": 9674
+    },
+    {
+      "epoch": 0.4457087575436495,
+      "grad_norm": 2.024057774864801,
+      "learning_rate": 0.0001222604057654407,
+      "loss": 1.3351,
+      "step": 9675
+    },
+    {
+      "epoch": 0.4457548256322845,
+      "grad_norm": 1.6767858076064077,
+      "learning_rate": 0.00012224585901211306,
+      "loss": 1.065,
+      "step": 9676
+    },
+    {
+      "epoch": 0.4458008937209195,
+      "grad_norm": 1.508750431812677,
+      "learning_rate": 0.00012223131176351945,
+      "loss": 1.6457,
+      "step": 9677
+    },
+    {
+      "epoch": 0.4458469618095545,
+      "grad_norm": 1.4308980966038218,
+      "learning_rate": 0.00012221676401998383,
+      "loss": 1.4065,
+      "step": 9678
+    },
+    {
+      "epoch": 0.4458930298981895,
+      "grad_norm": 5.099871474337913,
+      "learning_rate": 0.00012220221578183006,
+      "loss": 1.5663,
+      "step": 9679
+    },
+    {
+      "epoch": 0.4459390979868245,
+      "grad_norm": 1.2536286199191002,
+      "learning_rate": 0.000122187667049382,
+      "loss": 1.3952,
+      "step": 9680
+    },
+    {
+      "epoch": 0.4459851660754595,
+      "grad_norm": 1.429324434938138,
+      "learning_rate": 0.00012217311782296362,
+      "loss": 0.9639,
+      "step": 9681
+    },
+    {
+      "epoch": 0.4460312341640945,
+      "grad_norm": 1.5021626145172202,
+      "learning_rate": 0.00012215856810289877,
+      "loss": 1.4422,
+      "step": 9682
+    },
+    {
+      "epoch": 0.4460773022527295,
+      "grad_norm": 1.47535067570772,
+      "learning_rate": 0.00012214401788951138,
+      "loss": 1.3904,
+      "step": 9683
+    },
+    {
+      "epoch": 0.4461233703413645,
+      "grad_norm": 2.430678308172028,
+      "learning_rate": 0.00012212946718312542,
+      "loss": 1.3434,
+      "step": 9684
+    },
+    {
+      "epoch": 0.44616943842999957,
+      "grad_norm": 1.608759728958273,
+      "learning_rate": 0.0001221149159840648,
+      "loss": 1.3631,
+      "step": 9685
+    },
+    {
+      "epoch": 0.44621550651863456,
+      "grad_norm": 1.0482006095083427,
+      "learning_rate": 0.00012210036429265348,
+      "loss": 1.5181,
+      "step": 9686
+    },
+    {
+      "epoch": 0.44626157460726956,
+      "grad_norm": 2.635614949652399,
+      "learning_rate": 0.00012208581210921548,
+      "loss": 1.2895,
+      "step": 9687
+    },
+    {
+      "epoch": 0.44630764269590456,
+      "grad_norm": 1.7684519822734146,
+      "learning_rate": 0.00012207125943407475,
+      "loss": 1.4694,
+      "step": 9688
+    },
+    {
+      "epoch": 0.44635371078453956,
+      "grad_norm": 1.5227361532804955,
+      "learning_rate": 0.00012205670626755523,
+      "loss": 1.3229,
+      "step": 9689
+    },
+    {
+      "epoch": 0.44639977887317456,
+      "grad_norm": 2.321380435006986,
+      "learning_rate": 0.00012204215260998098,
+      "loss": 1.3102,
+      "step": 9690
+    },
+    {
+      "epoch": 0.44644584696180956,
+      "grad_norm": 1.3511815592904182,
+      "learning_rate": 0.000122027598461676,
+      "loss": 1.2666,
+      "step": 9691
+    },
+    {
+      "epoch": 0.44649191505044455,
+      "grad_norm": 0.9851042312503769,
+      "learning_rate": 0.0001220130438229643,
+      "loss": 1.4892,
+      "step": 9692
+    },
+    {
+      "epoch": 0.44653798313907955,
+      "grad_norm": 1.3988238364969965,
+      "learning_rate": 0.00012199848869416992,
+      "loss": 1.3356,
+      "step": 9693
+    },
+    {
+      "epoch": 0.44658405122771455,
+      "grad_norm": 1.5537030769376445,
+      "learning_rate": 0.00012198393307561695,
+      "loss": 1.3364,
+      "step": 9694
+    },
+    {
+      "epoch": 0.44663011931634955,
+      "grad_norm": 5.061908252671455,
+      "learning_rate": 0.00012196937696762935,
+      "loss": 1.1993,
+      "step": 9695
+    },
+    {
+      "epoch": 0.44667618740498455,
+      "grad_norm": 5.311317544625134,
+      "learning_rate": 0.00012195482037053128,
+      "loss": 1.6274,
+      "step": 9696
+    },
+    {
+      "epoch": 0.44672225549361955,
+      "grad_norm": 2.1874399501328345,
+      "learning_rate": 0.00012194026328464676,
+      "loss": 0.9673,
+      "step": 9697
+    },
+    {
+      "epoch": 0.44676832358225455,
+      "grad_norm": 2.4049592623829863,
+      "learning_rate": 0.00012192570571029991,
+      "loss": 1.504,
+      "step": 9698
+    },
+    {
+      "epoch": 0.4468143916708896,
+      "grad_norm": 0.9532871335753846,
+      "learning_rate": 0.00012191114764781481,
+      "loss": 1.2761,
+      "step": 9699
+    },
+    {
+      "epoch": 0.4468604597595246,
+      "grad_norm": 4.1480007251201805,
+      "learning_rate": 0.00012189658909751557,
+      "loss": 1.5731,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4469065278481596,
+      "grad_norm": 1.2662943401743325,
+      "learning_rate": 0.00012188203005972634,
+      "loss": 1.332,
+      "step": 9701
+    },
+    {
+      "epoch": 0.4469525959367946,
+      "grad_norm": 1.8891604969365692,
+      "learning_rate": 0.00012186747053477118,
+      "loss": 1.4046,
+      "step": 9702
+    },
+    {
+      "epoch": 0.4469986640254296,
+      "grad_norm": 5.1959044479107375,
+      "learning_rate": 0.00012185291052297434,
+      "loss": 1.3187,
+      "step": 9703
+    },
+    {
+      "epoch": 0.4470447321140646,
+      "grad_norm": 1.8210467365707075,
+      "learning_rate": 0.00012183835002465989,
+      "loss": 1.1139,
+      "step": 9704
+    },
+    {
+      "epoch": 0.4470908002026996,
+      "grad_norm": 1.0406419934243671,
+      "learning_rate": 0.00012182378904015204,
+      "loss": 1.5119,
+      "step": 9705
+    },
+    {
+      "epoch": 0.4471368682913346,
+      "grad_norm": 1.5256817642251876,
+      "learning_rate": 0.00012180922756977493,
+      "loss": 1.2503,
+      "step": 9706
+    },
+    {
+      "epoch": 0.4471829363799696,
+      "grad_norm": 1.4829583966061375,
+      "learning_rate": 0.00012179466561385278,
+      "loss": 1.5494,
+      "step": 9707
+    },
+    {
+      "epoch": 0.4472290044686046,
+      "grad_norm": 1.3983962791675038,
+      "learning_rate": 0.00012178010317270976,
+      "loss": 1.2394,
+      "step": 9708
+    },
+    {
+      "epoch": 0.4472750725572396,
+      "grad_norm": 1.1959287064638182,
+      "learning_rate": 0.00012176554024667006,
+      "loss": 1.5133,
+      "step": 9709
+    },
+    {
+      "epoch": 0.4473211406458746,
+      "grad_norm": 1.725097618198362,
+      "learning_rate": 0.00012175097683605796,
+      "loss": 1.5492,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4473672087345096,
+      "grad_norm": 1.4572424646828097,
+      "learning_rate": 0.00012173641294119764,
+      "loss": 1.3283,
+      "step": 9711
+    },
+    {
+      "epoch": 0.44741327682314463,
+      "grad_norm": 1.5901638558712614,
+      "learning_rate": 0.00012172184856241338,
+      "loss": 1.3773,
+      "step": 9712
+    },
+    {
+      "epoch": 0.44745934491177963,
+      "grad_norm": 1.4649703257145472,
+      "learning_rate": 0.00012170728370002938,
+      "loss": 1.1506,
+      "step": 9713
+    },
+    {
+      "epoch": 0.44750541300041463,
+      "grad_norm": 1.7998648386437635,
+      "learning_rate": 0.00012169271835436995,
+      "loss": 1.3382,
+      "step": 9714
+    },
+    {
+      "epoch": 0.44755148108904963,
+      "grad_norm": 0.8732451044664777,
+      "learning_rate": 0.00012167815252575931,
+      "loss": 1.4577,
+      "step": 9715
+    },
+    {
+      "epoch": 0.44759754917768463,
+      "grad_norm": 1.0314461261189283,
+      "learning_rate": 0.00012166358621452183,
+      "loss": 1.3874,
+      "step": 9716
+    },
+    {
+      "epoch": 0.4476436172663196,
+      "grad_norm": 1.1162454734494345,
+      "learning_rate": 0.00012164901942098168,
+      "loss": 1.4457,
+      "step": 9717
+    },
+    {
+      "epoch": 0.4476896853549546,
+      "grad_norm": 2.9483432905078013,
+      "learning_rate": 0.00012163445214546328,
+      "loss": 1.2973,
+      "step": 9718
+    },
+    {
+      "epoch": 0.4477357534435896,
+      "grad_norm": 1.4287828951579382,
+      "learning_rate": 0.00012161988438829089,
+      "loss": 1.2861,
+      "step": 9719
+    },
+    {
+      "epoch": 0.4477818215322246,
+      "grad_norm": 1.2856419174572078,
+      "learning_rate": 0.0001216053161497888,
+      "loss": 1.4365,
+      "step": 9720
+    },
+    {
+      "epoch": 0.4478278896208596,
+      "grad_norm": 1.4213518354086794,
+      "learning_rate": 0.00012159074743028145,
+      "loss": 1.3151,
+      "step": 9721
+    },
+    {
+      "epoch": 0.4478739577094946,
+      "grad_norm": 1.7861236213776257,
+      "learning_rate": 0.00012157617823009308,
+      "loss": 1.23,
+      "step": 9722
+    },
+    {
+      "epoch": 0.4479200257981296,
+      "grad_norm": 1.3185510065166515,
+      "learning_rate": 0.00012156160854954814,
+      "loss": 1.6187,
+      "step": 9723
+    },
+    {
+      "epoch": 0.4479660938867646,
+      "grad_norm": 1.5872562748134365,
+      "learning_rate": 0.00012154703838897092,
+      "loss": 1.3571,
+      "step": 9724
+    },
+    {
+      "epoch": 0.4480121619753996,
+      "grad_norm": 0.9549228141163612,
+      "learning_rate": 0.0001215324677486859,
+      "loss": 1.5112,
+      "step": 9725
+    },
+    {
+      "epoch": 0.44805823006403467,
+      "grad_norm": 0.8582782098021872,
+      "learning_rate": 0.00012151789662901734,
+      "loss": 1.3112,
+      "step": 9726
+    },
+    {
+      "epoch": 0.44810429815266967,
+      "grad_norm": 1.4789172278959615,
+      "learning_rate": 0.00012150332503028974,
+      "loss": 1.2959,
+      "step": 9727
+    },
+    {
+      "epoch": 0.44815036624130467,
+      "grad_norm": 1.5511573112737933,
+      "learning_rate": 0.00012148875295282745,
+      "loss": 1.4479,
+      "step": 9728
+    },
+    {
+      "epoch": 0.44819643432993966,
+      "grad_norm": 2.8375265133047667,
+      "learning_rate": 0.00012147418039695492,
+      "loss": 1.1839,
+      "step": 9729
+    },
+    {
+      "epoch": 0.44824250241857466,
+      "grad_norm": 2.7523298227516433,
+      "learning_rate": 0.0001214596073629966,
+      "loss": 1.396,
+      "step": 9730
+    },
+    {
+      "epoch": 0.44828857050720966,
+      "grad_norm": 1.4904381419380628,
+      "learning_rate": 0.00012144503385127693,
+      "loss": 1.4823,
+      "step": 9731
+    },
+    {
+      "epoch": 0.44833463859584466,
+      "grad_norm": 1.6478975132957538,
+      "learning_rate": 0.00012143045986212034,
+      "loss": 1.4691,
+      "step": 9732
+    },
+    {
+      "epoch": 0.44838070668447966,
+      "grad_norm": 1.5163707729565783,
+      "learning_rate": 0.00012141588539585131,
+      "loss": 1.4297,
+      "step": 9733
+    },
+    {
+      "epoch": 0.44842677477311466,
+      "grad_norm": 1.8568705135928085,
+      "learning_rate": 0.0001214013104527943,
+      "loss": 1.0946,
+      "step": 9734
+    },
+    {
+      "epoch": 0.44847284286174965,
+      "grad_norm": 1.184011314746724,
+      "learning_rate": 0.00012138673503327382,
+      "loss": 1.4019,
+      "step": 9735
+    },
+    {
+      "epoch": 0.44851891095038465,
+      "grad_norm": 1.651920920964758,
+      "learning_rate": 0.00012137215913761434,
+      "loss": 1.5933,
+      "step": 9736
+    },
+    {
+      "epoch": 0.44856497903901965,
+      "grad_norm": 1.4177900338618763,
+      "learning_rate": 0.0001213575827661404,
+      "loss": 1.3302,
+      "step": 9737
+    },
+    {
+      "epoch": 0.44861104712765465,
+      "grad_norm": 1.0632866292790757,
+      "learning_rate": 0.00012134300591917649,
+      "loss": 1.3165,
+      "step": 9738
+    },
+    {
+      "epoch": 0.4486571152162897,
+      "grad_norm": 1.0189895830381055,
+      "learning_rate": 0.00012132842859704714,
+      "loss": 1.1665,
+      "step": 9739
+    },
+    {
+      "epoch": 0.4487031833049247,
+      "grad_norm": 1.619909126157063,
+      "learning_rate": 0.00012131385080007694,
+      "loss": 1.2899,
+      "step": 9740
+    },
+    {
+      "epoch": 0.4487492513935597,
+      "grad_norm": 3.765526484594507,
+      "learning_rate": 0.00012129927252859034,
+      "loss": 1.6805,
+      "step": 9741
+    },
+    {
+      "epoch": 0.4487953194821947,
+      "grad_norm": 1.328914667128178,
+      "learning_rate": 0.00012128469378291198,
+      "loss": 1.3496,
+      "step": 9742
+    },
+    {
+      "epoch": 0.4488413875708297,
+      "grad_norm": 0.9659268211642872,
+      "learning_rate": 0.00012127011456336644,
+      "loss": 0.9096,
+      "step": 9743
+    },
+    {
+      "epoch": 0.4488874556594647,
+      "grad_norm": 2.7386957539712946,
+      "learning_rate": 0.00012125553487027823,
+      "loss": 1.0986,
+      "step": 9744
+    },
+    {
+      "epoch": 0.4489335237480997,
+      "grad_norm": 1.3896613066382466,
+      "learning_rate": 0.00012124095470397199,
+      "loss": 1.5676,
+      "step": 9745
+    },
+    {
+      "epoch": 0.4489795918367347,
+      "grad_norm": 1.200788983464746,
+      "learning_rate": 0.0001212263740647723,
+      "loss": 1.6275,
+      "step": 9746
+    },
+    {
+      "epoch": 0.4490256599253697,
+      "grad_norm": 1.3592548543397087,
+      "learning_rate": 0.00012121179295300383,
+      "loss": 1.2787,
+      "step": 9747
+    },
+    {
+      "epoch": 0.4490717280140047,
+      "grad_norm": 1.2416739145265854,
+      "learning_rate": 0.00012119721136899112,
+      "loss": 1.1794,
+      "step": 9748
+    },
+    {
+      "epoch": 0.4491177961026397,
+      "grad_norm": 1.7214089373380685,
+      "learning_rate": 0.00012118262931305886,
+      "loss": 1.4974,
+      "step": 9749
+    },
+    {
+      "epoch": 0.4491638641912747,
+      "grad_norm": 1.7353589590313008,
+      "learning_rate": 0.00012116804678553167,
+      "loss": 1.1942,
+      "step": 9750
+    },
+    {
+      "epoch": 0.4492099322799097,
+      "grad_norm": 1.3378225539378361,
+      "learning_rate": 0.00012115346378673425,
+      "loss": 1.4909,
+      "step": 9751
+    },
+    {
+      "epoch": 0.4492560003685447,
+      "grad_norm": 1.2747980629495703,
+      "learning_rate": 0.00012113888031699117,
+      "loss": 1.2678,
+      "step": 9752
+    },
+    {
+      "epoch": 0.44930206845717974,
+      "grad_norm": 1.609702146826006,
+      "learning_rate": 0.00012112429637662721,
+      "loss": 1.3201,
+      "step": 9753
+    },
+    {
+      "epoch": 0.44934813654581474,
+      "grad_norm": 1.046469538987009,
+      "learning_rate": 0.00012110971196596701,
+      "loss": 1.3237,
+      "step": 9754
+    },
+    {
+      "epoch": 0.44939420463444973,
+      "grad_norm": 1.6403727470065403,
+      "learning_rate": 0.00012109512708533523,
+      "loss": 1.2986,
+      "step": 9755
+    },
+    {
+      "epoch": 0.44944027272308473,
+      "grad_norm": 1.257684790461839,
+      "learning_rate": 0.00012108054173505665,
+      "loss": 1.6008,
+      "step": 9756
+    },
+    {
+      "epoch": 0.44948634081171973,
+      "grad_norm": 1.251460973493377,
+      "learning_rate": 0.00012106595591545593,
+      "loss": 1.4466,
+      "step": 9757
+    },
+    {
+      "epoch": 0.44953240890035473,
+      "grad_norm": 1.8042309283709081,
+      "learning_rate": 0.00012105136962685785,
+      "loss": 1.5251,
+      "step": 9758
+    },
+    {
+      "epoch": 0.44957847698898973,
+      "grad_norm": 3.7989360464079174,
+      "learning_rate": 0.00012103678286958711,
+      "loss": 1.4819,
+      "step": 9759
+    },
+    {
+      "epoch": 0.4496245450776247,
+      "grad_norm": 1.6824887088249756,
+      "learning_rate": 0.00012102219564396845,
+      "loss": 1.3124,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4496706131662597,
+      "grad_norm": 0.8803253710468903,
+      "learning_rate": 0.00012100760795032668,
+      "loss": 1.2398,
+      "step": 9761
+    },
+    {
+      "epoch": 0.4497166812548947,
+      "grad_norm": 1.6929456561676304,
+      "learning_rate": 0.00012099301978898656,
+      "loss": 1.3314,
+      "step": 9762
+    },
+    {
+      "epoch": 0.4497627493435297,
+      "grad_norm": 1.5038448204566264,
+      "learning_rate": 0.00012097843116027281,
+      "loss": 1.3703,
+      "step": 9763
+    },
+    {
+      "epoch": 0.4498088174321647,
+      "grad_norm": 2.60044059586645,
+      "learning_rate": 0.00012096384206451023,
+      "loss": 1.3014,
+      "step": 9764
+    },
+    {
+      "epoch": 0.4498548855207997,
+      "grad_norm": 0.9536767510178652,
+      "learning_rate": 0.00012094925250202371,
+      "loss": 1.4009,
+      "step": 9765
+    },
+    {
+      "epoch": 0.4499009536094347,
+      "grad_norm": 1.6265014528636794,
+      "learning_rate": 0.00012093466247313796,
+      "loss": 1.4533,
+      "step": 9766
+    },
+    {
+      "epoch": 0.44994702169806977,
+      "grad_norm": 1.9186336040597411,
+      "learning_rate": 0.00012092007197817787,
+      "loss": 1.4632,
+      "step": 9767
+    },
+    {
+      "epoch": 0.44999308978670477,
+      "grad_norm": 2.724408148439312,
+      "learning_rate": 0.00012090548101746826,
+      "loss": 1.2555,
+      "step": 9768
+    },
+    {
+      "epoch": 0.45003915787533977,
+      "grad_norm": 3.272280942891503,
+      "learning_rate": 0.00012089088959133395,
+      "loss": 1.4545,
+      "step": 9769
+    },
+    {
+      "epoch": 0.45008522596397477,
+      "grad_norm": 1.7878630825270618,
+      "learning_rate": 0.00012087629770009979,
+      "loss": 1.4863,
+      "step": 9770
+    },
+    {
+      "epoch": 0.45013129405260977,
+      "grad_norm": 2.7004148108329766,
+      "learning_rate": 0.00012086170534409069,
+      "loss": 1.5003,
+      "step": 9771
+    },
+    {
+      "epoch": 0.45017736214124476,
+      "grad_norm": 1.9212179842766184,
+      "learning_rate": 0.00012084711252363146,
+      "loss": 1.3645,
+      "step": 9772
+    },
+    {
+      "epoch": 0.45022343022987976,
+      "grad_norm": 1.5821367552268342,
+      "learning_rate": 0.00012083251923904702,
+      "loss": 1.4932,
+      "step": 9773
+    },
+    {
+      "epoch": 0.45026949831851476,
+      "grad_norm": 1.5864565397325223,
+      "learning_rate": 0.00012081792549066225,
+      "loss": 1.2153,
+      "step": 9774
+    },
+    {
+      "epoch": 0.45031556640714976,
+      "grad_norm": 1.5521130448951044,
+      "learning_rate": 0.00012080333127880208,
+      "loss": 1.3981,
+      "step": 9775
+    },
+    {
+      "epoch": 0.45036163449578476,
+      "grad_norm": 3.187696050394006,
+      "learning_rate": 0.0001207887366037914,
+      "loss": 1.4348,
+      "step": 9776
+    },
+    {
+      "epoch": 0.45040770258441976,
+      "grad_norm": 1.0045309889283511,
+      "learning_rate": 0.00012077414146595511,
+      "loss": 1.8092,
+      "step": 9777
+    },
+    {
+      "epoch": 0.45045377067305475,
+      "grad_norm": 1.1612416192139854,
+      "learning_rate": 0.00012075954586561824,
+      "loss": 1.4171,
+      "step": 9778
+    },
+    {
+      "epoch": 0.45049983876168975,
+      "grad_norm": 2.588543700985898,
+      "learning_rate": 0.00012074494980310564,
+      "loss": 1.5584,
+      "step": 9779
+    },
+    {
+      "epoch": 0.4505459068503248,
+      "grad_norm": 0.9006666270853829,
+      "learning_rate": 0.00012073035327874236,
+      "loss": 1.3743,
+      "step": 9780
+    },
+    {
+      "epoch": 0.4505919749389598,
+      "grad_norm": 0.9798387170495992,
+      "learning_rate": 0.00012071575629285325,
+      "loss": 1.4083,
+      "step": 9781
+    },
+    {
+      "epoch": 0.4506380430275948,
+      "grad_norm": 1.930014259472254,
+      "learning_rate": 0.00012070115884576336,
+      "loss": 1.3825,
+      "step": 9782
+    },
+    {
+      "epoch": 0.4506841111162298,
+      "grad_norm": 1.5259057078269118,
+      "learning_rate": 0.00012068656093779767,
+      "loss": 1.4153,
+      "step": 9783
+    },
+    {
+      "epoch": 0.4507301792048648,
+      "grad_norm": 0.9878956430335759,
+      "learning_rate": 0.00012067196256928117,
+      "loss": 1.4013,
+      "step": 9784
+    },
+    {
+      "epoch": 0.4507762472934998,
+      "grad_norm": 1.2602623253207137,
+      "learning_rate": 0.00012065736374053885,
+      "loss": 1.3279,
+      "step": 9785
+    },
+    {
+      "epoch": 0.4508223153821348,
+      "grad_norm": 9.60657703856958,
+      "learning_rate": 0.00012064276445189578,
+      "loss": 1.4149,
+      "step": 9786
+    },
+    {
+      "epoch": 0.4508683834707698,
+      "grad_norm": 3.8223042067539437,
+      "learning_rate": 0.00012062816470367691,
+      "loss": 1.6037,
+      "step": 9787
+    },
+    {
+      "epoch": 0.4509144515594048,
+      "grad_norm": 1.1330078415643465,
+      "learning_rate": 0.00012061356449620735,
+      "loss": 0.9982,
+      "step": 9788
+    },
+    {
+      "epoch": 0.4509605196480398,
+      "grad_norm": 1.6313652196288846,
+      "learning_rate": 0.00012059896382981213,
+      "loss": 1.2661,
+      "step": 9789
+    },
+    {
+      "epoch": 0.4510065877366748,
+      "grad_norm": 3.150222502903921,
+      "learning_rate": 0.00012058436270481628,
+      "loss": 1.5473,
+      "step": 9790
+    },
+    {
+      "epoch": 0.4510526558253098,
+      "grad_norm": 1.530399404379329,
+      "learning_rate": 0.0001205697611215449,
+      "loss": 1.4872,
+      "step": 9791
+    },
+    {
+      "epoch": 0.4510987239139448,
+      "grad_norm": 1.2343111242458242,
+      "learning_rate": 0.00012055515908032304,
+      "loss": 1.3585,
+      "step": 9792
+    },
+    {
+      "epoch": 0.4511447920025798,
+      "grad_norm": 2.109709266258926,
+      "learning_rate": 0.00012054055658147585,
+      "loss": 1.0689,
+      "step": 9793
+    },
+    {
+      "epoch": 0.45119086009121484,
+      "grad_norm": 1.5813191274660203,
+      "learning_rate": 0.00012052595362532835,
+      "loss": 1.4667,
+      "step": 9794
+    },
+    {
+      "epoch": 0.45123692817984984,
+      "grad_norm": 1.0307273533991668,
+      "learning_rate": 0.0001205113502122057,
+      "loss": 1.2777,
+      "step": 9795
+    },
+    {
+      "epoch": 0.45128299626848484,
+      "grad_norm": 1.0597633895708582,
+      "learning_rate": 0.000120496746342433,
+      "loss": 1.4868,
+      "step": 9796
+    },
+    {
+      "epoch": 0.45132906435711984,
+      "grad_norm": 3.7359522012624886,
+      "learning_rate": 0.00012048214201633538,
+      "loss": 1.0228,
+      "step": 9797
+    },
+    {
+      "epoch": 0.45137513244575483,
+      "grad_norm": 1.817723541689461,
+      "learning_rate": 0.00012046753723423801,
+      "loss": 1.4398,
+      "step": 9798
+    },
+    {
+      "epoch": 0.45142120053438983,
+      "grad_norm": 21.177841918197874,
+      "learning_rate": 0.00012045293199646598,
+      "loss": 1.1534,
+      "step": 9799
+    },
+    {
+      "epoch": 0.45146726862302483,
+      "grad_norm": 2.5835081376143916,
+      "learning_rate": 0.00012043832630334455,
+      "loss": 1.294,
+      "step": 9800
+    },
+    {
+      "epoch": 0.45151333671165983,
+      "grad_norm": 1.1004419633675715,
+      "learning_rate": 0.00012042372015519879,
+      "loss": 1.4419,
+      "step": 9801
+    },
+    {
+      "epoch": 0.45155940480029483,
+      "grad_norm": 1.3168025554361034,
+      "learning_rate": 0.00012040911355235391,
+      "loss": 1.503,
+      "step": 9802
+    },
+    {
+      "epoch": 0.4516054728889298,
+      "grad_norm": 7.004528462222024,
+      "learning_rate": 0.00012039450649513511,
+      "loss": 1.4087,
+      "step": 9803
+    },
+    {
+      "epoch": 0.4516515409775648,
+      "grad_norm": 0.9065815314380821,
+      "learning_rate": 0.00012037989898386763,
+      "loss": 1.6222,
+      "step": 9804
+    },
+    {
+      "epoch": 0.4516976090661998,
+      "grad_norm": 1.297768161331163,
+      "learning_rate": 0.00012036529101887662,
+      "loss": 1.4852,
+      "step": 9805
+    },
+    {
+      "epoch": 0.4517436771548348,
+      "grad_norm": 1.5986788792568176,
+      "learning_rate": 0.00012035068260048732,
+      "loss": 1.4373,
+      "step": 9806
+    },
+    {
+      "epoch": 0.4517897452434699,
+      "grad_norm": 0.9956228901791592,
+      "learning_rate": 0.00012033607372902498,
+      "loss": 1.3726,
+      "step": 9807
+    },
+    {
+      "epoch": 0.4518358133321049,
+      "grad_norm": 1.4227751886340816,
+      "learning_rate": 0.0001203214644048148,
+      "loss": 1.382,
+      "step": 9808
+    },
+    {
+      "epoch": 0.4518818814207399,
+      "grad_norm": 1.0914178503575627,
+      "learning_rate": 0.00012030685462818208,
+      "loss": 1.281,
+      "step": 9809
+    },
+    {
+      "epoch": 0.45192794950937487,
+      "grad_norm": 0.927165012686802,
+      "learning_rate": 0.00012029224439945203,
+      "loss": 1.0112,
+      "step": 9810
+    },
+    {
+      "epoch": 0.45197401759800987,
+      "grad_norm": 1.8366832373634518,
+      "learning_rate": 0.00012027763371895,
+      "loss": 1.4142,
+      "step": 9811
+    },
+    {
+      "epoch": 0.45202008568664487,
+      "grad_norm": 0.9151114937462257,
+      "learning_rate": 0.0001202630225870012,
+      "loss": 1.2912,
+      "step": 9812
+    },
+    {
+      "epoch": 0.45206615377527987,
+      "grad_norm": 1.5029304259760525,
+      "learning_rate": 0.00012024841100393097,
+      "loss": 1.3796,
+      "step": 9813
+    },
+    {
+      "epoch": 0.45211222186391486,
+      "grad_norm": 2.1276051069851483,
+      "learning_rate": 0.00012023379897006455,
+      "loss": 1.4959,
+      "step": 9814
+    },
+    {
+      "epoch": 0.45215828995254986,
+      "grad_norm": 1.1823761016720764,
+      "learning_rate": 0.0001202191864857273,
+      "loss": 1.4397,
+      "step": 9815
+    },
+    {
+      "epoch": 0.45220435804118486,
+      "grad_norm": 1.0214515907189963,
+      "learning_rate": 0.00012020457355124452,
+      "loss": 1.4293,
+      "step": 9816
+    },
+    {
+      "epoch": 0.45225042612981986,
+      "grad_norm": 1.6906190249006967,
+      "learning_rate": 0.00012018996016694161,
+      "loss": 1.3624,
+      "step": 9817
+    },
+    {
+      "epoch": 0.45229649421845486,
+      "grad_norm": 1.5044204326674262,
+      "learning_rate": 0.0001201753463331438,
+      "loss": 1.5116,
+      "step": 9818
+    },
+    {
+      "epoch": 0.45234256230708986,
+      "grad_norm": 0.8376211495341808,
+      "learning_rate": 0.0001201607320501765,
+      "loss": 1.6156,
+      "step": 9819
+    },
+    {
+      "epoch": 0.45238863039572486,
+      "grad_norm": 1.526054079779163,
+      "learning_rate": 0.00012014611731836509,
+      "loss": 1.2485,
+      "step": 9820
+    },
+    {
+      "epoch": 0.4524346984843599,
+      "grad_norm": 1.4168043323983344,
+      "learning_rate": 0.00012013150213803491,
+      "loss": 1.2648,
+      "step": 9821
+    },
+    {
+      "epoch": 0.4524807665729949,
+      "grad_norm": 1.0098262973170276,
+      "learning_rate": 0.00012011688650951136,
+      "loss": 1.5493,
+      "step": 9822
+    },
+    {
+      "epoch": 0.4525268346616299,
+      "grad_norm": 1.1330870483740132,
+      "learning_rate": 0.00012010227043311981,
+      "loss": 1.5688,
+      "step": 9823
+    },
+    {
+      "epoch": 0.4525729027502649,
+      "grad_norm": 2.012764310036743,
+      "learning_rate": 0.0001200876539091857,
+      "loss": 1.4049,
+      "step": 9824
+    },
+    {
+      "epoch": 0.4526189708388999,
+      "grad_norm": 1.0909290022471525,
+      "learning_rate": 0.00012007303693803438,
+      "loss": 1.235,
+      "step": 9825
+    },
+    {
+      "epoch": 0.4526650389275349,
+      "grad_norm": 1.5349750778299536,
+      "learning_rate": 0.00012005841951999137,
+      "loss": 1.3346,
+      "step": 9826
+    },
+    {
+      "epoch": 0.4527111070161699,
+      "grad_norm": 2.3438598158123383,
+      "learning_rate": 0.000120043801655382,
+      "loss": 1.4096,
+      "step": 9827
+    },
+    {
+      "epoch": 0.4527571751048049,
+      "grad_norm": 1.0643682245925845,
+      "learning_rate": 0.00012002918334453178,
+      "loss": 1.2671,
+      "step": 9828
+    },
+    {
+      "epoch": 0.4528032431934399,
+      "grad_norm": 2.2645394604235762,
+      "learning_rate": 0.0001200145645877661,
+      "loss": 1.2091,
+      "step": 9829
+    },
+    {
+      "epoch": 0.4528493112820749,
+      "grad_norm": 1.2108045399490341,
+      "learning_rate": 0.00011999994538541049,
+      "loss": 1.2587,
+      "step": 9830
+    },
+    {
+      "epoch": 0.4528953793707099,
+      "grad_norm": 1.0868236014230028,
+      "learning_rate": 0.00011998532573779035,
+      "loss": 1.538,
+      "step": 9831
+    },
+    {
+      "epoch": 0.4529414474593449,
+      "grad_norm": 1.3663563754726988,
+      "learning_rate": 0.00011997070564523123,
+      "loss": 1.5157,
+      "step": 9832
+    },
+    {
+      "epoch": 0.4529875155479799,
+      "grad_norm": 1.6564473691467467,
+      "learning_rate": 0.00011995608510805858,
+      "loss": 1.3513,
+      "step": 9833
+    },
+    {
+      "epoch": 0.4530335836366149,
+      "grad_norm": 1.9705334139775446,
+      "learning_rate": 0.00011994146412659791,
+      "loss": 1.4599,
+      "step": 9834
+    },
+    {
+      "epoch": 0.45307965172524994,
+      "grad_norm": 2.306235898135934,
+      "learning_rate": 0.00011992684270117478,
+      "loss": 1.357,
+      "step": 9835
+    },
+    {
+      "epoch": 0.45312571981388494,
+      "grad_norm": 1.9843239599829179,
+      "learning_rate": 0.0001199122208321146,
+      "loss": 1.109,
+      "step": 9836
+    },
+    {
+      "epoch": 0.45317178790251994,
+      "grad_norm": 1.0731111709593562,
+      "learning_rate": 0.000119897598519743,
+      "loss": 1.5005,
+      "step": 9837
+    },
+    {
+      "epoch": 0.45321785599115494,
+      "grad_norm": 1.7373061902530613,
+      "learning_rate": 0.00011988297576438547,
+      "loss": 1.1241,
+      "step": 9838
+    },
+    {
+      "epoch": 0.45326392407978994,
+      "grad_norm": 1.2102789055495995,
+      "learning_rate": 0.00011986835256636759,
+      "loss": 1.5973,
+      "step": 9839
+    },
+    {
+      "epoch": 0.45330999216842494,
+      "grad_norm": 1.7233221060683501,
+      "learning_rate": 0.00011985372892601489,
+      "loss": 1.5393,
+      "step": 9840
+    },
+    {
+      "epoch": 0.45335606025705993,
+      "grad_norm": 5.796093693640676,
+      "learning_rate": 0.00011983910484365297,
+      "loss": 1.3715,
+      "step": 9841
+    },
+    {
+      "epoch": 0.45340212834569493,
+      "grad_norm": 2.936827634711125,
+      "learning_rate": 0.00011982448031960738,
+      "loss": 1.1574,
+      "step": 9842
+    },
+    {
+      "epoch": 0.45344819643432993,
+      "grad_norm": 1.215328694829789,
+      "learning_rate": 0.00011980985535420374,
+      "loss": 1.1102,
+      "step": 9843
+    },
+    {
+      "epoch": 0.45349426452296493,
+      "grad_norm": 2.34305019357263,
+      "learning_rate": 0.00011979522994776766,
+      "loss": 1.0675,
+      "step": 9844
+    },
+    {
+      "epoch": 0.45354033261159993,
+      "grad_norm": 2.6230306498412794,
+      "learning_rate": 0.0001197806041006247,
+      "loss": 1.3187,
+      "step": 9845
+    },
+    {
+      "epoch": 0.4535864007002349,
+      "grad_norm": 2.2907530569654213,
+      "learning_rate": 0.00011976597781310052,
+      "loss": 1.0833,
+      "step": 9846
+    },
+    {
+      "epoch": 0.4536324687888699,
+      "grad_norm": 12.161316950585777,
+      "learning_rate": 0.00011975135108552071,
+      "loss": 1.4187,
+      "step": 9847
+    },
+    {
+      "epoch": 0.453678536877505,
+      "grad_norm": 1.7281087653288743,
+      "learning_rate": 0.00011973672391821097,
+      "loss": 1.2352,
+      "step": 9848
+    },
+    {
+      "epoch": 0.45372460496614,
+      "grad_norm": 1.776549386204755,
+      "learning_rate": 0.0001197220963114969,
+      "loss": 1.3048,
+      "step": 9849
+    },
+    {
+      "epoch": 0.453770673054775,
+      "grad_norm": 2.014743047963768,
+      "learning_rate": 0.00011970746826570416,
+      "loss": 1.1657,
+      "step": 9850
+    },
+    {
+      "epoch": 0.45381674114341,
+      "grad_norm": 3.3905159203654187,
+      "learning_rate": 0.00011969283978115844,
+      "loss": 1.3112,
+      "step": 9851
+    },
+    {
+      "epoch": 0.453862809232045,
+      "grad_norm": 3.0467333088662474,
+      "learning_rate": 0.0001196782108581854,
+      "loss": 1.4233,
+      "step": 9852
+    },
+    {
+      "epoch": 0.45390887732067997,
+      "grad_norm": 2.4287741166519083,
+      "learning_rate": 0.00011966358149711077,
+      "loss": 1.3514,
+      "step": 9853
+    },
+    {
+      "epoch": 0.45395494540931497,
+      "grad_norm": 1.6165335699660195,
+      "learning_rate": 0.00011964895169826016,
+      "loss": 1.2082,
+      "step": 9854
+    },
+    {
+      "epoch": 0.45400101349794997,
+      "grad_norm": 1.3885404203254519,
+      "learning_rate": 0.00011963432146195936,
+      "loss": 1.2351,
+      "step": 9855
+    },
+    {
+      "epoch": 0.45404708158658497,
+      "grad_norm": 2.5971298446668745,
+      "learning_rate": 0.00011961969078853403,
+      "loss": 1.0555,
+      "step": 9856
+    },
+    {
+      "epoch": 0.45409314967521996,
+      "grad_norm": 3.862661110146108,
+      "learning_rate": 0.00011960505967830997,
+      "loss": 1.3973,
+      "step": 9857
+    },
+    {
+      "epoch": 0.45413921776385496,
+      "grad_norm": 1.67584282490468,
+      "learning_rate": 0.00011959042813161286,
+      "loss": 1.451,
+      "step": 9858
+    },
+    {
+      "epoch": 0.45418528585248996,
+      "grad_norm": 1.385155579058689,
+      "learning_rate": 0.00011957579614876843,
+      "loss": 1.2926,
+      "step": 9859
+    },
+    {
+      "epoch": 0.45423135394112496,
+      "grad_norm": 1.8092049950949287,
+      "learning_rate": 0.00011956116373010249,
+      "loss": 1.1624,
+      "step": 9860
+    },
+    {
+      "epoch": 0.45427742202975996,
+      "grad_norm": 4.924530114492611,
+      "learning_rate": 0.00011954653087594079,
+      "loss": 1.2139,
+      "step": 9861
+    },
+    {
+      "epoch": 0.454323490118395,
+      "grad_norm": 1.872167730396831,
+      "learning_rate": 0.00011953189758660907,
+      "loss": 1.4035,
+      "step": 9862
+    },
+    {
+      "epoch": 0.45436955820703,
+      "grad_norm": 0.9582154264989596,
+      "learning_rate": 0.00011951726386243314,
+      "loss": 1.2073,
+      "step": 9863
+    },
+    {
+      "epoch": 0.454415626295665,
+      "grad_norm": 2.038902953367662,
+      "learning_rate": 0.00011950262970373883,
+      "loss": 1.4327,
+      "step": 9864
+    },
+    {
+      "epoch": 0.4544616943843,
+      "grad_norm": 1.4372599229910645,
+      "learning_rate": 0.00011948799511085185,
+      "loss": 1.6151,
+      "step": 9865
+    },
+    {
+      "epoch": 0.454507762472935,
+      "grad_norm": 1.467872003439764,
+      "learning_rate": 0.00011947336008409813,
+      "loss": 1.5044,
+      "step": 9866
+    },
+    {
+      "epoch": 0.45455383056157,
+      "grad_norm": 1.3743526282344918,
+      "learning_rate": 0.00011945872462380339,
+      "loss": 1.5756,
+      "step": 9867
+    },
+    {
+      "epoch": 0.454599898650205,
+      "grad_norm": 2.326430731684045,
+      "learning_rate": 0.00011944408873029356,
+      "loss": 1.3747,
+      "step": 9868
+    },
+    {
+      "epoch": 0.45464596673884,
+      "grad_norm": 2.3566688391802955,
+      "learning_rate": 0.00011942945240389439,
+      "loss": 1.3949,
+      "step": 9869
+    },
+    {
+      "epoch": 0.454692034827475,
+      "grad_norm": 2.972538310692512,
+      "learning_rate": 0.0001194148156449318,
+      "loss": 1.3532,
+      "step": 9870
+    },
+    {
+      "epoch": 0.45473810291611,
+      "grad_norm": 2.4316539699274933,
+      "learning_rate": 0.00011940017845373164,
+      "loss": 1.5333,
+      "step": 9871
+    },
+    {
+      "epoch": 0.454784171004745,
+      "grad_norm": 3.5764519499837153,
+      "learning_rate": 0.00011938554083061977,
+      "loss": 1.3604,
+      "step": 9872
+    },
+    {
+      "epoch": 0.45483023909338,
+      "grad_norm": 1.160533945148111,
+      "learning_rate": 0.00011937090277592204,
+      "loss": 1.3465,
+      "step": 9873
+    },
+    {
+      "epoch": 0.454876307182015,
+      "grad_norm": 2.2577336549327542,
+      "learning_rate": 0.0001193562642899644,
+      "loss": 1.4615,
+      "step": 9874
+    },
+    {
+      "epoch": 0.45492237527065005,
+      "grad_norm": 1.381332626154389,
+      "learning_rate": 0.00011934162537307273,
+      "loss": 1.2431,
+      "step": 9875
+    },
+    {
+      "epoch": 0.45496844335928505,
+      "grad_norm": 1.242559949944213,
+      "learning_rate": 0.00011932698602557288,
+      "loss": 1.3423,
+      "step": 9876
+    },
+    {
+      "epoch": 0.45501451144792004,
+      "grad_norm": 1.2618597172001842,
+      "learning_rate": 0.00011931234624779088,
+      "loss": 1.7033,
+      "step": 9877
+    },
+    {
+      "epoch": 0.45506057953655504,
+      "grad_norm": 2.695681795007296,
+      "learning_rate": 0.00011929770604005257,
+      "loss": 1.2398,
+      "step": 9878
+    },
+    {
+      "epoch": 0.45510664762519004,
+      "grad_norm": 1.4704360697711096,
+      "learning_rate": 0.00011928306540268396,
+      "loss": 1.0442,
+      "step": 9879
+    },
+    {
+      "epoch": 0.45515271571382504,
+      "grad_norm": 1.2233228541024916,
+      "learning_rate": 0.00011926842433601093,
+      "loss": 1.4315,
+      "step": 9880
+    },
+    {
+      "epoch": 0.45519878380246004,
+      "grad_norm": 2.213669155325796,
+      "learning_rate": 0.00011925378284035953,
+      "loss": 1.2739,
+      "step": 9881
+    },
+    {
+      "epoch": 0.45524485189109504,
+      "grad_norm": 1.6094023552675887,
+      "learning_rate": 0.00011923914091605561,
+      "loss": 1.5096,
+      "step": 9882
+    },
+    {
+      "epoch": 0.45529091997973004,
+      "grad_norm": 0.8524616451482357,
+      "learning_rate": 0.00011922449856342521,
+      "loss": 1.6716,
+      "step": 9883
+    },
+    {
+      "epoch": 0.45533698806836503,
+      "grad_norm": 1.6253024323819638,
+      "learning_rate": 0.00011920985578279434,
+      "loss": 1.2792,
+      "step": 9884
+    },
+    {
+      "epoch": 0.45538305615700003,
+      "grad_norm": 5.441282201827962,
+      "learning_rate": 0.00011919521257448894,
+      "loss": 1.1161,
+      "step": 9885
+    },
+    {
+      "epoch": 0.45542912424563503,
+      "grad_norm": 0.9449401305848256,
+      "learning_rate": 0.00011918056893883504,
+      "loss": 1.5258,
+      "step": 9886
+    },
+    {
+      "epoch": 0.45547519233427003,
+      "grad_norm": 1.3419752389214652,
+      "learning_rate": 0.00011916592487615869,
+      "loss": 1.2934,
+      "step": 9887
+    },
+    {
+      "epoch": 0.455521260422905,
+      "grad_norm": 3.5486626649172037,
+      "learning_rate": 0.00011915128038678586,
+      "loss": 1.0522,
+      "step": 9888
+    },
+    {
+      "epoch": 0.4555673285115401,
+      "grad_norm": 1.1757755341877996,
+      "learning_rate": 0.00011913663547104261,
+      "loss": 1.3615,
+      "step": 9889
+    },
+    {
+      "epoch": 0.4556133966001751,
+      "grad_norm": 7.767555612299987,
+      "learning_rate": 0.00011912199012925502,
+      "loss": 1.3034,
+      "step": 9890
+    },
+    {
+      "epoch": 0.4556594646888101,
+      "grad_norm": 1.498005440700444,
+      "learning_rate": 0.00011910734436174909,
+      "loss": 1.3993,
+      "step": 9891
+    },
+    {
+      "epoch": 0.4557055327774451,
+      "grad_norm": 2.7202206500789994,
+      "learning_rate": 0.00011909269816885089,
+      "loss": 1.4614,
+      "step": 9892
+    },
+    {
+      "epoch": 0.4557516008660801,
+      "grad_norm": 2.0002495216286165,
+      "learning_rate": 0.0001190780515508865,
+      "loss": 1.4808,
+      "step": 9893
+    },
+    {
+      "epoch": 0.4557976689547151,
+      "grad_norm": 1.2799730222232117,
+      "learning_rate": 0.000119063404508182,
+      "loss": 1.2663,
+      "step": 9894
+    },
+    {
+      "epoch": 0.4558437370433501,
+      "grad_norm": 1.2234334720426978,
+      "learning_rate": 0.00011904875704106348,
+      "loss": 1.281,
+      "step": 9895
+    },
+    {
+      "epoch": 0.45588980513198507,
+      "grad_norm": 0.9430495684317259,
+      "learning_rate": 0.0001190341091498571,
+      "loss": 1.3868,
+      "step": 9896
+    },
+    {
+      "epoch": 0.45593587322062007,
+      "grad_norm": 1.1221492506738229,
+      "learning_rate": 0.00011901946083488884,
+      "loss": 1.3173,
+      "step": 9897
+    },
+    {
+      "epoch": 0.45598194130925507,
+      "grad_norm": 0.9445456385607518,
+      "learning_rate": 0.00011900481209648495,
+      "loss": 1.535,
+      "step": 9898
+    },
+    {
+      "epoch": 0.45602800939789007,
+      "grad_norm": 1.3508815149800335,
+      "learning_rate": 0.00011899016293497149,
+      "loss": 1.3139,
+      "step": 9899
+    },
+    {
+      "epoch": 0.45607407748652506,
+      "grad_norm": 0.9528354819256563,
+      "learning_rate": 0.00011897551335067461,
+      "loss": 1.2377,
+      "step": 9900
+    },
+    {
+      "epoch": 0.45612014557516006,
+      "grad_norm": 0.9754991328624611,
+      "learning_rate": 0.00011896086334392044,
+      "loss": 1.5194,
+      "step": 9901
+    },
+    {
+      "epoch": 0.45616621366379506,
+      "grad_norm": 3.572644067334823,
+      "learning_rate": 0.00011894621291503518,
+      "loss": 1.1486,
+      "step": 9902
+    },
+    {
+      "epoch": 0.4562122817524301,
+      "grad_norm": 1.075736230614043,
+      "learning_rate": 0.000118931562064345,
+      "loss": 1.4179,
+      "step": 9903
+    },
+    {
+      "epoch": 0.4562583498410651,
+      "grad_norm": 1.2421880820556277,
+      "learning_rate": 0.000118916910792176,
+      "loss": 1.578,
+      "step": 9904
+    },
+    {
+      "epoch": 0.4563044179297001,
+      "grad_norm": 20.27380273258321,
+      "learning_rate": 0.00011890225909885447,
+      "loss": 0.9568,
+      "step": 9905
+    },
+    {
+      "epoch": 0.4563504860183351,
+      "grad_norm": 1.5958243260750937,
+      "learning_rate": 0.00011888760698470649,
+      "loss": 1.3852,
+      "step": 9906
+    },
+    {
+      "epoch": 0.4563965541069701,
+      "grad_norm": 1.161602519684626,
+      "learning_rate": 0.00011887295445005837,
+      "loss": 1.6573,
+      "step": 9907
+    },
+    {
+      "epoch": 0.4564426221956051,
+      "grad_norm": 1.166502915556164,
+      "learning_rate": 0.0001188583014952363,
+      "loss": 1.3068,
+      "step": 9908
+    },
+    {
+      "epoch": 0.4564886902842401,
+      "grad_norm": 1.797566273526769,
+      "learning_rate": 0.00011884364812056643,
+      "loss": 1.4594,
+      "step": 9909
+    },
+    {
+      "epoch": 0.4565347583728751,
+      "grad_norm": 1.6199312863047886,
+      "learning_rate": 0.00011882899432637507,
+      "loss": 1.3595,
+      "step": 9910
+    },
+    {
+      "epoch": 0.4565808264615101,
+      "grad_norm": 1.5382206405808805,
+      "learning_rate": 0.00011881434011298842,
+      "loss": 1.3577,
+      "step": 9911
+    },
+    {
+      "epoch": 0.4566268945501451,
+      "grad_norm": 1.6823719903216632,
+      "learning_rate": 0.00011879968548073278,
+      "loss": 1.5833,
+      "step": 9912
+    },
+    {
+      "epoch": 0.4566729626387801,
+      "grad_norm": 3.268565650766794,
+      "learning_rate": 0.00011878503042993434,
+      "loss": 1.4506,
+      "step": 9913
+    },
+    {
+      "epoch": 0.4567190307274151,
+      "grad_norm": 1.110279221666036,
+      "learning_rate": 0.00011877037496091943,
+      "loss": 1.3919,
+      "step": 9914
+    },
+    {
+      "epoch": 0.4567650988160501,
+      "grad_norm": 2.034401846663391,
+      "learning_rate": 0.00011875571907401428,
+      "loss": 1.3887,
+      "step": 9915
+    },
+    {
+      "epoch": 0.45681116690468515,
+      "grad_norm": 2.2280324447132136,
+      "learning_rate": 0.00011874106276954524,
+      "loss": 1.4399,
+      "step": 9916
+    },
+    {
+      "epoch": 0.45685723499332015,
+      "grad_norm": 2.534034712885555,
+      "learning_rate": 0.00011872640604783856,
+      "loss": 1.4593,
+      "step": 9917
+    },
+    {
+      "epoch": 0.45690330308195515,
+      "grad_norm": 2.0424489614873247,
+      "learning_rate": 0.00011871174890922056,
+      "loss": 1.322,
+      "step": 9918
+    },
+    {
+      "epoch": 0.45694937117059015,
+      "grad_norm": 1.8241567563233625,
+      "learning_rate": 0.00011869709135401754,
+      "loss": 1.6003,
+      "step": 9919
+    },
+    {
+      "epoch": 0.45699543925922514,
+      "grad_norm": 3.1047438538078147,
+      "learning_rate": 0.00011868243338255585,
+      "loss": 1.3675,
+      "step": 9920
+    },
+    {
+      "epoch": 0.45704150734786014,
+      "grad_norm": 1.2931151867908879,
+      "learning_rate": 0.00011866777499516182,
+      "loss": 1.3692,
+      "step": 9921
+    },
+    {
+      "epoch": 0.45708757543649514,
+      "grad_norm": 1.15225852999704,
+      "learning_rate": 0.00011865311619216174,
+      "loss": 1.7841,
+      "step": 9922
+    },
+    {
+      "epoch": 0.45713364352513014,
+      "grad_norm": 0.7840126845013297,
+      "learning_rate": 0.00011863845697388207,
+      "loss": 1.2736,
+      "step": 9923
+    },
+    {
+      "epoch": 0.45717971161376514,
+      "grad_norm": 2.2554363531372204,
+      "learning_rate": 0.00011862379734064908,
+      "loss": 1.0389,
+      "step": 9924
+    },
+    {
+      "epoch": 0.45722577970240014,
+      "grad_norm": 1.399042526981349,
+      "learning_rate": 0.0001186091372927892,
+      "loss": 1.3892,
+      "step": 9925
+    },
+    {
+      "epoch": 0.45727184779103514,
+      "grad_norm": 3.122443083759263,
+      "learning_rate": 0.00011859447683062878,
+      "loss": 0.9983,
+      "step": 9926
+    },
+    {
+      "epoch": 0.45731791587967013,
+      "grad_norm": 2.598174657263909,
+      "learning_rate": 0.00011857981595449423,
+      "loss": 1.5025,
+      "step": 9927
+    },
+    {
+      "epoch": 0.45736398396830513,
+      "grad_norm": 0.9944643986122,
+      "learning_rate": 0.00011856515466471192,
+      "loss": 1.4127,
+      "step": 9928
+    },
+    {
+      "epoch": 0.45741005205694013,
+      "grad_norm": 1.1102036598690797,
+      "learning_rate": 0.00011855049296160827,
+      "loss": 1.484,
+      "step": 9929
+    },
+    {
+      "epoch": 0.4574561201455752,
+      "grad_norm": 1.1604826752726984,
+      "learning_rate": 0.0001185358308455097,
+      "loss": 1.4689,
+      "step": 9930
+    },
+    {
+      "epoch": 0.4575021882342102,
+      "grad_norm": 1.2113066048109085,
+      "learning_rate": 0.00011852116831674265,
+      "loss": 1.4329,
+      "step": 9931
+    },
+    {
+      "epoch": 0.4575482563228452,
+      "grad_norm": 3.2455045277889725,
+      "learning_rate": 0.0001185065053756335,
+      "loss": 1.3728,
+      "step": 9932
+    },
+    {
+      "epoch": 0.4575943244114802,
+      "grad_norm": 2.664994455257975,
+      "learning_rate": 0.00011849184202250878,
+      "loss": 1.3496,
+      "step": 9933
+    },
+    {
+      "epoch": 0.4576403925001152,
+      "grad_norm": 1.068293417365494,
+      "learning_rate": 0.00011847717825769491,
+      "loss": 1.5146,
+      "step": 9934
+    },
+    {
+      "epoch": 0.4576864605887502,
+      "grad_norm": 1.2124131848881667,
+      "learning_rate": 0.00011846251408151831,
+      "loss": 1.1537,
+      "step": 9935
+    },
+    {
+      "epoch": 0.4577325286773852,
+      "grad_norm": 1.2667519413278574,
+      "learning_rate": 0.00011844784949430556,
+      "loss": 1.5254,
+      "step": 9936
+    },
+    {
+      "epoch": 0.4577785967660202,
+      "grad_norm": 1.596082472194755,
+      "learning_rate": 0.00011843318449638301,
+      "loss": 1.5435,
+      "step": 9937
+    },
+    {
+      "epoch": 0.4578246648546552,
+      "grad_norm": 1.3275406727777073,
+      "learning_rate": 0.00011841851908807724,
+      "loss": 1.2987,
+      "step": 9938
+    },
+    {
+      "epoch": 0.45787073294329017,
+      "grad_norm": 0.9422986755930561,
+      "learning_rate": 0.0001184038532697147,
+      "loss": 1.3222,
+      "step": 9939
+    },
+    {
+      "epoch": 0.45791680103192517,
+      "grad_norm": 1.306350754922399,
+      "learning_rate": 0.00011838918704162196,
+      "loss": 1.2494,
+      "step": 9940
+    },
+    {
+      "epoch": 0.45796286912056017,
+      "grad_norm": 1.84013112739567,
+      "learning_rate": 0.00011837452040412546,
+      "loss": 1.3765,
+      "step": 9941
+    },
+    {
+      "epoch": 0.45800893720919517,
+      "grad_norm": 2.061041813793217,
+      "learning_rate": 0.0001183598533575518,
+      "loss": 1.4807,
+      "step": 9942
+    },
+    {
+      "epoch": 0.4580550052978302,
+      "grad_norm": 1.4268709774569117,
+      "learning_rate": 0.00011834518590222746,
+      "loss": 1.4131,
+      "step": 9943
+    },
+    {
+      "epoch": 0.4581010733864652,
+      "grad_norm": 0.8440401723384114,
+      "learning_rate": 0.00011833051803847903,
+      "loss": 1.5659,
+      "step": 9944
+    },
+    {
+      "epoch": 0.4581471414751002,
+      "grad_norm": 1.856046913212666,
+      "learning_rate": 0.00011831584976663306,
+      "loss": 1.4168,
+      "step": 9945
+    },
+    {
+      "epoch": 0.4581932095637352,
+      "grad_norm": 1.7986062676635424,
+      "learning_rate": 0.00011830118108701606,
+      "loss": 0.8592,
+      "step": 9946
+    },
+    {
+      "epoch": 0.4582392776523702,
+      "grad_norm": 2.6655832563115243,
+      "learning_rate": 0.00011828651199995468,
+      "loss": 1.2993,
+      "step": 9947
+    },
+    {
+      "epoch": 0.4582853457410052,
+      "grad_norm": 1.1456114432721332,
+      "learning_rate": 0.00011827184250577544,
+      "loss": 1.5754,
+      "step": 9948
+    },
+    {
+      "epoch": 0.4583314138296402,
+      "grad_norm": 0.9763974328728314,
+      "learning_rate": 0.00011825717260480498,
+      "loss": 1.3307,
+      "step": 9949
+    },
+    {
+      "epoch": 0.4583774819182752,
+      "grad_norm": 1.0739616130853982,
+      "learning_rate": 0.00011824250229736985,
+      "loss": 1.4583,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4584235500069102,
+      "grad_norm": 1.6638450509611937,
+      "learning_rate": 0.00011822783158379669,
+      "loss": 1.0936,
+      "step": 9951
+    },
+    {
+      "epoch": 0.4584696180955452,
+      "grad_norm": 1.1282832424364324,
+      "learning_rate": 0.0001182131604644121,
+      "loss": 1.2479,
+      "step": 9952
+    },
+    {
+      "epoch": 0.4585156861841802,
+      "grad_norm": 3.2146528375032952,
+      "learning_rate": 0.00011819848893954278,
+      "loss": 1.1648,
+      "step": 9953
+    },
+    {
+      "epoch": 0.4585617542728152,
+      "grad_norm": 1.3376902749907404,
+      "learning_rate": 0.00011818381700951526,
+      "loss": 1.2728,
+      "step": 9954
+    },
+    {
+      "epoch": 0.4586078223614502,
+      "grad_norm": 2.4952575079411843,
+      "learning_rate": 0.00011816914467465623,
+      "loss": 1.0916,
+      "step": 9955
+    },
+    {
+      "epoch": 0.4586538904500852,
+      "grad_norm": 0.9957400656117575,
+      "learning_rate": 0.00011815447193529237,
+      "loss": 1.2713,
+      "step": 9956
+    },
+    {
+      "epoch": 0.45869995853872025,
+      "grad_norm": 0.8870596599054049,
+      "learning_rate": 0.0001181397987917503,
+      "loss": 1.1642,
+      "step": 9957
+    },
+    {
+      "epoch": 0.45874602662735525,
+      "grad_norm": 0.9547643491949077,
+      "learning_rate": 0.00011812512524435673,
+      "loss": 1.4816,
+      "step": 9958
+    },
+    {
+      "epoch": 0.45879209471599025,
+      "grad_norm": 1.5781309330734108,
+      "learning_rate": 0.00011811045129343832,
+      "loss": 1.1052,
+      "step": 9959
+    },
+    {
+      "epoch": 0.45883816280462525,
+      "grad_norm": 0.9415218144502255,
+      "learning_rate": 0.00011809577693932177,
+      "loss": 1.443,
+      "step": 9960
+    },
+    {
+      "epoch": 0.45888423089326025,
+      "grad_norm": 1.243917963769734,
+      "learning_rate": 0.00011808110218233374,
+      "loss": 1.5555,
+      "step": 9961
+    },
+    {
+      "epoch": 0.45893029898189525,
+      "grad_norm": 1.1350622664807266,
+      "learning_rate": 0.00011806642702280102,
+      "loss": 1.4092,
+      "step": 9962
+    },
+    {
+      "epoch": 0.45897636707053024,
+      "grad_norm": 1.344784178569031,
+      "learning_rate": 0.00011805175146105026,
+      "loss": 1.3278,
+      "step": 9963
+    },
+    {
+      "epoch": 0.45902243515916524,
+      "grad_norm": 1.9915985483015148,
+      "learning_rate": 0.00011803707549740819,
+      "loss": 1.2078,
+      "step": 9964
+    },
+    {
+      "epoch": 0.45906850324780024,
+      "grad_norm": 1.8991464870315347,
+      "learning_rate": 0.00011802239913220159,
+      "loss": 1.3408,
+      "step": 9965
+    },
+    {
+      "epoch": 0.45911457133643524,
+      "grad_norm": 1.5830874362140759,
+      "learning_rate": 0.00011800772236575715,
+      "loss": 1.1649,
+      "step": 9966
+    },
+    {
+      "epoch": 0.45916063942507024,
+      "grad_norm": 1.6220983648750384,
+      "learning_rate": 0.00011799304519840167,
+      "loss": 1.3312,
+      "step": 9967
+    },
+    {
+      "epoch": 0.45920670751370524,
+      "grad_norm": 1.8460228770875378,
+      "learning_rate": 0.00011797836763046187,
+      "loss": 1.0258,
+      "step": 9968
+    },
+    {
+      "epoch": 0.45925277560234024,
+      "grad_norm": 1.091078635168358,
+      "learning_rate": 0.00011796368966226455,
+      "loss": 1.5155,
+      "step": 9969
+    },
+    {
+      "epoch": 0.4592988436909753,
+      "grad_norm": 1.7256816823203853,
+      "learning_rate": 0.00011794901129413648,
+      "loss": 1.0408,
+      "step": 9970
+    },
+    {
+      "epoch": 0.4593449117796103,
+      "grad_norm": 1.0400272552643761,
+      "learning_rate": 0.00011793433252640444,
+      "loss": 1.2604,
+      "step": 9971
+    },
+    {
+      "epoch": 0.4593909798682453,
+      "grad_norm": 2.6834222254123437,
+      "learning_rate": 0.00011791965335939525,
+      "loss": 1.505,
+      "step": 9972
+    },
+    {
+      "epoch": 0.4594370479568803,
+      "grad_norm": 1.695688965334592,
+      "learning_rate": 0.00011790497379343573,
+      "loss": 1.0971,
+      "step": 9973
+    },
+    {
+      "epoch": 0.4594831160455153,
+      "grad_norm": 1.4318413264539203,
+      "learning_rate": 0.00011789029382885262,
+      "loss": 1.2991,
+      "step": 9974
+    },
+    {
+      "epoch": 0.4595291841341503,
+      "grad_norm": 4.526837157976475,
+      "learning_rate": 0.0001178756134659728,
+      "loss": 1.5571,
+      "step": 9975
+    },
+    {
+      "epoch": 0.4595752522227853,
+      "grad_norm": 1.5927237473572864,
+      "learning_rate": 0.00011786093270512311,
+      "loss": 1.3617,
+      "step": 9976
+    },
+    {
+      "epoch": 0.4596213203114203,
+      "grad_norm": 2.125292713865073,
+      "learning_rate": 0.00011784625154663036,
+      "loss": 1.2405,
+      "step": 9977
+    },
+    {
+      "epoch": 0.4596673884000553,
+      "grad_norm": 7.711868845961853,
+      "learning_rate": 0.00011783156999082146,
+      "loss": 1.4049,
+      "step": 9978
+    },
+    {
+      "epoch": 0.4597134564886903,
+      "grad_norm": 2.028128468954595,
+      "learning_rate": 0.00011781688803802317,
+      "loss": 1.5287,
+      "step": 9979
+    },
+    {
+      "epoch": 0.4597595245773253,
+      "grad_norm": 1.353018706203476,
+      "learning_rate": 0.00011780220568856248,
+      "loss": 1.4031,
+      "step": 9980
+    },
+    {
+      "epoch": 0.45980559266596027,
+      "grad_norm": 2.5172642182549407,
+      "learning_rate": 0.00011778752294276619,
+      "loss": 1.2593,
+      "step": 9981
+    },
+    {
+      "epoch": 0.45985166075459527,
+      "grad_norm": 14.369338561954603,
+      "learning_rate": 0.00011777283980096119,
+      "loss": 1.2904,
+      "step": 9982
+    },
+    {
+      "epoch": 0.45989772884323027,
+      "grad_norm": 3.5988341029524373,
+      "learning_rate": 0.00011775815626347438,
+      "loss": 1.2534,
+      "step": 9983
+    },
+    {
+      "epoch": 0.4599437969318653,
+      "grad_norm": 7.377068887295016,
+      "learning_rate": 0.00011774347233063268,
+      "loss": 1.2716,
+      "step": 9984
+    },
+    {
+      "epoch": 0.4599898650205003,
+      "grad_norm": 5.022845417534253,
+      "learning_rate": 0.00011772878800276297,
+      "loss": 1.6436,
+      "step": 9985
+    },
+    {
+      "epoch": 0.4600359331091353,
+      "grad_norm": 3.1183931120527064,
+      "learning_rate": 0.00011771410328019221,
+      "loss": 1.3905,
+      "step": 9986
+    },
+    {
+      "epoch": 0.4600820011977703,
+      "grad_norm": 2.4470331895005386,
+      "learning_rate": 0.0001176994181632473,
+      "loss": 1.7014,
+      "step": 9987
+    },
+    {
+      "epoch": 0.4601280692864053,
+      "grad_norm": 2.6153084609525017,
+      "learning_rate": 0.0001176847326522552,
+      "loss": 1.1676,
+      "step": 9988
+    },
+    {
+      "epoch": 0.4601741373750403,
+      "grad_norm": 1.6518544949441942,
+      "learning_rate": 0.00011767004674754285,
+      "loss": 1.4907,
+      "step": 9989
+    },
+    {
+      "epoch": 0.4602202054636753,
+      "grad_norm": 1.2085022949624897,
+      "learning_rate": 0.00011765536044943724,
+      "loss": 1.3851,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4602662735523103,
+      "grad_norm": 0.9205145066854741,
+      "learning_rate": 0.00011764067375826526,
+      "loss": 1.267,
+      "step": 9991
+    },
+    {
+      "epoch": 0.4603123416409453,
+      "grad_norm": 1.9093503482683276,
+      "learning_rate": 0.00011762598667435394,
+      "loss": 1.3454,
+      "step": 9992
+    },
+    {
+      "epoch": 0.4603584097295803,
+      "grad_norm": 1.3751085921877972,
+      "learning_rate": 0.00011761129919803027,
+      "loss": 1.3238,
+      "step": 9993
+    },
+    {
+      "epoch": 0.4604044778182153,
+      "grad_norm": 3.7159342564525244,
+      "learning_rate": 0.00011759661132962117,
+      "loss": 1.3281,
+      "step": 9994
+    },
+    {
+      "epoch": 0.4604505459068503,
+      "grad_norm": 3.985102725105661,
+      "learning_rate": 0.00011758192306945373,
+      "loss": 1.3082,
+      "step": 9995
+    },
+    {
+      "epoch": 0.4604966139954853,
+      "grad_norm": 16.322183194197095,
+      "learning_rate": 0.00011756723441785489,
+      "loss": 1.2765,
+      "step": 9996
+    },
+    {
+      "epoch": 0.4605426820841203,
+      "grad_norm": 1.629639674363125,
+      "learning_rate": 0.00011755254537515173,
+      "loss": 1.1884,
+      "step": 9997
+    },
+    {
+      "epoch": 0.46058875017275536,
+      "grad_norm": 17.13215732558758,
+      "learning_rate": 0.00011753785594167122,
+      "loss": 1.4156,
+      "step": 9998
+    },
+    {
+      "epoch": 0.46063481826139036,
+      "grad_norm": 4.281798547015992,
+      "learning_rate": 0.00011752316611774043,
+      "loss": 1.2246,
+      "step": 9999
+    },
+    {
+      "epoch": 0.46068088635002535,
+      "grad_norm": 2.7510215399114757,
+      "learning_rate": 0.00011750847590368641,
+      "loss": 1.1995,
+      "step": 10000
+    },
+    {
+      "epoch": 0.46072695443866035,
+      "grad_norm": 1.7164768053525772,
+      "learning_rate": 0.00011749378529983617,
+      "loss": 1.2085,
+      "step": 10001
+    },
+    {
+      "epoch": 0.46077302252729535,
+      "grad_norm": 2.657165151298816,
+      "learning_rate": 0.0001174790943065168,
+      "loss": 1.3848,
+      "step": 10002
+    },
+    {
+      "epoch": 0.46081909061593035,
+      "grad_norm": 7.531983770832245,
+      "learning_rate": 0.00011746440292405534,
+      "loss": 1.5351,
+      "step": 10003
+    },
+    {
+      "epoch": 0.46086515870456535,
+      "grad_norm": 1.2194770681034808,
+      "learning_rate": 0.00011744971115277894,
+      "loss": 1.2842,
+      "step": 10004
+    },
+    {
+      "epoch": 0.46091122679320035,
+      "grad_norm": 1.5233752183231433,
+      "learning_rate": 0.00011743501899301462,
+      "loss": 1.6558,
+      "step": 10005
+    },
+    {
+      "epoch": 0.46095729488183534,
+      "grad_norm": 4.480719987244074,
+      "learning_rate": 0.00011742032644508949,
+      "loss": 1.5345,
+      "step": 10006
+    },
+    {
+      "epoch": 0.46100336297047034,
+      "grad_norm": 1.448174640657947,
+      "learning_rate": 0.00011740563350933065,
+      "loss": 1.632,
+      "step": 10007
+    },
+    {
+      "epoch": 0.46104943105910534,
+      "grad_norm": 6.369695628032357,
+      "learning_rate": 0.00011739094018606527,
+      "loss": 1.6191,
+      "step": 10008
+    },
+    {
+      "epoch": 0.46109549914774034,
+      "grad_norm": 1.7941210051664829,
+      "learning_rate": 0.0001173762464756204,
+      "loss": 1.6542,
+      "step": 10009
+    },
+    {
+      "epoch": 0.46114156723637534,
+      "grad_norm": 1.1035713025026737,
+      "learning_rate": 0.00011736155237832318,
+      "loss": 1.2504,
+      "step": 10010
+    },
+    {
+      "epoch": 0.4611876353250104,
+      "grad_norm": 2.2111408383227094,
+      "learning_rate": 0.0001173468578945008,
+      "loss": 1.4068,
+      "step": 10011
+    },
+    {
+      "epoch": 0.4612337034136454,
+      "grad_norm": 3.027094647546069,
+      "learning_rate": 0.00011733216302448033,
+      "loss": 1.5284,
+      "step": 10012
+    },
+    {
+      "epoch": 0.4612797715022804,
+      "grad_norm": 5.340959726980379,
+      "learning_rate": 0.000117317467768589,
+      "loss": 1.3864,
+      "step": 10013
+    },
+    {
+      "epoch": 0.4613258395909154,
+      "grad_norm": 2.477819284455143,
+      "learning_rate": 0.00011730277212715393,
+      "loss": 1.5057,
+      "step": 10014
+    },
+    {
+      "epoch": 0.4613719076795504,
+      "grad_norm": 3.498082785096898,
+      "learning_rate": 0.0001172880761005023,
+      "loss": 1.3336,
+      "step": 10015
+    },
+    {
+      "epoch": 0.4614179757681854,
+      "grad_norm": 2.5103186474540418,
+      "learning_rate": 0.0001172733796889613,
+      "loss": 1.2223,
+      "step": 10016
+    },
+    {
+      "epoch": 0.4614640438568204,
+      "grad_norm": 1.3766313839150135,
+      "learning_rate": 0.00011725868289285813,
+      "loss": 1.4238,
+      "step": 10017
+    },
+    {
+      "epoch": 0.4615101119454554,
+      "grad_norm": 2.7448033966878502,
+      "learning_rate": 0.00011724398571252,
+      "loss": 1.4008,
+      "step": 10018
+    },
+    {
+      "epoch": 0.4615561800340904,
+      "grad_norm": 1.726283161838041,
+      "learning_rate": 0.00011722928814827405,
+      "loss": 1.4177,
+      "step": 10019
+    },
+    {
+      "epoch": 0.4616022481227254,
+      "grad_norm": 3.6876751123914997,
+      "learning_rate": 0.00011721459020044756,
+      "loss": 1.4085,
+      "step": 10020
+    },
+    {
+      "epoch": 0.4616483162113604,
+      "grad_norm": 1.8718605207274739,
+      "learning_rate": 0.00011719989186936773,
+      "loss": 1.389,
+      "step": 10021
+    },
+    {
+      "epoch": 0.4616943842999954,
+      "grad_norm": 2.2901637766262315,
+      "learning_rate": 0.00011718519315536183,
+      "loss": 1.3937,
+      "step": 10022
+    },
+    {
+      "epoch": 0.4617404523886304,
+      "grad_norm": 1.3243626657569996,
+      "learning_rate": 0.00011717049405875702,
+      "loss": 1.3858,
+      "step": 10023
+    },
+    {
+      "epoch": 0.46178652047726537,
+      "grad_norm": 2.9110285161908447,
+      "learning_rate": 0.00011715579457988064,
+      "loss": 1.3534,
+      "step": 10024
+    },
+    {
+      "epoch": 0.4618325885659004,
+      "grad_norm": 3.2946216468992637,
+      "learning_rate": 0.0001171410947190599,
+      "loss": 1.2595,
+      "step": 10025
+    },
+    {
+      "epoch": 0.4618786566545354,
+      "grad_norm": 1.2356071386263214,
+      "learning_rate": 0.00011712639447662205,
+      "loss": 1.4956,
+      "step": 10026
+    },
+    {
+      "epoch": 0.4619247247431704,
+      "grad_norm": 1.557317743098579,
+      "learning_rate": 0.00011711169385289445,
+      "loss": 1.387,
+      "step": 10027
+    },
+    {
+      "epoch": 0.4619707928318054,
+      "grad_norm": 3.08625447060677,
+      "learning_rate": 0.0001170969928482043,
+      "loss": 1.4208,
+      "step": 10028
+    },
+    {
+      "epoch": 0.4620168609204404,
+      "grad_norm": 4.9807643897191625,
+      "learning_rate": 0.0001170822914628789,
+      "loss": 1.3226,
+      "step": 10029
+    },
+    {
+      "epoch": 0.4620629290090754,
+      "grad_norm": 1.2442178063722165,
+      "learning_rate": 0.00011706758969724558,
+      "loss": 1.6777,
+      "step": 10030
+    },
+    {
+      "epoch": 0.4621089970977104,
+      "grad_norm": 2.3652581111420936,
+      "learning_rate": 0.00011705288755163164,
+      "loss": 1.2482,
+      "step": 10031
+    },
+    {
+      "epoch": 0.4621550651863454,
+      "grad_norm": 2.9178026138905118,
+      "learning_rate": 0.00011703818502636439,
+      "loss": 1.2476,
+      "step": 10032
+    },
+    {
+      "epoch": 0.4622011332749804,
+      "grad_norm": 3.619351602500418,
+      "learning_rate": 0.00011702348212177118,
+      "loss": 1.4196,
+      "step": 10033
+    },
+    {
+      "epoch": 0.4622472013636154,
+      "grad_norm": 1.236559902063726,
+      "learning_rate": 0.00011700877883817931,
+      "loss": 1.4615,
+      "step": 10034
+    },
+    {
+      "epoch": 0.4622932694522504,
+      "grad_norm": 2.7956225681711984,
+      "learning_rate": 0.00011699407517591616,
+      "loss": 1.3521,
+      "step": 10035
+    },
+    {
+      "epoch": 0.4623393375408854,
+      "grad_norm": 5.013782404187989,
+      "learning_rate": 0.00011697937113530908,
+      "loss": 1.4643,
+      "step": 10036
+    },
+    {
+      "epoch": 0.4623854056295204,
+      "grad_norm": 2.1014419321478455,
+      "learning_rate": 0.00011696466671668543,
+      "loss": 1.3486,
+      "step": 10037
+    },
+    {
+      "epoch": 0.46243147371815546,
+      "grad_norm": 3.142888039210154,
+      "learning_rate": 0.0001169499619203725,
+      "loss": 1.5933,
+      "step": 10038
+    },
+    {
+      "epoch": 0.46247754180679046,
+      "grad_norm": 2.233815603506551,
+      "learning_rate": 0.00011693525674669779,
+      "loss": 1.4437,
+      "step": 10039
+    },
+    {
+      "epoch": 0.46252360989542546,
+      "grad_norm": 5.199760157530706,
+      "learning_rate": 0.00011692055119598858,
+      "loss": 1.2152,
+      "step": 10040
+    },
+    {
+      "epoch": 0.46256967798406046,
+      "grad_norm": 2.3847379875972203,
+      "learning_rate": 0.00011690584526857233,
+      "loss": 1.2061,
+      "step": 10041
+    },
+    {
+      "epoch": 0.46261574607269546,
+      "grad_norm": 3.2040554214767396,
+      "learning_rate": 0.00011689113896477642,
+      "loss": 1.4803,
+      "step": 10042
+    },
+    {
+      "epoch": 0.46266181416133045,
+      "grad_norm": 4.580796423788875,
+      "learning_rate": 0.00011687643228492826,
+      "loss": 1.4974,
+      "step": 10043
+    },
+    {
+      "epoch": 0.46270788224996545,
+      "grad_norm": 2.984483702165032,
+      "learning_rate": 0.00011686172522935525,
+      "loss": 1.5618,
+      "step": 10044
+    },
+    {
+      "epoch": 0.46275395033860045,
+      "grad_norm": 1.7991329901767532,
+      "learning_rate": 0.00011684701779838489,
+      "loss": 1.6652,
+      "step": 10045
+    },
+    {
+      "epoch": 0.46280001842723545,
+      "grad_norm": 5.268007467001773,
+      "learning_rate": 0.00011683230999234456,
+      "loss": 1.5547,
+      "step": 10046
+    },
+    {
+      "epoch": 0.46284608651587045,
+      "grad_norm": 3.5777137917882977,
+      "learning_rate": 0.00011681760181156167,
+      "loss": 1.487,
+      "step": 10047
+    },
+    {
+      "epoch": 0.46289215460450545,
+      "grad_norm": 2.546790677216773,
+      "learning_rate": 0.00011680289325636373,
+      "loss": 1.5276,
+      "step": 10048
+    },
+    {
+      "epoch": 0.46293822269314044,
+      "grad_norm": 1.5570044320610728,
+      "learning_rate": 0.00011678818432707819,
+      "loss": 1.5108,
+      "step": 10049
+    },
+    {
+      "epoch": 0.46298429078177544,
+      "grad_norm": 0.8084511361441942,
+      "learning_rate": 0.00011677347502403251,
+      "loss": 1.5836,
+      "step": 10050
+    },
+    {
+      "epoch": 0.46303035887041044,
+      "grad_norm": 2.8419155286318576,
+      "learning_rate": 0.00011675876534755414,
+      "loss": 1.4066,
+      "step": 10051
+    },
+    {
+      "epoch": 0.4630764269590455,
+      "grad_norm": 2.326220219013797,
+      "learning_rate": 0.00011674405529797064,
+      "loss": 1.3273,
+      "step": 10052
+    },
+    {
+      "epoch": 0.4631224950476805,
+      "grad_norm": 3.8517441399146253,
+      "learning_rate": 0.00011672934487560942,
+      "loss": 1.0107,
+      "step": 10053
+    },
+    {
+      "epoch": 0.4631685631363155,
+      "grad_norm": 2.6336213661992076,
+      "learning_rate": 0.00011671463408079806,
+      "loss": 1.3585,
+      "step": 10054
+    },
+    {
+      "epoch": 0.4632146312249505,
+      "grad_norm": 7.369230805473311,
+      "learning_rate": 0.000116699922913864,
+      "loss": 1.3063,
+      "step": 10055
+    },
+    {
+      "epoch": 0.4632606993135855,
+      "grad_norm": 2.056637697754626,
+      "learning_rate": 0.0001166852113751348,
+      "loss": 1.2209,
+      "step": 10056
+    },
+    {
+      "epoch": 0.4633067674022205,
+      "grad_norm": 4.928228201361815,
+      "learning_rate": 0.000116670499464938,
+      "loss": 1.3985,
+      "step": 10057
+    },
+    {
+      "epoch": 0.4633528354908555,
+      "grad_norm": 4.299599144763147,
+      "learning_rate": 0.00011665578718360106,
+      "loss": 1.3759,
+      "step": 10058
+    },
+    {
+      "epoch": 0.4633989035794905,
+      "grad_norm": 1.2998480542194903,
+      "learning_rate": 0.00011664107453145163,
+      "loss": 1.2554,
+      "step": 10059
+    },
+    {
+      "epoch": 0.4634449716681255,
+      "grad_norm": 1.1806319191178696,
+      "learning_rate": 0.00011662636150881719,
+      "loss": 1.4865,
+      "step": 10060
+    },
+    {
+      "epoch": 0.4634910397567605,
+      "grad_norm": 1.1178556370936508,
+      "learning_rate": 0.00011661164811602532,
+      "loss": 1.4005,
+      "step": 10061
+    },
+    {
+      "epoch": 0.4635371078453955,
+      "grad_norm": 1.048519280444596,
+      "learning_rate": 0.00011659693435340355,
+      "loss": 1.2755,
+      "step": 10062
+    },
+    {
+      "epoch": 0.4635831759340305,
+      "grad_norm": 3.435688876003208,
+      "learning_rate": 0.00011658222022127954,
+      "loss": 1.267,
+      "step": 10063
+    },
+    {
+      "epoch": 0.4636292440226655,
+      "grad_norm": 1.3398843248663754,
+      "learning_rate": 0.00011656750571998084,
+      "loss": 1.3962,
+      "step": 10064
+    },
+    {
+      "epoch": 0.4636753121113005,
+      "grad_norm": 1.6631205564940763,
+      "learning_rate": 0.00011655279084983499,
+      "loss": 1.3057,
+      "step": 10065
+    },
+    {
+      "epoch": 0.46372138019993553,
+      "grad_norm": 1.0854287976040176,
+      "learning_rate": 0.00011653807561116967,
+      "loss": 1.2212,
+      "step": 10066
+    },
+    {
+      "epoch": 0.4637674482885705,
+      "grad_norm": 2.760071506828404,
+      "learning_rate": 0.00011652336000431243,
+      "loss": 1.187,
+      "step": 10067
+    },
+    {
+      "epoch": 0.4638135163772055,
+      "grad_norm": 2.6482027754100077,
+      "learning_rate": 0.00011650864402959092,
+      "loss": 1.28,
+      "step": 10068
+    },
+    {
+      "epoch": 0.4638595844658405,
+      "grad_norm": 1.5051244935660724,
+      "learning_rate": 0.00011649392768733274,
+      "loss": 1.3376,
+      "step": 10069
+    },
+    {
+      "epoch": 0.4639056525544755,
+      "grad_norm": 5.6453732239104815,
+      "learning_rate": 0.00011647921097786556,
+      "loss": 1.528,
+      "step": 10070
+    },
+    {
+      "epoch": 0.4639517206431105,
+      "grad_norm": 1.4934097745768207,
+      "learning_rate": 0.00011646449390151699,
+      "loss": 1.4583,
+      "step": 10071
+    },
+    {
+      "epoch": 0.4639977887317455,
+      "grad_norm": 11.290254761719561,
+      "learning_rate": 0.00011644977645861473,
+      "loss": 1.4632,
+      "step": 10072
+    },
+    {
+      "epoch": 0.4640438568203805,
+      "grad_norm": 1.4826929916069187,
+      "learning_rate": 0.00011643505864948636,
+      "loss": 1.2314,
+      "step": 10073
+    },
+    {
+      "epoch": 0.4640899249090155,
+      "grad_norm": 1.5246358040928638,
+      "learning_rate": 0.0001164203404744596,
+      "loss": 1.4258,
+      "step": 10074
+    },
+    {
+      "epoch": 0.4641359929976505,
+      "grad_norm": 1.6986458498691028,
+      "learning_rate": 0.00011640562193386213,
+      "loss": 1.3741,
+      "step": 10075
+    },
+    {
+      "epoch": 0.4641820610862855,
+      "grad_norm": 3.5422570145619714,
+      "learning_rate": 0.0001163909030280216,
+      "loss": 1.2022,
+      "step": 10076
+    },
+    {
+      "epoch": 0.4642281291749205,
+      "grad_norm": 9.869995559377491,
+      "learning_rate": 0.00011637618375726574,
+      "loss": 1.3787,
+      "step": 10077
+    },
+    {
+      "epoch": 0.4642741972635555,
+      "grad_norm": 1.4905988531758156,
+      "learning_rate": 0.00011636146412192219,
+      "loss": 1.0955,
+      "step": 10078
+    },
+    {
+      "epoch": 0.46432026535219056,
+      "grad_norm": 1.5724733810298515,
+      "learning_rate": 0.00011634674412231874,
+      "loss": 1.26,
+      "step": 10079
+    },
+    {
+      "epoch": 0.46436633344082556,
+      "grad_norm": 1.1279269615911336,
+      "learning_rate": 0.00011633202375878302,
+      "loss": 1.5505,
+      "step": 10080
+    },
+    {
+      "epoch": 0.46441240152946056,
+      "grad_norm": 23.46640692008284,
+      "learning_rate": 0.0001163173030316428,
+      "loss": 1.0063,
+      "step": 10081
+    },
+    {
+      "epoch": 0.46445846961809556,
+      "grad_norm": 2.0080295912347057,
+      "learning_rate": 0.00011630258194122585,
+      "loss": 1.3265,
+      "step": 10082
+    },
+    {
+      "epoch": 0.46450453770673056,
+      "grad_norm": 4.701698286469957,
+      "learning_rate": 0.00011628786048785985,
+      "loss": 1.5386,
+      "step": 10083
+    },
+    {
+      "epoch": 0.46455060579536556,
+      "grad_norm": 6.2951150761072325,
+      "learning_rate": 0.00011627313867187253,
+      "loss": 1.4857,
+      "step": 10084
+    },
+    {
+      "epoch": 0.46459667388400055,
+      "grad_norm": 9.632109646762226,
+      "learning_rate": 0.0001162584164935917,
+      "loss": 1.6507,
+      "step": 10085
+    },
+    {
+      "epoch": 0.46464274197263555,
+      "grad_norm": 1.815261398828935,
+      "learning_rate": 0.00011624369395334511,
+      "loss": 1.6196,
+      "step": 10086
+    },
+    {
+      "epoch": 0.46468881006127055,
+      "grad_norm": 2.0313235838003028,
+      "learning_rate": 0.00011622897105146049,
+      "loss": 1.5735,
+      "step": 10087
+    },
+    {
+      "epoch": 0.46473487814990555,
+      "grad_norm": 2.9683759459424093,
+      "learning_rate": 0.0001162142477882657,
+      "loss": 1.4798,
+      "step": 10088
+    },
+    {
+      "epoch": 0.46478094623854055,
+      "grad_norm": 1.7007303555478122,
+      "learning_rate": 0.00011619952416408845,
+      "loss": 1.6301,
+      "step": 10089
+    },
+    {
+      "epoch": 0.46482701432717555,
+      "grad_norm": 1.283252355228339,
+      "learning_rate": 0.00011618480017925662,
+      "loss": 1.6067,
+      "step": 10090
+    },
+    {
+      "epoch": 0.46487308241581055,
+      "grad_norm": 1.4368845664891943,
+      "learning_rate": 0.00011617007583409793,
+      "loss": 1.3955,
+      "step": 10091
+    },
+    {
+      "epoch": 0.46491915050444554,
+      "grad_norm": 2.0016523740450185,
+      "learning_rate": 0.00011615535112894024,
+      "loss": 1.2333,
+      "step": 10092
+    },
+    {
+      "epoch": 0.4649652185930806,
+      "grad_norm": 0.9127927301059425,
+      "learning_rate": 0.00011614062606411136,
+      "loss": 1.7072,
+      "step": 10093
+    },
+    {
+      "epoch": 0.4650112866817156,
+      "grad_norm": 1.9304175658793556,
+      "learning_rate": 0.00011612590063993911,
+      "loss": 1.28,
+      "step": 10094
+    },
+    {
+      "epoch": 0.4650573547703506,
+      "grad_norm": 1.7741362088241628,
+      "learning_rate": 0.00011611117485675131,
+      "loss": 1.0897,
+      "step": 10095
+    },
+    {
+      "epoch": 0.4651034228589856,
+      "grad_norm": 1.9382643688355117,
+      "learning_rate": 0.00011609644871487586,
+      "loss": 1.4652,
+      "step": 10096
+    },
+    {
+      "epoch": 0.4651494909476206,
+      "grad_norm": 1.1103187757323583,
+      "learning_rate": 0.00011608172221464055,
+      "loss": 1.3832,
+      "step": 10097
+    },
+    {
+      "epoch": 0.4651955590362556,
+      "grad_norm": 1.3198666064042126,
+      "learning_rate": 0.00011606699535637329,
+      "loss": 1.3757,
+      "step": 10098
+    },
+    {
+      "epoch": 0.4652416271248906,
+      "grad_norm": 1.2508126318614912,
+      "learning_rate": 0.00011605226814040191,
+      "loss": 1.3204,
+      "step": 10099
+    },
+    {
+      "epoch": 0.4652876952135256,
+      "grad_norm": 1.1264270450379468,
+      "learning_rate": 0.00011603754056705431,
+      "loss": 1.4959,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4653337633021606,
+      "grad_norm": 1.4283251319032777,
+      "learning_rate": 0.0001160228126366584,
+      "loss": 1.2714,
+      "step": 10101
+    },
+    {
+      "epoch": 0.4653798313907956,
+      "grad_norm": 1.185149499105681,
+      "learning_rate": 0.000116008084349542,
+      "loss": 1.472,
+      "step": 10102
+    },
+    {
+      "epoch": 0.4654258994794306,
+      "grad_norm": 1.0446827307045417,
+      "learning_rate": 0.00011599335570603307,
+      "loss": 1.6838,
+      "step": 10103
+    },
+    {
+      "epoch": 0.4654719675680656,
+      "grad_norm": 1.537176799836618,
+      "learning_rate": 0.00011597862670645947,
+      "loss": 1.3205,
+      "step": 10104
+    },
+    {
+      "epoch": 0.4655180356567006,
+      "grad_norm": 1.1905748151298539,
+      "learning_rate": 0.00011596389735114916,
+      "loss": 1.2878,
+      "step": 10105
+    },
+    {
+      "epoch": 0.46556410374533563,
+      "grad_norm": 0.9832471509202181,
+      "learning_rate": 0.00011594916764043005,
+      "loss": 1.5488,
+      "step": 10106
+    },
+    {
+      "epoch": 0.46561017183397063,
+      "grad_norm": 1.1479426910714532,
+      "learning_rate": 0.00011593443757463007,
+      "loss": 1.4685,
+      "step": 10107
+    },
+    {
+      "epoch": 0.46565623992260563,
+      "grad_norm": 2.103450379757559,
+      "learning_rate": 0.00011591970715407716,
+      "loss": 1.2864,
+      "step": 10108
+    },
+    {
+      "epoch": 0.46570230801124063,
+      "grad_norm": 1.995533454724618,
+      "learning_rate": 0.00011590497637909927,
+      "loss": 1.6967,
+      "step": 10109
+    },
+    {
+      "epoch": 0.4657483760998756,
+      "grad_norm": 2.017139485839729,
+      "learning_rate": 0.00011589024525002437,
+      "loss": 1.2209,
+      "step": 10110
+    },
+    {
+      "epoch": 0.4657944441885106,
+      "grad_norm": 1.9758516362032919,
+      "learning_rate": 0.00011587551376718038,
+      "loss": 1.3452,
+      "step": 10111
+    },
+    {
+      "epoch": 0.4658405122771456,
+      "grad_norm": 2.113965178946743,
+      "learning_rate": 0.0001158607819308953,
+      "loss": 1.1887,
+      "step": 10112
+    },
+    {
+      "epoch": 0.4658865803657806,
+      "grad_norm": 1.2878941088985043,
+      "learning_rate": 0.00011584604974149708,
+      "loss": 1.6684,
+      "step": 10113
+    },
+    {
+      "epoch": 0.4659326484544156,
+      "grad_norm": 1.6728988758552372,
+      "learning_rate": 0.00011583131719931377,
+      "loss": 1.3458,
+      "step": 10114
+    },
+    {
+      "epoch": 0.4659787165430506,
+      "grad_norm": 1.4959502278297436,
+      "learning_rate": 0.00011581658430467332,
+      "loss": 1.2783,
+      "step": 10115
+    },
+    {
+      "epoch": 0.4660247846316856,
+      "grad_norm": 3.6599159146526805,
+      "learning_rate": 0.00011580185105790373,
+      "loss": 1.8118,
+      "step": 10116
+    },
+    {
+      "epoch": 0.4660708527203206,
+      "grad_norm": 2.0788978746299143,
+      "learning_rate": 0.00011578711745933301,
+      "loss": 1.2055,
+      "step": 10117
+    },
+    {
+      "epoch": 0.4661169208089556,
+      "grad_norm": 2.012310458641499,
+      "learning_rate": 0.00011577238350928923,
+      "loss": 1.3712,
+      "step": 10118
+    },
+    {
+      "epoch": 0.4661629888975906,
+      "grad_norm": 2.1836521328225253,
+      "learning_rate": 0.00011575764920810037,
+      "loss": 1.2329,
+      "step": 10119
+    },
+    {
+      "epoch": 0.46620905698622567,
+      "grad_norm": 2.0516919583949598,
+      "learning_rate": 0.00011574291455609443,
+      "loss": 1.275,
+      "step": 10120
+    },
+    {
+      "epoch": 0.46625512507486067,
+      "grad_norm": 1.028457610018946,
+      "learning_rate": 0.00011572817955359951,
+      "loss": 1.4539,
+      "step": 10121
+    },
+    {
+      "epoch": 0.46630119316349566,
+      "grad_norm": 1.4228131690654344,
+      "learning_rate": 0.00011571344420094363,
+      "loss": 1.3662,
+      "step": 10122
+    },
+    {
+      "epoch": 0.46634726125213066,
+      "grad_norm": 0.8258264649489735,
+      "learning_rate": 0.00011569870849845488,
+      "loss": 1.4559,
+      "step": 10123
+    },
+    {
+      "epoch": 0.46639332934076566,
+      "grad_norm": 1.6190347881709073,
+      "learning_rate": 0.0001156839724464613,
+      "loss": 1.2172,
+      "step": 10124
+    },
+    {
+      "epoch": 0.46643939742940066,
+      "grad_norm": 2.9151107937229415,
+      "learning_rate": 0.00011566923604529095,
+      "loss": 1.4432,
+      "step": 10125
+    },
+    {
+      "epoch": 0.46648546551803566,
+      "grad_norm": 1.0806404687703624,
+      "learning_rate": 0.00011565449929527192,
+      "loss": 1.3946,
+      "step": 10126
+    },
+    {
+      "epoch": 0.46653153360667066,
+      "grad_norm": 2.1974324348235563,
+      "learning_rate": 0.00011563976219673233,
+      "loss": 1.3864,
+      "step": 10127
+    },
+    {
+      "epoch": 0.46657760169530565,
+      "grad_norm": 2.423078083076882,
+      "learning_rate": 0.00011562502475000026,
+      "loss": 1.4998,
+      "step": 10128
+    },
+    {
+      "epoch": 0.46662366978394065,
+      "grad_norm": 4.893105593473111,
+      "learning_rate": 0.00011561028695540379,
+      "loss": 1.4573,
+      "step": 10129
+    },
+    {
+      "epoch": 0.46666973787257565,
+      "grad_norm": 1.3048868682459644,
+      "learning_rate": 0.00011559554881327104,
+      "loss": 1.4702,
+      "step": 10130
+    },
+    {
+      "epoch": 0.46671580596121065,
+      "grad_norm": 2.2374138883292254,
+      "learning_rate": 0.00011558081032393011,
+      "loss": 1.1668,
+      "step": 10131
+    },
+    {
+      "epoch": 0.46676187404984565,
+      "grad_norm": 1.08043830470586,
+      "learning_rate": 0.0001155660714877092,
+      "loss": 1.3917,
+      "step": 10132
+    },
+    {
+      "epoch": 0.46680794213848065,
+      "grad_norm": 2.0780636566281947,
+      "learning_rate": 0.00011555133230493635,
+      "loss": 1.3286,
+      "step": 10133
+    },
+    {
+      "epoch": 0.4668540102271157,
+      "grad_norm": 3.911543321251834,
+      "learning_rate": 0.0001155365927759398,
+      "loss": 1.4551,
+      "step": 10134
+    },
+    {
+      "epoch": 0.4669000783157507,
+      "grad_norm": 2.7692468939221713,
+      "learning_rate": 0.00011552185290104762,
+      "loss": 1.2346,
+      "step": 10135
+    },
+    {
+      "epoch": 0.4669461464043857,
+      "grad_norm": 1.3624173625149658,
+      "learning_rate": 0.00011550711268058802,
+      "loss": 1.6002,
+      "step": 10136
+    },
+    {
+      "epoch": 0.4669922144930207,
+      "grad_norm": 1.7328148503908654,
+      "learning_rate": 0.00011549237211488915,
+      "loss": 1.5891,
+      "step": 10137
+    },
+    {
+      "epoch": 0.4670382825816557,
+      "grad_norm": 1.5830302037644455,
+      "learning_rate": 0.00011547763120427917,
+      "loss": 1.2906,
+      "step": 10138
+    },
+    {
+      "epoch": 0.4670843506702907,
+      "grad_norm": 2.708179171180622,
+      "learning_rate": 0.00011546288994908625,
+      "loss": 1.3951,
+      "step": 10139
+    },
+    {
+      "epoch": 0.4671304187589257,
+      "grad_norm": 1.0059782725437594,
+      "learning_rate": 0.00011544814834963861,
+      "loss": 1.5398,
+      "step": 10140
+    },
+    {
+      "epoch": 0.4671764868475607,
+      "grad_norm": 2.671115145400043,
+      "learning_rate": 0.00011543340640626445,
+      "loss": 1.3637,
+      "step": 10141
+    },
+    {
+      "epoch": 0.4672225549361957,
+      "grad_norm": 2.188375077619174,
+      "learning_rate": 0.00011541866411929195,
+      "loss": 1.3276,
+      "step": 10142
+    },
+    {
+      "epoch": 0.4672686230248307,
+      "grad_norm": 4.149503280649352,
+      "learning_rate": 0.00011540392148904931,
+      "loss": 1.1022,
+      "step": 10143
+    },
+    {
+      "epoch": 0.4673146911134657,
+      "grad_norm": 1.8772055424197507,
+      "learning_rate": 0.00011538917851586477,
+      "loss": 1.5501,
+      "step": 10144
+    },
+    {
+      "epoch": 0.4673607592021007,
+      "grad_norm": 2.804335501569969,
+      "learning_rate": 0.00011537443520006659,
+      "loss": 1.4321,
+      "step": 10145
+    },
+    {
+      "epoch": 0.4674068272907357,
+      "grad_norm": 1.7994625200768246,
+      "learning_rate": 0.00011535969154198298,
+      "loss": 1.0126,
+      "step": 10146
+    },
+    {
+      "epoch": 0.46745289537937074,
+      "grad_norm": 4.716934617802347,
+      "learning_rate": 0.00011534494754194217,
+      "loss": 1.2125,
+      "step": 10147
+    },
+    {
+      "epoch": 0.46749896346800573,
+      "grad_norm": 3.019587139946688,
+      "learning_rate": 0.00011533020320027237,
+      "loss": 1.0642,
+      "step": 10148
+    },
+    {
+      "epoch": 0.46754503155664073,
+      "grad_norm": 3.500104743464604,
+      "learning_rate": 0.00011531545851730192,
+      "loss": 1.4793,
+      "step": 10149
+    },
+    {
+      "epoch": 0.46759109964527573,
+      "grad_norm": 1.3829912491206064,
+      "learning_rate": 0.00011530071349335902,
+      "loss": 1.6199,
+      "step": 10150
+    },
+    {
+      "epoch": 0.46763716773391073,
+      "grad_norm": 2.425407973194627,
+      "learning_rate": 0.00011528596812877199,
+      "loss": 1.5719,
+      "step": 10151
+    },
+    {
+      "epoch": 0.46768323582254573,
+      "grad_norm": 3.202089320833655,
+      "learning_rate": 0.00011527122242386908,
+      "loss": 1.5165,
+      "step": 10152
+    },
+    {
+      "epoch": 0.4677293039111807,
+      "grad_norm": 23.670612664247138,
+      "learning_rate": 0.00011525647637897858,
+      "loss": 1.3477,
+      "step": 10153
+    },
+    {
+      "epoch": 0.4677753719998157,
+      "grad_norm": 8.671105294167068,
+      "learning_rate": 0.00011524172999442879,
+      "loss": 0.9694,
+      "step": 10154
+    },
+    {
+      "epoch": 0.4678214400884507,
+      "grad_norm": 1.2543269812085567,
+      "learning_rate": 0.00011522698327054804,
+      "loss": 1.6907,
+      "step": 10155
+    },
+    {
+      "epoch": 0.4678675081770857,
+      "grad_norm": 1.5591560810905363,
+      "learning_rate": 0.00011521223620766461,
+      "loss": 1.2579,
+      "step": 10156
+    },
+    {
+      "epoch": 0.4679135762657207,
+      "grad_norm": 7.351939530907395,
+      "learning_rate": 0.00011519748880610681,
+      "loss": 1.2555,
+      "step": 10157
+    },
+    {
+      "epoch": 0.4679596443543557,
+      "grad_norm": 3.784460545228609,
+      "learning_rate": 0.00011518274106620299,
+      "loss": 1.6921,
+      "step": 10158
+    },
+    {
+      "epoch": 0.4680057124429907,
+      "grad_norm": 1.984516140124279,
+      "learning_rate": 0.00011516799298828148,
+      "loss": 1.4313,
+      "step": 10159
+    },
+    {
+      "epoch": 0.4680517805316257,
+      "grad_norm": 1.0334670198068872,
+      "learning_rate": 0.0001151532445726706,
+      "loss": 1.517,
+      "step": 10160
+    },
+    {
+      "epoch": 0.46809784862026077,
+      "grad_norm": 3.5028048364471807,
+      "learning_rate": 0.00011513849581969869,
+      "loss": 1.5381,
+      "step": 10161
+    },
+    {
+      "epoch": 0.46814391670889577,
+      "grad_norm": 3.028034030855675,
+      "learning_rate": 0.00011512374672969417,
+      "loss": 1.2525,
+      "step": 10162
+    },
+    {
+      "epoch": 0.46818998479753077,
+      "grad_norm": 3.3275977969367267,
+      "learning_rate": 0.00011510899730298535,
+      "loss": 1.2112,
+      "step": 10163
+    },
+    {
+      "epoch": 0.46823605288616577,
+      "grad_norm": 2.278918734110247,
+      "learning_rate": 0.00011509424753990061,
+      "loss": 1.1558,
+      "step": 10164
+    },
+    {
+      "epoch": 0.46828212097480076,
+      "grad_norm": 4.49215540975257,
+      "learning_rate": 0.00011507949744076833,
+      "loss": 1.3338,
+      "step": 10165
+    },
+    {
+      "epoch": 0.46832818906343576,
+      "grad_norm": 1.894986598823649,
+      "learning_rate": 0.00011506474700591688,
+      "loss": 1.5965,
+      "step": 10166
+    },
+    {
+      "epoch": 0.46837425715207076,
+      "grad_norm": 3.9144887681508975,
+      "learning_rate": 0.00011504999623567467,
+      "loss": 1.2851,
+      "step": 10167
+    },
+    {
+      "epoch": 0.46842032524070576,
+      "grad_norm": 2.4567554517969783,
+      "learning_rate": 0.00011503524513037009,
+      "loss": 1.1931,
+      "step": 10168
+    },
+    {
+      "epoch": 0.46846639332934076,
+      "grad_norm": 3.55818420767834,
+      "learning_rate": 0.00011502049369033161,
+      "loss": 1.3675,
+      "step": 10169
+    },
+    {
+      "epoch": 0.46851246141797576,
+      "grad_norm": 2.0718544688494194,
+      "learning_rate": 0.00011500574191588753,
+      "loss": 1.5733,
+      "step": 10170
+    },
+    {
+      "epoch": 0.46855852950661075,
+      "grad_norm": 0.994440272664122,
+      "learning_rate": 0.00011499098980736638,
+      "loss": 1.6083,
+      "step": 10171
+    },
+    {
+      "epoch": 0.46860459759524575,
+      "grad_norm": 3.997207394449105,
+      "learning_rate": 0.00011497623736509652,
+      "loss": 1.3474,
+      "step": 10172
+    },
+    {
+      "epoch": 0.46865066568388075,
+      "grad_norm": 2.0598510269714594,
+      "learning_rate": 0.00011496148458940643,
+      "loss": 1.6777,
+      "step": 10173
+    },
+    {
+      "epoch": 0.4686967337725158,
+      "grad_norm": 5.731854759807782,
+      "learning_rate": 0.00011494673148062453,
+      "loss": 1.6243,
+      "step": 10174
+    },
+    {
+      "epoch": 0.4687428018611508,
+      "grad_norm": 3.496586789681375,
+      "learning_rate": 0.00011493197803907929,
+      "loss": 1.427,
+      "step": 10175
+    },
+    {
+      "epoch": 0.4687888699497858,
+      "grad_norm": 1.254344105500684,
+      "learning_rate": 0.00011491722426509916,
+      "loss": 1.5752,
+      "step": 10176
+    },
+    {
+      "epoch": 0.4688349380384208,
+      "grad_norm": 3.302841617785694,
+      "learning_rate": 0.00011490247015901258,
+      "loss": 1.6864,
+      "step": 10177
+    },
+    {
+      "epoch": 0.4688810061270558,
+      "grad_norm": 1.5580818204921154,
+      "learning_rate": 0.00011488771572114809,
+      "loss": 1.3852,
+      "step": 10178
+    },
+    {
+      "epoch": 0.4689270742156908,
+      "grad_norm": 2.2966885445746703,
+      "learning_rate": 0.00011487296095183413,
+      "loss": 1.5165,
+      "step": 10179
+    },
+    {
+      "epoch": 0.4689731423043258,
+      "grad_norm": 1.4845405955676514,
+      "learning_rate": 0.00011485820585139919,
+      "loss": 1.4975,
+      "step": 10180
+    },
+    {
+      "epoch": 0.4690192103929608,
+      "grad_norm": 3.0726282678666936,
+      "learning_rate": 0.00011484345042017177,
+      "loss": 1.2254,
+      "step": 10181
+    },
+    {
+      "epoch": 0.4690652784815958,
+      "grad_norm": 2.505875061603965,
+      "learning_rate": 0.00011482869465848039,
+      "loss": 1.5326,
+      "step": 10182
+    },
+    {
+      "epoch": 0.4691113465702308,
+      "grad_norm": 2.09186305119046,
+      "learning_rate": 0.00011481393856665354,
+      "loss": 1.4871,
+      "step": 10183
+    },
+    {
+      "epoch": 0.4691574146588658,
+      "grad_norm": 1.6799927482362542,
+      "learning_rate": 0.00011479918214501977,
+      "loss": 1.4165,
+      "step": 10184
+    },
+    {
+      "epoch": 0.4692034827475008,
+      "grad_norm": 1.252107534039415,
+      "learning_rate": 0.00011478442539390755,
+      "loss": 1.3849,
+      "step": 10185
+    },
+    {
+      "epoch": 0.4692495508361358,
+      "grad_norm": 3.474834821724161,
+      "learning_rate": 0.00011476966831364544,
+      "loss": 1.3434,
+      "step": 10186
+    },
+    {
+      "epoch": 0.4692956189247708,
+      "grad_norm": 2.505553601955984,
+      "learning_rate": 0.00011475491090456203,
+      "loss": 1.3625,
+      "step": 10187
+    },
+    {
+      "epoch": 0.46934168701340584,
+      "grad_norm": 3.0361604704583915,
+      "learning_rate": 0.00011474015316698578,
+      "loss": 1.5622,
+      "step": 10188
+    },
+    {
+      "epoch": 0.46938775510204084,
+      "grad_norm": 1.7173731421149396,
+      "learning_rate": 0.00011472539510124536,
+      "loss": 1.4317,
+      "step": 10189
+    },
+    {
+      "epoch": 0.46943382319067584,
+      "grad_norm": 2.5179703614348834,
+      "learning_rate": 0.00011471063670766922,
+      "loss": 1.4301,
+      "step": 10190
+    },
+    {
+      "epoch": 0.46947989127931083,
+      "grad_norm": 1.8731961760079485,
+      "learning_rate": 0.00011469587798658598,
+      "loss": 1.652,
+      "step": 10191
+    },
+    {
+      "epoch": 0.46952595936794583,
+      "grad_norm": 6.886973610107093,
+      "learning_rate": 0.00011468111893832425,
+      "loss": 1.3437,
+      "step": 10192
+    },
+    {
+      "epoch": 0.46957202745658083,
+      "grad_norm": 0.920466336791604,
+      "learning_rate": 0.00011466635956321254,
+      "loss": 1.58,
+      "step": 10193
+    },
+    {
+      "epoch": 0.46961809554521583,
+      "grad_norm": 1.7188538653218741,
+      "learning_rate": 0.0001146515998615795,
+      "loss": 1.3505,
+      "step": 10194
+    },
+    {
+      "epoch": 0.46966416363385083,
+      "grad_norm": 3.153585132101209,
+      "learning_rate": 0.0001146368398337537,
+      "loss": 1.4661,
+      "step": 10195
+    },
+    {
+      "epoch": 0.4697102317224858,
+      "grad_norm": 6.310283925778098,
+      "learning_rate": 0.00011462207948006375,
+      "loss": 1.2015,
+      "step": 10196
+    },
+    {
+      "epoch": 0.4697562998111208,
+      "grad_norm": 2.1699975641104357,
+      "learning_rate": 0.0001146073188008383,
+      "loss": 1.3557,
+      "step": 10197
+    },
+    {
+      "epoch": 0.4698023678997558,
+      "grad_norm": 1.647002201836965,
+      "learning_rate": 0.00011459255779640592,
+      "loss": 1.1629,
+      "step": 10198
+    },
+    {
+      "epoch": 0.4698484359883908,
+      "grad_norm": 1.5432271819921402,
+      "learning_rate": 0.00011457779646709525,
+      "loss": 1.5712,
+      "step": 10199
+    },
+    {
+      "epoch": 0.4698945040770258,
+      "grad_norm": 1.6348524554632162,
+      "learning_rate": 0.00011456303481323496,
+      "loss": 1.5526,
+      "step": 10200
+    },
+    {
+      "epoch": 0.4699405721656608,
+      "grad_norm": 2.9898855234624264,
+      "learning_rate": 0.00011454827283515368,
+      "loss": 1.3559,
+      "step": 10201
+    },
+    {
+      "epoch": 0.4699866402542959,
+      "grad_norm": 3.7663199028600323,
+      "learning_rate": 0.00011453351053318004,
+      "loss": 1.6859,
+      "step": 10202
+    },
+    {
+      "epoch": 0.47003270834293087,
+      "grad_norm": 2.2963193531014294,
+      "learning_rate": 0.0001145187479076427,
+      "loss": 1.306,
+      "step": 10203
+    },
+    {
+      "epoch": 0.47007877643156587,
+      "grad_norm": 1.9455136712904733,
+      "learning_rate": 0.00011450398495887036,
+      "loss": 1.7117,
+      "step": 10204
+    },
+    {
+      "epoch": 0.47012484452020087,
+      "grad_norm": 3.0928480251903787,
+      "learning_rate": 0.0001144892216871916,
+      "loss": 1.3681,
+      "step": 10205
+    },
+    {
+      "epoch": 0.47017091260883587,
+      "grad_norm": 3.6196039074760282,
+      "learning_rate": 0.00011447445809293522,
+      "loss": 1.2767,
+      "step": 10206
+    },
+    {
+      "epoch": 0.47021698069747087,
+      "grad_norm": 26.120415742274602,
+      "learning_rate": 0.00011445969417642984,
+      "loss": 1.6451,
+      "step": 10207
+    },
+    {
+      "epoch": 0.47026304878610586,
+      "grad_norm": 1.9242462026000973,
+      "learning_rate": 0.00011444492993800416,
+      "loss": 1.354,
+      "step": 10208
+    },
+    {
+      "epoch": 0.47030911687474086,
+      "grad_norm": 1.1437876056600786,
+      "learning_rate": 0.00011443016537798686,
+      "loss": 1.5497,
+      "step": 10209
+    },
+    {
+      "epoch": 0.47035518496337586,
+      "grad_norm": 3.054501551208422,
+      "learning_rate": 0.00011441540049670669,
+      "loss": 1.3802,
+      "step": 10210
+    },
+    {
+      "epoch": 0.47040125305201086,
+      "grad_norm": 1.5016711923034833,
+      "learning_rate": 0.00011440063529449236,
+      "loss": 1.5346,
+      "step": 10211
+    },
+    {
+      "epoch": 0.47044732114064586,
+      "grad_norm": 1.4069502617585854,
+      "learning_rate": 0.00011438586977167254,
+      "loss": 1.2848,
+      "step": 10212
+    },
+    {
+      "epoch": 0.47049338922928086,
+      "grad_norm": 1.985805824306041,
+      "learning_rate": 0.00011437110392857601,
+      "loss": 1.1407,
+      "step": 10213
+    },
+    {
+      "epoch": 0.47053945731791585,
+      "grad_norm": 1.8227600596824107,
+      "learning_rate": 0.0001143563377655315,
+      "loss": 1.318,
+      "step": 10214
+    },
+    {
+      "epoch": 0.4705855254065509,
+      "grad_norm": 7.029475418667048,
+      "learning_rate": 0.00011434157128286775,
+      "loss": 1.4038,
+      "step": 10215
+    },
+    {
+      "epoch": 0.4706315934951859,
+      "grad_norm": 3.475499628366646,
+      "learning_rate": 0.00011432680448091349,
+      "loss": 1.2469,
+      "step": 10216
+    },
+    {
+      "epoch": 0.4706776615838209,
+      "grad_norm": 2.7585296105064168,
+      "learning_rate": 0.00011431203735999751,
+      "loss": 1.3388,
+      "step": 10217
+    },
+    {
+      "epoch": 0.4707237296724559,
+      "grad_norm": 2.0307363535576073,
+      "learning_rate": 0.00011429726992044853,
+      "loss": 1.3576,
+      "step": 10218
+    },
+    {
+      "epoch": 0.4707697977610909,
+      "grad_norm": 4.032769968002264,
+      "learning_rate": 0.00011428250216259539,
+      "loss": 1.2597,
+      "step": 10219
+    },
+    {
+      "epoch": 0.4708158658497259,
+      "grad_norm": 1.4630428241498035,
+      "learning_rate": 0.00011426773408676683,
+      "loss": 1.3975,
+      "step": 10220
+    },
+    {
+      "epoch": 0.4708619339383609,
+      "grad_norm": 2.539086449523499,
+      "learning_rate": 0.00011425296569329158,
+      "loss": 1.3893,
+      "step": 10221
+    },
+    {
+      "epoch": 0.4709080020269959,
+      "grad_norm": 1.5179268455611126,
+      "learning_rate": 0.00011423819698249854,
+      "loss": 1.5481,
+      "step": 10222
+    },
+    {
+      "epoch": 0.4709540701156309,
+      "grad_norm": 2.5426813292891435,
+      "learning_rate": 0.00011422342795471643,
+      "loss": 1.8378,
+      "step": 10223
+    },
+    {
+      "epoch": 0.4710001382042659,
+      "grad_norm": 4.338251484711063,
+      "learning_rate": 0.0001142086586102741,
+      "loss": 1.6037,
+      "step": 10224
+    },
+    {
+      "epoch": 0.4710462062929009,
+      "grad_norm": 2.6970068839629153,
+      "learning_rate": 0.00011419388894950031,
+      "loss": 1.4468,
+      "step": 10225
+    },
+    {
+      "epoch": 0.4710922743815359,
+      "grad_norm": 1.621862936103705,
+      "learning_rate": 0.00011417911897272397,
+      "loss": 1.1492,
+      "step": 10226
+    },
+    {
+      "epoch": 0.4711383424701709,
+      "grad_norm": 1.4906686517950878,
+      "learning_rate": 0.00011416434868027384,
+      "loss": 1.1879,
+      "step": 10227
+    },
+    {
+      "epoch": 0.4711844105588059,
+      "grad_norm": 1.8336973082310668,
+      "learning_rate": 0.0001141495780724788,
+      "loss": 1.6137,
+      "step": 10228
+    },
+    {
+      "epoch": 0.47123047864744094,
+      "grad_norm": 4.229871293473279,
+      "learning_rate": 0.00011413480714966762,
+      "loss": 1.3104,
+      "step": 10229
+    },
+    {
+      "epoch": 0.47127654673607594,
+      "grad_norm": 2.399867217207665,
+      "learning_rate": 0.00011412003591216919,
+      "loss": 1.4654,
+      "step": 10230
+    },
+    {
+      "epoch": 0.47132261482471094,
+      "grad_norm": 1.8554626328763155,
+      "learning_rate": 0.00011410526436031241,
+      "loss": 1.2149,
+      "step": 10231
+    },
+    {
+      "epoch": 0.47136868291334594,
+      "grad_norm": 26.451664102353014,
+      "learning_rate": 0.00011409049249442609,
+      "loss": 1.4134,
+      "step": 10232
+    },
+    {
+      "epoch": 0.47141475100198094,
+      "grad_norm": 1.0776292127021967,
+      "learning_rate": 0.00011407572031483913,
+      "loss": 1.5287,
+      "step": 10233
+    },
+    {
+      "epoch": 0.47146081909061593,
+      "grad_norm": 2.6162071709341714,
+      "learning_rate": 0.00011406094782188035,
+      "loss": 1.5053,
+      "step": 10234
+    },
+    {
+      "epoch": 0.47150688717925093,
+      "grad_norm": 1.649678875391279,
+      "learning_rate": 0.00011404617501587871,
+      "loss": 1.4508,
+      "step": 10235
+    },
+    {
+      "epoch": 0.47155295526788593,
+      "grad_norm": 2.0672920793912266,
+      "learning_rate": 0.00011403140189716306,
+      "loss": 1.4112,
+      "step": 10236
+    },
+    {
+      "epoch": 0.47159902335652093,
+      "grad_norm": 1.3279712932857637,
+      "learning_rate": 0.00011401662846606229,
+      "loss": 1.7411,
+      "step": 10237
+    },
+    {
+      "epoch": 0.47164509144515593,
+      "grad_norm": 1.8122594675827521,
+      "learning_rate": 0.00011400185472290536,
+      "loss": 1.165,
+      "step": 10238
+    },
+    {
+      "epoch": 0.4716911595337909,
+      "grad_norm": 3.43512993470242,
+      "learning_rate": 0.00011398708066802112,
+      "loss": 1.3833,
+      "step": 10239
+    },
+    {
+      "epoch": 0.4717372276224259,
+      "grad_norm": 1.5712522109720057,
+      "learning_rate": 0.0001139723063017385,
+      "loss": 1.5713,
+      "step": 10240
+    },
+    {
+      "epoch": 0.4717832957110609,
+      "grad_norm": 1.4446577929909168,
+      "learning_rate": 0.00011395753162438645,
+      "loss": 1.3137,
+      "step": 10241
+    },
+    {
+      "epoch": 0.471829363799696,
+      "grad_norm": 1.5578161031073678,
+      "learning_rate": 0.00011394275663629388,
+      "loss": 1.4507,
+      "step": 10242
+    },
+    {
+      "epoch": 0.471875431888331,
+      "grad_norm": 1.6551507880355074,
+      "learning_rate": 0.00011392798133778972,
+      "loss": 1.5698,
+      "step": 10243
+    },
+    {
+      "epoch": 0.471921499976966,
+      "grad_norm": 3.086992924686067,
+      "learning_rate": 0.00011391320572920299,
+      "loss": 1.6524,
+      "step": 10244
+    },
+    {
+      "epoch": 0.471967568065601,
+      "grad_norm": 1.652163077302791,
+      "learning_rate": 0.00011389842981086253,
+      "loss": 1.3687,
+      "step": 10245
+    },
+    {
+      "epoch": 0.47201363615423597,
+      "grad_norm": 1.2932555777632078,
+      "learning_rate": 0.00011388365358309743,
+      "loss": 1.5467,
+      "step": 10246
+    },
+    {
+      "epoch": 0.47205970424287097,
+      "grad_norm": 1.1077854500247257,
+      "learning_rate": 0.00011386887704623654,
+      "loss": 1.4027,
+      "step": 10247
+    },
+    {
+      "epoch": 0.47210577233150597,
+      "grad_norm": 1.0476560584688832,
+      "learning_rate": 0.00011385410020060892,
+      "loss": 1.3328,
+      "step": 10248
+    },
+    {
+      "epoch": 0.47215184042014097,
+      "grad_norm": 1.5145342647788642,
+      "learning_rate": 0.00011383932304654347,
+      "loss": 1.3228,
+      "step": 10249
+    },
+    {
+      "epoch": 0.47219790850877597,
+      "grad_norm": 0.9175654986121108,
+      "learning_rate": 0.00011382454558436925,
+      "loss": 1.2824,
+      "step": 10250
+    },
+    {
+      "epoch": 0.47224397659741096,
+      "grad_norm": 1.4082458320095559,
+      "learning_rate": 0.00011380976781441522,
+      "loss": 1.7985,
+      "step": 10251
+    },
+    {
+      "epoch": 0.47229004468604596,
+      "grad_norm": 1.234756369588889,
+      "learning_rate": 0.0001137949897370104,
+      "loss": 1.5648,
+      "step": 10252
+    },
+    {
+      "epoch": 0.47233611277468096,
+      "grad_norm": 2.2971960294825715,
+      "learning_rate": 0.00011378021135248376,
+      "loss": 1.2144,
+      "step": 10253
+    },
+    {
+      "epoch": 0.47238218086331596,
+      "grad_norm": 1.3474984675278896,
+      "learning_rate": 0.00011376543266116435,
+      "loss": 1.411,
+      "step": 10254
+    },
+    {
+      "epoch": 0.47242824895195096,
+      "grad_norm": 3.3734202424512367,
+      "learning_rate": 0.00011375065366338118,
+      "loss": 1.287,
+      "step": 10255
+    },
+    {
+      "epoch": 0.472474317040586,
+      "grad_norm": 1.5480798419693265,
+      "learning_rate": 0.00011373587435946331,
+      "loss": 1.1903,
+      "step": 10256
+    },
+    {
+      "epoch": 0.472520385129221,
+      "grad_norm": 1.8025115338718636,
+      "learning_rate": 0.00011372109474973973,
+      "loss": 1.2872,
+      "step": 10257
+    },
+    {
+      "epoch": 0.472566453217856,
+      "grad_norm": 1.5997452815775732,
+      "learning_rate": 0.0001137063148345395,
+      "loss": 1.3184,
+      "step": 10258
+    },
+    {
+      "epoch": 0.472612521306491,
+      "grad_norm": 1.155837359667116,
+      "learning_rate": 0.00011369153461419168,
+      "loss": 1.4859,
+      "step": 10259
+    },
+    {
+      "epoch": 0.472658589395126,
+      "grad_norm": 1.3480783228167543,
+      "learning_rate": 0.0001136767540890253,
+      "loss": 1.6266,
+      "step": 10260
+    },
+    {
+      "epoch": 0.472704657483761,
+      "grad_norm": 1.3310408106293872,
+      "learning_rate": 0.00011366197325936947,
+      "loss": 1.4698,
+      "step": 10261
+    },
+    {
+      "epoch": 0.472750725572396,
+      "grad_norm": 1.7869820693725516,
+      "learning_rate": 0.00011364719212555321,
+      "loss": 1.4756,
+      "step": 10262
+    },
+    {
+      "epoch": 0.472796793661031,
+      "grad_norm": 3.035361825613116,
+      "learning_rate": 0.00011363241068790562,
+      "loss": 1.1875,
+      "step": 10263
+    },
+    {
+      "epoch": 0.472842861749666,
+      "grad_norm": 1.0111889262000884,
+      "learning_rate": 0.00011361762894675578,
+      "loss": 1.3034,
+      "step": 10264
+    },
+    {
+      "epoch": 0.472888929838301,
+      "grad_norm": 1.3821378560167163,
+      "learning_rate": 0.00011360284690243279,
+      "loss": 1.3882,
+      "step": 10265
+    },
+    {
+      "epoch": 0.472934997926936,
+      "grad_norm": 1.7609074850074078,
+      "learning_rate": 0.00011358806455526574,
+      "loss": 1.4252,
+      "step": 10266
+    },
+    {
+      "epoch": 0.472981066015571,
+      "grad_norm": 0.86069547929898,
+      "learning_rate": 0.00011357328190558372,
+      "loss": 1.4348,
+      "step": 10267
+    },
+    {
+      "epoch": 0.473027134104206,
+      "grad_norm": 2.015260527903865,
+      "learning_rate": 0.00011355849895371587,
+      "loss": 1.5658,
+      "step": 10268
+    },
+    {
+      "epoch": 0.473073202192841,
+      "grad_norm": 1.3409798716924515,
+      "learning_rate": 0.00011354371569999125,
+      "loss": 1.408,
+      "step": 10269
+    },
+    {
+      "epoch": 0.47311927028147605,
+      "grad_norm": 1.3110063394378377,
+      "learning_rate": 0.00011352893214473906,
+      "loss": 1.4729,
+      "step": 10270
+    },
+    {
+      "epoch": 0.47316533837011104,
+      "grad_norm": 1.2103087015033203,
+      "learning_rate": 0.00011351414828828834,
+      "loss": 1.614,
+      "step": 10271
+    },
+    {
+      "epoch": 0.47321140645874604,
+      "grad_norm": 2.3929215857815986,
+      "learning_rate": 0.00011349936413096833,
+      "loss": 1.6416,
+      "step": 10272
+    },
+    {
+      "epoch": 0.47325747454738104,
+      "grad_norm": 3.0101874988942194,
+      "learning_rate": 0.0001134845796731081,
+      "loss": 1.3848,
+      "step": 10273
+    },
+    {
+      "epoch": 0.47330354263601604,
+      "grad_norm": 3.198878632630441,
+      "learning_rate": 0.00011346979491503686,
+      "loss": 1.4777,
+      "step": 10274
+    },
+    {
+      "epoch": 0.47334961072465104,
+      "grad_norm": 2.6374712893596413,
+      "learning_rate": 0.00011345500985708373,
+      "loss": 1.5185,
+      "step": 10275
+    },
+    {
+      "epoch": 0.47339567881328604,
+      "grad_norm": 1.0663898572659203,
+      "learning_rate": 0.00011344022449957783,
+      "loss": 1.5475,
+      "step": 10276
+    },
+    {
+      "epoch": 0.47344174690192103,
+      "grad_norm": 2.4600610989407112,
+      "learning_rate": 0.00011342543884284842,
+      "loss": 1.3527,
+      "step": 10277
+    },
+    {
+      "epoch": 0.47348781499055603,
+      "grad_norm": 3.0104215804837446,
+      "learning_rate": 0.00011341065288722459,
+      "loss": 1.3828,
+      "step": 10278
+    },
+    {
+      "epoch": 0.47353388307919103,
+      "grad_norm": 1.378816454224933,
+      "learning_rate": 0.0001133958666330356,
+      "loss": 1.2413,
+      "step": 10279
+    },
+    {
+      "epoch": 0.47357995116782603,
+      "grad_norm": 8.498052815775182,
+      "learning_rate": 0.00011338108008061059,
+      "loss": 1.5013,
+      "step": 10280
+    },
+    {
+      "epoch": 0.47362601925646103,
+      "grad_norm": 1.2524891831310294,
+      "learning_rate": 0.0001133662932302788,
+      "loss": 1.466,
+      "step": 10281
+    },
+    {
+      "epoch": 0.473672087345096,
+      "grad_norm": 2.8429461185630927,
+      "learning_rate": 0.00011335150608236939,
+      "loss": 1.4238,
+      "step": 10282
+    },
+    {
+      "epoch": 0.4737181554337311,
+      "grad_norm": 2.394087534357363,
+      "learning_rate": 0.00011333671863721163,
+      "loss": 1.3826,
+      "step": 10283
+    },
+    {
+      "epoch": 0.4737642235223661,
+      "grad_norm": 2.0656261415761583,
+      "learning_rate": 0.00011332193089513465,
+      "loss": 1.3234,
+      "step": 10284
+    },
+    {
+      "epoch": 0.4738102916110011,
+      "grad_norm": 2.354073157758907,
+      "learning_rate": 0.00011330714285646774,
+      "loss": 1.4249,
+      "step": 10285
+    },
+    {
+      "epoch": 0.4738563596996361,
+      "grad_norm": 1.9063383436890684,
+      "learning_rate": 0.0001132923545215401,
+      "loss": 1.2791,
+      "step": 10286
+    },
+    {
+      "epoch": 0.4739024277882711,
+      "grad_norm": 2.9781927903222516,
+      "learning_rate": 0.00011327756589068097,
+      "loss": 1.4978,
+      "step": 10287
+    },
+    {
+      "epoch": 0.4739484958769061,
+      "grad_norm": 1.5772858640255751,
+      "learning_rate": 0.00011326277696421963,
+      "loss": 1.4375,
+      "step": 10288
+    },
+    {
+      "epoch": 0.47399456396554107,
+      "grad_norm": 1.5026538652095975,
+      "learning_rate": 0.00011324798774248526,
+      "loss": 1.5867,
+      "step": 10289
+    },
+    {
+      "epoch": 0.47404063205417607,
+      "grad_norm": 1.4229854574096112,
+      "learning_rate": 0.0001132331982258072,
+      "loss": 1.4735,
+      "step": 10290
+    },
+    {
+      "epoch": 0.47408670014281107,
+      "grad_norm": 1.7003505830507768,
+      "learning_rate": 0.00011321840841451465,
+      "loss": 1.3227,
+      "step": 10291
+    },
+    {
+      "epoch": 0.47413276823144607,
+      "grad_norm": 1.1639657250068538,
+      "learning_rate": 0.00011320361830893694,
+      "loss": 1.3278,
+      "step": 10292
+    },
+    {
+      "epoch": 0.47417883632008107,
+      "grad_norm": 2.412020260370894,
+      "learning_rate": 0.00011318882790940327,
+      "loss": 1.1484,
+      "step": 10293
+    },
+    {
+      "epoch": 0.47422490440871606,
+      "grad_norm": 1.6554099463310152,
+      "learning_rate": 0.00011317403721624298,
+      "loss": 1.6973,
+      "step": 10294
+    },
+    {
+      "epoch": 0.47427097249735106,
+      "grad_norm": 1.1360527218777252,
+      "learning_rate": 0.0001131592462297853,
+      "loss": 1.4351,
+      "step": 10295
+    },
+    {
+      "epoch": 0.47431704058598606,
+      "grad_norm": 1.9833509557083546,
+      "learning_rate": 0.00011314445495035961,
+      "loss": 1.2754,
+      "step": 10296
+    },
+    {
+      "epoch": 0.4743631086746211,
+      "grad_norm": 0.7808498614600689,
+      "learning_rate": 0.00011312966337829513,
+      "loss": 1.6588,
+      "step": 10297
+    },
+    {
+      "epoch": 0.4744091767632561,
+      "grad_norm": 0.9130369406803029,
+      "learning_rate": 0.00011311487151392121,
+      "loss": 1.5459,
+      "step": 10298
+    },
+    {
+      "epoch": 0.4744552448518911,
+      "grad_norm": 1.3618978911072466,
+      "learning_rate": 0.00011310007935756721,
+      "loss": 1.2732,
+      "step": 10299
+    },
+    {
+      "epoch": 0.4745013129405261,
+      "grad_norm": 1.3190001766006425,
+      "learning_rate": 0.00011308528690956234,
+      "loss": 1.2769,
+      "step": 10300
+    },
+    {
+      "epoch": 0.4745473810291611,
+      "grad_norm": 0.9822834895627472,
+      "learning_rate": 0.00011307049417023605,
+      "loss": 1.2217,
+      "step": 10301
+    },
+    {
+      "epoch": 0.4745934491177961,
+      "grad_norm": 2.0250522349070406,
+      "learning_rate": 0.00011305570113991762,
+      "loss": 1.62,
+      "step": 10302
+    },
+    {
+      "epoch": 0.4746395172064311,
+      "grad_norm": 1.0226348285852103,
+      "learning_rate": 0.00011304090781893637,
+      "loss": 1.5334,
+      "step": 10303
+    },
+    {
+      "epoch": 0.4746855852950661,
+      "grad_norm": 1.6122433257649451,
+      "learning_rate": 0.00011302611420762166,
+      "loss": 1.3488,
+      "step": 10304
+    },
+    {
+      "epoch": 0.4747316533837011,
+      "grad_norm": 1.3068217353966296,
+      "learning_rate": 0.00011301132030630286,
+      "loss": 1.4939,
+      "step": 10305
+    },
+    {
+      "epoch": 0.4747777214723361,
+      "grad_norm": 2.152187715611883,
+      "learning_rate": 0.0001129965261153093,
+      "loss": 1.3049,
+      "step": 10306
+    },
+    {
+      "epoch": 0.4748237895609711,
+      "grad_norm": 1.1432803523232007,
+      "learning_rate": 0.00011298173163497042,
+      "loss": 1.2557,
+      "step": 10307
+    },
+    {
+      "epoch": 0.4748698576496061,
+      "grad_norm": 1.1456924623872333,
+      "learning_rate": 0.0001129669368656155,
+      "loss": 1.1289,
+      "step": 10308
+    },
+    {
+      "epoch": 0.4749159257382411,
+      "grad_norm": 1.007532688546475,
+      "learning_rate": 0.00011295214180757398,
+      "loss": 1.4531,
+      "step": 10309
+    },
+    {
+      "epoch": 0.47496199382687615,
+      "grad_norm": 16.366983148795946,
+      "learning_rate": 0.00011293734646117523,
+      "loss": 1.1726,
+      "step": 10310
+    },
+    {
+      "epoch": 0.47500806191551115,
+      "grad_norm": 1.3817852820867058,
+      "learning_rate": 0.00011292255082674865,
+      "loss": 1.1469,
+      "step": 10311
+    },
+    {
+      "epoch": 0.47505413000414615,
+      "grad_norm": 1.0442135102973564,
+      "learning_rate": 0.00011290775490462364,
+      "loss": 1.3252,
+      "step": 10312
+    },
+    {
+      "epoch": 0.47510019809278115,
+      "grad_norm": 1.1209264339923994,
+      "learning_rate": 0.00011289295869512957,
+      "loss": 1.6022,
+      "step": 10313
+    },
+    {
+      "epoch": 0.47514626618141614,
+      "grad_norm": 0.8738396372476886,
+      "learning_rate": 0.0001128781621985959,
+      "loss": 1.4833,
+      "step": 10314
+    },
+    {
+      "epoch": 0.47519233427005114,
+      "grad_norm": 0.9109504992667147,
+      "learning_rate": 0.00011286336541535202,
+      "loss": 1.4191,
+      "step": 10315
+    },
+    {
+      "epoch": 0.47523840235868614,
+      "grad_norm": 0.8993736983148237,
+      "learning_rate": 0.00011284856834572739,
+      "loss": 1.512,
+      "step": 10316
+    },
+    {
+      "epoch": 0.47528447044732114,
+      "grad_norm": 1.2925633410498296,
+      "learning_rate": 0.00011283377099005139,
+      "loss": 1.3202,
+      "step": 10317
+    },
+    {
+      "epoch": 0.47533053853595614,
+      "grad_norm": 1.492313917101912,
+      "learning_rate": 0.0001128189733486535,
+      "loss": 1.0911,
+      "step": 10318
+    },
+    {
+      "epoch": 0.47537660662459114,
+      "grad_norm": 0.8639515258709909,
+      "learning_rate": 0.00011280417542186312,
+      "loss": 1.456,
+      "step": 10319
+    },
+    {
+      "epoch": 0.47542267471322613,
+      "grad_norm": 1.086747732342518,
+      "learning_rate": 0.00011278937721000978,
+      "loss": 1.2126,
+      "step": 10320
+    },
+    {
+      "epoch": 0.47546874280186113,
+      "grad_norm": 1.1957992713517434,
+      "learning_rate": 0.00011277457871342285,
+      "loss": 1.4712,
+      "step": 10321
+    },
+    {
+      "epoch": 0.47551481089049613,
+      "grad_norm": 0.941304170156002,
+      "learning_rate": 0.00011275977993243183,
+      "loss": 1.4619,
+      "step": 10322
+    },
+    {
+      "epoch": 0.47556087897913113,
+      "grad_norm": 1.8188175319961268,
+      "learning_rate": 0.0001127449808673662,
+      "loss": 1.2312,
+      "step": 10323
+    },
+    {
+      "epoch": 0.4756069470677662,
+      "grad_norm": 0.8393588465501057,
+      "learning_rate": 0.00011273018151855541,
+      "loss": 1.2867,
+      "step": 10324
+    },
+    {
+      "epoch": 0.4756530151564012,
+      "grad_norm": 1.281814107693604,
+      "learning_rate": 0.00011271538188632895,
+      "loss": 1.5054,
+      "step": 10325
+    },
+    {
+      "epoch": 0.4756990832450362,
+      "grad_norm": 0.9911929917622281,
+      "learning_rate": 0.00011270058197101634,
+      "loss": 1.6622,
+      "step": 10326
+    },
+    {
+      "epoch": 0.4757451513336712,
+      "grad_norm": 0.742015137035493,
+      "learning_rate": 0.00011268578177294705,
+      "loss": 1.4151,
+      "step": 10327
+    },
+    {
+      "epoch": 0.4757912194223062,
+      "grad_norm": 0.895214207248755,
+      "learning_rate": 0.00011267098129245055,
+      "loss": 1.4945,
+      "step": 10328
+    },
+    {
+      "epoch": 0.4758372875109412,
+      "grad_norm": 1.305667197168308,
+      "learning_rate": 0.0001126561805298564,
+      "loss": 1.6067,
+      "step": 10329
+    },
+    {
+      "epoch": 0.4758833555995762,
+      "grad_norm": 0.8714145049351649,
+      "learning_rate": 0.00011264137948549409,
+      "loss": 1.4815,
+      "step": 10330
+    },
+    {
+      "epoch": 0.4759294236882112,
+      "grad_norm": 1.0234005635734054,
+      "learning_rate": 0.00011262657815969312,
+      "loss": 1.2612,
+      "step": 10331
+    },
+    {
+      "epoch": 0.47597549177684617,
+      "grad_norm": 2.3331048319425336,
+      "learning_rate": 0.00011261177655278308,
+      "loss": 1.0539,
+      "step": 10332
+    },
+    {
+      "epoch": 0.47602155986548117,
+      "grad_norm": 1.991668615984187,
+      "learning_rate": 0.0001125969746650934,
+      "loss": 1.0871,
+      "step": 10333
+    },
+    {
+      "epoch": 0.47606762795411617,
+      "grad_norm": 1.074597213219965,
+      "learning_rate": 0.00011258217249695373,
+      "loss": 1.1699,
+      "step": 10334
+    },
+    {
+      "epoch": 0.47611369604275117,
+      "grad_norm": 1.6710725912256232,
+      "learning_rate": 0.00011256737004869353,
+      "loss": 1.3475,
+      "step": 10335
+    },
+    {
+      "epoch": 0.47615976413138617,
+      "grad_norm": 1.163098630531715,
+      "learning_rate": 0.00011255256732064241,
+      "loss": 1.7039,
+      "step": 10336
+    },
+    {
+      "epoch": 0.4762058322200212,
+      "grad_norm": 0.881684553014863,
+      "learning_rate": 0.00011253776431312987,
+      "loss": 1.4716,
+      "step": 10337
+    },
+    {
+      "epoch": 0.4762519003086562,
+      "grad_norm": 1.1539058031540643,
+      "learning_rate": 0.00011252296102648557,
+      "loss": 1.2487,
+      "step": 10338
+    },
+    {
+      "epoch": 0.4762979683972912,
+      "grad_norm": 1.9139572511080103,
+      "learning_rate": 0.00011250815746103896,
+      "loss": 1.6641,
+      "step": 10339
+    },
+    {
+      "epoch": 0.4763440364859262,
+      "grad_norm": 8.534621614284081,
+      "learning_rate": 0.00011249335361711968,
+      "loss": 1.5022,
+      "step": 10340
+    },
+    {
+      "epoch": 0.4763901045745612,
+      "grad_norm": 2.2326276511197958,
+      "learning_rate": 0.0001124785494950573,
+      "loss": 1.4627,
+      "step": 10341
+    },
+    {
+      "epoch": 0.4764361726631962,
+      "grad_norm": 1.7372159944187149,
+      "learning_rate": 0.0001124637450951814,
+      "loss": 1.4425,
+      "step": 10342
+    },
+    {
+      "epoch": 0.4764822407518312,
+      "grad_norm": 2.0701980087589,
+      "learning_rate": 0.00011244894041782161,
+      "loss": 1.5002,
+      "step": 10343
+    },
+    {
+      "epoch": 0.4765283088404662,
+      "grad_norm": 1.4875453418495361,
+      "learning_rate": 0.0001124341354633075,
+      "loss": 1.4136,
+      "step": 10344
+    },
+    {
+      "epoch": 0.4765743769291012,
+      "grad_norm": 2.0013465175435208,
+      "learning_rate": 0.00011241933023196869,
+      "loss": 1.401,
+      "step": 10345
+    },
+    {
+      "epoch": 0.4766204450177362,
+      "grad_norm": 2.46796755975492,
+      "learning_rate": 0.00011240452472413477,
+      "loss": 1.2311,
+      "step": 10346
+    },
+    {
+      "epoch": 0.4766665131063712,
+      "grad_norm": 1.3842210337738807,
+      "learning_rate": 0.00011238971894013541,
+      "loss": 1.6097,
+      "step": 10347
+    },
+    {
+      "epoch": 0.4767125811950062,
+      "grad_norm": 2.2380755480932066,
+      "learning_rate": 0.00011237491288030018,
+      "loss": 1.4834,
+      "step": 10348
+    },
+    {
+      "epoch": 0.4767586492836412,
+      "grad_norm": 1.8431417158645917,
+      "learning_rate": 0.00011236010654495874,
+      "loss": 1.4462,
+      "step": 10349
+    },
+    {
+      "epoch": 0.4768047173722762,
+      "grad_norm": 1.2749070995147198,
+      "learning_rate": 0.0001123452999344407,
+      "loss": 1.518,
+      "step": 10350
+    },
+    {
+      "epoch": 0.47685078546091125,
+      "grad_norm": 3.789533521321906,
+      "learning_rate": 0.00011233049304907575,
+      "loss": 1.5214,
+      "step": 10351
+    },
+    {
+      "epoch": 0.47689685354954625,
+      "grad_norm": 1.9948614845503092,
+      "learning_rate": 0.00011231568588919348,
+      "loss": 1.2632,
+      "step": 10352
+    },
+    {
+      "epoch": 0.47694292163818125,
+      "grad_norm": 1.618227895760991,
+      "learning_rate": 0.0001123008784551236,
+      "loss": 1.566,
+      "step": 10353
+    },
+    {
+      "epoch": 0.47698898972681625,
+      "grad_norm": 1.4956828718357915,
+      "learning_rate": 0.00011228607074719574,
+      "loss": 1.3518,
+      "step": 10354
+    },
+    {
+      "epoch": 0.47703505781545125,
+      "grad_norm": 1.5902223423526005,
+      "learning_rate": 0.00011227126276573961,
+      "loss": 1.3789,
+      "step": 10355
+    },
+    {
+      "epoch": 0.47708112590408625,
+      "grad_norm": 1.486242720226757,
+      "learning_rate": 0.00011225645451108486,
+      "loss": 1.4174,
+      "step": 10356
+    },
+    {
+      "epoch": 0.47712719399272124,
+      "grad_norm": 1.0415546297368419,
+      "learning_rate": 0.00011224164598356113,
+      "loss": 1.4451,
+      "step": 10357
+    },
+    {
+      "epoch": 0.47717326208135624,
+      "grad_norm": 1.8267345675678817,
+      "learning_rate": 0.00011222683718349818,
+      "loss": 1.3974,
+      "step": 10358
+    },
+    {
+      "epoch": 0.47721933016999124,
+      "grad_norm": 1.8538748768010793,
+      "learning_rate": 0.00011221202811122562,
+      "loss": 1.3393,
+      "step": 10359
+    },
+    {
+      "epoch": 0.47726539825862624,
+      "grad_norm": 1.6186056831737008,
+      "learning_rate": 0.00011219721876707323,
+      "loss": 1.4045,
+      "step": 10360
+    },
+    {
+      "epoch": 0.47731146634726124,
+      "grad_norm": 6.6211291417369456,
+      "learning_rate": 0.00011218240915137067,
+      "loss": 1.5313,
+      "step": 10361
+    },
+    {
+      "epoch": 0.47735753443589624,
+      "grad_norm": 1.760290078686046,
+      "learning_rate": 0.00011216759926444765,
+      "loss": 1.5204,
+      "step": 10362
+    },
+    {
+      "epoch": 0.47740360252453123,
+      "grad_norm": 3.380452280787394,
+      "learning_rate": 0.0001121527891066339,
+      "loss": 1.6402,
+      "step": 10363
+    },
+    {
+      "epoch": 0.47744967061316623,
+      "grad_norm": 2.8517786730143753,
+      "learning_rate": 0.00011213797867825912,
+      "loss": 1.7477,
+      "step": 10364
+    },
+    {
+      "epoch": 0.4774957387018013,
+      "grad_norm": 6.504597060208349,
+      "learning_rate": 0.00011212316797965308,
+      "loss": 1.1289,
+      "step": 10365
+    },
+    {
+      "epoch": 0.4775418067904363,
+      "grad_norm": 2.4973107054952424,
+      "learning_rate": 0.00011210835701114548,
+      "loss": 1.5021,
+      "step": 10366
+    },
+    {
+      "epoch": 0.4775878748790713,
+      "grad_norm": 3.2311229243412884,
+      "learning_rate": 0.0001120935457730661,
+      "loss": 1.1499,
+      "step": 10367
+    },
+    {
+      "epoch": 0.4776339429677063,
+      "grad_norm": 9.174060421794064,
+      "learning_rate": 0.00011207873426574462,
+      "loss": 1.349,
+      "step": 10368
+    },
+    {
+      "epoch": 0.4776800110563413,
+      "grad_norm": 2.282552194391805,
+      "learning_rate": 0.00011206392248951085,
+      "loss": 1.2622,
+      "step": 10369
+    },
+    {
+      "epoch": 0.4777260791449763,
+      "grad_norm": 6.173352484841366,
+      "learning_rate": 0.00011204911044469452,
+      "loss": 1.5687,
+      "step": 10370
+    },
+    {
+      "epoch": 0.4777721472336113,
+      "grad_norm": 2.316204239002366,
+      "learning_rate": 0.00011203429813162543,
+      "loss": 1.4219,
+      "step": 10371
+    },
+    {
+      "epoch": 0.4778182153222463,
+      "grad_norm": 3.5063478116068065,
+      "learning_rate": 0.00011201948555063331,
+      "loss": 0.9398,
+      "step": 10372
+    },
+    {
+      "epoch": 0.4778642834108813,
+      "grad_norm": 1.6996182345269641,
+      "learning_rate": 0.00011200467270204795,
+      "loss": 1.3755,
+      "step": 10373
+    },
+    {
+      "epoch": 0.4779103514995163,
+      "grad_norm": 7.191792391510674,
+      "learning_rate": 0.00011198985958619914,
+      "loss": 1.1237,
+      "step": 10374
+    },
+    {
+      "epoch": 0.47795641958815127,
+      "grad_norm": 2.3336915277272277,
+      "learning_rate": 0.00011197504620341667,
+      "loss": 1.3987,
+      "step": 10375
+    },
+    {
+      "epoch": 0.47800248767678627,
+      "grad_norm": 1.5513165658926353,
+      "learning_rate": 0.00011196023255403035,
+      "loss": 1.3921,
+      "step": 10376
+    },
+    {
+      "epoch": 0.47804855576542127,
+      "grad_norm": 2.555589833783889,
+      "learning_rate": 0.00011194541863836991,
+      "loss": 1.273,
+      "step": 10377
+    },
+    {
+      "epoch": 0.4780946238540563,
+      "grad_norm": 3.372510823880155,
+      "learning_rate": 0.00011193060445676525,
+      "loss": 1.4348,
+      "step": 10378
+    },
+    {
+      "epoch": 0.4781406919426913,
+      "grad_norm": 1.4475452355358283,
+      "learning_rate": 0.00011191579000954608,
+      "loss": 1.1676,
+      "step": 10379
+    },
+    {
+      "epoch": 0.4781867600313263,
+      "grad_norm": 1.5539067304325678,
+      "learning_rate": 0.00011190097529704233,
+      "loss": 1.4285,
+      "step": 10380
+    },
+    {
+      "epoch": 0.4782328281199613,
+      "grad_norm": 3.449461544186981,
+      "learning_rate": 0.00011188616031958375,
+      "loss": 1.3658,
+      "step": 10381
+    },
+    {
+      "epoch": 0.4782788962085963,
+      "grad_norm": 1.1277142959621895,
+      "learning_rate": 0.00011187134507750021,
+      "loss": 1.4333,
+      "step": 10382
+    },
+    {
+      "epoch": 0.4783249642972313,
+      "grad_norm": 1.1899005397773386,
+      "learning_rate": 0.00011185652957112149,
+      "loss": 1.4881,
+      "step": 10383
+    },
+    {
+      "epoch": 0.4783710323858663,
+      "grad_norm": 1.0480335365500686,
+      "learning_rate": 0.00011184171380077752,
+      "loss": 1.3736,
+      "step": 10384
+    },
+    {
+      "epoch": 0.4784171004745013,
+      "grad_norm": 2.08208081976231,
+      "learning_rate": 0.00011182689776679807,
+      "loss": 1.4442,
+      "step": 10385
+    },
+    {
+      "epoch": 0.4784631685631363,
+      "grad_norm": 1.153443060367444,
+      "learning_rate": 0.00011181208146951301,
+      "loss": 1.0554,
+      "step": 10386
+    },
+    {
+      "epoch": 0.4785092366517713,
+      "grad_norm": 2.23282578663843,
+      "learning_rate": 0.00011179726490925222,
+      "loss": 1.3494,
+      "step": 10387
+    },
+    {
+      "epoch": 0.4785553047404063,
+      "grad_norm": 2.135572330069842,
+      "learning_rate": 0.00011178244808634554,
+      "loss": 1.3637,
+      "step": 10388
+    },
+    {
+      "epoch": 0.4786013728290413,
+      "grad_norm": 0.9075986178647045,
+      "learning_rate": 0.00011176763100112288,
+      "loss": 1.6458,
+      "step": 10389
+    },
+    {
+      "epoch": 0.4786474409176763,
+      "grad_norm": 1.71675567972933,
+      "learning_rate": 0.00011175281365391408,
+      "loss": 1.2882,
+      "step": 10390
+    },
+    {
+      "epoch": 0.4786935090063113,
+      "grad_norm": 1.8598546866416426,
+      "learning_rate": 0.00011173799604504904,
+      "loss": 1.266,
+      "step": 10391
+    },
+    {
+      "epoch": 0.47873957709494636,
+      "grad_norm": 0.8227589935263562,
+      "learning_rate": 0.00011172317817485764,
+      "loss": 1.6206,
+      "step": 10392
+    },
+    {
+      "epoch": 0.47878564518358135,
+      "grad_norm": 0.96770630022417,
+      "learning_rate": 0.0001117083600436698,
+      "loss": 1.4911,
+      "step": 10393
+    },
+    {
+      "epoch": 0.47883171327221635,
+      "grad_norm": 2.823400907502656,
+      "learning_rate": 0.00011169354165181538,
+      "loss": 1.315,
+      "step": 10394
+    },
+    {
+      "epoch": 0.47887778136085135,
+      "grad_norm": 1.5780072065699038,
+      "learning_rate": 0.00011167872299962432,
+      "loss": 1.2649,
+      "step": 10395
+    },
+    {
+      "epoch": 0.47892384944948635,
+      "grad_norm": 1.4833475569952161,
+      "learning_rate": 0.00011166390408742648,
+      "loss": 1.5839,
+      "step": 10396
+    },
+    {
+      "epoch": 0.47896991753812135,
+      "grad_norm": 1.2388224778623729,
+      "learning_rate": 0.00011164908491555183,
+      "loss": 1.2754,
+      "step": 10397
+    },
+    {
+      "epoch": 0.47901598562675635,
+      "grad_norm": 3.31775624270097,
+      "learning_rate": 0.00011163426548433029,
+      "loss": 1.3408,
+      "step": 10398
+    },
+    {
+      "epoch": 0.47906205371539134,
+      "grad_norm": 1.012280165927587,
+      "learning_rate": 0.00011161944579409176,
+      "loss": 1.4473,
+      "step": 10399
+    },
+    {
+      "epoch": 0.47910812180402634,
+      "grad_norm": 1.106727957386655,
+      "learning_rate": 0.00011160462584516622,
+      "loss": 1.1372,
+      "step": 10400
+    },
+    {
+      "epoch": 0.47915418989266134,
+      "grad_norm": 1.1671722630390045,
+      "learning_rate": 0.00011158980563788357,
+      "loss": 1.0633,
+      "step": 10401
+    },
+    {
+      "epoch": 0.47920025798129634,
+      "grad_norm": 1.1765640557422492,
+      "learning_rate": 0.00011157498517257378,
+      "loss": 1.3581,
+      "step": 10402
+    },
+    {
+      "epoch": 0.47924632606993134,
+      "grad_norm": 1.105913272376492,
+      "learning_rate": 0.00011156016444956678,
+      "loss": 1.4376,
+      "step": 10403
+    },
+    {
+      "epoch": 0.47929239415856634,
+      "grad_norm": 0.9993210934300527,
+      "learning_rate": 0.00011154534346919254,
+      "loss": 1.2122,
+      "step": 10404
+    },
+    {
+      "epoch": 0.4793384622472014,
+      "grad_norm": 1.4280097124477282,
+      "learning_rate": 0.00011153052223178102,
+      "loss": 1.4011,
+      "step": 10405
+    },
+    {
+      "epoch": 0.4793845303358364,
+      "grad_norm": 2.372353756390111,
+      "learning_rate": 0.00011151570073766222,
+      "loss": 1.3858,
+      "step": 10406
+    },
+    {
+      "epoch": 0.4794305984244714,
+      "grad_norm": 0.9975601390289837,
+      "learning_rate": 0.00011150087898716604,
+      "loss": 1.3436,
+      "step": 10407
+    },
+    {
+      "epoch": 0.4794766665131064,
+      "grad_norm": 2.9118068854942596,
+      "learning_rate": 0.00011148605698062255,
+      "loss": 1.1789,
+      "step": 10408
+    },
+    {
+      "epoch": 0.4795227346017414,
+      "grad_norm": 0.9195446566484464,
+      "learning_rate": 0.00011147123471836167,
+      "loss": 1.331,
+      "step": 10409
+    },
+    {
+      "epoch": 0.4795688026903764,
+      "grad_norm": 1.4503957752497543,
+      "learning_rate": 0.00011145641220071343,
+      "loss": 1.415,
+      "step": 10410
+    },
+    {
+      "epoch": 0.4796148707790114,
+      "grad_norm": 0.8587386682163811,
+      "learning_rate": 0.00011144158942800782,
+      "loss": 1.5062,
+      "step": 10411
+    },
+    {
+      "epoch": 0.4796609388676464,
+      "grad_norm": 1.7851929491792016,
+      "learning_rate": 0.00011142676640057484,
+      "loss": 1.4054,
+      "step": 10412
+    },
+    {
+      "epoch": 0.4797070069562814,
+      "grad_norm": 0.942097985068053,
+      "learning_rate": 0.0001114119431187445,
+      "loss": 1.4926,
+      "step": 10413
+    },
+    {
+      "epoch": 0.4797530750449164,
+      "grad_norm": 1.0308922836382362,
+      "learning_rate": 0.0001113971195828468,
+      "loss": 1.4248,
+      "step": 10414
+    },
+    {
+      "epoch": 0.4797991431335514,
+      "grad_norm": 1.4514627898416332,
+      "learning_rate": 0.00011138229579321178,
+      "loss": 1.4059,
+      "step": 10415
+    },
+    {
+      "epoch": 0.4798452112221864,
+      "grad_norm": 1.6797082807113144,
+      "learning_rate": 0.00011136747175016947,
+      "loss": 1.1486,
+      "step": 10416
+    },
+    {
+      "epoch": 0.4798912793108214,
+      "grad_norm": 1.6114934904812497,
+      "learning_rate": 0.00011135264745404987,
+      "loss": 1.4335,
+      "step": 10417
+    },
+    {
+      "epoch": 0.47993734739945637,
+      "grad_norm": 1.630959052774913,
+      "learning_rate": 0.00011133782290518304,
+      "loss": 1.309,
+      "step": 10418
+    },
+    {
+      "epoch": 0.4799834154880914,
+      "grad_norm": 1.1867249255085888,
+      "learning_rate": 0.00011132299810389905,
+      "loss": 1.3185,
+      "step": 10419
+    },
+    {
+      "epoch": 0.4800294835767264,
+      "grad_norm": 3.0011157107530755,
+      "learning_rate": 0.0001113081730505279,
+      "loss": 1.2614,
+      "step": 10420
+    },
+    {
+      "epoch": 0.4800755516653614,
+      "grad_norm": 3.556924313968856,
+      "learning_rate": 0.00011129334774539968,
+      "loss": 1.192,
+      "step": 10421
+    },
+    {
+      "epoch": 0.4801216197539964,
+      "grad_norm": 4.4242281965881345,
+      "learning_rate": 0.00011127852218884445,
+      "loss": 1.2737,
+      "step": 10422
+    },
+    {
+      "epoch": 0.4801676878426314,
+      "grad_norm": 1.0304217996480778,
+      "learning_rate": 0.00011126369638119222,
+      "loss": 1.3476,
+      "step": 10423
+    },
+    {
+      "epoch": 0.4802137559312664,
+      "grad_norm": 4.818682458014455,
+      "learning_rate": 0.00011124887032277316,
+      "loss": 1.0995,
+      "step": 10424
+    },
+    {
+      "epoch": 0.4802598240199014,
+      "grad_norm": 2.0837154719363653,
+      "learning_rate": 0.00011123404401391724,
+      "loss": 1.1324,
+      "step": 10425
+    },
+    {
+      "epoch": 0.4803058921085364,
+      "grad_norm": 1.357167308282677,
+      "learning_rate": 0.0001112192174549546,
+      "loss": 1.0524,
+      "step": 10426
+    },
+    {
+      "epoch": 0.4803519601971714,
+      "grad_norm": 1.8076916401706904,
+      "learning_rate": 0.00011120439064621533,
+      "loss": 1.4469,
+      "step": 10427
+    },
+    {
+      "epoch": 0.4803980282858064,
+      "grad_norm": 1.16861745831192,
+      "learning_rate": 0.00011118956358802953,
+      "loss": 1.4149,
+      "step": 10428
+    },
+    {
+      "epoch": 0.4804440963744414,
+      "grad_norm": 3.5938293294503,
+      "learning_rate": 0.00011117473628072725,
+      "loss": 1.5062,
+      "step": 10429
+    },
+    {
+      "epoch": 0.4804901644630764,
+      "grad_norm": 3.0323103288040487,
+      "learning_rate": 0.00011115990872463864,
+      "loss": 1.507,
+      "step": 10430
+    },
+    {
+      "epoch": 0.4805362325517114,
+      "grad_norm": 1.4678795415895007,
+      "learning_rate": 0.00011114508092009381,
+      "loss": 1.4193,
+      "step": 10431
+    },
+    {
+      "epoch": 0.4805823006403464,
+      "grad_norm": 1.0388350582805013,
+      "learning_rate": 0.00011113025286742284,
+      "loss": 1.2845,
+      "step": 10432
+    },
+    {
+      "epoch": 0.48062836872898146,
+      "grad_norm": 1.1183990005763074,
+      "learning_rate": 0.00011111542456695592,
+      "loss": 1.3866,
+      "step": 10433
+    },
+    {
+      "epoch": 0.48067443681761646,
+      "grad_norm": 1.230709690880666,
+      "learning_rate": 0.00011110059601902308,
+      "loss": 1.433,
+      "step": 10434
+    },
+    {
+      "epoch": 0.48072050490625146,
+      "grad_norm": 1.4794572272894828,
+      "learning_rate": 0.00011108576722395452,
+      "loss": 1.5449,
+      "step": 10435
+    },
+    {
+      "epoch": 0.48076657299488645,
+      "grad_norm": 2.0583688788463563,
+      "learning_rate": 0.00011107093818208037,
+      "loss": 1.3175,
+      "step": 10436
+    },
+    {
+      "epoch": 0.48081264108352145,
+      "grad_norm": 1.9919190769564579,
+      "learning_rate": 0.00011105610889373075,
+      "loss": 1.4135,
+      "step": 10437
+    },
+    {
+      "epoch": 0.48085870917215645,
+      "grad_norm": 2.408148355138044,
+      "learning_rate": 0.00011104127935923583,
+      "loss": 1.1095,
+      "step": 10438
+    },
+    {
+      "epoch": 0.48090477726079145,
+      "grad_norm": 1.0065122251590217,
+      "learning_rate": 0.00011102644957892578,
+      "loss": 1.3432,
+      "step": 10439
+    },
+    {
+      "epoch": 0.48095084534942645,
+      "grad_norm": 7.771656066327663,
+      "learning_rate": 0.00011101161955313071,
+      "loss": 1.5258,
+      "step": 10440
+    },
+    {
+      "epoch": 0.48099691343806145,
+      "grad_norm": 1.5130765647763527,
+      "learning_rate": 0.00011099678928218078,
+      "loss": 1.2887,
+      "step": 10441
+    },
+    {
+      "epoch": 0.48104298152669644,
+      "grad_norm": 1.621430699035244,
+      "learning_rate": 0.00011098195876640625,
+      "loss": 1.0955,
+      "step": 10442
+    },
+    {
+      "epoch": 0.48108904961533144,
+      "grad_norm": 1.3813049285068137,
+      "learning_rate": 0.0001109671280061372,
+      "loss": 1.313,
+      "step": 10443
+    },
+    {
+      "epoch": 0.48113511770396644,
+      "grad_norm": 1.2696067184060575,
+      "learning_rate": 0.00011095229700170386,
+      "loss": 1.5915,
+      "step": 10444
+    },
+    {
+      "epoch": 0.48118118579260144,
+      "grad_norm": 0.878225191569485,
+      "learning_rate": 0.00011093746575343641,
+      "loss": 1.6952,
+      "step": 10445
+    },
+    {
+      "epoch": 0.4812272538812365,
+      "grad_norm": 1.086252761107273,
+      "learning_rate": 0.00011092263426166505,
+      "loss": 1.3091,
+      "step": 10446
+    },
+    {
+      "epoch": 0.4812733219698715,
+      "grad_norm": 1.2630366963707242,
+      "learning_rate": 0.00011090780252671995,
+      "loss": 1.5935,
+      "step": 10447
+    },
+    {
+      "epoch": 0.4813193900585065,
+      "grad_norm": 0.7798695985049814,
+      "learning_rate": 0.00011089297054893135,
+      "loss": 1.6221,
+      "step": 10448
+    },
+    {
+      "epoch": 0.4813654581471415,
+      "grad_norm": 1.2645264428924008,
+      "learning_rate": 0.00011087813832862941,
+      "loss": 1.369,
+      "step": 10449
+    },
+    {
+      "epoch": 0.4814115262357765,
+      "grad_norm": 1.048243755357603,
+      "learning_rate": 0.0001108633058661444,
+      "loss": 1.3492,
+      "step": 10450
+    },
+    {
+      "epoch": 0.4814575943244115,
+      "grad_norm": 2.3427353509086286,
+      "learning_rate": 0.00011084847316180648,
+      "loss": 1.2734,
+      "step": 10451
+    },
+    {
+      "epoch": 0.4815036624130465,
+      "grad_norm": 1.5028491513655562,
+      "learning_rate": 0.00011083364021594593,
+      "loss": 1.2186,
+      "step": 10452
+    },
+    {
+      "epoch": 0.4815497305016815,
+      "grad_norm": 0.96380565178616,
+      "learning_rate": 0.00011081880702889295,
+      "loss": 1.3914,
+      "step": 10453
+    },
+    {
+      "epoch": 0.4815957985903165,
+      "grad_norm": 1.9463559157110049,
+      "learning_rate": 0.00011080397360097777,
+      "loss": 1.4474,
+      "step": 10454
+    },
+    {
+      "epoch": 0.4816418666789515,
+      "grad_norm": 1.5526006945823487,
+      "learning_rate": 0.00011078913993253066,
+      "loss": 1.4175,
+      "step": 10455
+    },
+    {
+      "epoch": 0.4816879347675865,
+      "grad_norm": 1.2364979889070649,
+      "learning_rate": 0.00011077430602388183,
+      "loss": 0.9217,
+      "step": 10456
+    },
+    {
+      "epoch": 0.4817340028562215,
+      "grad_norm": 0.9999345011131614,
+      "learning_rate": 0.0001107594718753616,
+      "loss": 1.5809,
+      "step": 10457
+    },
+    {
+      "epoch": 0.4817800709448565,
+      "grad_norm": 2.5725141256194957,
+      "learning_rate": 0.00011074463748730013,
+      "loss": 1.2804,
+      "step": 10458
+    },
+    {
+      "epoch": 0.4818261390334915,
+      "grad_norm": 1.1886597162859576,
+      "learning_rate": 0.00011072980286002774,
+      "loss": 1.5609,
+      "step": 10459
+    },
+    {
+      "epoch": 0.48187220712212653,
+      "grad_norm": 0.9510044519087194,
+      "learning_rate": 0.00011071496799387468,
+      "loss": 1.2023,
+      "step": 10460
+    },
+    {
+      "epoch": 0.4819182752107615,
+      "grad_norm": 10.586255841473896,
+      "learning_rate": 0.00011070013288917125,
+      "loss": 1.5013,
+      "step": 10461
+    },
+    {
+      "epoch": 0.4819643432993965,
+      "grad_norm": 0.794453294010941,
+      "learning_rate": 0.00011068529754624768,
+      "loss": 1.5186,
+      "step": 10462
+    },
+    {
+      "epoch": 0.4820104113880315,
+      "grad_norm": 1.2043517184928811,
+      "learning_rate": 0.00011067046196543431,
+      "loss": 1.3371,
+      "step": 10463
+    },
+    {
+      "epoch": 0.4820564794766665,
+      "grad_norm": 1.3326128464194025,
+      "learning_rate": 0.00011065562614706138,
+      "loss": 1.2131,
+      "step": 10464
+    },
+    {
+      "epoch": 0.4821025475653015,
+      "grad_norm": 1.0479501375707,
+      "learning_rate": 0.0001106407900914592,
+      "loss": 1.3839,
+      "step": 10465
+    },
+    {
+      "epoch": 0.4821486156539365,
+      "grad_norm": 1.5977079191200092,
+      "learning_rate": 0.0001106259537989581,
+      "loss": 1.2603,
+      "step": 10466
+    },
+    {
+      "epoch": 0.4821946837425715,
+      "grad_norm": 1.0922734391328597,
+      "learning_rate": 0.00011061111726988833,
+      "loss": 1.3175,
+      "step": 10467
+    },
+    {
+      "epoch": 0.4822407518312065,
+      "grad_norm": 1.6889912041478865,
+      "learning_rate": 0.00011059628050458024,
+      "loss": 1.4899,
+      "step": 10468
+    },
+    {
+      "epoch": 0.4822868199198415,
+      "grad_norm": 1.2460508845951463,
+      "learning_rate": 0.00011058144350336412,
+      "loss": 1.4036,
+      "step": 10469
+    },
+    {
+      "epoch": 0.4823328880084765,
+      "grad_norm": 1.1458964539023935,
+      "learning_rate": 0.00011056660626657034,
+      "loss": 1.3147,
+      "step": 10470
+    },
+    {
+      "epoch": 0.4823789560971115,
+      "grad_norm": 3.6684028643058246,
+      "learning_rate": 0.00011055176879452916,
+      "loss": 1.2416,
+      "step": 10471
+    },
+    {
+      "epoch": 0.4824250241857465,
+      "grad_norm": 3.755681165063442,
+      "learning_rate": 0.00011053693108757095,
+      "loss": 1.3391,
+      "step": 10472
+    },
+    {
+      "epoch": 0.48247109227438156,
+      "grad_norm": 2.2144270280202183,
+      "learning_rate": 0.00011052209314602602,
+      "loss": 1.4604,
+      "step": 10473
+    },
+    {
+      "epoch": 0.48251716036301656,
+      "grad_norm": 2.5157612831849043,
+      "learning_rate": 0.00011050725497022476,
+      "loss": 1.4673,
+      "step": 10474
+    },
+    {
+      "epoch": 0.48256322845165156,
+      "grad_norm": 1.052144667166574,
+      "learning_rate": 0.00011049241656049747,
+      "loss": 1.3899,
+      "step": 10475
+    },
+    {
+      "epoch": 0.48260929654028656,
+      "grad_norm": 1.054067660501581,
+      "learning_rate": 0.0001104775779171745,
+      "loss": 1.3566,
+      "step": 10476
+    },
+    {
+      "epoch": 0.48265536462892156,
+      "grad_norm": 1.2262603751344514,
+      "learning_rate": 0.00011046273904058624,
+      "loss": 1.4378,
+      "step": 10477
+    },
+    {
+      "epoch": 0.48270143271755656,
+      "grad_norm": 3.037581228303704,
+      "learning_rate": 0.00011044789993106301,
+      "loss": 1.5311,
+      "step": 10478
+    },
+    {
+      "epoch": 0.48274750080619155,
+      "grad_norm": 1.2516385668042083,
+      "learning_rate": 0.00011043306058893523,
+      "loss": 1.3765,
+      "step": 10479
+    },
+    {
+      "epoch": 0.48279356889482655,
+      "grad_norm": 1.1027444368323045,
+      "learning_rate": 0.00011041822101453323,
+      "loss": 1.3739,
+      "step": 10480
+    },
+    {
+      "epoch": 0.48283963698346155,
+      "grad_norm": 2.3150707571600355,
+      "learning_rate": 0.00011040338120818743,
+      "loss": 1.3159,
+      "step": 10481
+    },
+    {
+      "epoch": 0.48288570507209655,
+      "grad_norm": 0.9709949991377738,
+      "learning_rate": 0.00011038854117022814,
+      "loss": 1.1614,
+      "step": 10482
+    },
+    {
+      "epoch": 0.48293177316073155,
+      "grad_norm": 1.0500293902726716,
+      "learning_rate": 0.00011037370090098582,
+      "loss": 1.2251,
+      "step": 10483
+    },
+    {
+      "epoch": 0.48297784124936655,
+      "grad_norm": 1.2537348846232192,
+      "learning_rate": 0.00011035886040079082,
+      "loss": 1.597,
+      "step": 10484
+    },
+    {
+      "epoch": 0.48302390933800154,
+      "grad_norm": 2.015321664579229,
+      "learning_rate": 0.00011034401966997359,
+      "loss": 1.2484,
+      "step": 10485
+    },
+    {
+      "epoch": 0.48306997742663654,
+      "grad_norm": 1.7133368617177227,
+      "learning_rate": 0.00011032917870886447,
+      "loss": 1.3571,
+      "step": 10486
+    },
+    {
+      "epoch": 0.4831160455152716,
+      "grad_norm": 0.8296221154522877,
+      "learning_rate": 0.0001103143375177939,
+      "loss": 1.5184,
+      "step": 10487
+    },
+    {
+      "epoch": 0.4831621136039066,
+      "grad_norm": 1.4035624628711372,
+      "learning_rate": 0.00011029949609709232,
+      "loss": 1.3297,
+      "step": 10488
+    },
+    {
+      "epoch": 0.4832081816925416,
+      "grad_norm": 2.557378442688712,
+      "learning_rate": 0.00011028465444709006,
+      "loss": 1.382,
+      "step": 10489
+    },
+    {
+      "epoch": 0.4832542497811766,
+      "grad_norm": 1.9753450052553514,
+      "learning_rate": 0.00011026981256811766,
+      "loss": 1.2472,
+      "step": 10490
+    },
+    {
+      "epoch": 0.4833003178698116,
+      "grad_norm": 1.7198435872666287,
+      "learning_rate": 0.00011025497046050546,
+      "loss": 1.4681,
+      "step": 10491
+    },
+    {
+      "epoch": 0.4833463859584466,
+      "grad_norm": 1.8152113830386456,
+      "learning_rate": 0.00011024012812458393,
+      "loss": 1.4827,
+      "step": 10492
+    },
+    {
+      "epoch": 0.4833924540470816,
+      "grad_norm": 2.3676608400484707,
+      "learning_rate": 0.0001102252855606835,
+      "loss": 1.7069,
+      "step": 10493
+    },
+    {
+      "epoch": 0.4834385221357166,
+      "grad_norm": 2.880085304550783,
+      "learning_rate": 0.00011021044276913466,
+      "loss": 1.2942,
+      "step": 10494
+    },
+    {
+      "epoch": 0.4834845902243516,
+      "grad_norm": 1.1963372880222405,
+      "learning_rate": 0.00011019559975026778,
+      "loss": 1.4968,
+      "step": 10495
+    },
+    {
+      "epoch": 0.4835306583129866,
+      "grad_norm": 1.0509254400339363,
+      "learning_rate": 0.00011018075650441336,
+      "loss": 1.27,
+      "step": 10496
+    },
+    {
+      "epoch": 0.4835767264016216,
+      "grad_norm": 14.341779078962713,
+      "learning_rate": 0.00011016591303190187,
+      "loss": 1.3952,
+      "step": 10497
+    },
+    {
+      "epoch": 0.4836227944902566,
+      "grad_norm": 1.376590274171555,
+      "learning_rate": 0.00011015106933306371,
+      "loss": 1.2264,
+      "step": 10498
+    },
+    {
+      "epoch": 0.4836688625788916,
+      "grad_norm": 1.1977058730314951,
+      "learning_rate": 0.00011013622540822945,
+      "loss": 1.3424,
+      "step": 10499
+    },
+    {
+      "epoch": 0.4837149306675266,
+      "grad_norm": 1.9121991409304888,
+      "learning_rate": 0.00011012138125772949,
+      "loss": 1.2741,
+      "step": 10500
+    },
+    {
+      "epoch": 0.48376099875616163,
+      "grad_norm": 2.7589363042791533,
+      "learning_rate": 0.00011010653688189432,
+      "loss": 1.1939,
+      "step": 10501
+    },
+    {
+      "epoch": 0.48380706684479663,
+      "grad_norm": 2.3967024683585443,
+      "learning_rate": 0.00011009169228105445,
+      "loss": 1.5465,
+      "step": 10502
+    },
+    {
+      "epoch": 0.48385313493343163,
+      "grad_norm": 1.5677960167042941,
+      "learning_rate": 0.00011007684745554036,
+      "loss": 1.3313,
+      "step": 10503
+    },
+    {
+      "epoch": 0.4838992030220666,
+      "grad_norm": 2.2768044687806053,
+      "learning_rate": 0.00011006200240568252,
+      "loss": 1.5108,
+      "step": 10504
+    },
+    {
+      "epoch": 0.4839452711107016,
+      "grad_norm": 1.7147221671127533,
+      "learning_rate": 0.00011004715713181148,
+      "loss": 1.3139,
+      "step": 10505
+    },
+    {
+      "epoch": 0.4839913391993366,
+      "grad_norm": 3.791235376663215,
+      "learning_rate": 0.00011003231163425769,
+      "loss": 1.2176,
+      "step": 10506
+    },
+    {
+      "epoch": 0.4840374072879716,
+      "grad_norm": 2.2473327371954492,
+      "learning_rate": 0.0001100174659133517,
+      "loss": 1.4186,
+      "step": 10507
+    },
+    {
+      "epoch": 0.4840834753766066,
+      "grad_norm": 1.2467388893910352,
+      "learning_rate": 0.00011000261996942396,
+      "loss": 1.7091,
+      "step": 10508
+    },
+    {
+      "epoch": 0.4841295434652416,
+      "grad_norm": 1.157438833808255,
+      "learning_rate": 0.00010998777380280509,
+      "loss": 1.5186,
+      "step": 10509
+    },
+    {
+      "epoch": 0.4841756115538766,
+      "grad_norm": 1.9435944638981268,
+      "learning_rate": 0.00010997292741382553,
+      "loss": 1.1259,
+      "step": 10510
+    },
+    {
+      "epoch": 0.4842216796425116,
+      "grad_norm": 1.1076330515643902,
+      "learning_rate": 0.00010995808080281585,
+      "loss": 1.3615,
+      "step": 10511
+    },
+    {
+      "epoch": 0.4842677477311466,
+      "grad_norm": 1.2351389404333737,
+      "learning_rate": 0.0001099432339701066,
+      "loss": 1.2089,
+      "step": 10512
+    },
+    {
+      "epoch": 0.4843138158197816,
+      "grad_norm": 1.1333528679678901,
+      "learning_rate": 0.00010992838691602826,
+      "loss": 1.4041,
+      "step": 10513
+    },
+    {
+      "epoch": 0.48435988390841667,
+      "grad_norm": 1.9299840973789029,
+      "learning_rate": 0.00010991353964091143,
+      "loss": 1.3393,
+      "step": 10514
+    },
+    {
+      "epoch": 0.48440595199705166,
+      "grad_norm": 0.9901562308149732,
+      "learning_rate": 0.00010989869214508662,
+      "loss": 1.1561,
+      "step": 10515
+    },
+    {
+      "epoch": 0.48445202008568666,
+      "grad_norm": 1.1383713651972356,
+      "learning_rate": 0.00010988384442888442,
+      "loss": 1.4988,
+      "step": 10516
+    },
+    {
+      "epoch": 0.48449808817432166,
+      "grad_norm": 2.4094861442573396,
+      "learning_rate": 0.00010986899649263536,
+      "loss": 1.8191,
+      "step": 10517
+    },
+    {
+      "epoch": 0.48454415626295666,
+      "grad_norm": 1.1773347537821537,
+      "learning_rate": 0.00010985414833667002,
+      "loss": 1.3523,
+      "step": 10518
+    },
+    {
+      "epoch": 0.48459022435159166,
+      "grad_norm": 2.0251696266609205,
+      "learning_rate": 0.00010983929996131894,
+      "loss": 1.3281,
+      "step": 10519
+    },
+    {
+      "epoch": 0.48463629244022666,
+      "grad_norm": 1.4161870214862384,
+      "learning_rate": 0.00010982445136691276,
+      "loss": 1.6282,
+      "step": 10520
+    },
+    {
+      "epoch": 0.48468236052886166,
+      "grad_norm": 3.388638904638721,
+      "learning_rate": 0.00010980960255378201,
+      "loss": 1.1362,
+      "step": 10521
+    },
+    {
+      "epoch": 0.48472842861749665,
+      "grad_norm": 2.35436629012633,
+      "learning_rate": 0.00010979475352225726,
+      "loss": 1.3676,
+      "step": 10522
+    },
+    {
+      "epoch": 0.48477449670613165,
+      "grad_norm": 1.2208288097610227,
+      "learning_rate": 0.00010977990427266911,
+      "loss": 1.5708,
+      "step": 10523
+    },
+    {
+      "epoch": 0.48482056479476665,
+      "grad_norm": 1.643133745065472,
+      "learning_rate": 0.00010976505480534816,
+      "loss": 1.3735,
+      "step": 10524
+    },
+    {
+      "epoch": 0.48486663288340165,
+      "grad_norm": 1.7705795073895163,
+      "learning_rate": 0.00010975020512062503,
+      "loss": 1.049,
+      "step": 10525
+    },
+    {
+      "epoch": 0.48491270097203665,
+      "grad_norm": 2.464161150636634,
+      "learning_rate": 0.00010973535521883027,
+      "loss": 1.4932,
+      "step": 10526
+    },
+    {
+      "epoch": 0.48495876906067165,
+      "grad_norm": 1.5751288871104943,
+      "learning_rate": 0.00010972050510029457,
+      "loss": 1.3808,
+      "step": 10527
+    },
+    {
+      "epoch": 0.4850048371493067,
+      "grad_norm": 1.1625598714358973,
+      "learning_rate": 0.00010970565476534845,
+      "loss": 1.2676,
+      "step": 10528
+    },
+    {
+      "epoch": 0.4850509052379417,
+      "grad_norm": 1.4855930292873747,
+      "learning_rate": 0.00010969080421432258,
+      "loss": 1.2657,
+      "step": 10529
+    },
+    {
+      "epoch": 0.4850969733265767,
+      "grad_norm": 1.461593618253591,
+      "learning_rate": 0.00010967595344754758,
+      "loss": 1.4039,
+      "step": 10530
+    },
+    {
+      "epoch": 0.4851430414152117,
+      "grad_norm": 1.0138697727018966,
+      "learning_rate": 0.00010966110246535404,
+      "loss": 1.3458,
+      "step": 10531
+    },
+    {
+      "epoch": 0.4851891095038467,
+      "grad_norm": 2.075417527574932,
+      "learning_rate": 0.00010964625126807265,
+      "loss": 1.4827,
+      "step": 10532
+    },
+    {
+      "epoch": 0.4852351775924817,
+      "grad_norm": 1.6441441529161398,
+      "learning_rate": 0.00010963139985603399,
+      "loss": 1.3183,
+      "step": 10533
+    },
+    {
+      "epoch": 0.4852812456811167,
+      "grad_norm": 1.9538904067676013,
+      "learning_rate": 0.00010961654822956873,
+      "loss": 1.5723,
+      "step": 10534
+    },
+    {
+      "epoch": 0.4853273137697517,
+      "grad_norm": 0.8751535615515152,
+      "learning_rate": 0.0001096016963890075,
+      "loss": 1.4949,
+      "step": 10535
+    },
+    {
+      "epoch": 0.4853733818583867,
+      "grad_norm": 1.4960998409185584,
+      "learning_rate": 0.000109586844334681,
+      "loss": 0.9625,
+      "step": 10536
+    },
+    {
+      "epoch": 0.4854194499470217,
+      "grad_norm": 0.9100569186978302,
+      "learning_rate": 0.00010957199206691981,
+      "loss": 1.5984,
+      "step": 10537
+    },
+    {
+      "epoch": 0.4854655180356567,
+      "grad_norm": 1.3905885902266875,
+      "learning_rate": 0.00010955713958605466,
+      "loss": 1.0427,
+      "step": 10538
+    },
+    {
+      "epoch": 0.4855115861242917,
+      "grad_norm": 1.6508566687262334,
+      "learning_rate": 0.00010954228689241618,
+      "loss": 1.3331,
+      "step": 10539
+    },
+    {
+      "epoch": 0.4855576542129267,
+      "grad_norm": 1.7173223843954024,
+      "learning_rate": 0.00010952743398633505,
+      "loss": 1.2849,
+      "step": 10540
+    },
+    {
+      "epoch": 0.48560372230156174,
+      "grad_norm": 1.9239065112744114,
+      "learning_rate": 0.00010951258086814193,
+      "loss": 1.4169,
+      "step": 10541
+    },
+    {
+      "epoch": 0.48564979039019673,
+      "grad_norm": 0.9349519196897093,
+      "learning_rate": 0.0001094977275381675,
+      "loss": 1.3533,
+      "step": 10542
+    },
+    {
+      "epoch": 0.48569585847883173,
+      "grad_norm": 0.999493189302293,
+      "learning_rate": 0.00010948287399674246,
+      "loss": 1.3075,
+      "step": 10543
+    },
+    {
+      "epoch": 0.48574192656746673,
+      "grad_norm": 1.4668142507198816,
+      "learning_rate": 0.00010946802024419746,
+      "loss": 1.2025,
+      "step": 10544
+    },
+    {
+      "epoch": 0.48578799465610173,
+      "grad_norm": 0.8273968223266157,
+      "learning_rate": 0.00010945316628086326,
+      "loss": 1.3147,
+      "step": 10545
+    },
+    {
+      "epoch": 0.4858340627447367,
+      "grad_norm": 2.282911269081329,
+      "learning_rate": 0.00010943831210707051,
+      "loss": 1.3308,
+      "step": 10546
+    },
+    {
+      "epoch": 0.4858801308333717,
+      "grad_norm": 1.0753212550250717,
+      "learning_rate": 0.00010942345772314992,
+      "loss": 1.1616,
+      "step": 10547
+    },
+    {
+      "epoch": 0.4859261989220067,
+      "grad_norm": 6.21449684706523,
+      "learning_rate": 0.0001094086031294322,
+      "loss": 1.3371,
+      "step": 10548
+    },
+    {
+      "epoch": 0.4859722670106417,
+      "grad_norm": 1.636807896174514,
+      "learning_rate": 0.00010939374832624811,
+      "loss": 1.2095,
+      "step": 10549
+    },
+    {
+      "epoch": 0.4860183350992767,
+      "grad_norm": 1.7758878623938046,
+      "learning_rate": 0.00010937889331392825,
+      "loss": 1.0986,
+      "step": 10550
+    },
+    {
+      "epoch": 0.4860644031879117,
+      "grad_norm": 1.6167048119204814,
+      "learning_rate": 0.00010936403809280347,
+      "loss": 1.7842,
+      "step": 10551
+    },
+    {
+      "epoch": 0.4861104712765467,
+      "grad_norm": 1.2944689935581297,
+      "learning_rate": 0.0001093491826632044,
+      "loss": 1.4431,
+      "step": 10552
+    },
+    {
+      "epoch": 0.4861565393651817,
+      "grad_norm": 1.075700987476402,
+      "learning_rate": 0.0001093343270254618,
+      "loss": 1.5198,
+      "step": 10553
+    },
+    {
+      "epoch": 0.4862026074538167,
+      "grad_norm": 1.2543724094351048,
+      "learning_rate": 0.00010931947117990644,
+      "loss": 1.352,
+      "step": 10554
+    },
+    {
+      "epoch": 0.48624867554245177,
+      "grad_norm": 1.1026827613048886,
+      "learning_rate": 0.000109304615126869,
+      "loss": 1.4512,
+      "step": 10555
+    },
+    {
+      "epoch": 0.48629474363108677,
+      "grad_norm": 2.1906843485332437,
+      "learning_rate": 0.00010928975886668031,
+      "loss": 1.4458,
+      "step": 10556
+    },
+    {
+      "epoch": 0.48634081171972177,
+      "grad_norm": 1.3969155189615465,
+      "learning_rate": 0.00010927490239967102,
+      "loss": 1.4901,
+      "step": 10557
+    },
+    {
+      "epoch": 0.48638687980835676,
+      "grad_norm": 1.1488301679397237,
+      "learning_rate": 0.00010926004572617199,
+      "loss": 1.3736,
+      "step": 10558
+    },
+    {
+      "epoch": 0.48643294789699176,
+      "grad_norm": 1.4074785856467984,
+      "learning_rate": 0.00010924518884651387,
+      "loss": 1.28,
+      "step": 10559
+    },
+    {
+      "epoch": 0.48647901598562676,
+      "grad_norm": 1.026983460609275,
+      "learning_rate": 0.00010923033176102749,
+      "loss": 1.123,
+      "step": 10560
+    },
+    {
+      "epoch": 0.48652508407426176,
+      "grad_norm": 1.2608942978432287,
+      "learning_rate": 0.0001092154744700436,
+      "loss": 1.3954,
+      "step": 10561
+    },
+    {
+      "epoch": 0.48657115216289676,
+      "grad_norm": 1.7146445025557246,
+      "learning_rate": 0.00010920061697389297,
+      "loss": 1.481,
+      "step": 10562
+    },
+    {
+      "epoch": 0.48661722025153176,
+      "grad_norm": 1.018603500847239,
+      "learning_rate": 0.00010918575927290635,
+      "loss": 1.3014,
+      "step": 10563
+    },
+    {
+      "epoch": 0.48666328834016676,
+      "grad_norm": 1.0080052917206392,
+      "learning_rate": 0.00010917090136741459,
+      "loss": 1.7897,
+      "step": 10564
+    },
+    {
+      "epoch": 0.48670935642880175,
+      "grad_norm": 1.9591443189235052,
+      "learning_rate": 0.0001091560432577484,
+      "loss": 1.4197,
+      "step": 10565
+    },
+    {
+      "epoch": 0.48675542451743675,
+      "grad_norm": 1.479727825832125,
+      "learning_rate": 0.00010914118494423861,
+      "loss": 1.6104,
+      "step": 10566
+    },
+    {
+      "epoch": 0.48680149260607175,
+      "grad_norm": 1.2411755079592195,
+      "learning_rate": 0.00010912632642721608,
+      "loss": 1.5941,
+      "step": 10567
+    },
+    {
+      "epoch": 0.48684756069470675,
+      "grad_norm": 1.1301304848808924,
+      "learning_rate": 0.00010911146770701146,
+      "loss": 1.4255,
+      "step": 10568
+    },
+    {
+      "epoch": 0.4868936287833418,
+      "grad_norm": 0.8329094616759608,
+      "learning_rate": 0.00010909660878395566,
+      "loss": 1.3232,
+      "step": 10569
+    },
+    {
+      "epoch": 0.4869396968719768,
+      "grad_norm": 2.582477514075206,
+      "learning_rate": 0.00010908174965837948,
+      "loss": 1.3311,
+      "step": 10570
+    },
+    {
+      "epoch": 0.4869857649606118,
+      "grad_norm": 1.5343608471593062,
+      "learning_rate": 0.00010906689033061371,
+      "loss": 1.2638,
+      "step": 10571
+    },
+    {
+      "epoch": 0.4870318330492468,
+      "grad_norm": 10.780561850834516,
+      "learning_rate": 0.00010905203080098913,
+      "loss": 1.3672,
+      "step": 10572
+    },
+    {
+      "epoch": 0.4870779011378818,
+      "grad_norm": 0.9769151480624142,
+      "learning_rate": 0.00010903717106983666,
+      "loss": 1.2812,
+      "step": 10573
+    },
+    {
+      "epoch": 0.4871239692265168,
+      "grad_norm": 1.0028295974424017,
+      "learning_rate": 0.00010902231113748705,
+      "loss": 1.467,
+      "step": 10574
+    },
+    {
+      "epoch": 0.4871700373151518,
+      "grad_norm": 0.9053121852392166,
+      "learning_rate": 0.00010900745100427114,
+      "loss": 1.4815,
+      "step": 10575
+    },
+    {
+      "epoch": 0.4872161054037868,
+      "grad_norm": 1.1106053782939018,
+      "learning_rate": 0.00010899259067051981,
+      "loss": 1.557,
+      "step": 10576
+    },
+    {
+      "epoch": 0.4872621734924218,
+      "grad_norm": 1.296652445621068,
+      "learning_rate": 0.00010897773013656385,
+      "loss": 1.3266,
+      "step": 10577
+    },
+    {
+      "epoch": 0.4873082415810568,
+      "grad_norm": 1.3150825994005744,
+      "learning_rate": 0.00010896286940273412,
+      "loss": 1.5138,
+      "step": 10578
+    },
+    {
+      "epoch": 0.4873543096696918,
+      "grad_norm": 1.1917068032804654,
+      "learning_rate": 0.00010894800846936146,
+      "loss": 1.3467,
+      "step": 10579
+    },
+    {
+      "epoch": 0.4874003777583268,
+      "grad_norm": 0.8880725883960047,
+      "learning_rate": 0.00010893314733677675,
+      "loss": 1.5655,
+      "step": 10580
+    },
+    {
+      "epoch": 0.4874464458469618,
+      "grad_norm": 1.1970690230727676,
+      "learning_rate": 0.00010891828600531083,
+      "loss": 1.4483,
+      "step": 10581
+    },
+    {
+      "epoch": 0.48749251393559684,
+      "grad_norm": 3.253283828991604,
+      "learning_rate": 0.00010890342447529455,
+      "loss": 1.1978,
+      "step": 10582
+    },
+    {
+      "epoch": 0.48753858202423184,
+      "grad_norm": 1.8837668415320292,
+      "learning_rate": 0.00010888856274705878,
+      "loss": 1.3219,
+      "step": 10583
+    },
+    {
+      "epoch": 0.48758465011286684,
+      "grad_norm": 4.374721685947259,
+      "learning_rate": 0.00010887370082093445,
+      "loss": 1.3751,
+      "step": 10584
+    },
+    {
+      "epoch": 0.48763071820150183,
+      "grad_norm": 1.2760851528196857,
+      "learning_rate": 0.00010885883869725237,
+      "loss": 1.5065,
+      "step": 10585
+    },
+    {
+      "epoch": 0.48767678629013683,
+      "grad_norm": 1.2662606590613372,
+      "learning_rate": 0.0001088439763763434,
+      "loss": 1.2099,
+      "step": 10586
+    },
+    {
+      "epoch": 0.48772285437877183,
+      "grad_norm": 1.007124339987443,
+      "learning_rate": 0.00010882911385853851,
+      "loss": 1.3891,
+      "step": 10587
+    },
+    {
+      "epoch": 0.48776892246740683,
+      "grad_norm": 0.9450881733014584,
+      "learning_rate": 0.0001088142511441685,
+      "loss": 1.4996,
+      "step": 10588
+    },
+    {
+      "epoch": 0.4878149905560418,
+      "grad_norm": 0.9829692411708713,
+      "learning_rate": 0.00010879938823356433,
+      "loss": 1.4309,
+      "step": 10589
+    },
+    {
+      "epoch": 0.4878610586446768,
+      "grad_norm": 0.7830584495569806,
+      "learning_rate": 0.00010878452512705684,
+      "loss": 1.4032,
+      "step": 10590
+    },
+    {
+      "epoch": 0.4879071267333118,
+      "grad_norm": 5.666961255202074,
+      "learning_rate": 0.00010876966182497697,
+      "loss": 1.0808,
+      "step": 10591
+    },
+    {
+      "epoch": 0.4879531948219468,
+      "grad_norm": 1.6077402592547039,
+      "learning_rate": 0.0001087547983276556,
+      "loss": 1.3508,
+      "step": 10592
+    },
+    {
+      "epoch": 0.4879992629105818,
+      "grad_norm": 0.9741543728368482,
+      "learning_rate": 0.00010873993463542369,
+      "loss": 1.303,
+      "step": 10593
+    },
+    {
+      "epoch": 0.4880453309992168,
+      "grad_norm": 2.673209707851917,
+      "learning_rate": 0.00010872507074861211,
+      "loss": 1.3568,
+      "step": 10594
+    },
+    {
+      "epoch": 0.4880913990878518,
+      "grad_norm": 1.4632924784140677,
+      "learning_rate": 0.0001087102066675518,
+      "loss": 1.2521,
+      "step": 10595
+    },
+    {
+      "epoch": 0.48813746717648687,
+      "grad_norm": 6.987522522031504,
+      "learning_rate": 0.00010869534239257366,
+      "loss": 1.1273,
+      "step": 10596
+    },
+    {
+      "epoch": 0.48818353526512187,
+      "grad_norm": 1.186305214110043,
+      "learning_rate": 0.00010868047792400862,
+      "loss": 1.1004,
+      "step": 10597
+    },
+    {
+      "epoch": 0.48822960335375687,
+      "grad_norm": 1.3168131598148256,
+      "learning_rate": 0.00010866561326218765,
+      "loss": 1.2982,
+      "step": 10598
+    },
+    {
+      "epoch": 0.48827567144239187,
+      "grad_norm": 1.1199520996203856,
+      "learning_rate": 0.00010865074840744162,
+      "loss": 1.515,
+      "step": 10599
+    },
+    {
+      "epoch": 0.48832173953102687,
+      "grad_norm": 1.105564943483963,
+      "learning_rate": 0.00010863588336010156,
+      "loss": 1.4734,
+      "step": 10600
+    },
+    {
+      "epoch": 0.48836780761966186,
+      "grad_norm": 1.765774706046889,
+      "learning_rate": 0.00010862101812049832,
+      "loss": 1.5553,
+      "step": 10601
+    },
+    {
+      "epoch": 0.48841387570829686,
+      "grad_norm": 1.4059263030377398,
+      "learning_rate": 0.0001086061526889629,
+      "loss": 1.3981,
+      "step": 10602
+    },
+    {
+      "epoch": 0.48845994379693186,
+      "grad_norm": 3.2371337006262952,
+      "learning_rate": 0.00010859128706582626,
+      "loss": 1.4339,
+      "step": 10603
+    },
+    {
+      "epoch": 0.48850601188556686,
+      "grad_norm": 1.6269188332554985,
+      "learning_rate": 0.00010857642125141938,
+      "loss": 1.4388,
+      "step": 10604
+    },
+    {
+      "epoch": 0.48855207997420186,
+      "grad_norm": 0.9249669381138047,
+      "learning_rate": 0.00010856155524607312,
+      "loss": 1.4588,
+      "step": 10605
+    },
+    {
+      "epoch": 0.48859814806283686,
+      "grad_norm": 1.2245895004530938,
+      "learning_rate": 0.00010854668905011856,
+      "loss": 1.2504,
+      "step": 10606
+    },
+    {
+      "epoch": 0.48864421615147186,
+      "grad_norm": 1.4881327123516035,
+      "learning_rate": 0.00010853182266388659,
+      "loss": 1.2753,
+      "step": 10607
+    },
+    {
+      "epoch": 0.48869028424010685,
+      "grad_norm": 2.9675958738212063,
+      "learning_rate": 0.00010851695608770825,
+      "loss": 1.2615,
+      "step": 10608
+    },
+    {
+      "epoch": 0.4887363523287419,
+      "grad_norm": 0.9552683697376498,
+      "learning_rate": 0.00010850208932191443,
+      "loss": 1.2507,
+      "step": 10609
+    },
+    {
+      "epoch": 0.4887824204173769,
+      "grad_norm": 1.0708901707436629,
+      "learning_rate": 0.00010848722236683621,
+      "loss": 1.4539,
+      "step": 10610
+    },
+    {
+      "epoch": 0.4888284885060119,
+      "grad_norm": 0.9099594083666844,
+      "learning_rate": 0.00010847235522280454,
+      "loss": 1.5221,
+      "step": 10611
+    },
+    {
+      "epoch": 0.4888745565946469,
+      "grad_norm": 1.2222206818912893,
+      "learning_rate": 0.0001084574878901504,
+      "loss": 1.2358,
+      "step": 10612
+    },
+    {
+      "epoch": 0.4889206246832819,
+      "grad_norm": 1.8066560654569135,
+      "learning_rate": 0.00010844262036920481,
+      "loss": 1.1034,
+      "step": 10613
+    },
+    {
+      "epoch": 0.4889666927719169,
+      "grad_norm": 0.9720111700598064,
+      "learning_rate": 0.00010842775266029872,
+      "loss": 1.8284,
+      "step": 10614
+    },
+    {
+      "epoch": 0.4890127608605519,
+      "grad_norm": 2.330540967723197,
+      "learning_rate": 0.00010841288476376321,
+      "loss": 1.1844,
+      "step": 10615
+    },
+    {
+      "epoch": 0.4890588289491869,
+      "grad_norm": 1.3999632246822156,
+      "learning_rate": 0.00010839801667992922,
+      "loss": 1.178,
+      "step": 10616
+    },
+    {
+      "epoch": 0.4891048970378219,
+      "grad_norm": 1.4896047416163778,
+      "learning_rate": 0.00010838314840912778,
+      "loss": 1.429,
+      "step": 10617
+    },
+    {
+      "epoch": 0.4891509651264569,
+      "grad_norm": 1.7587731957756696,
+      "learning_rate": 0.00010836827995168993,
+      "loss": 1.1467,
+      "step": 10618
+    },
+    {
+      "epoch": 0.4891970332150919,
+      "grad_norm": 1.517739380339018,
+      "learning_rate": 0.00010835341130794667,
+      "loss": 1.3861,
+      "step": 10619
+    },
+    {
+      "epoch": 0.4892431013037269,
+      "grad_norm": 0.9029389178592538,
+      "learning_rate": 0.00010833854247822901,
+      "loss": 1.1688,
+      "step": 10620
+    },
+    {
+      "epoch": 0.4892891693923619,
+      "grad_norm": 0.9317958961665607,
+      "learning_rate": 0.00010832367346286802,
+      "loss": 1.5522,
+      "step": 10621
+    },
+    {
+      "epoch": 0.4893352374809969,
+      "grad_norm": 0.861465426013367,
+      "learning_rate": 0.00010830880426219475,
+      "loss": 1.2831,
+      "step": 10622
+    },
+    {
+      "epoch": 0.48938130556963194,
+      "grad_norm": 2.9925165801533757,
+      "learning_rate": 0.00010829393487654014,
+      "loss": 1.2011,
+      "step": 10623
+    },
+    {
+      "epoch": 0.48942737365826694,
+      "grad_norm": 2.11703556509778,
+      "learning_rate": 0.00010827906530623531,
+      "loss": 1.2249,
+      "step": 10624
+    },
+    {
+      "epoch": 0.48947344174690194,
+      "grad_norm": 2.0779161423502353,
+      "learning_rate": 0.00010826419555161127,
+      "loss": 1.5341,
+      "step": 10625
+    },
+    {
+      "epoch": 0.48951950983553694,
+      "grad_norm": 1.3269915906524767,
+      "learning_rate": 0.00010824932561299911,
+      "loss": 1.4297,
+      "step": 10626
+    },
+    {
+      "epoch": 0.48956557792417194,
+      "grad_norm": 2.2204245981074737,
+      "learning_rate": 0.00010823445549072986,
+      "loss": 1.387,
+      "step": 10627
+    },
+    {
+      "epoch": 0.48961164601280693,
+      "grad_norm": 1.5373814673695714,
+      "learning_rate": 0.00010821958518513458,
+      "loss": 1.2144,
+      "step": 10628
+    },
+    {
+      "epoch": 0.48965771410144193,
+      "grad_norm": 2.12508664759872,
+      "learning_rate": 0.00010820471469654429,
+      "loss": 1.6292,
+      "step": 10629
+    },
+    {
+      "epoch": 0.48970378219007693,
+      "grad_norm": 3.231595376774529,
+      "learning_rate": 0.00010818984402529016,
+      "loss": 1.5241,
+      "step": 10630
+    },
+    {
+      "epoch": 0.48974985027871193,
+      "grad_norm": 1.2160093642442436,
+      "learning_rate": 0.00010817497317170316,
+      "loss": 1.4776,
+      "step": 10631
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 2.9225913857878343,
+      "learning_rate": 0.00010816010213611438,
+      "loss": 1.5834,
+      "step": 10632
+    },
+    {
+      "epoch": 0.4898419864559819,
+      "grad_norm": 2.0378750972213044,
+      "learning_rate": 0.00010814523091885496,
+      "loss": 1.2939,
+      "step": 10633
+    },
+    {
+      "epoch": 0.4898880545446169,
+      "grad_norm": 1.1145601549207156,
+      "learning_rate": 0.0001081303595202559,
+      "loss": 1.1542,
+      "step": 10634
+    },
+    {
+      "epoch": 0.4899341226332519,
+      "grad_norm": 0.8652331385399568,
+      "learning_rate": 0.00010811548794064837,
+      "loss": 1.2011,
+      "step": 10635
+    },
+    {
+      "epoch": 0.4899801907218869,
+      "grad_norm": 1.279749376428962,
+      "learning_rate": 0.00010810061618036335,
+      "loss": 1.3141,
+      "step": 10636
+    },
+    {
+      "epoch": 0.490026258810522,
+      "grad_norm": 0.770241883463002,
+      "learning_rate": 0.00010808574423973205,
+      "loss": 1.559,
+      "step": 10637
+    },
+    {
+      "epoch": 0.490072326899157,
+      "grad_norm": 1.271213467335016,
+      "learning_rate": 0.00010807087211908551,
+      "loss": 1.2172,
+      "step": 10638
+    },
+    {
+      "epoch": 0.49011839498779197,
+      "grad_norm": 1.9726586070697736,
+      "learning_rate": 0.00010805599981875485,
+      "loss": 1.1166,
+      "step": 10639
+    },
+    {
+      "epoch": 0.49016446307642697,
+      "grad_norm": 1.0979248662457486,
+      "learning_rate": 0.00010804112733907116,
+      "loss": 1.3,
+      "step": 10640
+    },
+    {
+      "epoch": 0.49021053116506197,
+      "grad_norm": 0.8813341948906459,
+      "learning_rate": 0.00010802625468036555,
+      "loss": 1.4024,
+      "step": 10641
+    },
+    {
+      "epoch": 0.49025659925369697,
+      "grad_norm": 1.166324365801717,
+      "learning_rate": 0.00010801138184296916,
+      "loss": 1.4332,
+      "step": 10642
+    },
+    {
+      "epoch": 0.49030266734233197,
+      "grad_norm": 1.2794795229423863,
+      "learning_rate": 0.00010799650882721306,
+      "loss": 1.3843,
+      "step": 10643
+    },
+    {
+      "epoch": 0.49034873543096696,
+      "grad_norm": 1.3306247654035008,
+      "learning_rate": 0.00010798163563342843,
+      "loss": 1.3559,
+      "step": 10644
+    },
+    {
+      "epoch": 0.49039480351960196,
+      "grad_norm": 1.2587802625594178,
+      "learning_rate": 0.00010796676226194636,
+      "loss": 1.5206,
+      "step": 10645
+    },
+    {
+      "epoch": 0.49044087160823696,
+      "grad_norm": 9.389456734584787,
+      "learning_rate": 0.00010795188871309798,
+      "loss": 1.4226,
+      "step": 10646
+    },
+    {
+      "epoch": 0.49048693969687196,
+      "grad_norm": 1.762872083533871,
+      "learning_rate": 0.00010793701498721443,
+      "loss": 1.2035,
+      "step": 10647
+    },
+    {
+      "epoch": 0.49053300778550696,
+      "grad_norm": 0.9579919110812525,
+      "learning_rate": 0.00010792214108462687,
+      "loss": 1.5758,
+      "step": 10648
+    },
+    {
+      "epoch": 0.49057907587414196,
+      "grad_norm": 1.2060379451589807,
+      "learning_rate": 0.0001079072670056664,
+      "loss": 1.3852,
+      "step": 10649
+    },
+    {
+      "epoch": 0.490625143962777,
+      "grad_norm": 1.1388688412759311,
+      "learning_rate": 0.0001078923927506642,
+      "loss": 1.2194,
+      "step": 10650
+    },
+    {
+      "epoch": 0.490671212051412,
+      "grad_norm": 3.3237545269568396,
+      "learning_rate": 0.00010787751831995141,
+      "loss": 1.3827,
+      "step": 10651
+    },
+    {
+      "epoch": 0.490717280140047,
+      "grad_norm": 1.539593342310283,
+      "learning_rate": 0.00010786264371385917,
+      "loss": 1.5744,
+      "step": 10652
+    },
+    {
+      "epoch": 0.490763348228682,
+      "grad_norm": 1.2595277619119674,
+      "learning_rate": 0.00010784776893271866,
+      "loss": 1.3766,
+      "step": 10653
+    },
+    {
+      "epoch": 0.490809416317317,
+      "grad_norm": 1.3646219086320528,
+      "learning_rate": 0.00010783289397686101,
+      "loss": 1.3154,
+      "step": 10654
+    },
+    {
+      "epoch": 0.490855484405952,
+      "grad_norm": 1.739722454151946,
+      "learning_rate": 0.00010781801884661744,
+      "loss": 1.4291,
+      "step": 10655
+    },
+    {
+      "epoch": 0.490901552494587,
+      "grad_norm": 1.1699494290168406,
+      "learning_rate": 0.00010780314354231906,
+      "loss": 1.3156,
+      "step": 10656
+    },
+    {
+      "epoch": 0.490947620583222,
+      "grad_norm": 1.11922515754867,
+      "learning_rate": 0.00010778826806429707,
+      "loss": 1.439,
+      "step": 10657
+    },
+    {
+      "epoch": 0.490993688671857,
+      "grad_norm": 1.4835985841984602,
+      "learning_rate": 0.00010777339241288266,
+      "loss": 1.4074,
+      "step": 10658
+    },
+    {
+      "epoch": 0.491039756760492,
+      "grad_norm": 0.9826292279792695,
+      "learning_rate": 0.00010775851658840701,
+      "loss": 1.7687,
+      "step": 10659
+    },
+    {
+      "epoch": 0.491085824849127,
+      "grad_norm": 1.17844374932938,
+      "learning_rate": 0.00010774364059120126,
+      "loss": 1.4793,
+      "step": 10660
+    },
+    {
+      "epoch": 0.491131892937762,
+      "grad_norm": 2.5953855601935434,
+      "learning_rate": 0.00010772876442159666,
+      "loss": 1.5,
+      "step": 10661
+    },
+    {
+      "epoch": 0.491177961026397,
+      "grad_norm": 1.3504892112007794,
+      "learning_rate": 0.00010771388807992433,
+      "loss": 1.6021,
+      "step": 10662
+    },
+    {
+      "epoch": 0.491224029115032,
+      "grad_norm": 1.232204508487574,
+      "learning_rate": 0.00010769901156651554,
+      "loss": 1.2247,
+      "step": 10663
+    },
+    {
+      "epoch": 0.49127009720366704,
+      "grad_norm": 0.8647165012235166,
+      "learning_rate": 0.00010768413488170144,
+      "loss": 1.485,
+      "step": 10664
+    },
+    {
+      "epoch": 0.49131616529230204,
+      "grad_norm": 1.1332957053972716,
+      "learning_rate": 0.00010766925802581324,
+      "loss": 1.1139,
+      "step": 10665
+    },
+    {
+      "epoch": 0.49136223338093704,
+      "grad_norm": 1.004623167420078,
+      "learning_rate": 0.00010765438099918219,
+      "loss": 1.4387,
+      "step": 10666
+    },
+    {
+      "epoch": 0.49140830146957204,
+      "grad_norm": 0.931608488937346,
+      "learning_rate": 0.00010763950380213946,
+      "loss": 1.4275,
+      "step": 10667
+    },
+    {
+      "epoch": 0.49145436955820704,
+      "grad_norm": 1.0223700821761506,
+      "learning_rate": 0.00010762462643501631,
+      "loss": 1.2864,
+      "step": 10668
+    },
+    {
+      "epoch": 0.49150043764684204,
+      "grad_norm": 1.7428489694264706,
+      "learning_rate": 0.00010760974889814388,
+      "loss": 1.0244,
+      "step": 10669
+    },
+    {
+      "epoch": 0.49154650573547704,
+      "grad_norm": 1.112644619686333,
+      "learning_rate": 0.00010759487119185348,
+      "loss": 1.2234,
+      "step": 10670
+    },
+    {
+      "epoch": 0.49159257382411203,
+      "grad_norm": 0.724222728335408,
+      "learning_rate": 0.00010757999331647625,
+      "loss": 1.4667,
+      "step": 10671
+    },
+    {
+      "epoch": 0.49163864191274703,
+      "grad_norm": 1.2987372905181442,
+      "learning_rate": 0.0001075651152723435,
+      "loss": 1.7431,
+      "step": 10672
+    },
+    {
+      "epoch": 0.49168471000138203,
+      "grad_norm": 1.243876946753801,
+      "learning_rate": 0.00010755023705978639,
+      "loss": 1.6663,
+      "step": 10673
+    },
+    {
+      "epoch": 0.49173077809001703,
+      "grad_norm": 2.348472273899485,
+      "learning_rate": 0.00010753535867913624,
+      "loss": 1.3592,
+      "step": 10674
+    },
+    {
+      "epoch": 0.491776846178652,
+      "grad_norm": 2.3856992089180036,
+      "learning_rate": 0.00010752048013072423,
+      "loss": 1.2221,
+      "step": 10675
+    },
+    {
+      "epoch": 0.491822914267287,
+      "grad_norm": 1.2951196437069872,
+      "learning_rate": 0.00010750560141488162,
+      "loss": 1.4048,
+      "step": 10676
+    },
+    {
+      "epoch": 0.4918689823559221,
+      "grad_norm": 1.6714200709555211,
+      "learning_rate": 0.00010749072253193968,
+      "loss": 1.4299,
+      "step": 10677
+    },
+    {
+      "epoch": 0.4919150504445571,
+      "grad_norm": 1.2167054644930375,
+      "learning_rate": 0.00010747584348222963,
+      "loss": 1.1857,
+      "step": 10678
+    },
+    {
+      "epoch": 0.4919611185331921,
+      "grad_norm": 3.8421228973400914,
+      "learning_rate": 0.00010746096426608277,
+      "loss": 1.6901,
+      "step": 10679
+    },
+    {
+      "epoch": 0.4920071866218271,
+      "grad_norm": 2.419997561316357,
+      "learning_rate": 0.0001074460848838303,
+      "loss": 1.2458,
+      "step": 10680
+    },
+    {
+      "epoch": 0.4920532547104621,
+      "grad_norm": 1.181115582966668,
+      "learning_rate": 0.00010743120533580353,
+      "loss": 1.4619,
+      "step": 10681
+    },
+    {
+      "epoch": 0.49209932279909707,
+      "grad_norm": 1.649828742088622,
+      "learning_rate": 0.00010741632562233372,
+      "loss": 1.5189,
+      "step": 10682
+    },
+    {
+      "epoch": 0.49214539088773207,
+      "grad_norm": 2.7758209052338083,
+      "learning_rate": 0.00010740144574375213,
+      "loss": 1.1985,
+      "step": 10683
+    },
+    {
+      "epoch": 0.49219145897636707,
+      "grad_norm": 1.8227400936780735,
+      "learning_rate": 0.00010738656570039003,
+      "loss": 1.5168,
+      "step": 10684
+    },
+    {
+      "epoch": 0.49223752706500207,
+      "grad_norm": 2.6976138838256194,
+      "learning_rate": 0.00010737168549257874,
+      "loss": 1.2016,
+      "step": 10685
+    },
+    {
+      "epoch": 0.49228359515363707,
+      "grad_norm": 1.8341244414722748,
+      "learning_rate": 0.00010735680512064952,
+      "loss": 1.324,
+      "step": 10686
+    },
+    {
+      "epoch": 0.49232966324227206,
+      "grad_norm": 1.2129154090012562,
+      "learning_rate": 0.0001073419245849336,
+      "loss": 1.3276,
+      "step": 10687
+    },
+    {
+      "epoch": 0.49237573133090706,
+      "grad_norm": 1.1324660900629804,
+      "learning_rate": 0.00010732704388576236,
+      "loss": 1.3532,
+      "step": 10688
+    },
+    {
+      "epoch": 0.49242179941954206,
+      "grad_norm": 1.1841783630626153,
+      "learning_rate": 0.000107312163023467,
+      "loss": 1.4511,
+      "step": 10689
+    },
+    {
+      "epoch": 0.49246786750817706,
+      "grad_norm": 1.3613596702120168,
+      "learning_rate": 0.00010729728199837891,
+      "loss": 1.2744,
+      "step": 10690
+    },
+    {
+      "epoch": 0.4925139355968121,
+      "grad_norm": 0.7282070156630046,
+      "learning_rate": 0.00010728240081082934,
+      "loss": 1.4901,
+      "step": 10691
+    },
+    {
+      "epoch": 0.4925600036854471,
+      "grad_norm": 1.6422574387051028,
+      "learning_rate": 0.0001072675194611496,
+      "loss": 1.405,
+      "step": 10692
+    },
+    {
+      "epoch": 0.4926060717740821,
+      "grad_norm": 1.8434914126895803,
+      "learning_rate": 0.000107252637949671,
+      "loss": 1.4808,
+      "step": 10693
+    },
+    {
+      "epoch": 0.4926521398627171,
+      "grad_norm": 1.086710438468606,
+      "learning_rate": 0.00010723775627672488,
+      "loss": 1.4256,
+      "step": 10694
+    },
+    {
+      "epoch": 0.4926982079513521,
+      "grad_norm": 0.9082883856546786,
+      "learning_rate": 0.00010722287444264249,
+      "loss": 1.7485,
+      "step": 10695
+    },
+    {
+      "epoch": 0.4927442760399871,
+      "grad_norm": 1.5566570905164983,
+      "learning_rate": 0.0001072079924477552,
+      "loss": 1.165,
+      "step": 10696
+    },
+    {
+      "epoch": 0.4927903441286221,
+      "grad_norm": 0.7739144725272229,
+      "learning_rate": 0.00010719311029239432,
+      "loss": 1.4485,
+      "step": 10697
+    },
+    {
+      "epoch": 0.4928364122172571,
+      "grad_norm": 1.1579389496464183,
+      "learning_rate": 0.00010717822797689115,
+      "loss": 1.7207,
+      "step": 10698
+    },
+    {
+      "epoch": 0.4928824803058921,
+      "grad_norm": 1.1714228822359194,
+      "learning_rate": 0.00010716334550157707,
+      "loss": 1.3064,
+      "step": 10699
+    },
+    {
+      "epoch": 0.4929285483945271,
+      "grad_norm": 2.198178935866248,
+      "learning_rate": 0.00010714846286678338,
+      "loss": 1.3181,
+      "step": 10700
+    },
+    {
+      "epoch": 0.4929746164831621,
+      "grad_norm": 0.7970317549508794,
+      "learning_rate": 0.0001071335800728414,
+      "loss": 1.1877,
+      "step": 10701
+    },
+    {
+      "epoch": 0.4930206845717971,
+      "grad_norm": 1.258493255810198,
+      "learning_rate": 0.00010711869712008249,
+      "loss": 1.3686,
+      "step": 10702
+    },
+    {
+      "epoch": 0.4930667526604321,
+      "grad_norm": 1.0038511531632786,
+      "learning_rate": 0.00010710381400883802,
+      "loss": 1.3097,
+      "step": 10703
+    },
+    {
+      "epoch": 0.49311282074906715,
+      "grad_norm": 1.5722594895349882,
+      "learning_rate": 0.00010708893073943933,
+      "loss": 1.2035,
+      "step": 10704
+    },
+    {
+      "epoch": 0.49315888883770215,
+      "grad_norm": 0.986514506515005,
+      "learning_rate": 0.00010707404731221773,
+      "loss": 1.3969,
+      "step": 10705
+    },
+    {
+      "epoch": 0.49320495692633715,
+      "grad_norm": 1.4641670530562834,
+      "learning_rate": 0.00010705916372750457,
+      "loss": 1.1027,
+      "step": 10706
+    },
+    {
+      "epoch": 0.49325102501497214,
+      "grad_norm": 2.046878265287035,
+      "learning_rate": 0.00010704427998563126,
+      "loss": 1.3987,
+      "step": 10707
+    },
+    {
+      "epoch": 0.49329709310360714,
+      "grad_norm": 6.467516896095321,
+      "learning_rate": 0.00010702939608692912,
+      "loss": 1.0045,
+      "step": 10708
+    },
+    {
+      "epoch": 0.49334316119224214,
+      "grad_norm": 1.2804807921920607,
+      "learning_rate": 0.00010701451203172952,
+      "loss": 1.4404,
+      "step": 10709
+    },
+    {
+      "epoch": 0.49338922928087714,
+      "grad_norm": 1.105996804255754,
+      "learning_rate": 0.00010699962782036388,
+      "loss": 1.2983,
+      "step": 10710
+    },
+    {
+      "epoch": 0.49343529736951214,
+      "grad_norm": 1.846566723352692,
+      "learning_rate": 0.00010698474345316347,
+      "loss": 1.6894,
+      "step": 10711
+    },
+    {
+      "epoch": 0.49348136545814714,
+      "grad_norm": 0.8251930596324639,
+      "learning_rate": 0.00010696985893045977,
+      "loss": 1.2465,
+      "step": 10712
+    },
+    {
+      "epoch": 0.49352743354678213,
+      "grad_norm": 0.8899224316126549,
+      "learning_rate": 0.00010695497425258412,
+      "loss": 1.3507,
+      "step": 10713
+    },
+    {
+      "epoch": 0.49357350163541713,
+      "grad_norm": 1.1001738155387637,
+      "learning_rate": 0.00010694008941986786,
+      "loss": 1.5448,
+      "step": 10714
+    },
+    {
+      "epoch": 0.49361956972405213,
+      "grad_norm": 0.8977669649935981,
+      "learning_rate": 0.00010692520443264243,
+      "loss": 1.4198,
+      "step": 10715
+    },
+    {
+      "epoch": 0.49366563781268713,
+      "grad_norm": 1.373409098504037,
+      "learning_rate": 0.00010691031929123918,
+      "loss": 1.4988,
+      "step": 10716
+    },
+    {
+      "epoch": 0.49371170590132213,
+      "grad_norm": 1.078339048588751,
+      "learning_rate": 0.00010689543399598951,
+      "loss": 1.2901,
+      "step": 10717
+    },
+    {
+      "epoch": 0.4937577739899572,
+      "grad_norm": 1.2920392475193123,
+      "learning_rate": 0.00010688054854722486,
+      "loss": 1.3973,
+      "step": 10718
+    },
+    {
+      "epoch": 0.4938038420785922,
+      "grad_norm": 0.8942795702786193,
+      "learning_rate": 0.00010686566294527656,
+      "loss": 1.4085,
+      "step": 10719
+    },
+    {
+      "epoch": 0.4938499101672272,
+      "grad_norm": 2.564231728068108,
+      "learning_rate": 0.00010685077719047607,
+      "loss": 1.2912,
+      "step": 10720
+    },
+    {
+      "epoch": 0.4938959782558622,
+      "grad_norm": 0.9540191776869287,
+      "learning_rate": 0.00010683589128315476,
+      "loss": 1.3218,
+      "step": 10721
+    },
+    {
+      "epoch": 0.4939420463444972,
+      "grad_norm": 2.2432886164743158,
+      "learning_rate": 0.00010682100522364407,
+      "loss": 1.3327,
+      "step": 10722
+    },
+    {
+      "epoch": 0.4939881144331322,
+      "grad_norm": 1.386711273715031,
+      "learning_rate": 0.00010680611901227537,
+      "loss": 1.413,
+      "step": 10723
+    },
+    {
+      "epoch": 0.4940341825217672,
+      "grad_norm": 0.9589093195505214,
+      "learning_rate": 0.00010679123264938011,
+      "loss": 1.292,
+      "step": 10724
+    },
+    {
+      "epoch": 0.49408025061040217,
+      "grad_norm": 0.9970638998023775,
+      "learning_rate": 0.00010677634613528971,
+      "loss": 1.4252,
+      "step": 10725
+    },
+    {
+      "epoch": 0.49412631869903717,
+      "grad_norm": 1.8949379636432264,
+      "learning_rate": 0.00010676145947033556,
+      "loss": 1.1757,
+      "step": 10726
+    },
+    {
+      "epoch": 0.49417238678767217,
+      "grad_norm": 1.9849559305906586,
+      "learning_rate": 0.00010674657265484913,
+      "loss": 1.5441,
+      "step": 10727
+    },
+    {
+      "epoch": 0.49421845487630717,
+      "grad_norm": 1.4106868986590386,
+      "learning_rate": 0.00010673168568916181,
+      "loss": 1.3423,
+      "step": 10728
+    },
+    {
+      "epoch": 0.49426452296494217,
+      "grad_norm": 4.302118351622759,
+      "learning_rate": 0.00010671679857360506,
+      "loss": 1.2682,
+      "step": 10729
+    },
+    {
+      "epoch": 0.49431059105357716,
+      "grad_norm": 2.2868899968221696,
+      "learning_rate": 0.00010670191130851031,
+      "loss": 1.2559,
+      "step": 10730
+    },
+    {
+      "epoch": 0.49435665914221216,
+      "grad_norm": 1.1665660372885325,
+      "learning_rate": 0.00010668702389420899,
+      "loss": 1.5461,
+      "step": 10731
+    },
+    {
+      "epoch": 0.4944027272308472,
+      "grad_norm": 1.359808286737838,
+      "learning_rate": 0.00010667213633103257,
+      "loss": 1.4366,
+      "step": 10732
+    },
+    {
+      "epoch": 0.4944487953194822,
+      "grad_norm": 1.2712195790609255,
+      "learning_rate": 0.00010665724861931244,
+      "loss": 1.2626,
+      "step": 10733
+    },
+    {
+      "epoch": 0.4944948634081172,
+      "grad_norm": 1.464794248591135,
+      "learning_rate": 0.0001066423607593801,
+      "loss": 1.258,
+      "step": 10734
+    },
+    {
+      "epoch": 0.4945409314967522,
+      "grad_norm": 2.6567839252802776,
+      "learning_rate": 0.00010662747275156699,
+      "loss": 1.5122,
+      "step": 10735
+    },
+    {
+      "epoch": 0.4945869995853872,
+      "grad_norm": 2.8869621755165373,
+      "learning_rate": 0.00010661258459620456,
+      "loss": 1.662,
+      "step": 10736
+    },
+    {
+      "epoch": 0.4946330676740222,
+      "grad_norm": 1.755612503737357,
+      "learning_rate": 0.00010659769629362426,
+      "loss": 1.4697,
+      "step": 10737
+    },
+    {
+      "epoch": 0.4946791357626572,
+      "grad_norm": 2.3804081448804633,
+      "learning_rate": 0.0001065828078441576,
+      "loss": 1.3908,
+      "step": 10738
+    },
+    {
+      "epoch": 0.4947252038512922,
+      "grad_norm": 0.7835666386685648,
+      "learning_rate": 0.00010656791924813596,
+      "loss": 1.6212,
+      "step": 10739
+    },
+    {
+      "epoch": 0.4947712719399272,
+      "grad_norm": 6.077947848295047,
+      "learning_rate": 0.0001065530305058909,
+      "loss": 1.1461,
+      "step": 10740
+    },
+    {
+      "epoch": 0.4948173400285622,
+      "grad_norm": 0.972419014154353,
+      "learning_rate": 0.00010653814161775384,
+      "loss": 1.6175,
+      "step": 10741
+    },
+    {
+      "epoch": 0.4948634081171972,
+      "grad_norm": 2.0268596311035285,
+      "learning_rate": 0.00010652325258405626,
+      "loss": 1.1851,
+      "step": 10742
+    },
+    {
+      "epoch": 0.4949094762058322,
+      "grad_norm": 2.511539363990448,
+      "learning_rate": 0.00010650836340512967,
+      "loss": 1.6095,
+      "step": 10743
+    },
+    {
+      "epoch": 0.4949555442944672,
+      "grad_norm": 0.9480692996186617,
+      "learning_rate": 0.00010649347408130547,
+      "loss": 1.3404,
+      "step": 10744
+    },
+    {
+      "epoch": 0.49500161238310225,
+      "grad_norm": 2.813651364045123,
+      "learning_rate": 0.00010647858461291527,
+      "loss": 1.2688,
+      "step": 10745
+    },
+    {
+      "epoch": 0.49504768047173725,
+      "grad_norm": 0.7107329270243296,
+      "learning_rate": 0.00010646369500029046,
+      "loss": 1.4091,
+      "step": 10746
+    },
+    {
+      "epoch": 0.49509374856037225,
+      "grad_norm": 0.8347558925340697,
+      "learning_rate": 0.00010644880524376257,
+      "loss": 1.6701,
+      "step": 10747
+    },
+    {
+      "epoch": 0.49513981664900725,
+      "grad_norm": 1.3829533693526719,
+      "learning_rate": 0.00010643391534366306,
+      "loss": 1.6417,
+      "step": 10748
+    },
+    {
+      "epoch": 0.49518588473764225,
+      "grad_norm": 2.605385432699396,
+      "learning_rate": 0.00010641902530032351,
+      "loss": 1.3006,
+      "step": 10749
+    },
+    {
+      "epoch": 0.49523195282627724,
+      "grad_norm": 1.7054706068250913,
+      "learning_rate": 0.00010640413511407534,
+      "loss": 1.4032,
+      "step": 10750
+    },
+    {
+      "epoch": 0.49527802091491224,
+      "grad_norm": 1.8426675590284254,
+      "learning_rate": 0.00010638924478525008,
+      "loss": 1.3672,
+      "step": 10751
+    },
+    {
+      "epoch": 0.49532408900354724,
+      "grad_norm": 0.9299673365816946,
+      "learning_rate": 0.00010637435431417926,
+      "loss": 1.4451,
+      "step": 10752
+    },
+    {
+      "epoch": 0.49537015709218224,
+      "grad_norm": 5.8562829870795765,
+      "learning_rate": 0.00010635946370119433,
+      "loss": 1.3762,
+      "step": 10753
+    },
+    {
+      "epoch": 0.49541622518081724,
+      "grad_norm": 0.8715550951527697,
+      "learning_rate": 0.00010634457294662689,
+      "loss": 1.7387,
+      "step": 10754
+    },
+    {
+      "epoch": 0.49546229326945224,
+      "grad_norm": 1.7321415293529605,
+      "learning_rate": 0.00010632968205080837,
+      "loss": 1.539,
+      "step": 10755
+    },
+    {
+      "epoch": 0.49550836135808723,
+      "grad_norm": 1.031125701111714,
+      "learning_rate": 0.00010631479101407038,
+      "loss": 1.4919,
+      "step": 10756
+    },
+    {
+      "epoch": 0.49555442944672223,
+      "grad_norm": 2.4718874913099573,
+      "learning_rate": 0.00010629989983674435,
+      "loss": 1.0401,
+      "step": 10757
+    },
+    {
+      "epoch": 0.49560049753535723,
+      "grad_norm": 1.548487817498335,
+      "learning_rate": 0.00010628500851916187,
+      "loss": 1.4911,
+      "step": 10758
+    },
+    {
+      "epoch": 0.4956465656239923,
+      "grad_norm": 1.9627062203486352,
+      "learning_rate": 0.00010627011706165449,
+      "loss": 1.3564,
+      "step": 10759
+    },
+    {
+      "epoch": 0.4956926337126273,
+      "grad_norm": 3.82927312503753,
+      "learning_rate": 0.00010625522546455366,
+      "loss": 1.142,
+      "step": 10760
+    },
+    {
+      "epoch": 0.4957387018012623,
+      "grad_norm": 4.312497720712324,
+      "learning_rate": 0.00010624033372819096,
+      "loss": 1.3565,
+      "step": 10761
+    },
+    {
+      "epoch": 0.4957847698898973,
+      "grad_norm": 2.09362880056475,
+      "learning_rate": 0.00010622544185289796,
+      "loss": 1.5512,
+      "step": 10762
+    },
+    {
+      "epoch": 0.4958308379785323,
+      "grad_norm": 2.279760277239824,
+      "learning_rate": 0.00010621054983900614,
+      "loss": 1.0627,
+      "step": 10763
+    },
+    {
+      "epoch": 0.4958769060671673,
+      "grad_norm": 3.243669813239603,
+      "learning_rate": 0.00010619565768684708,
+      "loss": 1.5052,
+      "step": 10764
+    },
+    {
+      "epoch": 0.4959229741558023,
+      "grad_norm": 1.0053673786758393,
+      "learning_rate": 0.00010618076539675233,
+      "loss": 1.2502,
+      "step": 10765
+    },
+    {
+      "epoch": 0.4959690422444373,
+      "grad_norm": 1.3085703847902346,
+      "learning_rate": 0.00010616587296905342,
+      "loss": 1.7495,
+      "step": 10766
+    },
+    {
+      "epoch": 0.4960151103330723,
+      "grad_norm": 2.2995067169352432,
+      "learning_rate": 0.00010615098040408195,
+      "loss": 1.1783,
+      "step": 10767
+    },
+    {
+      "epoch": 0.49606117842170727,
+      "grad_norm": 1.1812916429883225,
+      "learning_rate": 0.00010613608770216945,
+      "loss": 1.1893,
+      "step": 10768
+    },
+    {
+      "epoch": 0.49610724651034227,
+      "grad_norm": 1.4422563600088325,
+      "learning_rate": 0.00010612119486364747,
+      "loss": 1.301,
+      "step": 10769
+    },
+    {
+      "epoch": 0.49615331459897727,
+      "grad_norm": 2.4135203920244757,
+      "learning_rate": 0.00010610630188884756,
+      "loss": 1.3377,
+      "step": 10770
+    },
+    {
+      "epoch": 0.49619938268761227,
+      "grad_norm": 1.3592943010628724,
+      "learning_rate": 0.00010609140877810133,
+      "loss": 1.0298,
+      "step": 10771
+    },
+    {
+      "epoch": 0.4962454507762473,
+      "grad_norm": 1.281884110824213,
+      "learning_rate": 0.00010607651553174032,
+      "loss": 1.3694,
+      "step": 10772
+    },
+    {
+      "epoch": 0.4962915188648823,
+      "grad_norm": 1.4626831848516317,
+      "learning_rate": 0.00010606162215009612,
+      "loss": 1.4095,
+      "step": 10773
+    },
+    {
+      "epoch": 0.4963375869535173,
+      "grad_norm": 1.1308454283127398,
+      "learning_rate": 0.00010604672863350026,
+      "loss": 1.1899,
+      "step": 10774
+    },
+    {
+      "epoch": 0.4963836550421523,
+      "grad_norm": 4.564129290775602,
+      "learning_rate": 0.00010603183498228442,
+      "loss": 1.3549,
+      "step": 10775
+    },
+    {
+      "epoch": 0.4964297231307873,
+      "grad_norm": 1.024735049626362,
+      "learning_rate": 0.00010601694119678007,
+      "loss": 1.1652,
+      "step": 10776
+    },
+    {
+      "epoch": 0.4964757912194223,
+      "grad_norm": 1.17556010310311,
+      "learning_rate": 0.00010600204727731887,
+      "loss": 1.1771,
+      "step": 10777
+    },
+    {
+      "epoch": 0.4965218593080573,
+      "grad_norm": 1.0881206063500906,
+      "learning_rate": 0.00010598715322423236,
+      "loss": 1.6749,
+      "step": 10778
+    },
+    {
+      "epoch": 0.4965679273966923,
+      "grad_norm": 1.5155845632793359,
+      "learning_rate": 0.00010597225903785213,
+      "loss": 1.4653,
+      "step": 10779
+    },
+    {
+      "epoch": 0.4966139954853273,
+      "grad_norm": 2.5498376989430103,
+      "learning_rate": 0.00010595736471850983,
+      "loss": 1.3044,
+      "step": 10780
+    },
+    {
+      "epoch": 0.4966600635739623,
+      "grad_norm": 1.683535471509609,
+      "learning_rate": 0.00010594247026653698,
+      "loss": 1.3417,
+      "step": 10781
+    },
+    {
+      "epoch": 0.4967061316625973,
+      "grad_norm": 0.9143611788322403,
+      "learning_rate": 0.00010592757568226526,
+      "loss": 1.4614,
+      "step": 10782
+    },
+    {
+      "epoch": 0.4967521997512323,
+      "grad_norm": 0.9580033437185332,
+      "learning_rate": 0.0001059126809660262,
+      "loss": 1.352,
+      "step": 10783
+    },
+    {
+      "epoch": 0.4967982678398673,
+      "grad_norm": 1.8893531935203802,
+      "learning_rate": 0.00010589778611815147,
+      "loss": 1.4691,
+      "step": 10784
+    },
+    {
+      "epoch": 0.4968443359285023,
+      "grad_norm": 2.445672483771418,
+      "learning_rate": 0.0001058828911389726,
+      "loss": 1.3235,
+      "step": 10785
+    },
+    {
+      "epoch": 0.49689040401713735,
+      "grad_norm": 1.099172195097685,
+      "learning_rate": 0.00010586799602882126,
+      "loss": 1.3935,
+      "step": 10786
+    },
+    {
+      "epoch": 0.49693647210577235,
+      "grad_norm": 1.438840823515841,
+      "learning_rate": 0.00010585310078802908,
+      "loss": 1.4168,
+      "step": 10787
+    },
+    {
+      "epoch": 0.49698254019440735,
+      "grad_norm": 1.0501498659691038,
+      "learning_rate": 0.00010583820541692762,
+      "loss": 1.3571,
+      "step": 10788
+    },
+    {
+      "epoch": 0.49702860828304235,
+      "grad_norm": 0.9361458750866234,
+      "learning_rate": 0.00010582330991584855,
+      "loss": 1.4669,
+      "step": 10789
+    },
+    {
+      "epoch": 0.49707467637167735,
+      "grad_norm": 4.9500643926768975,
+      "learning_rate": 0.00010580841428512342,
+      "loss": 1.1847,
+      "step": 10790
+    },
+    {
+      "epoch": 0.49712074446031235,
+      "grad_norm": 2.5020146921303295,
+      "learning_rate": 0.00010579351852508396,
+      "loss": 1.321,
+      "step": 10791
+    },
+    {
+      "epoch": 0.49716681254894735,
+      "grad_norm": 2.7892978408077465,
+      "learning_rate": 0.0001057786226360617,
+      "loss": 1.5286,
+      "step": 10792
+    },
+    {
+      "epoch": 0.49721288063758234,
+      "grad_norm": 2.004908294666025,
+      "learning_rate": 0.00010576372661838835,
+      "loss": 1.2478,
+      "step": 10793
+    },
+    {
+      "epoch": 0.49725894872621734,
+      "grad_norm": 2.462141861488937,
+      "learning_rate": 0.00010574883047239547,
+      "loss": 1.6288,
+      "step": 10794
+    },
+    {
+      "epoch": 0.49730501681485234,
+      "grad_norm": 1.4663447097499749,
+      "learning_rate": 0.00010573393419841479,
+      "loss": 1.333,
+      "step": 10795
+    },
+    {
+      "epoch": 0.49735108490348734,
+      "grad_norm": 1.7813093427306685,
+      "learning_rate": 0.00010571903779677787,
+      "loss": 1.2452,
+      "step": 10796
+    },
+    {
+      "epoch": 0.49739715299212234,
+      "grad_norm": 1.6949394448205626,
+      "learning_rate": 0.00010570414126781638,
+      "loss": 1.2311,
+      "step": 10797
+    },
+    {
+      "epoch": 0.49744322108075734,
+      "grad_norm": 1.00882861274134,
+      "learning_rate": 0.00010568924461186196,
+      "loss": 1.2175,
+      "step": 10798
+    },
+    {
+      "epoch": 0.49748928916939233,
+      "grad_norm": 1.9122218078952509,
+      "learning_rate": 0.00010567434782924626,
+      "loss": 1.4096,
+      "step": 10799
+    },
+    {
+      "epoch": 0.4975353572580274,
+      "grad_norm": 1.5415339567408357,
+      "learning_rate": 0.00010565945092030096,
+      "loss": 1.3053,
+      "step": 10800
+    },
+    {
+      "epoch": 0.4975814253466624,
+      "grad_norm": 1.1869046797069822,
+      "learning_rate": 0.00010564455388535766,
+      "loss": 1.4476,
+      "step": 10801
+    },
+    {
+      "epoch": 0.4976274934352974,
+      "grad_norm": 0.9480960790097642,
+      "learning_rate": 0.00010562965672474807,
+      "loss": 1.3353,
+      "step": 10802
+    },
+    {
+      "epoch": 0.4976735615239324,
+      "grad_norm": 1.3636307366762868,
+      "learning_rate": 0.00010561475943880379,
+      "loss": 1.2935,
+      "step": 10803
+    },
+    {
+      "epoch": 0.4977196296125674,
+      "grad_norm": 1.818799305514569,
+      "learning_rate": 0.00010559986202785654,
+      "loss": 1.4504,
+      "step": 10804
+    },
+    {
+      "epoch": 0.4977656977012024,
+      "grad_norm": 1.174804198325976,
+      "learning_rate": 0.00010558496449223799,
+      "loss": 1.2884,
+      "step": 10805
+    },
+    {
+      "epoch": 0.4978117657898374,
+      "grad_norm": 0.8517270387555259,
+      "learning_rate": 0.00010557006683227976,
+      "loss": 1.3143,
+      "step": 10806
+    },
+    {
+      "epoch": 0.4978578338784724,
+      "grad_norm": 3.89104218402183,
+      "learning_rate": 0.00010555516904831353,
+      "loss": 1.0875,
+      "step": 10807
+    },
+    {
+      "epoch": 0.4979039019671074,
+      "grad_norm": 0.9105757830447692,
+      "learning_rate": 0.000105540271140671,
+      "loss": 1.6031,
+      "step": 10808
+    },
+    {
+      "epoch": 0.4979499700557424,
+      "grad_norm": 1.460201466474485,
+      "learning_rate": 0.00010552537310968382,
+      "loss": 1.3702,
+      "step": 10809
+    },
+    {
+      "epoch": 0.4979960381443774,
+      "grad_norm": 0.957782799835104,
+      "learning_rate": 0.0001055104749556837,
+      "loss": 1.5857,
+      "step": 10810
+    },
+    {
+      "epoch": 0.49804210623301237,
+      "grad_norm": 2.915049828966653,
+      "learning_rate": 0.0001054955766790023,
+      "loss": 1.0534,
+      "step": 10811
+    },
+    {
+      "epoch": 0.49808817432164737,
+      "grad_norm": 1.9583819999210348,
+      "learning_rate": 0.00010548067827997128,
+      "loss": 1.3139,
+      "step": 10812
+    },
+    {
+      "epoch": 0.4981342424102824,
+      "grad_norm": 1.2158543952175982,
+      "learning_rate": 0.00010546577975892238,
+      "loss": 1.3283,
+      "step": 10813
+    },
+    {
+      "epoch": 0.4981803104989174,
+      "grad_norm": 1.0126777395696398,
+      "learning_rate": 0.00010545088111618728,
+      "loss": 1.343,
+      "step": 10814
+    },
+    {
+      "epoch": 0.4982263785875524,
+      "grad_norm": 2.4021254199343147,
+      "learning_rate": 0.00010543598235209764,
+      "loss": 1.6498,
+      "step": 10815
+    },
+    {
+      "epoch": 0.4982724466761874,
+      "grad_norm": 0.7648894107931913,
+      "learning_rate": 0.00010542108346698519,
+      "loss": 1.6167,
+      "step": 10816
+    },
+    {
+      "epoch": 0.4983185147648224,
+      "grad_norm": 1.0295333060997909,
+      "learning_rate": 0.00010540618446118157,
+      "loss": 1.3772,
+      "step": 10817
+    },
+    {
+      "epoch": 0.4983645828534574,
+      "grad_norm": 0.8420543914660421,
+      "learning_rate": 0.00010539128533501856,
+      "loss": 1.7334,
+      "step": 10818
+    },
+    {
+      "epoch": 0.4984106509420924,
+      "grad_norm": 1.396716800177747,
+      "learning_rate": 0.00010537638608882782,
+      "loss": 1.6976,
+      "step": 10819
+    },
+    {
+      "epoch": 0.4984567190307274,
+      "grad_norm": 1.9824118741920005,
+      "learning_rate": 0.00010536148672294104,
+      "loss": 1.3123,
+      "step": 10820
+    },
+    {
+      "epoch": 0.4985027871193624,
+      "grad_norm": 1.4165547041890485,
+      "learning_rate": 0.00010534658723768996,
+      "loss": 1.5658,
+      "step": 10821
+    },
+    {
+      "epoch": 0.4985488552079974,
+      "grad_norm": 3.925507577755055,
+      "learning_rate": 0.00010533168763340631,
+      "loss": 1.2857,
+      "step": 10822
+    },
+    {
+      "epoch": 0.4985949232966324,
+      "grad_norm": 1.6189888433079942,
+      "learning_rate": 0.00010531678791042177,
+      "loss": 1.0802,
+      "step": 10823
+    },
+    {
+      "epoch": 0.4986409913852674,
+      "grad_norm": 1.6772336343461915,
+      "learning_rate": 0.00010530188806906805,
+      "loss": 1.5495,
+      "step": 10824
+    },
+    {
+      "epoch": 0.4986870594739024,
+      "grad_norm": 1.2268268004040375,
+      "learning_rate": 0.00010528698810967688,
+      "loss": 1.4838,
+      "step": 10825
+    },
+    {
+      "epoch": 0.4987331275625374,
+      "grad_norm": 1.1654999173167435,
+      "learning_rate": 0.00010527208803258,
+      "loss": 1.1897,
+      "step": 10826
+    },
+    {
+      "epoch": 0.49877919565117246,
+      "grad_norm": 7.0981550851961535,
+      "learning_rate": 0.00010525718783810909,
+      "loss": 1.1309,
+      "step": 10827
+    },
+    {
+      "epoch": 0.49882526373980746,
+      "grad_norm": 11.17771388337106,
+      "learning_rate": 0.00010524228752659592,
+      "loss": 1.4901,
+      "step": 10828
+    },
+    {
+      "epoch": 0.49887133182844245,
+      "grad_norm": 0.868647846064967,
+      "learning_rate": 0.00010522738709837221,
+      "loss": 1.3675,
+      "step": 10829
+    },
+    {
+      "epoch": 0.49891739991707745,
+      "grad_norm": 1.4929311493575983,
+      "learning_rate": 0.00010521248655376972,
+      "loss": 1.4196,
+      "step": 10830
+    },
+    {
+      "epoch": 0.49896346800571245,
+      "grad_norm": 1.510198716373177,
+      "learning_rate": 0.0001051975858931201,
+      "loss": 1.456,
+      "step": 10831
+    },
+    {
+      "epoch": 0.49900953609434745,
+      "grad_norm": 0.9768849232740056,
+      "learning_rate": 0.00010518268511675518,
+      "loss": 1.3071,
+      "step": 10832
+    },
+    {
+      "epoch": 0.49905560418298245,
+      "grad_norm": 1.432143257658062,
+      "learning_rate": 0.00010516778422500666,
+      "loss": 1.1497,
+      "step": 10833
+    },
+    {
+      "epoch": 0.49910167227161745,
+      "grad_norm": 0.9235716535225604,
+      "learning_rate": 0.00010515288321820627,
+      "loss": 1.4645,
+      "step": 10834
+    },
+    {
+      "epoch": 0.49914774036025245,
+      "grad_norm": 0.8288716622698602,
+      "learning_rate": 0.00010513798209668577,
+      "loss": 1.433,
+      "step": 10835
+    },
+    {
+      "epoch": 0.49919380844888744,
+      "grad_norm": 1.4946965772992886,
+      "learning_rate": 0.00010512308086077691,
+      "loss": 1.363,
+      "step": 10836
+    },
+    {
+      "epoch": 0.49923987653752244,
+      "grad_norm": 2.488276244409353,
+      "learning_rate": 0.00010510817951081145,
+      "loss": 1.471,
+      "step": 10837
+    },
+    {
+      "epoch": 0.49928594462615744,
+      "grad_norm": 2.3002662328278505,
+      "learning_rate": 0.00010509327804712113,
+      "loss": 1.1637,
+      "step": 10838
+    },
+    {
+      "epoch": 0.49933201271479244,
+      "grad_norm": 1.1733574599052978,
+      "learning_rate": 0.0001050783764700377,
+      "loss": 1.3801,
+      "step": 10839
+    },
+    {
+      "epoch": 0.4993780808034275,
+      "grad_norm": 0.7801604716661282,
+      "learning_rate": 0.00010506347477989291,
+      "loss": 1.4933,
+      "step": 10840
+    },
+    {
+      "epoch": 0.4994241488920625,
+      "grad_norm": 1.2754942471815756,
+      "learning_rate": 0.00010504857297701858,
+      "loss": 1.4816,
+      "step": 10841
+    },
+    {
+      "epoch": 0.4994702169806975,
+      "grad_norm": 0.9053970857590876,
+      "learning_rate": 0.0001050336710617464,
+      "loss": 1.6497,
+      "step": 10842
+    },
+    {
+      "epoch": 0.4995162850693325,
+      "grad_norm": 5.6009710757814455,
+      "learning_rate": 0.00010501876903440815,
+      "loss": 1.2866,
+      "step": 10843
+    },
+    {
+      "epoch": 0.4995623531579675,
+      "grad_norm": 0.9455483377110037,
+      "learning_rate": 0.00010500386689533565,
+      "loss": 1.1215,
+      "step": 10844
+    },
+    {
+      "epoch": 0.4996084212466025,
+      "grad_norm": 0.9858596102594599,
+      "learning_rate": 0.0001049889646448606,
+      "loss": 1.5261,
+      "step": 10845
+    },
+    {
+      "epoch": 0.4996544893352375,
+      "grad_norm": 1.633255728247412,
+      "learning_rate": 0.00010497406228331483,
+      "loss": 1.425,
+      "step": 10846
+    },
+    {
+      "epoch": 0.4997005574238725,
+      "grad_norm": 0.8835095041417115,
+      "learning_rate": 0.00010495915981103006,
+      "loss": 1.1115,
+      "step": 10847
+    },
+    {
+      "epoch": 0.4997466255125075,
+      "grad_norm": 1.7419252789433242,
+      "learning_rate": 0.00010494425722833813,
+      "loss": 1.3022,
+      "step": 10848
+    },
+    {
+      "epoch": 0.4997926936011425,
+      "grad_norm": 1.0344120456094625,
+      "learning_rate": 0.00010492935453557078,
+      "loss": 1.5457,
+      "step": 10849
+    },
+    {
+      "epoch": 0.4998387616897775,
+      "grad_norm": 1.092326809465023,
+      "learning_rate": 0.00010491445173305983,
+      "loss": 1.3779,
+      "step": 10850
+    },
+    {
+      "epoch": 0.4998848297784125,
+      "grad_norm": 1.0961261577799422,
+      "learning_rate": 0.00010489954882113701,
+      "loss": 1.4015,
+      "step": 10851
+    },
+    {
+      "epoch": 0.4999308978670475,
+      "grad_norm": 1.4571632778741694,
+      "learning_rate": 0.00010488464580013413,
+      "loss": 1.3377,
+      "step": 10852
+    },
+    {
+      "epoch": 0.4999769659556825,
+      "grad_norm": 0.8607115745362839,
+      "learning_rate": 0.00010486974267038301,
+      "loss": 1.5368,
+      "step": 10853
+    },
+    {
+      "epoch": 0.5000230340443175,
+      "grad_norm": 1.2308616580981064,
+      "learning_rate": 0.00010485483943221538,
+      "loss": 1.391,
+      "step": 10854
+    },
+    {
+      "epoch": 0.5000691021329525,
+      "grad_norm": 1.0276364949850865,
+      "learning_rate": 0.00010483993608596311,
+      "loss": 1.2318,
+      "step": 10855
+    },
+    {
+      "epoch": 0.5001151702215875,
+      "grad_norm": 0.933193923794309,
+      "learning_rate": 0.00010482503263195795,
+      "loss": 1.271,
+      "step": 10856
+    },
+    {
+      "epoch": 0.5001612383102225,
+      "grad_norm": 0.9047063175872234,
+      "learning_rate": 0.00010481012907053174,
+      "loss": 1.3073,
+      "step": 10857
+    },
+    {
+      "epoch": 0.5002073063988575,
+      "grad_norm": 0.9810962997777762,
+      "learning_rate": 0.0001047952254020162,
+      "loss": 1.1412,
+      "step": 10858
+    },
+    {
+      "epoch": 0.5002533744874925,
+      "grad_norm": 1.2593868636258683,
+      "learning_rate": 0.00010478032162674323,
+      "loss": 1.3692,
+      "step": 10859
+    },
+    {
+      "epoch": 0.5002994425761275,
+      "grad_norm": 1.43501726680406,
+      "learning_rate": 0.0001047654177450446,
+      "loss": 1.4305,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5003455106647625,
+      "grad_norm": 1.6020776422394434,
+      "learning_rate": 0.00010475051375725211,
+      "loss": 1.5593,
+      "step": 10861
+    },
+    {
+      "epoch": 0.5003915787533976,
+      "grad_norm": 1.0097553791800735,
+      "learning_rate": 0.00010473560966369756,
+      "loss": 1.2208,
+      "step": 10862
+    },
+    {
+      "epoch": 0.5004376468420325,
+      "grad_norm": 1.445631960053288,
+      "learning_rate": 0.00010472070546471276,
+      "loss": 1.4814,
+      "step": 10863
+    },
+    {
+      "epoch": 0.5004837149306676,
+      "grad_norm": 1.2200100435416321,
+      "learning_rate": 0.00010470580116062959,
+      "loss": 1.2538,
+      "step": 10864
+    },
+    {
+      "epoch": 0.5005297830193025,
+      "grad_norm": 1.4419305086948182,
+      "learning_rate": 0.00010469089675177978,
+      "loss": 1.4213,
+      "step": 10865
+    },
+    {
+      "epoch": 0.5005758511079376,
+      "grad_norm": 0.8489953261803411,
+      "learning_rate": 0.00010467599223849525,
+      "loss": 1.4826,
+      "step": 10866
+    },
+    {
+      "epoch": 0.5006219191965725,
+      "grad_norm": 1.456065343367883,
+      "learning_rate": 0.00010466108762110774,
+      "loss": 1.2967,
+      "step": 10867
+    },
+    {
+      "epoch": 0.5006679872852076,
+      "grad_norm": 1.2074782408752436,
+      "learning_rate": 0.00010464618289994909,
+      "loss": 1.1451,
+      "step": 10868
+    },
+    {
+      "epoch": 0.5007140553738425,
+      "grad_norm": 1.4119704717643173,
+      "learning_rate": 0.00010463127807535118,
+      "loss": 1.4231,
+      "step": 10869
+    },
+    {
+      "epoch": 0.5007601234624776,
+      "grad_norm": 1.0806442095596835,
+      "learning_rate": 0.0001046163731476458,
+      "loss": 1.3567,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5008061915511125,
+      "grad_norm": 2.582838483564239,
+      "learning_rate": 0.00010460146811716478,
+      "loss": 1.6221,
+      "step": 10871
+    },
+    {
+      "epoch": 0.5008522596397476,
+      "grad_norm": 1.7577311268303237,
+      "learning_rate": 0.00010458656298423996,
+      "loss": 1.2779,
+      "step": 10872
+    },
+    {
+      "epoch": 0.5008983277283825,
+      "grad_norm": 1.311146160483957,
+      "learning_rate": 0.00010457165774920317,
+      "loss": 1.5002,
+      "step": 10873
+    },
+    {
+      "epoch": 0.5009443958170176,
+      "grad_norm": 1.2774482555353621,
+      "learning_rate": 0.00010455675241238627,
+      "loss": 1.2376,
+      "step": 10874
+    },
+    {
+      "epoch": 0.5009904639056526,
+      "grad_norm": 1.7797021084744034,
+      "learning_rate": 0.00010454184697412109,
+      "loss": 1.4675,
+      "step": 10875
+    },
+    {
+      "epoch": 0.5010365319942875,
+      "grad_norm": 1.040779734330837,
+      "learning_rate": 0.00010452694143473945,
+      "loss": 1.2955,
+      "step": 10876
+    },
+    {
+      "epoch": 0.5010826000829226,
+      "grad_norm": 0.7695084532736445,
+      "learning_rate": 0.00010451203579457324,
+      "loss": 1.3842,
+      "step": 10877
+    },
+    {
+      "epoch": 0.5011286681715575,
+      "grad_norm": 2.4462609351944784,
+      "learning_rate": 0.00010449713005395431,
+      "loss": 1.2908,
+      "step": 10878
+    },
+    {
+      "epoch": 0.5011747362601926,
+      "grad_norm": 1.4371332252107332,
+      "learning_rate": 0.0001044822242132145,
+      "loss": 1.0356,
+      "step": 10879
+    },
+    {
+      "epoch": 0.5012208043488275,
+      "grad_norm": 0.8406709357806929,
+      "learning_rate": 0.0001044673182726856,
+      "loss": 1.2962,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5012668724374626,
+      "grad_norm": 0.9305500458417906,
+      "learning_rate": 0.00010445241223269956,
+      "loss": 1.4397,
+      "step": 10881
+    },
+    {
+      "epoch": 0.5013129405260975,
+      "grad_norm": 1.3528683690703291,
+      "learning_rate": 0.00010443750609358818,
+      "loss": 1.3041,
+      "step": 10882
+    },
+    {
+      "epoch": 0.5013590086147326,
+      "grad_norm": 1.151732062123427,
+      "learning_rate": 0.00010442259985568335,
+      "loss": 1.5021,
+      "step": 10883
+    },
+    {
+      "epoch": 0.5014050767033675,
+      "grad_norm": 0.7944593025663012,
+      "learning_rate": 0.00010440769351931688,
+      "loss": 1.4413,
+      "step": 10884
+    },
+    {
+      "epoch": 0.5014511447920026,
+      "grad_norm": 1.049816388471166,
+      "learning_rate": 0.00010439278708482073,
+      "loss": 1.4601,
+      "step": 10885
+    },
+    {
+      "epoch": 0.5014972128806375,
+      "grad_norm": 1.1500571878637171,
+      "learning_rate": 0.00010437788055252664,
+      "loss": 1.5439,
+      "step": 10886
+    },
+    {
+      "epoch": 0.5015432809692726,
+      "grad_norm": 0.9943866810668721,
+      "learning_rate": 0.0001043629739227666,
+      "loss": 1.3035,
+      "step": 10887
+    },
+    {
+      "epoch": 0.5015893490579076,
+      "grad_norm": 1.2700890292230143,
+      "learning_rate": 0.0001043480671958724,
+      "loss": 1.5457,
+      "step": 10888
+    },
+    {
+      "epoch": 0.5016354171465426,
+      "grad_norm": 1.0493949952243424,
+      "learning_rate": 0.00010433316037217593,
+      "loss": 1.3658,
+      "step": 10889
+    },
+    {
+      "epoch": 0.5016814852351776,
+      "grad_norm": 0.9041742573640602,
+      "learning_rate": 0.00010431825345200912,
+      "loss": 1.3157,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5017275533238126,
+      "grad_norm": 1.1057654600256868,
+      "learning_rate": 0.00010430334643570374,
+      "loss": 0.8657,
+      "step": 10891
+    },
+    {
+      "epoch": 0.5017736214124476,
+      "grad_norm": 1.8812954187871391,
+      "learning_rate": 0.00010428843932359178,
+      "loss": 1.5161,
+      "step": 10892
+    },
+    {
+      "epoch": 0.5018196895010826,
+      "grad_norm": 3.1277435669562794,
+      "learning_rate": 0.00010427353211600504,
+      "loss": 1.1269,
+      "step": 10893
+    },
+    {
+      "epoch": 0.5018657575897176,
+      "grad_norm": 1.3942079104029155,
+      "learning_rate": 0.00010425862481327546,
+      "loss": 1.452,
+      "step": 10894
+    },
+    {
+      "epoch": 0.5019118256783526,
+      "grad_norm": 0.8487826687687099,
+      "learning_rate": 0.00010424371741573486,
+      "loss": 1.3655,
+      "step": 10895
+    },
+    {
+      "epoch": 0.5019578937669876,
+      "grad_norm": 3.4519411760690266,
+      "learning_rate": 0.00010422880992371522,
+      "loss": 1.3536,
+      "step": 10896
+    },
+    {
+      "epoch": 0.5020039618556226,
+      "grad_norm": 1.070477045494276,
+      "learning_rate": 0.00010421390233754835,
+      "loss": 1.4131,
+      "step": 10897
+    },
+    {
+      "epoch": 0.5020500299442576,
+      "grad_norm": 10.10504944464806,
+      "learning_rate": 0.00010419899465756617,
+      "loss": 1.2839,
+      "step": 10898
+    },
+    {
+      "epoch": 0.5020960980328926,
+      "grad_norm": 0.8845394114605517,
+      "learning_rate": 0.0001041840868841006,
+      "loss": 1.1896,
+      "step": 10899
+    },
+    {
+      "epoch": 0.5021421661215276,
+      "grad_norm": 4.943903906665891,
+      "learning_rate": 0.00010416917901748347,
+      "loss": 1.6225,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5021882342101626,
+      "grad_norm": 0.7071306697774985,
+      "learning_rate": 0.00010415427105804674,
+      "loss": 1.5983,
+      "step": 10901
+    },
+    {
+      "epoch": 0.5022343022987976,
+      "grad_norm": 1.37693383840533,
+      "learning_rate": 0.00010413936300612226,
+      "loss": 1.3898,
+      "step": 10902
+    },
+    {
+      "epoch": 0.5022803703874327,
+      "grad_norm": 1.2361667909064171,
+      "learning_rate": 0.00010412445486204199,
+      "loss": 1.4092,
+      "step": 10903
+    },
+    {
+      "epoch": 0.5023264384760676,
+      "grad_norm": 1.9987617455355355,
+      "learning_rate": 0.00010410954662613779,
+      "loss": 1.1894,
+      "step": 10904
+    },
+    {
+      "epoch": 0.5023725065647027,
+      "grad_norm": 1.3011391455872796,
+      "learning_rate": 0.00010409463829874162,
+      "loss": 1.4237,
+      "step": 10905
+    },
+    {
+      "epoch": 0.5024185746533376,
+      "grad_norm": 1.084212778905891,
+      "learning_rate": 0.00010407972988018531,
+      "loss": 1.4513,
+      "step": 10906
+    },
+    {
+      "epoch": 0.5024646427419727,
+      "grad_norm": 0.7581813888416966,
+      "learning_rate": 0.0001040648213708008,
+      "loss": 1.5554,
+      "step": 10907
+    },
+    {
+      "epoch": 0.5025107108306076,
+      "grad_norm": 1.3847674802729122,
+      "learning_rate": 0.00010404991277092004,
+      "loss": 1.0172,
+      "step": 10908
+    },
+    {
+      "epoch": 0.5025567789192427,
+      "grad_norm": 0.9338897546813353,
+      "learning_rate": 0.0001040350040808749,
+      "loss": 1.5245,
+      "step": 10909
+    },
+    {
+      "epoch": 0.5026028470078776,
+      "grad_norm": 3.032731275603297,
+      "learning_rate": 0.00010402009530099733,
+      "loss": 1.2503,
+      "step": 10910
+    },
+    {
+      "epoch": 0.5026489150965127,
+      "grad_norm": 1.1549521669745653,
+      "learning_rate": 0.00010400518643161918,
+      "loss": 1.6069,
+      "step": 10911
+    },
+    {
+      "epoch": 0.5026949831851476,
+      "grad_norm": 2.2795116760043093,
+      "learning_rate": 0.00010399027747307249,
+      "loss": 1.1023,
+      "step": 10912
+    },
+    {
+      "epoch": 0.5027410512737827,
+      "grad_norm": 0.9156020208067484,
+      "learning_rate": 0.00010397536842568907,
+      "loss": 1.2581,
+      "step": 10913
+    },
+    {
+      "epoch": 0.5027871193624176,
+      "grad_norm": 0.8545901349201858,
+      "learning_rate": 0.00010396045928980092,
+      "loss": 1.2992,
+      "step": 10914
+    },
+    {
+      "epoch": 0.5028331874510527,
+      "grad_norm": 1.0062644425550853,
+      "learning_rate": 0.00010394555006573991,
+      "loss": 1.3678,
+      "step": 10915
+    },
+    {
+      "epoch": 0.5028792555396877,
+      "grad_norm": 0.9920312623996735,
+      "learning_rate": 0.000103930640753838,
+      "loss": 1.5023,
+      "step": 10916
+    },
+    {
+      "epoch": 0.5029253236283227,
+      "grad_norm": 1.7473412170084917,
+      "learning_rate": 0.00010391573135442712,
+      "loss": 1.3665,
+      "step": 10917
+    },
+    {
+      "epoch": 0.5029713917169577,
+      "grad_norm": 0.9757196139384737,
+      "learning_rate": 0.00010390082186783918,
+      "loss": 1.4479,
+      "step": 10918
+    },
+    {
+      "epoch": 0.5030174598055926,
+      "grad_norm": 3.7155705797399174,
+      "learning_rate": 0.00010388591229440614,
+      "loss": 0.9396,
+      "step": 10919
+    },
+    {
+      "epoch": 0.5030635278942277,
+      "grad_norm": 1.7955700288075536,
+      "learning_rate": 0.0001038710026344599,
+      "loss": 1.4764,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5031095959828626,
+      "grad_norm": 3.0084497263318695,
+      "learning_rate": 0.00010385609288833247,
+      "loss": 1.2644,
+      "step": 10921
+    },
+    {
+      "epoch": 0.5031556640714977,
+      "grad_norm": 1.9667481860751654,
+      "learning_rate": 0.0001038411830563557,
+      "loss": 1.2845,
+      "step": 10922
+    },
+    {
+      "epoch": 0.5032017321601326,
+      "grad_norm": 1.2961419403759085,
+      "learning_rate": 0.00010382627313886159,
+      "loss": 1.2068,
+      "step": 10923
+    },
+    {
+      "epoch": 0.5032478002487677,
+      "grad_norm": 1.4618861550849018,
+      "learning_rate": 0.0001038113631361821,
+      "loss": 1.172,
+      "step": 10924
+    },
+    {
+      "epoch": 0.5032938683374026,
+      "grad_norm": 2.5950445352148876,
+      "learning_rate": 0.00010379645304864914,
+      "loss": 1.1325,
+      "step": 10925
+    },
+    {
+      "epoch": 0.5033399364260377,
+      "grad_norm": 1.0227933405819094,
+      "learning_rate": 0.00010378154287659463,
+      "loss": 1.674,
+      "step": 10926
+    },
+    {
+      "epoch": 0.5033860045146726,
+      "grad_norm": 1.294471008221843,
+      "learning_rate": 0.00010376663262035055,
+      "loss": 1.4685,
+      "step": 10927
+    },
+    {
+      "epoch": 0.5034320726033077,
+      "grad_norm": 1.2305223697387135,
+      "learning_rate": 0.00010375172228024888,
+      "loss": 1.3133,
+      "step": 10928
+    },
+    {
+      "epoch": 0.5034781406919427,
+      "grad_norm": 1.8516312118131517,
+      "learning_rate": 0.00010373681185662155,
+      "loss": 1.5316,
+      "step": 10929
+    },
+    {
+      "epoch": 0.5035242087805777,
+      "grad_norm": 1.3735892056290622,
+      "learning_rate": 0.00010372190134980047,
+      "loss": 1.3768,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5035702768692127,
+      "grad_norm": 0.8538586556371759,
+      "learning_rate": 0.00010370699076011768,
+      "loss": 1.4817,
+      "step": 10931
+    },
+    {
+      "epoch": 0.5036163449578477,
+      "grad_norm": 1.9607124541241947,
+      "learning_rate": 0.00010369208008790505,
+      "loss": 1.15,
+      "step": 10932
+    },
+    {
+      "epoch": 0.5036624130464827,
+      "grad_norm": 1.0857989398540278,
+      "learning_rate": 0.00010367716933349463,
+      "loss": 1.2942,
+      "step": 10933
+    },
+    {
+      "epoch": 0.5037084811351177,
+      "grad_norm": 1.6244545041316547,
+      "learning_rate": 0.00010366225849721834,
+      "loss": 1.1536,
+      "step": 10934
+    },
+    {
+      "epoch": 0.5037545492237527,
+      "grad_norm": 1.279565843230893,
+      "learning_rate": 0.00010364734757940813,
+      "loss": 1.3545,
+      "step": 10935
+    },
+    {
+      "epoch": 0.5038006173123877,
+      "grad_norm": 1.483243535632786,
+      "learning_rate": 0.00010363243658039598,
+      "loss": 1.3824,
+      "step": 10936
+    },
+    {
+      "epoch": 0.5038466854010227,
+      "grad_norm": 1.306297744987059,
+      "learning_rate": 0.00010361752550051384,
+      "loss": 1.1175,
+      "step": 10937
+    },
+    {
+      "epoch": 0.5038927534896577,
+      "grad_norm": 5.6142750703052915,
+      "learning_rate": 0.00010360261434009373,
+      "loss": 1.3707,
+      "step": 10938
+    },
+    {
+      "epoch": 0.5039388215782927,
+      "grad_norm": 7.292035588955238,
+      "learning_rate": 0.00010358770309946757,
+      "loss": 1.1742,
+      "step": 10939
+    },
+    {
+      "epoch": 0.5039848896669277,
+      "grad_norm": 1.140606571787169,
+      "learning_rate": 0.00010357279177896737,
+      "loss": 1.2348,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5040309577555627,
+      "grad_norm": 1.3392865489082713,
+      "learning_rate": 0.00010355788037892506,
+      "loss": 1.5426,
+      "step": 10941
+    },
+    {
+      "epoch": 0.5040770258441978,
+      "grad_norm": 2.354829727746675,
+      "learning_rate": 0.00010354296889967267,
+      "loss": 1.5613,
+      "step": 10942
+    },
+    {
+      "epoch": 0.5041230939328327,
+      "grad_norm": 0.9659013040330351,
+      "learning_rate": 0.00010352805734154213,
+      "loss": 1.5802,
+      "step": 10943
+    },
+    {
+      "epoch": 0.5041691620214678,
+      "grad_norm": 1.2512474411845547,
+      "learning_rate": 0.00010351314570486546,
+      "loss": 1.1814,
+      "step": 10944
+    },
+    {
+      "epoch": 0.5042152301101027,
+      "grad_norm": 0.8734491682731617,
+      "learning_rate": 0.00010349823398997463,
+      "loss": 1.7572,
+      "step": 10945
+    },
+    {
+      "epoch": 0.5042612981987378,
+      "grad_norm": 1.8110478789675541,
+      "learning_rate": 0.00010348332219720159,
+      "loss": 1.4078,
+      "step": 10946
+    },
+    {
+      "epoch": 0.5043073662873727,
+      "grad_norm": 0.981954992347959,
+      "learning_rate": 0.00010346841032687837,
+      "loss": 1.2159,
+      "step": 10947
+    },
+    {
+      "epoch": 0.5043534343760078,
+      "grad_norm": 1.80455466267913,
+      "learning_rate": 0.00010345349837933693,
+      "loss": 1.5283,
+      "step": 10948
+    },
+    {
+      "epoch": 0.5043995024646427,
+      "grad_norm": 5.390070896482145,
+      "learning_rate": 0.00010343858635490931,
+      "loss": 1.3534,
+      "step": 10949
+    },
+    {
+      "epoch": 0.5044455705532778,
+      "grad_norm": 12.2682664626608,
+      "learning_rate": 0.0001034236742539274,
+      "loss": 1.3771,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5044916386419127,
+      "grad_norm": 1.2708818173773488,
+      "learning_rate": 0.00010340876207672332,
+      "loss": 1.5969,
+      "step": 10951
+    },
+    {
+      "epoch": 0.5045377067305478,
+      "grad_norm": 0.852131268624991,
+      "learning_rate": 0.00010339384982362897,
+      "loss": 1.3383,
+      "step": 10952
+    },
+    {
+      "epoch": 0.5045837748191827,
+      "grad_norm": 1.3365368135487679,
+      "learning_rate": 0.00010337893749497634,
+      "loss": 1.292,
+      "step": 10953
+    },
+    {
+      "epoch": 0.5046298429078178,
+      "grad_norm": 1.3545319919071726,
+      "learning_rate": 0.0001033640250910975,
+      "loss": 1.3926,
+      "step": 10954
+    },
+    {
+      "epoch": 0.5046759109964527,
+      "grad_norm": 1.1424421493409735,
+      "learning_rate": 0.00010334911261232439,
+      "loss": 1.2661,
+      "step": 10955
+    },
+    {
+      "epoch": 0.5047219790850878,
+      "grad_norm": 1.5071915404400582,
+      "learning_rate": 0.00010333420005898904,
+      "loss": 1.4802,
+      "step": 10956
+    },
+    {
+      "epoch": 0.5047680471737228,
+      "grad_norm": 1.0357704386075586,
+      "learning_rate": 0.00010331928743142343,
+      "loss": 1.2663,
+      "step": 10957
+    },
+    {
+      "epoch": 0.5048141152623578,
+      "grad_norm": 1.577883337993588,
+      "learning_rate": 0.0001033043747299596,
+      "loss": 1.3346,
+      "step": 10958
+    },
+    {
+      "epoch": 0.5048601833509928,
+      "grad_norm": 1.8357471945898096,
+      "learning_rate": 0.0001032894619549295,
+      "loss": 1.2756,
+      "step": 10959
+    },
+    {
+      "epoch": 0.5049062514396278,
+      "grad_norm": 3.4868806997479043,
+      "learning_rate": 0.00010327454910666521,
+      "loss": 1.0873,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5049523195282628,
+      "grad_norm": 2.0795410419821434,
+      "learning_rate": 0.00010325963618549866,
+      "loss": 1.5452,
+      "step": 10961
+    },
+    {
+      "epoch": 0.5049983876168977,
+      "grad_norm": 1.0142454174385025,
+      "learning_rate": 0.0001032447231917619,
+      "loss": 1.2861,
+      "step": 10962
+    },
+    {
+      "epoch": 0.5050444557055328,
+      "grad_norm": 1.6102940834999868,
+      "learning_rate": 0.00010322981012578697,
+      "loss": 1.0574,
+      "step": 10963
+    },
+    {
+      "epoch": 0.5050905237941677,
+      "grad_norm": 1.9294289535596596,
+      "learning_rate": 0.00010321489698790583,
+      "loss": 1.1565,
+      "step": 10964
+    },
+    {
+      "epoch": 0.5051365918828028,
+      "grad_norm": 1.44030370297198,
+      "learning_rate": 0.00010319998377845055,
+      "loss": 1.2415,
+      "step": 10965
+    },
+    {
+      "epoch": 0.5051826599714377,
+      "grad_norm": 2.4800483228858043,
+      "learning_rate": 0.00010318507049775308,
+      "loss": 1.1421,
+      "step": 10966
+    },
+    {
+      "epoch": 0.5052287280600728,
+      "grad_norm": 1.7585479574097937,
+      "learning_rate": 0.0001031701571461455,
+      "loss": 1.4388,
+      "step": 10967
+    },
+    {
+      "epoch": 0.5052747961487077,
+      "grad_norm": 1.1769811015358786,
+      "learning_rate": 0.0001031552437239598,
+      "loss": 1.346,
+      "step": 10968
+    },
+    {
+      "epoch": 0.5053208642373428,
+      "grad_norm": 1.7052504440884786,
+      "learning_rate": 0.00010314033023152803,
+      "loss": 1.0798,
+      "step": 10969
+    },
+    {
+      "epoch": 0.5053669323259778,
+      "grad_norm": 1.676063827261615,
+      "learning_rate": 0.00010312541666918219,
+      "loss": 1.5048,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5054130004146128,
+      "grad_norm": 1.4946010638561198,
+      "learning_rate": 0.00010311050303725429,
+      "loss": 1.3252,
+      "step": 10971
+    },
+    {
+      "epoch": 0.5054590685032478,
+      "grad_norm": 1.1553742833120093,
+      "learning_rate": 0.00010309558933607637,
+      "loss": 1.3551,
+      "step": 10972
+    },
+    {
+      "epoch": 0.5055051365918828,
+      "grad_norm": 1.1158514099618724,
+      "learning_rate": 0.00010308067556598047,
+      "loss": 1.4658,
+      "step": 10973
+    },
+    {
+      "epoch": 0.5055512046805178,
+      "grad_norm": 0.9456238942846615,
+      "learning_rate": 0.0001030657617272986,
+      "loss": 1.2936,
+      "step": 10974
+    },
+    {
+      "epoch": 0.5055972727691528,
+      "grad_norm": 2.5597175628881628,
+      "learning_rate": 0.00010305084782036278,
+      "loss": 1.4517,
+      "step": 10975
+    },
+    {
+      "epoch": 0.5056433408577878,
+      "grad_norm": 1.261556082365633,
+      "learning_rate": 0.00010303593384550512,
+      "loss": 1.4882,
+      "step": 10976
+    },
+    {
+      "epoch": 0.5056894089464228,
+      "grad_norm": 1.6672237933808065,
+      "learning_rate": 0.00010302101980305757,
+      "loss": 1.4956,
+      "step": 10977
+    },
+    {
+      "epoch": 0.5057354770350578,
+      "grad_norm": 3.2915633889185827,
+      "learning_rate": 0.00010300610569335218,
+      "loss": 1.1254,
+      "step": 10978
+    },
+    {
+      "epoch": 0.5057815451236928,
+      "grad_norm": 0.8655907751711394,
+      "learning_rate": 0.00010299119151672102,
+      "loss": 1.2583,
+      "step": 10979
+    },
+    {
+      "epoch": 0.5058276132123278,
+      "grad_norm": 2.6927927381317547,
+      "learning_rate": 0.00010297627727349613,
+      "loss": 1.1987,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5058736813009628,
+      "grad_norm": 1.6408667400012056,
+      "learning_rate": 0.0001029613629640095,
+      "loss": 1.3203,
+      "step": 10981
+    },
+    {
+      "epoch": 0.5059197493895978,
+      "grad_norm": 3.0679600515469887,
+      "learning_rate": 0.00010294644858859322,
+      "loss": 1.1491,
+      "step": 10982
+    },
+    {
+      "epoch": 0.5059658174782329,
+      "grad_norm": 1.1060500162393347,
+      "learning_rate": 0.00010293153414757928,
+      "loss": 1.3614,
+      "step": 10983
+    },
+    {
+      "epoch": 0.5060118855668678,
+      "grad_norm": 0.9759282335362569,
+      "learning_rate": 0.00010291661964129979,
+      "loss": 1.5412,
+      "step": 10984
+    },
+    {
+      "epoch": 0.5060579536555029,
+      "grad_norm": 1.1693487712720219,
+      "learning_rate": 0.00010290170507008674,
+      "loss": 1.195,
+      "step": 10985
+    },
+    {
+      "epoch": 0.5061040217441378,
+      "grad_norm": 1.6298561068168282,
+      "learning_rate": 0.00010288679043427222,
+      "loss": 1.3938,
+      "step": 10986
+    },
+    {
+      "epoch": 0.5061500898327729,
+      "grad_norm": 1.3203574581651016,
+      "learning_rate": 0.00010287187573418826,
+      "loss": 1.1815,
+      "step": 10987
+    },
+    {
+      "epoch": 0.5061961579214078,
+      "grad_norm": 0.9285179764751791,
+      "learning_rate": 0.0001028569609701669,
+      "loss": 0.8735,
+      "step": 10988
+    },
+    {
+      "epoch": 0.5062422260100429,
+      "grad_norm": 1.2071684797104563,
+      "learning_rate": 0.00010284204614254023,
+      "loss": 1.0333,
+      "step": 10989
+    },
+    {
+      "epoch": 0.5062882940986778,
+      "grad_norm": 1.1007060243017828,
+      "learning_rate": 0.00010282713125164025,
+      "loss": 1.5684,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5063343621873129,
+      "grad_norm": 1.3573355828925149,
+      "learning_rate": 0.00010281221629779905,
+      "loss": 0.8747,
+      "step": 10991
+    },
+    {
+      "epoch": 0.5063804302759478,
+      "grad_norm": 1.1756276124726985,
+      "learning_rate": 0.00010279730128134868,
+      "loss": 1.2913,
+      "step": 10992
+    },
+    {
+      "epoch": 0.5064264983645829,
+      "grad_norm": 1.0553984159569842,
+      "learning_rate": 0.0001027823862026212,
+      "loss": 1.316,
+      "step": 10993
+    },
+    {
+      "epoch": 0.5064725664532178,
+      "grad_norm": 2.69597215238365,
+      "learning_rate": 0.00010276747106194862,
+      "loss": 1.3069,
+      "step": 10994
+    },
+    {
+      "epoch": 0.5065186345418529,
+      "grad_norm": 1.4520546075013794,
+      "learning_rate": 0.00010275255585966308,
+      "loss": 1.2746,
+      "step": 10995
+    },
+    {
+      "epoch": 0.5065647026304878,
+      "grad_norm": 0.7617239052442102,
+      "learning_rate": 0.00010273764059609658,
+      "loss": 1.5431,
+      "step": 10996
+    },
+    {
+      "epoch": 0.5066107707191229,
+      "grad_norm": 0.8290655207473203,
+      "learning_rate": 0.00010272272527158122,
+      "loss": 1.4883,
+      "step": 10997
+    },
+    {
+      "epoch": 0.5066568388077579,
+      "grad_norm": 0.8704066105961741,
+      "learning_rate": 0.00010270780988644907,
+      "loss": 1.3925,
+      "step": 10998
+    },
+    {
+      "epoch": 0.5067029068963929,
+      "grad_norm": 1.299954410170227,
+      "learning_rate": 0.00010269289444103214,
+      "loss": 1.1966,
+      "step": 10999
+    },
+    {
+      "epoch": 0.5067489749850279,
+      "grad_norm": 1.5522395191755922,
+      "learning_rate": 0.00010267797893566256,
+      "loss": 1.423,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5067950430736629,
+      "grad_norm": 1.9260846555199376,
+      "learning_rate": 0.00010266306337067235,
+      "loss": 1.3563,
+      "step": 11001
+    },
+    {
+      "epoch": 0.5068411111622979,
+      "grad_norm": 0.9564420378471317,
+      "learning_rate": 0.00010264814774639362,
+      "loss": 1.4125,
+      "step": 11002
+    },
+    {
+      "epoch": 0.5068871792509329,
+      "grad_norm": 1.1228510544730232,
+      "learning_rate": 0.00010263323206315841,
+      "loss": 1.4922,
+      "step": 11003
+    },
+    {
+      "epoch": 0.5069332473395679,
+      "grad_norm": 0.9366297808524553,
+      "learning_rate": 0.00010261831632129881,
+      "loss": 0.9027,
+      "step": 11004
+    },
+    {
+      "epoch": 0.5069793154282028,
+      "grad_norm": 4.551141328069053,
+      "learning_rate": 0.00010260340052114685,
+      "loss": 1.3681,
+      "step": 11005
+    },
+    {
+      "epoch": 0.5070253835168379,
+      "grad_norm": 5.847430714753284,
+      "learning_rate": 0.0001025884846630347,
+      "loss": 1.5836,
+      "step": 11006
+    },
+    {
+      "epoch": 0.5070714516054728,
+      "grad_norm": 4.7628521183878005,
+      "learning_rate": 0.00010257356874729436,
+      "loss": 1.3244,
+      "step": 11007
+    },
+    {
+      "epoch": 0.5071175196941079,
+      "grad_norm": 0.9100418899535908,
+      "learning_rate": 0.0001025586527742579,
+      "loss": 1.4816,
+      "step": 11008
+    },
+    {
+      "epoch": 0.5071635877827428,
+      "grad_norm": 2.100167749164139,
+      "learning_rate": 0.00010254373674425744,
+      "loss": 1.2377,
+      "step": 11009
+    },
+    {
+      "epoch": 0.5072096558713779,
+      "grad_norm": 2.594082882958596,
+      "learning_rate": 0.00010252882065762503,
+      "loss": 1.0985,
+      "step": 11010
+    },
+    {
+      "epoch": 0.507255723960013,
+      "grad_norm": 1.478684709956316,
+      "learning_rate": 0.00010251390451469279,
+      "loss": 1.1044,
+      "step": 11011
+    },
+    {
+      "epoch": 0.5073017920486479,
+      "grad_norm": 1.4947336755982974,
+      "learning_rate": 0.00010249898831579274,
+      "loss": 1.1734,
+      "step": 11012
+    },
+    {
+      "epoch": 0.507347860137283,
+      "grad_norm": 0.8619156581770718,
+      "learning_rate": 0.00010248407206125703,
+      "loss": 1.4833,
+      "step": 11013
+    },
+    {
+      "epoch": 0.5073939282259179,
+      "grad_norm": 1.7849637016369615,
+      "learning_rate": 0.00010246915575141769,
+      "loss": 1.2684,
+      "step": 11014
+    },
+    {
+      "epoch": 0.5074399963145529,
+      "grad_norm": 1.1200946130151803,
+      "learning_rate": 0.00010245423938660687,
+      "loss": 1.3196,
+      "step": 11015
+    },
+    {
+      "epoch": 0.5074860644031879,
+      "grad_norm": 0.8469360465546214,
+      "learning_rate": 0.0001024393229671566,
+      "loss": 1.3639,
+      "step": 11016
+    },
+    {
+      "epoch": 0.5075321324918229,
+      "grad_norm": 1.084121920902777,
+      "learning_rate": 0.00010242440649339897,
+      "loss": 1.4862,
+      "step": 11017
+    },
+    {
+      "epoch": 0.5075782005804579,
+      "grad_norm": 0.8858949068585862,
+      "learning_rate": 0.0001024094899656661,
+      "loss": 1.3628,
+      "step": 11018
+    },
+    {
+      "epoch": 0.5076242686690929,
+      "grad_norm": 0.7625377005277245,
+      "learning_rate": 0.00010239457338429004,
+      "loss": 1.3132,
+      "step": 11019
+    },
+    {
+      "epoch": 0.5076703367577279,
+      "grad_norm": 3.215900331866927,
+      "learning_rate": 0.00010237965674960295,
+      "loss": 1.1272,
+      "step": 11020
+    },
+    {
+      "epoch": 0.5077164048463629,
+      "grad_norm": 1.4455265597784304,
+      "learning_rate": 0.00010236474006193686,
+      "loss": 1.4618,
+      "step": 11021
+    },
+    {
+      "epoch": 0.5077624729349979,
+      "grad_norm": 3.509681063078072,
+      "learning_rate": 0.00010234982332162391,
+      "loss": 1.218,
+      "step": 11022
+    },
+    {
+      "epoch": 0.5078085410236329,
+      "grad_norm": 1.0303030903613128,
+      "learning_rate": 0.00010233490652899613,
+      "loss": 1.136,
+      "step": 11023
+    },
+    {
+      "epoch": 0.507854609112268,
+      "grad_norm": 0.9818275438159005,
+      "learning_rate": 0.00010231998968438572,
+      "loss": 1.2393,
+      "step": 11024
+    },
+    {
+      "epoch": 0.5079006772009029,
+      "grad_norm": 0.8302867263551345,
+      "learning_rate": 0.00010230507278812467,
+      "loss": 0.959,
+      "step": 11025
+    },
+    {
+      "epoch": 0.507946745289538,
+      "grad_norm": 1.3682328968620978,
+      "learning_rate": 0.00010229015584054515,
+      "loss": 1.2243,
+      "step": 11026
+    },
+    {
+      "epoch": 0.5079928133781729,
+      "grad_norm": 1.4738264116210016,
+      "learning_rate": 0.00010227523884197923,
+      "loss": 1.1025,
+      "step": 11027
+    },
+    {
+      "epoch": 0.508038881466808,
+      "grad_norm": 0.9512253521592335,
+      "learning_rate": 0.00010226032179275901,
+      "loss": 1.3587,
+      "step": 11028
+    },
+    {
+      "epoch": 0.5080849495554429,
+      "grad_norm": 0.6912458151849536,
+      "learning_rate": 0.0001022454046932166,
+      "loss": 1.5451,
+      "step": 11029
+    },
+    {
+      "epoch": 0.508131017644078,
+      "grad_norm": 0.8663648234940161,
+      "learning_rate": 0.00010223048754368411,
+      "loss": 1.3179,
+      "step": 11030
+    },
+    {
+      "epoch": 0.5081770857327129,
+      "grad_norm": 1.1453596453181252,
+      "learning_rate": 0.00010221557034449362,
+      "loss": 1.1152,
+      "step": 11031
+    },
+    {
+      "epoch": 0.508223153821348,
+      "grad_norm": 1.1239863539320119,
+      "learning_rate": 0.00010220065309597729,
+      "loss": 1.0823,
+      "step": 11032
+    },
+    {
+      "epoch": 0.5082692219099829,
+      "grad_norm": 1.2594006988969968,
+      "learning_rate": 0.00010218573579846717,
+      "loss": 1.4038,
+      "step": 11033
+    },
+    {
+      "epoch": 0.508315289998618,
+      "grad_norm": 1.2449116719158737,
+      "learning_rate": 0.00010217081845229542,
+      "loss": 1.5309,
+      "step": 11034
+    },
+    {
+      "epoch": 0.5083613580872529,
+      "grad_norm": 1.3495399966149784,
+      "learning_rate": 0.0001021559010577941,
+      "loss": 1.3898,
+      "step": 11035
+    },
+    {
+      "epoch": 0.508407426175888,
+      "grad_norm": 0.7471926565834188,
+      "learning_rate": 0.00010214098361529531,
+      "loss": 1.5714,
+      "step": 11036
+    },
+    {
+      "epoch": 0.508453494264523,
+      "grad_norm": 1.2133545617608295,
+      "learning_rate": 0.00010212606612513125,
+      "loss": 1.2773,
+      "step": 11037
+    },
+    {
+      "epoch": 0.508499562353158,
+      "grad_norm": 1.0940620763254358,
+      "learning_rate": 0.00010211114858763394,
+      "loss": 1.4872,
+      "step": 11038
+    },
+    {
+      "epoch": 0.508545630441793,
+      "grad_norm": 2.503362521534598,
+      "learning_rate": 0.00010209623100313552,
+      "loss": 1.3037,
+      "step": 11039
+    },
+    {
+      "epoch": 0.508591698530428,
+      "grad_norm": 0.8779126973149022,
+      "learning_rate": 0.00010208131337196812,
+      "loss": 1.2586,
+      "step": 11040
+    },
+    {
+      "epoch": 0.508637766619063,
+      "grad_norm": 0.6708585251056943,
+      "learning_rate": 0.00010206639569446383,
+      "loss": 1.2325,
+      "step": 11041
+    },
+    {
+      "epoch": 0.508683834707698,
+      "grad_norm": 1.0846942670289617,
+      "learning_rate": 0.0001020514779709548,
+      "loss": 1.458,
+      "step": 11042
+    },
+    {
+      "epoch": 0.508729902796333,
+      "grad_norm": 1.229312529290487,
+      "learning_rate": 0.00010203656020177312,
+      "loss": 1.3597,
+      "step": 11043
+    },
+    {
+      "epoch": 0.508775970884968,
+      "grad_norm": 1.3477987586475701,
+      "learning_rate": 0.00010202164238725096,
+      "loss": 1.429,
+      "step": 11044
+    },
+    {
+      "epoch": 0.508822038973603,
+      "grad_norm": 1.2080213822931936,
+      "learning_rate": 0.00010200672452772036,
+      "loss": 1.3613,
+      "step": 11045
+    },
+    {
+      "epoch": 0.508868107062238,
+      "grad_norm": 1.586600201851058,
+      "learning_rate": 0.00010199180662351349,
+      "loss": 1.4136,
+      "step": 11046
+    },
+    {
+      "epoch": 0.508914175150873,
+      "grad_norm": 1.1265881309586754,
+      "learning_rate": 0.00010197688867496242,
+      "loss": 1.1783,
+      "step": 11047
+    },
+    {
+      "epoch": 0.508960243239508,
+      "grad_norm": 0.9329814975518897,
+      "learning_rate": 0.00010196197068239936,
+      "loss": 1.2532,
+      "step": 11048
+    },
+    {
+      "epoch": 0.509006311328143,
+      "grad_norm": 1.6440466483494367,
+      "learning_rate": 0.00010194705264615639,
+      "loss": 1.6162,
+      "step": 11049
+    },
+    {
+      "epoch": 0.509052379416778,
+      "grad_norm": 1.6640866994619283,
+      "learning_rate": 0.00010193213456656561,
+      "loss": 1.141,
+      "step": 11050
+    },
+    {
+      "epoch": 0.509098447505413,
+      "grad_norm": 1.1075342121778373,
+      "learning_rate": 0.00010191721644395917,
+      "loss": 1.255,
+      "step": 11051
+    },
+    {
+      "epoch": 0.509144515594048,
+      "grad_norm": 0.9450923457217506,
+      "learning_rate": 0.0001019022982786692,
+      "loss": 1.4718,
+      "step": 11052
+    },
+    {
+      "epoch": 0.509190583682683,
+      "grad_norm": 0.9015389679974368,
+      "learning_rate": 0.00010188738007102783,
+      "loss": 1.48,
+      "step": 11053
+    },
+    {
+      "epoch": 0.509236651771318,
+      "grad_norm": 1.1237704035559233,
+      "learning_rate": 0.00010187246182136716,
+      "loss": 1.2698,
+      "step": 11054
+    },
+    {
+      "epoch": 0.509282719859953,
+      "grad_norm": 1.248825336223233,
+      "learning_rate": 0.00010185754353001936,
+      "loss": 1.2827,
+      "step": 11055
+    },
+    {
+      "epoch": 0.509328787948588,
+      "grad_norm": 1.1850493144959209,
+      "learning_rate": 0.00010184262519731649,
+      "loss": 0.9979,
+      "step": 11056
+    },
+    {
+      "epoch": 0.509374856037223,
+      "grad_norm": 0.8122086572629267,
+      "learning_rate": 0.00010182770682359077,
+      "loss": 1.1291,
+      "step": 11057
+    },
+    {
+      "epoch": 0.509420924125858,
+      "grad_norm": 0.6737813717977192,
+      "learning_rate": 0.00010181278840917427,
+      "loss": 1.4796,
+      "step": 11058
+    },
+    {
+      "epoch": 0.509466992214493,
+      "grad_norm": 0.9519520271243161,
+      "learning_rate": 0.00010179786995439917,
+      "loss": 1.4994,
+      "step": 11059
+    },
+    {
+      "epoch": 0.509513060303128,
+      "grad_norm": 0.7410007286914254,
+      "learning_rate": 0.00010178295145959755,
+      "loss": 1.0625,
+      "step": 11060
+    },
+    {
+      "epoch": 0.509559128391763,
+      "grad_norm": 0.7480290942591118,
+      "learning_rate": 0.00010176803292510162,
+      "loss": 1.2812,
+      "step": 11061
+    },
+    {
+      "epoch": 0.509605196480398,
+      "grad_norm": 0.9308614429845079,
+      "learning_rate": 0.00010175311435124344,
+      "loss": 1.3754,
+      "step": 11062
+    },
+    {
+      "epoch": 0.509651264569033,
+      "grad_norm": 0.9934650729313752,
+      "learning_rate": 0.00010173819573835517,
+      "loss": 1.3944,
+      "step": 11063
+    },
+    {
+      "epoch": 0.509697332657668,
+      "grad_norm": 1.5955232724627684,
+      "learning_rate": 0.00010172327708676896,
+      "loss": 1.2128,
+      "step": 11064
+    },
+    {
+      "epoch": 0.5097434007463031,
+      "grad_norm": 0.9139698677042325,
+      "learning_rate": 0.00010170835839681691,
+      "loss": 1.1134,
+      "step": 11065
+    },
+    {
+      "epoch": 0.509789468834938,
+      "grad_norm": 5.048091233130882,
+      "learning_rate": 0.00010169343966883123,
+      "loss": 1.5485,
+      "step": 11066
+    },
+    {
+      "epoch": 0.5098355369235731,
+      "grad_norm": 0.8387719766685177,
+      "learning_rate": 0.00010167852090314401,
+      "loss": 1.2483,
+      "step": 11067
+    },
+    {
+      "epoch": 0.509881605012208,
+      "grad_norm": 1.2634900251896584,
+      "learning_rate": 0.00010166360210008739,
+      "loss": 1.3834,
+      "step": 11068
+    },
+    {
+      "epoch": 0.5099276731008431,
+      "grad_norm": 0.9204577239058069,
+      "learning_rate": 0.0001016486832599935,
+      "loss": 1.4108,
+      "step": 11069
+    },
+    {
+      "epoch": 0.509973741189478,
+      "grad_norm": 1.398882252227071,
+      "learning_rate": 0.00010163376438319458,
+      "loss": 1.4605,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5100198092781131,
+      "grad_norm": 0.9468080766144877,
+      "learning_rate": 0.00010161884547002262,
+      "loss": 1.5203,
+      "step": 11071
+    },
+    {
+      "epoch": 0.510065877366748,
+      "grad_norm": 1.2094413482868387,
+      "learning_rate": 0.0001016039265208099,
+      "loss": 1.1835,
+      "step": 11072
+    },
+    {
+      "epoch": 0.5101119454553831,
+      "grad_norm": 1.0785167220024698,
+      "learning_rate": 0.00010158900753588845,
+      "loss": 1.1823,
+      "step": 11073
+    },
+    {
+      "epoch": 0.510158013544018,
+      "grad_norm": 1.5329245734906647,
+      "learning_rate": 0.00010157408851559049,
+      "loss": 1.2783,
+      "step": 11074
+    },
+    {
+      "epoch": 0.5102040816326531,
+      "grad_norm": 1.3056400681852787,
+      "learning_rate": 0.00010155916946024817,
+      "loss": 1.468,
+      "step": 11075
+    },
+    {
+      "epoch": 0.510250149721288,
+      "grad_norm": 1.2859405984775973,
+      "learning_rate": 0.00010154425037019357,
+      "loss": 1.3935,
+      "step": 11076
+    },
+    {
+      "epoch": 0.5102962178099231,
+      "grad_norm": 1.3819765736026868,
+      "learning_rate": 0.00010152933124575894,
+      "loss": 1.1671,
+      "step": 11077
+    },
+    {
+      "epoch": 0.5103422858985581,
+      "grad_norm": 0.8439274760757265,
+      "learning_rate": 0.00010151441208727632,
+      "loss": 1.6064,
+      "step": 11078
+    },
+    {
+      "epoch": 0.5103883539871931,
+      "grad_norm": 1.3778843574430757,
+      "learning_rate": 0.00010149949289507798,
+      "loss": 1.2399,
+      "step": 11079
+    },
+    {
+      "epoch": 0.5104344220758281,
+      "grad_norm": 0.9074321460295658,
+      "learning_rate": 0.00010148457366949595,
+      "loss": 1.2853,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5104804901644631,
+      "grad_norm": 0.97582047734565,
+      "learning_rate": 0.00010146965441086244,
+      "loss": 1.6486,
+      "step": 11081
+    },
+    {
+      "epoch": 0.5105265582530981,
+      "grad_norm": 2.7074353138092215,
+      "learning_rate": 0.00010145473511950958,
+      "loss": 1.631,
+      "step": 11082
+    },
+    {
+      "epoch": 0.5105726263417331,
+      "grad_norm": 0.882081781427042,
+      "learning_rate": 0.00010143981579576958,
+      "loss": 1.3705,
+      "step": 11083
+    },
+    {
+      "epoch": 0.5106186944303681,
+      "grad_norm": 0.8288725344152774,
+      "learning_rate": 0.00010142489643997451,
+      "loss": 1.3997,
+      "step": 11084
+    },
+    {
+      "epoch": 0.5106647625190031,
+      "grad_norm": 2.6237139760791126,
+      "learning_rate": 0.00010140997705245658,
+      "loss": 1.5833,
+      "step": 11085
+    },
+    {
+      "epoch": 0.5107108306076381,
+      "grad_norm": 2.5420242411241736,
+      "learning_rate": 0.0001013950576335479,
+      "loss": 1.3717,
+      "step": 11086
+    },
+    {
+      "epoch": 0.510756898696273,
+      "grad_norm": 0.8396846126495552,
+      "learning_rate": 0.0001013801381835807,
+      "loss": 1.2082,
+      "step": 11087
+    },
+    {
+      "epoch": 0.5108029667849081,
+      "grad_norm": 0.8602665230064035,
+      "learning_rate": 0.00010136521870288709,
+      "loss": 1.2664,
+      "step": 11088
+    },
+    {
+      "epoch": 0.510849034873543,
+      "grad_norm": 1.0578916746714773,
+      "learning_rate": 0.0001013502991917992,
+      "loss": 1.4374,
+      "step": 11089
+    },
+    {
+      "epoch": 0.5108951029621781,
+      "grad_norm": 0.950768183974374,
+      "learning_rate": 0.00010133537965064921,
+      "loss": 1.5458,
+      "step": 11090
+    },
+    {
+      "epoch": 0.510941171050813,
+      "grad_norm": 0.7930285108871474,
+      "learning_rate": 0.00010132046007976927,
+      "loss": 1.4588,
+      "step": 11091
+    },
+    {
+      "epoch": 0.5109872391394481,
+      "grad_norm": 1.4282479170933904,
+      "learning_rate": 0.00010130554047949159,
+      "loss": 1.2861,
+      "step": 11092
+    },
+    {
+      "epoch": 0.5110333072280832,
+      "grad_norm": 0.8646975758011081,
+      "learning_rate": 0.00010129062085014827,
+      "loss": 1.6056,
+      "step": 11093
+    },
+    {
+      "epoch": 0.5110793753167181,
+      "grad_norm": 1.5222383113813167,
+      "learning_rate": 0.00010127570119207148,
+      "loss": 1.6704,
+      "step": 11094
+    },
+    {
+      "epoch": 0.5111254434053532,
+      "grad_norm": 1.5116001628492621,
+      "learning_rate": 0.0001012607815055934,
+      "loss": 0.9507,
+      "step": 11095
+    },
+    {
+      "epoch": 0.5111715114939881,
+      "grad_norm": 1.3341977299222254,
+      "learning_rate": 0.00010124586179104621,
+      "loss": 1.4395,
+      "step": 11096
+    },
+    {
+      "epoch": 0.5112175795826231,
+      "grad_norm": 2.5167212101347816,
+      "learning_rate": 0.00010123094204876197,
+      "loss": 1.116,
+      "step": 11097
+    },
+    {
+      "epoch": 0.5112636476712581,
+      "grad_norm": 1.333914856892811,
+      "learning_rate": 0.00010121602227907298,
+      "loss": 1.4706,
+      "step": 11098
+    },
+    {
+      "epoch": 0.5113097157598931,
+      "grad_norm": 1.4991402098552231,
+      "learning_rate": 0.00010120110248231134,
+      "loss": 1.2331,
+      "step": 11099
+    },
+    {
+      "epoch": 0.5113557838485281,
+      "grad_norm": 1.8539050587847357,
+      "learning_rate": 0.0001011861826588092,
+      "loss": 1.4805,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5114018519371631,
+      "grad_norm": 1.4643642845612024,
+      "learning_rate": 0.00010117126280889873,
+      "loss": 1.1433,
+      "step": 11101
+    },
+    {
+      "epoch": 0.5114479200257981,
+      "grad_norm": 0.771342736687604,
+      "learning_rate": 0.00010115634293291208,
+      "loss": 1.2877,
+      "step": 11102
+    },
+    {
+      "epoch": 0.5114939881144331,
+      "grad_norm": 1.0439898882278065,
+      "learning_rate": 0.00010114142303118148,
+      "loss": 1.4164,
+      "step": 11103
+    },
+    {
+      "epoch": 0.5115400562030681,
+      "grad_norm": 1.3274446495806826,
+      "learning_rate": 0.00010112650310403901,
+      "loss": 1.4147,
+      "step": 11104
+    },
+    {
+      "epoch": 0.5115861242917031,
+      "grad_norm": 1.3933554738723946,
+      "learning_rate": 0.00010111158315181693,
+      "loss": 1.2677,
+      "step": 11105
+    },
+    {
+      "epoch": 0.5116321923803382,
+      "grad_norm": 3.200321677680787,
+      "learning_rate": 0.00010109666317484732,
+      "loss": 1.2627,
+      "step": 11106
+    },
+    {
+      "epoch": 0.5116782604689731,
+      "grad_norm": 1.4554926080168311,
+      "learning_rate": 0.00010108174317346239,
+      "loss": 1.4077,
+      "step": 11107
+    },
+    {
+      "epoch": 0.5117243285576082,
+      "grad_norm": 1.5788174403422797,
+      "learning_rate": 0.00010106682314799432,
+      "loss": 1.346,
+      "step": 11108
+    },
+    {
+      "epoch": 0.5117703966462431,
+      "grad_norm": 0.7043966792983769,
+      "learning_rate": 0.00010105190309877524,
+      "loss": 1.4931,
+      "step": 11109
+    },
+    {
+      "epoch": 0.5118164647348782,
+      "grad_norm": 0.902449521965525,
+      "learning_rate": 0.00010103698302613735,
+      "loss": 1.1398,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5118625328235131,
+      "grad_norm": 1.6948492945059888,
+      "learning_rate": 0.0001010220629304128,
+      "loss": 1.2699,
+      "step": 11111
+    },
+    {
+      "epoch": 0.5119086009121482,
+      "grad_norm": 1.070634179600118,
+      "learning_rate": 0.00010100714281193379,
+      "loss": 1.0912,
+      "step": 11112
+    },
+    {
+      "epoch": 0.5119546690007831,
+      "grad_norm": 2.5542364223388367,
+      "learning_rate": 0.00010099222267103245,
+      "loss": 1.2319,
+      "step": 11113
+    },
+    {
+      "epoch": 0.5120007370894182,
+      "grad_norm": 1.2668749696334207,
+      "learning_rate": 0.000100977302508041,
+      "loss": 1.6351,
+      "step": 11114
+    },
+    {
+      "epoch": 0.5120468051780531,
+      "grad_norm": 1.0497302878628045,
+      "learning_rate": 0.00010096238232329158,
+      "loss": 1.5577,
+      "step": 11115
+    },
+    {
+      "epoch": 0.5120928732666882,
+      "grad_norm": 1.0181673018983588,
+      "learning_rate": 0.00010094746211711639,
+      "loss": 1.3225,
+      "step": 11116
+    },
+    {
+      "epoch": 0.5121389413553231,
+      "grad_norm": 0.912527529826781,
+      "learning_rate": 0.00010093254188984753,
+      "loss": 1.4977,
+      "step": 11117
+    },
+    {
+      "epoch": 0.5121850094439582,
+      "grad_norm": 1.9706420633204766,
+      "learning_rate": 0.00010091762164181724,
+      "loss": 1.5849,
+      "step": 11118
+    },
+    {
+      "epoch": 0.5122310775325932,
+      "grad_norm": 1.9287849197013873,
+      "learning_rate": 0.00010090270137335768,
+      "loss": 1.2748,
+      "step": 11119
+    },
+    {
+      "epoch": 0.5122771456212282,
+      "grad_norm": 1.470919492816078,
+      "learning_rate": 0.00010088778108480103,
+      "loss": 1.4826,
+      "step": 11120
+    },
+    {
+      "epoch": 0.5123232137098632,
+      "grad_norm": 0.9201807178274952,
+      "learning_rate": 0.00010087286077647946,
+      "loss": 1.3286,
+      "step": 11121
+    },
+    {
+      "epoch": 0.5123692817984982,
+      "grad_norm": 0.8862128012715045,
+      "learning_rate": 0.00010085794044872512,
+      "loss": 1.4376,
+      "step": 11122
+    },
+    {
+      "epoch": 0.5124153498871332,
+      "grad_norm": 1.319865680598931,
+      "learning_rate": 0.00010084302010187024,
+      "loss": 1.168,
+      "step": 11123
+    },
+    {
+      "epoch": 0.5124614179757682,
+      "grad_norm": 2.121973917160093,
+      "learning_rate": 0.00010082809973624692,
+      "loss": 1.4536,
+      "step": 11124
+    },
+    {
+      "epoch": 0.5125074860644032,
+      "grad_norm": 0.8214258141630215,
+      "learning_rate": 0.00010081317935218744,
+      "loss": 1.5093,
+      "step": 11125
+    },
+    {
+      "epoch": 0.5125535541530382,
+      "grad_norm": 2.958432363064522,
+      "learning_rate": 0.00010079825895002386,
+      "loss": 0.9778,
+      "step": 11126
+    },
+    {
+      "epoch": 0.5125996222416732,
+      "grad_norm": 1.082478036927752,
+      "learning_rate": 0.00010078333853008846,
+      "loss": 1.1324,
+      "step": 11127
+    },
+    {
+      "epoch": 0.5126456903303082,
+      "grad_norm": 0.7775643024252752,
+      "learning_rate": 0.00010076841809271332,
+      "loss": 1.5743,
+      "step": 11128
+    },
+    {
+      "epoch": 0.5126917584189432,
+      "grad_norm": 0.817617747422754,
+      "learning_rate": 0.0001007534976382307,
+      "loss": 1.4525,
+      "step": 11129
+    },
+    {
+      "epoch": 0.5127378265075782,
+      "grad_norm": 1.6526177943557225,
+      "learning_rate": 0.00010073857716697273,
+      "loss": 1.3079,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5127838945962132,
+      "grad_norm": 1.4911745430786596,
+      "learning_rate": 0.00010072365667927159,
+      "loss": 1.1759,
+      "step": 11131
+    },
+    {
+      "epoch": 0.5128299626848482,
+      "grad_norm": 0.9868739542479752,
+      "learning_rate": 0.00010070873617545953,
+      "loss": 1.4432,
+      "step": 11132
+    },
+    {
+      "epoch": 0.5128760307734832,
+      "grad_norm": 1.3783080267534669,
+      "learning_rate": 0.00010069381565586862,
+      "loss": 1.0876,
+      "step": 11133
+    },
+    {
+      "epoch": 0.5129220988621183,
+      "grad_norm": 7.854283180603459,
+      "learning_rate": 0.00010067889512083115,
+      "loss": 1.1074,
+      "step": 11134
+    },
+    {
+      "epoch": 0.5129681669507532,
+      "grad_norm": 0.7968972527190166,
+      "learning_rate": 0.0001006639745706792,
+      "loss": 1.4502,
+      "step": 11135
+    },
+    {
+      "epoch": 0.5130142350393883,
+      "grad_norm": 1.3940597624389783,
+      "learning_rate": 0.000100649054005745,
+      "loss": 1.0865,
+      "step": 11136
+    },
+    {
+      "epoch": 0.5130603031280232,
+      "grad_norm": 4.8265775875792665,
+      "learning_rate": 0.00010063413342636074,
+      "loss": 1.1502,
+      "step": 11137
+    },
+    {
+      "epoch": 0.5131063712166583,
+      "grad_norm": 1.218376881042088,
+      "learning_rate": 0.00010061921283285858,
+      "loss": 1.6079,
+      "step": 11138
+    },
+    {
+      "epoch": 0.5131524393052932,
+      "grad_norm": 1.2963820872133824,
+      "learning_rate": 0.00010060429222557069,
+      "loss": 1.0875,
+      "step": 11139
+    },
+    {
+      "epoch": 0.5131985073939282,
+      "grad_norm": 0.9936670410965587,
+      "learning_rate": 0.00010058937160482931,
+      "loss": 1.4837,
+      "step": 11140
+    },
+    {
+      "epoch": 0.5132445754825632,
+      "grad_norm": 1.253862370130284,
+      "learning_rate": 0.00010057445097096653,
+      "loss": 1.3347,
+      "step": 11141
+    },
+    {
+      "epoch": 0.5132906435711982,
+      "grad_norm": 0.7622898500789015,
+      "learning_rate": 0.00010055953032431462,
+      "loss": 1.3306,
+      "step": 11142
+    },
+    {
+      "epoch": 0.5133367116598332,
+      "grad_norm": 1.5078248699664263,
+      "learning_rate": 0.00010054460966520574,
+      "loss": 1.3686,
+      "step": 11143
+    },
+    {
+      "epoch": 0.5133827797484682,
+      "grad_norm": 1.0604393066405045,
+      "learning_rate": 0.00010052968899397201,
+      "loss": 1.285,
+      "step": 11144
+    },
+    {
+      "epoch": 0.5134288478371032,
+      "grad_norm": 1.3209617384728072,
+      "learning_rate": 0.0001005147683109457,
+      "loss": 1.5256,
+      "step": 11145
+    },
+    {
+      "epoch": 0.5134749159257382,
+      "grad_norm": 1.3695331353880975,
+      "learning_rate": 0.00010049984761645894,
+      "loss": 1.3617,
+      "step": 11146
+    },
+    {
+      "epoch": 0.5135209840143733,
+      "grad_norm": 1.0109374574928778,
+      "learning_rate": 0.00010048492691084395,
+      "loss": 1.4205,
+      "step": 11147
+    },
+    {
+      "epoch": 0.5135670521030082,
+      "grad_norm": 1.5855214016596186,
+      "learning_rate": 0.00010047000619443287,
+      "loss": 1.5527,
+      "step": 11148
+    },
+    {
+      "epoch": 0.5136131201916433,
+      "grad_norm": 1.1126042387897925,
+      "learning_rate": 0.00010045508546755793,
+      "loss": 1.2617,
+      "step": 11149
+    },
+    {
+      "epoch": 0.5136591882802782,
+      "grad_norm": 1.032024757439021,
+      "learning_rate": 0.00010044016473055127,
+      "loss": 1.1875,
+      "step": 11150
+    },
+    {
+      "epoch": 0.5137052563689133,
+      "grad_norm": 1.8578869890457845,
+      "learning_rate": 0.00010042524398374511,
+      "loss": 1.2858,
+      "step": 11151
+    },
+    {
+      "epoch": 0.5137513244575482,
+      "grad_norm": 1.0146980739386493,
+      "learning_rate": 0.00010041032322747161,
+      "loss": 1.3104,
+      "step": 11152
+    },
+    {
+      "epoch": 0.5137973925461833,
+      "grad_norm": 1.0572136796663016,
+      "learning_rate": 0.000100395402462063,
+      "loss": 1.1668,
+      "step": 11153
+    },
+    {
+      "epoch": 0.5138434606348182,
+      "grad_norm": 1.1513094593535964,
+      "learning_rate": 0.00010038048168785141,
+      "loss": 1.4967,
+      "step": 11154
+    },
+    {
+      "epoch": 0.5138895287234533,
+      "grad_norm": 1.3831697081329741,
+      "learning_rate": 0.00010036556090516904,
+      "loss": 1.2953,
+      "step": 11155
+    },
+    {
+      "epoch": 0.5139355968120882,
+      "grad_norm": 1.2491655354542204,
+      "learning_rate": 0.0001003506401143481,
+      "loss": 1.3949,
+      "step": 11156
+    },
+    {
+      "epoch": 0.5139816649007233,
+      "grad_norm": 1.0157902680613553,
+      "learning_rate": 0.00010033571931572072,
+      "loss": 1.2373,
+      "step": 11157
+    },
+    {
+      "epoch": 0.5140277329893582,
+      "grad_norm": 1.065096689921284,
+      "learning_rate": 0.00010032079850961917,
+      "loss": 1.2979,
+      "step": 11158
+    },
+    {
+      "epoch": 0.5140738010779933,
+      "grad_norm": 1.192187489714932,
+      "learning_rate": 0.00010030587769637556,
+      "loss": 1.3333,
+      "step": 11159
+    },
+    {
+      "epoch": 0.5141198691666283,
+      "grad_norm": 1.4308186993689238,
+      "learning_rate": 0.00010029095687632215,
+      "loss": 1.2799,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5141659372552633,
+      "grad_norm": 7.520233429887766,
+      "learning_rate": 0.00010027603604979103,
+      "loss": 1.1993,
+      "step": 11161
+    },
+    {
+      "epoch": 0.5142120053438983,
+      "grad_norm": 1.708720220713829,
+      "learning_rate": 0.00010026111521711448,
+      "loss": 1.2419,
+      "step": 11162
+    },
+    {
+      "epoch": 0.5142580734325333,
+      "grad_norm": 0.7574200775209831,
+      "learning_rate": 0.00010024619437862466,
+      "loss": 1.5268,
+      "step": 11163
+    },
+    {
+      "epoch": 0.5143041415211683,
+      "grad_norm": 2.511183057229957,
+      "learning_rate": 0.00010023127353465369,
+      "loss": 1.1181,
+      "step": 11164
+    },
+    {
+      "epoch": 0.5143502096098033,
+      "grad_norm": 0.9809781992474004,
+      "learning_rate": 0.00010021635268553387,
+      "loss": 1.57,
+      "step": 11165
+    },
+    {
+      "epoch": 0.5143962776984383,
+      "grad_norm": 1.4477719033180791,
+      "learning_rate": 0.00010020143183159727,
+      "loss": 1.1559,
+      "step": 11166
+    },
+    {
+      "epoch": 0.5144423457870733,
+      "grad_norm": 1.1168657785362999,
+      "learning_rate": 0.00010018651097317617,
+      "loss": 1.5143,
+      "step": 11167
+    },
+    {
+      "epoch": 0.5144884138757083,
+      "grad_norm": 0.9026118700964598,
+      "learning_rate": 0.0001001715901106027,
+      "loss": 1.5012,
+      "step": 11168
+    },
+    {
+      "epoch": 0.5145344819643433,
+      "grad_norm": 1.5477651800868644,
+      "learning_rate": 0.0001001566692442091,
+      "loss": 1.4102,
+      "step": 11169
+    },
+    {
+      "epoch": 0.5145805500529783,
+      "grad_norm": 2.5724302119306057,
+      "learning_rate": 0.0001001417483743275,
+      "loss": 1.2507,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5146266181416133,
+      "grad_norm": 5.748581729488669,
+      "learning_rate": 0.00010012682750129018,
+      "loss": 1.344,
+      "step": 11171
+    },
+    {
+      "epoch": 0.5146726862302483,
+      "grad_norm": 1.2563698870787874,
+      "learning_rate": 0.0001001119066254292,
+      "loss": 1.376,
+      "step": 11172
+    },
+    {
+      "epoch": 0.5147187543188834,
+      "grad_norm": 1.3381136345669626,
+      "learning_rate": 0.00010009698574707682,
+      "loss": 1.1978,
+      "step": 11173
+    },
+    {
+      "epoch": 0.5147648224075183,
+      "grad_norm": 0.9373619284950411,
+      "learning_rate": 0.00010008206486656521,
+      "loss": 1.2702,
+      "step": 11174
+    },
+    {
+      "epoch": 0.5148108904961534,
+      "grad_norm": 1.0914679086767065,
+      "learning_rate": 0.00010006714398422659,
+      "loss": 1.2347,
+      "step": 11175
+    },
+    {
+      "epoch": 0.5148569585847883,
+      "grad_norm": 1.120711235513972,
+      "learning_rate": 0.00010005222310039312,
+      "loss": 1.2786,
+      "step": 11176
+    },
+    {
+      "epoch": 0.5149030266734234,
+      "grad_norm": 0.853586964414347,
+      "learning_rate": 0.00010003730221539697,
+      "loss": 1.6337,
+      "step": 11177
+    },
+    {
+      "epoch": 0.5149490947620583,
+      "grad_norm": 1.4095078086855872,
+      "learning_rate": 0.00010002238132957038,
+      "loss": 1.2429,
+      "step": 11178
+    },
+    {
+      "epoch": 0.5149951628506934,
+      "grad_norm": 1.994565735584192,
+      "learning_rate": 0.0001000074604432455,
+      "loss": 1.3151,
+      "step": 11179
+    },
+    {
+      "epoch": 0.5150412309393283,
+      "grad_norm": 2.209616074531316,
+      "learning_rate": 9.999253955675451e-05,
+      "loss": 1.4583,
+      "step": 11180
+    },
+    {
+      "epoch": 0.5150872990279634,
+      "grad_norm": 1.137236255836438,
+      "learning_rate": 9.997761867042965e-05,
+      "loss": 1.3315,
+      "step": 11181
+    },
+    {
+      "epoch": 0.5151333671165983,
+      "grad_norm": 1.3472437652694311,
+      "learning_rate": 9.996269778460302e-05,
+      "loss": 1.333,
+      "step": 11182
+    },
+    {
+      "epoch": 0.5151794352052333,
+      "grad_norm": 0.7260393758956382,
+      "learning_rate": 9.99477768996069e-05,
+      "loss": 1.2795,
+      "step": 11183
+    },
+    {
+      "epoch": 0.5152255032938683,
+      "grad_norm": 1.2524964516706407,
+      "learning_rate": 9.99328560157734e-05,
+      "loss": 1.2418,
+      "step": 11184
+    },
+    {
+      "epoch": 0.5152715713825033,
+      "grad_norm": 3.0493428652776844,
+      "learning_rate": 9.99179351334348e-05,
+      "loss": 1.089,
+      "step": 11185
+    },
+    {
+      "epoch": 0.5153176394711383,
+      "grad_norm": 2.750041288912949,
+      "learning_rate": 9.99030142529232e-05,
+      "loss": 1.4052,
+      "step": 11186
+    },
+    {
+      "epoch": 0.5153637075597733,
+      "grad_norm": 2.4860562211512542,
+      "learning_rate": 9.988809337457083e-05,
+      "loss": 1.2964,
+      "step": 11187
+    },
+    {
+      "epoch": 0.5154097756484084,
+      "grad_norm": 0.8271991792416876,
+      "learning_rate": 9.987317249870987e-05,
+      "loss": 1.4929,
+      "step": 11188
+    },
+    {
+      "epoch": 0.5154558437370433,
+      "grad_norm": 3.173154571261265,
+      "learning_rate": 9.985825162567248e-05,
+      "loss": 1.0295,
+      "step": 11189
+    },
+    {
+      "epoch": 0.5155019118256784,
+      "grad_norm": 1.8311154376915375,
+      "learning_rate": 9.98433307557909e-05,
+      "loss": 1.2934,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5155479799143133,
+      "grad_norm": 1.684047013733341,
+      "learning_rate": 9.982840988939729e-05,
+      "loss": 1.5685,
+      "step": 11191
+    },
+    {
+      "epoch": 0.5155940480029484,
+      "grad_norm": 1.4479151964790122,
+      "learning_rate": 9.981348902682385e-05,
+      "loss": 1.2155,
+      "step": 11192
+    },
+    {
+      "epoch": 0.5156401160915833,
+      "grad_norm": 2.528709957722016,
+      "learning_rate": 9.979856816840273e-05,
+      "loss": 1.3479,
+      "step": 11193
+    },
+    {
+      "epoch": 0.5156861841802184,
+      "grad_norm": 1.2461912898680734,
+      "learning_rate": 9.978364731446617e-05,
+      "loss": 1.375,
+      "step": 11194
+    },
+    {
+      "epoch": 0.5157322522688533,
+      "grad_norm": 0.9417491166097199,
+      "learning_rate": 9.976872646534633e-05,
+      "loss": 1.2106,
+      "step": 11195
+    },
+    {
+      "epoch": 0.5157783203574884,
+      "grad_norm": 1.172835315915364,
+      "learning_rate": 9.975380562137537e-05,
+      "loss": 1.4808,
+      "step": 11196
+    },
+    {
+      "epoch": 0.5158243884461233,
+      "grad_norm": 0.9970084988863704,
+      "learning_rate": 9.973888478288553e-05,
+      "loss": 1.5928,
+      "step": 11197
+    },
+    {
+      "epoch": 0.5158704565347584,
+      "grad_norm": 1.8131012973187708,
+      "learning_rate": 9.972396395020897e-05,
+      "loss": 1.3556,
+      "step": 11198
+    },
+    {
+      "epoch": 0.5159165246233933,
+      "grad_norm": 1.9989201259726084,
+      "learning_rate": 9.970904312367788e-05,
+      "loss": 1.3273,
+      "step": 11199
+    },
+    {
+      "epoch": 0.5159625927120284,
+      "grad_norm": 1.5742670062136594,
+      "learning_rate": 9.969412230362442e-05,
+      "loss": 1.7066,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5160086608006634,
+      "grad_norm": 2.4095598159187,
+      "learning_rate": 9.967920149038085e-05,
+      "loss": 1.4104,
+      "step": 11201
+    },
+    {
+      "epoch": 0.5160547288892984,
+      "grad_norm": 1.9638031433975462,
+      "learning_rate": 9.966428068427926e-05,
+      "loss": 1.4112,
+      "step": 11202
+    },
+    {
+      "epoch": 0.5161007969779334,
+      "grad_norm": 0.8215912738835061,
+      "learning_rate": 9.964935988565193e-05,
+      "loss": 1.3994,
+      "step": 11203
+    },
+    {
+      "epoch": 0.5161468650665684,
+      "grad_norm": 1.012661501926209,
+      "learning_rate": 9.963443909483099e-05,
+      "loss": 1.4742,
+      "step": 11204
+    },
+    {
+      "epoch": 0.5161929331552034,
+      "grad_norm": 0.7509498194402608,
+      "learning_rate": 9.961951831214861e-05,
+      "loss": 1.5779,
+      "step": 11205
+    },
+    {
+      "epoch": 0.5162390012438384,
+      "grad_norm": 1.4031488769485956,
+      "learning_rate": 9.960459753793702e-05,
+      "loss": 1.2195,
+      "step": 11206
+    },
+    {
+      "epoch": 0.5162850693324734,
+      "grad_norm": 1.1423936720769643,
+      "learning_rate": 9.958967677252838e-05,
+      "loss": 1.3802,
+      "step": 11207
+    },
+    {
+      "epoch": 0.5163311374211084,
+      "grad_norm": 1.8965083333379589,
+      "learning_rate": 9.95747560162549e-05,
+      "loss": 1.1345,
+      "step": 11208
+    },
+    {
+      "epoch": 0.5163772055097434,
+      "grad_norm": 1.379888244656383,
+      "learning_rate": 9.955983526944873e-05,
+      "loss": 1.1709,
+      "step": 11209
+    },
+    {
+      "epoch": 0.5164232735983784,
+      "grad_norm": 1.4805847172205542,
+      "learning_rate": 9.954491453244209e-05,
+      "loss": 1.3896,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5164693416870134,
+      "grad_norm": 2.7155001371644993,
+      "learning_rate": 9.952999380556713e-05,
+      "loss": 1.3184,
+      "step": 11211
+    },
+    {
+      "epoch": 0.5165154097756484,
+      "grad_norm": 1.147619816872156,
+      "learning_rate": 9.951507308915606e-05,
+      "loss": 1.3683,
+      "step": 11212
+    },
+    {
+      "epoch": 0.5165614778642834,
+      "grad_norm": 1.2329813638001461,
+      "learning_rate": 9.95001523835411e-05,
+      "loss": 1.1574,
+      "step": 11213
+    },
+    {
+      "epoch": 0.5166075459529185,
+      "grad_norm": 5.349791338039331,
+      "learning_rate": 9.948523168905431e-05,
+      "loss": 1.3436,
+      "step": 11214
+    },
+    {
+      "epoch": 0.5166536140415534,
+      "grad_norm": 1.2913848840272268,
+      "learning_rate": 9.947031100602801e-05,
+      "loss": 1.571,
+      "step": 11215
+    },
+    {
+      "epoch": 0.5166996821301885,
+      "grad_norm": 1.0546240498165023,
+      "learning_rate": 9.945539033479429e-05,
+      "loss": 1.244,
+      "step": 11216
+    },
+    {
+      "epoch": 0.5167457502188234,
+      "grad_norm": 1.6310256206861933,
+      "learning_rate": 9.94404696756854e-05,
+      "loss": 1.2949,
+      "step": 11217
+    },
+    {
+      "epoch": 0.5167918183074585,
+      "grad_norm": 0.9788534274225474,
+      "learning_rate": 9.942554902903346e-05,
+      "loss": 1.4629,
+      "step": 11218
+    },
+    {
+      "epoch": 0.5168378863960934,
+      "grad_norm": 4.721680009197417,
+      "learning_rate": 9.941062839517072e-05,
+      "loss": 1.2585,
+      "step": 11219
+    },
+    {
+      "epoch": 0.5168839544847285,
+      "grad_norm": 1.0987636709647266,
+      "learning_rate": 9.939570777442931e-05,
+      "loss": 1.3293,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5169300225733634,
+      "grad_norm": 1.9668341158915792,
+      "learning_rate": 9.938078716714145e-05,
+      "loss": 0.9713,
+      "step": 11221
+    },
+    {
+      "epoch": 0.5169760906619985,
+      "grad_norm": 1.1158211193381484,
+      "learning_rate": 9.936586657363931e-05,
+      "loss": 1.2246,
+      "step": 11222
+    },
+    {
+      "epoch": 0.5170221587506334,
+      "grad_norm": 1.3652408805560197,
+      "learning_rate": 9.9350945994255e-05,
+      "loss": 1.4482,
+      "step": 11223
+    },
+    {
+      "epoch": 0.5170682268392685,
+      "grad_norm": 7.663008007878047,
+      "learning_rate": 9.933602542932084e-05,
+      "loss": 1.483,
+      "step": 11224
+    },
+    {
+      "epoch": 0.5171142949279034,
+      "grad_norm": 1.22665591361728,
+      "learning_rate": 9.932110487916887e-05,
+      "loss": 1.4848,
+      "step": 11225
+    },
+    {
+      "epoch": 0.5171603630165384,
+      "grad_norm": 0.8081077866346124,
+      "learning_rate": 9.93061843441314e-05,
+      "loss": 1.5514,
+      "step": 11226
+    },
+    {
+      "epoch": 0.5172064311051734,
+      "grad_norm": 1.1358992319302796,
+      "learning_rate": 9.929126382454048e-05,
+      "loss": 1.3934,
+      "step": 11227
+    },
+    {
+      "epoch": 0.5172524991938084,
+      "grad_norm": 0.8929064644803327,
+      "learning_rate": 9.927634332072843e-05,
+      "loss": 1.3652,
+      "step": 11228
+    },
+    {
+      "epoch": 0.5172985672824435,
+      "grad_norm": 2.4859819730791433,
+      "learning_rate": 9.926142283302727e-05,
+      "loss": 1.0913,
+      "step": 11229
+    },
+    {
+      "epoch": 0.5173446353710784,
+      "grad_norm": 1.004733004286181,
+      "learning_rate": 9.924650236176932e-05,
+      "loss": 1.525,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5173907034597135,
+      "grad_norm": 5.846776184102391,
+      "learning_rate": 9.923158190728672e-05,
+      "loss": 1.0908,
+      "step": 11231
+    },
+    {
+      "epoch": 0.5174367715483484,
+      "grad_norm": 1.751856403678761,
+      "learning_rate": 9.921666146991158e-05,
+      "loss": 1.0205,
+      "step": 11232
+    },
+    {
+      "epoch": 0.5174828396369835,
+      "grad_norm": 0.9872266018231465,
+      "learning_rate": 9.920174104997617e-05,
+      "loss": 1.2846,
+      "step": 11233
+    },
+    {
+      "epoch": 0.5175289077256184,
+      "grad_norm": 0.9639910671914994,
+      "learning_rate": 9.918682064781259e-05,
+      "loss": 1.5564,
+      "step": 11234
+    },
+    {
+      "epoch": 0.5175749758142535,
+      "grad_norm": 0.968054424139321,
+      "learning_rate": 9.917190026375309e-05,
+      "loss": 1.6079,
+      "step": 11235
+    },
+    {
+      "epoch": 0.5176210439028884,
+      "grad_norm": 1.826062940589701,
+      "learning_rate": 9.915697989812977e-05,
+      "loss": 1.3025,
+      "step": 11236
+    },
+    {
+      "epoch": 0.5176671119915235,
+      "grad_norm": 2.2544606867696424,
+      "learning_rate": 9.91420595512749e-05,
+      "loss": 1.4126,
+      "step": 11237
+    },
+    {
+      "epoch": 0.5177131800801584,
+      "grad_norm": 0.9041360931955134,
+      "learning_rate": 9.912713922352053e-05,
+      "loss": 1.5161,
+      "step": 11238
+    },
+    {
+      "epoch": 0.5177592481687935,
+      "grad_norm": 1.216604478158914,
+      "learning_rate": 9.911221891519899e-05,
+      "loss": 1.1858,
+      "step": 11239
+    },
+    {
+      "epoch": 0.5178053162574284,
+      "grad_norm": 1.0278117886558509,
+      "learning_rate": 9.909729862664234e-05,
+      "loss": 1.1844,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5178513843460635,
+      "grad_norm": 1.215741324198853,
+      "learning_rate": 9.908237835818278e-05,
+      "loss": 1.3247,
+      "step": 11241
+    },
+    {
+      "epoch": 0.5178974524346985,
+      "grad_norm": 1.2877371840685952,
+      "learning_rate": 9.90674581101525e-05,
+      "loss": 1.384,
+      "step": 11242
+    },
+    {
+      "epoch": 0.5179435205233335,
+      "grad_norm": 1.0299643348583811,
+      "learning_rate": 9.905253788288363e-05,
+      "loss": 1.6227,
+      "step": 11243
+    },
+    {
+      "epoch": 0.5179895886119685,
+      "grad_norm": 1.00493803622698,
+      "learning_rate": 9.903761767670845e-05,
+      "loss": 1.5165,
+      "step": 11244
+    },
+    {
+      "epoch": 0.5180356567006035,
+      "grad_norm": 1.4389366146910134,
+      "learning_rate": 9.902269749195898e-05,
+      "loss": 1.4927,
+      "step": 11245
+    },
+    {
+      "epoch": 0.5180817247892385,
+      "grad_norm": 1.2104287325167178,
+      "learning_rate": 9.900777732896756e-05,
+      "loss": 1.3089,
+      "step": 11246
+    },
+    {
+      "epoch": 0.5181277928778735,
+      "grad_norm": 4.060951264535966,
+      "learning_rate": 9.89928571880662e-05,
+      "loss": 1.414,
+      "step": 11247
+    },
+    {
+      "epoch": 0.5181738609665085,
+      "grad_norm": 1.0949748057877993,
+      "learning_rate": 9.89779370695872e-05,
+      "loss": 1.2829,
+      "step": 11248
+    },
+    {
+      "epoch": 0.5182199290551435,
+      "grad_norm": 1.330467407160653,
+      "learning_rate": 9.896301697386269e-05,
+      "loss": 1.2964,
+      "step": 11249
+    },
+    {
+      "epoch": 0.5182659971437785,
+      "grad_norm": 3.7404568767505992,
+      "learning_rate": 9.894809690122478e-05,
+      "loss": 1.0693,
+      "step": 11250
+    },
+    {
+      "epoch": 0.5183120652324135,
+      "grad_norm": 1.106013878017463,
+      "learning_rate": 9.893317685200573e-05,
+      "loss": 1.1489,
+      "step": 11251
+    },
+    {
+      "epoch": 0.5183581333210485,
+      "grad_norm": 0.9529314380129962,
+      "learning_rate": 9.891825682653762e-05,
+      "loss": 1.4958,
+      "step": 11252
+    },
+    {
+      "epoch": 0.5184042014096835,
+      "grad_norm": 1.0938260990173907,
+      "learning_rate": 9.890333682515271e-05,
+      "loss": 1.1549,
+      "step": 11253
+    },
+    {
+      "epoch": 0.5184502694983185,
+      "grad_norm": 3.3958469940647085,
+      "learning_rate": 9.88884168481831e-05,
+      "loss": 1.0524,
+      "step": 11254
+    },
+    {
+      "epoch": 0.5184963375869536,
+      "grad_norm": 2.9700547067622045,
+      "learning_rate": 9.887349689596101e-05,
+      "loss": 1.4831,
+      "step": 11255
+    },
+    {
+      "epoch": 0.5185424056755885,
+      "grad_norm": 1.103296612183963,
+      "learning_rate": 9.885857696881853e-05,
+      "loss": 1.4025,
+      "step": 11256
+    },
+    {
+      "epoch": 0.5185884737642236,
+      "grad_norm": 1.0423523955402696,
+      "learning_rate": 9.884365706708793e-05,
+      "loss": 1.4277,
+      "step": 11257
+    },
+    {
+      "epoch": 0.5186345418528585,
+      "grad_norm": 1.4537161347658665,
+      "learning_rate": 9.882873719110132e-05,
+      "loss": 1.3187,
+      "step": 11258
+    },
+    {
+      "epoch": 0.5186806099414936,
+      "grad_norm": 1.7385261613380283,
+      "learning_rate": 9.881381734119084e-05,
+      "loss": 0.9413,
+      "step": 11259
+    },
+    {
+      "epoch": 0.5187266780301285,
+      "grad_norm": 0.918257150667235,
+      "learning_rate": 9.879889751768871e-05,
+      "loss": 1.4881,
+      "step": 11260
+    },
+    {
+      "epoch": 0.5187727461187636,
+      "grad_norm": 1.3873740789680864,
+      "learning_rate": 9.878397772092704e-05,
+      "loss": 1.3323,
+      "step": 11261
+    },
+    {
+      "epoch": 0.5188188142073985,
+      "grad_norm": 1.0804100612591196,
+      "learning_rate": 9.876905795123804e-05,
+      "loss": 1.0197,
+      "step": 11262
+    },
+    {
+      "epoch": 0.5188648822960336,
+      "grad_norm": 1.1269662860488567,
+      "learning_rate": 9.875413820895383e-05,
+      "loss": 1.0243,
+      "step": 11263
+    },
+    {
+      "epoch": 0.5189109503846685,
+      "grad_norm": 0.8431927336484839,
+      "learning_rate": 9.873921849440661e-05,
+      "loss": 1.5318,
+      "step": 11264
+    },
+    {
+      "epoch": 0.5189570184733036,
+      "grad_norm": 4.025078534683316,
+      "learning_rate": 9.872429880792853e-05,
+      "loss": 1.0315,
+      "step": 11265
+    },
+    {
+      "epoch": 0.5190030865619385,
+      "grad_norm": 1.4017189813898507,
+      "learning_rate": 9.870937914985175e-05,
+      "loss": 1.5793,
+      "step": 11266
+    },
+    {
+      "epoch": 0.5190491546505736,
+      "grad_norm": 0.915645505622921,
+      "learning_rate": 9.869445952050845e-05,
+      "loss": 1.6076,
+      "step": 11267
+    },
+    {
+      "epoch": 0.5190952227392085,
+      "grad_norm": 1.2755704573756619,
+      "learning_rate": 9.867953992023075e-05,
+      "loss": 1.3642,
+      "step": 11268
+    },
+    {
+      "epoch": 0.5191412908278435,
+      "grad_norm": 2.080510755029959,
+      "learning_rate": 9.866462034935083e-05,
+      "loss": 1.1453,
+      "step": 11269
+    },
+    {
+      "epoch": 0.5191873589164786,
+      "grad_norm": 5.856428392149633,
+      "learning_rate": 9.864970080820084e-05,
+      "loss": 1.3958,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5192334270051135,
+      "grad_norm": 1.090970910972426,
+      "learning_rate": 9.863478129711296e-05,
+      "loss": 1.3548,
+      "step": 11271
+    },
+    {
+      "epoch": 0.5192794950937486,
+      "grad_norm": 2.505060588934346,
+      "learning_rate": 9.861986181641932e-05,
+      "loss": 1.2772,
+      "step": 11272
+    },
+    {
+      "epoch": 0.5193255631823835,
+      "grad_norm": 0.71327283079739,
+      "learning_rate": 9.860494236645212e-05,
+      "loss": 1.5025,
+      "step": 11273
+    },
+    {
+      "epoch": 0.5193716312710186,
+      "grad_norm": 0.9567053916926739,
+      "learning_rate": 9.859002294754345e-05,
+      "loss": 1.3954,
+      "step": 11274
+    },
+    {
+      "epoch": 0.5194176993596535,
+      "grad_norm": 1.803679608028874,
+      "learning_rate": 9.857510356002552e-05,
+      "loss": 1.2447,
+      "step": 11275
+    },
+    {
+      "epoch": 0.5194637674482886,
+      "grad_norm": 0.7964416933270878,
+      "learning_rate": 9.856018420423047e-05,
+      "loss": 1.3046,
+      "step": 11276
+    },
+    {
+      "epoch": 0.5195098355369235,
+      "grad_norm": 1.0102669339812982,
+      "learning_rate": 9.854526488049042e-05,
+      "loss": 1.6085,
+      "step": 11277
+    },
+    {
+      "epoch": 0.5195559036255586,
+      "grad_norm": 1.4279099853741193,
+      "learning_rate": 9.85303455891376e-05,
+      "loss": 1.228,
+      "step": 11278
+    },
+    {
+      "epoch": 0.5196019717141935,
+      "grad_norm": 1.2187769179636003,
+      "learning_rate": 9.851542633050408e-05,
+      "loss": 1.3487,
+      "step": 11279
+    },
+    {
+      "epoch": 0.5196480398028286,
+      "grad_norm": 1.2186992468443305,
+      "learning_rate": 9.850050710492207e-05,
+      "loss": 1.4407,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5196941078914635,
+      "grad_norm": 1.3779422780432817,
+      "learning_rate": 9.848558791272367e-05,
+      "loss": 1.3846,
+      "step": 11281
+    },
+    {
+      "epoch": 0.5197401759800986,
+      "grad_norm": 1.8269126676659355,
+      "learning_rate": 9.84706687542411e-05,
+      "loss": 1.1647,
+      "step": 11282
+    },
+    {
+      "epoch": 0.5197862440687336,
+      "grad_norm": 1.1525058989917056,
+      "learning_rate": 9.845574962980641e-05,
+      "loss": 1.3181,
+      "step": 11283
+    },
+    {
+      "epoch": 0.5198323121573686,
+      "grad_norm": 1.3960031443661278,
+      "learning_rate": 9.844083053975185e-05,
+      "loss": 1.4911,
+      "step": 11284
+    },
+    {
+      "epoch": 0.5198783802460036,
+      "grad_norm": 1.631628975143931,
+      "learning_rate": 9.842591148440954e-05,
+      "loss": 1.1019,
+      "step": 11285
+    },
+    {
+      "epoch": 0.5199244483346386,
+      "grad_norm": 1.1957273376541493,
+      "learning_rate": 9.841099246411157e-05,
+      "loss": 1.3394,
+      "step": 11286
+    },
+    {
+      "epoch": 0.5199705164232736,
+      "grad_norm": 2.5006139403966614,
+      "learning_rate": 9.839607347919015e-05,
+      "loss": 1.1663,
+      "step": 11287
+    },
+    {
+      "epoch": 0.5200165845119086,
+      "grad_norm": 1.0666834258996647,
+      "learning_rate": 9.838115452997739e-05,
+      "loss": 1.4072,
+      "step": 11288
+    },
+    {
+      "epoch": 0.5200626526005436,
+      "grad_norm": 1.1438507704016248,
+      "learning_rate": 9.836623561680547e-05,
+      "loss": 1.23,
+      "step": 11289
+    },
+    {
+      "epoch": 0.5201087206891786,
+      "grad_norm": 0.9005846385972271,
+      "learning_rate": 9.835131674000648e-05,
+      "loss": 1.1519,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5201547887778136,
+      "grad_norm": 1.1001309544646585,
+      "learning_rate": 9.833639789991263e-05,
+      "loss": 1.4555,
+      "step": 11291
+    },
+    {
+      "epoch": 0.5202008568664486,
+      "grad_norm": 1.215943869396262,
+      "learning_rate": 9.8321479096856e-05,
+      "loss": 1.3574,
+      "step": 11292
+    },
+    {
+      "epoch": 0.5202469249550836,
+      "grad_norm": 1.6673538218213129,
+      "learning_rate": 9.830656033116878e-05,
+      "loss": 1.3687,
+      "step": 11293
+    },
+    {
+      "epoch": 0.5202929930437186,
+      "grad_norm": 0.9145777155478186,
+      "learning_rate": 9.829164160318311e-05,
+      "loss": 1.2931,
+      "step": 11294
+    },
+    {
+      "epoch": 0.5203390611323536,
+      "grad_norm": 1.181017898017831,
+      "learning_rate": 9.827672291323107e-05,
+      "loss": 1.3886,
+      "step": 11295
+    },
+    {
+      "epoch": 0.5203851292209887,
+      "grad_norm": 1.2538588318431074,
+      "learning_rate": 9.826180426164486e-05,
+      "loss": 1.1801,
+      "step": 11296
+    },
+    {
+      "epoch": 0.5204311973096236,
+      "grad_norm": 1.8304394540565283,
+      "learning_rate": 9.824688564875658e-05,
+      "loss": 1.5566,
+      "step": 11297
+    },
+    {
+      "epoch": 0.5204772653982587,
+      "grad_norm": 2.049425900201652,
+      "learning_rate": 9.823196707489841e-05,
+      "loss": 1.2245,
+      "step": 11298
+    },
+    {
+      "epoch": 0.5205233334868936,
+      "grad_norm": 1.2612163999262926,
+      "learning_rate": 9.821704854040244e-05,
+      "loss": 1.621,
+      "step": 11299
+    },
+    {
+      "epoch": 0.5205694015755287,
+      "grad_norm": 0.8645007697021939,
+      "learning_rate": 9.820213004560084e-05,
+      "loss": 1.2938,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5206154696641636,
+      "grad_norm": 1.057115555333668,
+      "learning_rate": 9.818721159082573e-05,
+      "loss": 1.1325,
+      "step": 11301
+    },
+    {
+      "epoch": 0.5206615377527987,
+      "grad_norm": 2.2803825348509714,
+      "learning_rate": 9.817229317640925e-05,
+      "loss": 1.189,
+      "step": 11302
+    },
+    {
+      "epoch": 0.5207076058414336,
+      "grad_norm": 3.0355149461331417,
+      "learning_rate": 9.81573748026835e-05,
+      "loss": 0.9189,
+      "step": 11303
+    },
+    {
+      "epoch": 0.5207536739300687,
+      "grad_norm": 2.3131698089971424,
+      "learning_rate": 9.814245646998067e-05,
+      "loss": 1.1378,
+      "step": 11304
+    },
+    {
+      "epoch": 0.5207997420187036,
+      "grad_norm": 0.9249445124722719,
+      "learning_rate": 9.812753817863287e-05,
+      "loss": 1.2564,
+      "step": 11305
+    },
+    {
+      "epoch": 0.5208458101073387,
+      "grad_norm": 1.6037025916297176,
+      "learning_rate": 9.81126199289722e-05,
+      "loss": 1.1877,
+      "step": 11306
+    },
+    {
+      "epoch": 0.5208918781959736,
+      "grad_norm": 1.7109684220379648,
+      "learning_rate": 9.809770172133083e-05,
+      "loss": 1.0482,
+      "step": 11307
+    },
+    {
+      "epoch": 0.5209379462846087,
+      "grad_norm": 1.133451116737102,
+      "learning_rate": 9.808278355604084e-05,
+      "loss": 1.3372,
+      "step": 11308
+    },
+    {
+      "epoch": 0.5209840143732437,
+      "grad_norm": 0.9912238665091004,
+      "learning_rate": 9.806786543343441e-05,
+      "loss": 1.3923,
+      "step": 11309
+    },
+    {
+      "epoch": 0.5210300824618787,
+      "grad_norm": 2.1554812306804463,
+      "learning_rate": 9.805294735384362e-05,
+      "loss": 1.189,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5210761505505137,
+      "grad_norm": 0.6990030900987817,
+      "learning_rate": 9.803802931760066e-05,
+      "loss": 1.5967,
+      "step": 11311
+    },
+    {
+      "epoch": 0.5211222186391486,
+      "grad_norm": 1.1046288355249876,
+      "learning_rate": 9.802311132503756e-05,
+      "loss": 0.7742,
+      "step": 11312
+    },
+    {
+      "epoch": 0.5211682867277837,
+      "grad_norm": 1.077334314701889,
+      "learning_rate": 9.800819337648655e-05,
+      "loss": 1.3843,
+      "step": 11313
+    },
+    {
+      "epoch": 0.5212143548164186,
+      "grad_norm": 1.3515364923278155,
+      "learning_rate": 9.79932754722797e-05,
+      "loss": 1.3506,
+      "step": 11314
+    },
+    {
+      "epoch": 0.5212604229050537,
+      "grad_norm": 0.8931928314689366,
+      "learning_rate": 9.797835761274907e-05,
+      "loss": 1.494,
+      "step": 11315
+    },
+    {
+      "epoch": 0.5213064909936886,
+      "grad_norm": 2.6235569821462135,
+      "learning_rate": 9.796343979822689e-05,
+      "loss": 1.6757,
+      "step": 11316
+    },
+    {
+      "epoch": 0.5213525590823237,
+      "grad_norm": 1.5876652355174654,
+      "learning_rate": 9.794852202904521e-05,
+      "loss": 1.0575,
+      "step": 11317
+    },
+    {
+      "epoch": 0.5213986271709586,
+      "grad_norm": 0.9266795068631883,
+      "learning_rate": 9.793360430553618e-05,
+      "loss": 1.3912,
+      "step": 11318
+    },
+    {
+      "epoch": 0.5214446952595937,
+      "grad_norm": 0.8399378235629624,
+      "learning_rate": 9.791868662803189e-05,
+      "loss": 1.4719,
+      "step": 11319
+    },
+    {
+      "epoch": 0.5214907633482286,
+      "grad_norm": 1.5803536373728955,
+      "learning_rate": 9.79037689968645e-05,
+      "loss": 1.2557,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5215368314368637,
+      "grad_norm": 0.8981343495009675,
+      "learning_rate": 9.788885141236607e-05,
+      "loss": 1.5025,
+      "step": 11321
+    },
+    {
+      "epoch": 0.5215828995254986,
+      "grad_norm": 1.2237483657440225,
+      "learning_rate": 9.787393387486878e-05,
+      "loss": 1.3132,
+      "step": 11322
+    },
+    {
+      "epoch": 0.5216289676141337,
+      "grad_norm": 1.3285938835643718,
+      "learning_rate": 9.785901638470472e-05,
+      "loss": 1.3204,
+      "step": 11323
+    },
+    {
+      "epoch": 0.5216750357027687,
+      "grad_norm": 0.8742535907525694,
+      "learning_rate": 9.784409894220592e-05,
+      "loss": 1.3239,
+      "step": 11324
+    },
+    {
+      "epoch": 0.5217211037914037,
+      "grad_norm": 1.316741656822367,
+      "learning_rate": 9.782918154770463e-05,
+      "loss": 1.4701,
+      "step": 11325
+    },
+    {
+      "epoch": 0.5217671718800387,
+      "grad_norm": 0.845944782369068,
+      "learning_rate": 9.781426420153283e-05,
+      "loss": 1.2433,
+      "step": 11326
+    },
+    {
+      "epoch": 0.5218132399686737,
+      "grad_norm": 2.0642381424699856,
+      "learning_rate": 9.779934690402275e-05,
+      "loss": 1.5491,
+      "step": 11327
+    },
+    {
+      "epoch": 0.5218593080573087,
+      "grad_norm": 1.2313941219045244,
+      "learning_rate": 9.778442965550636e-05,
+      "loss": 1.386,
+      "step": 11328
+    },
+    {
+      "epoch": 0.5219053761459437,
+      "grad_norm": 1.3724026631247892,
+      "learning_rate": 9.77695124563159e-05,
+      "loss": 1.3545,
+      "step": 11329
+    },
+    {
+      "epoch": 0.5219514442345787,
+      "grad_norm": 0.9425876629507038,
+      "learning_rate": 9.77545953067834e-05,
+      "loss": 1.5894,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5219975123232137,
+      "grad_norm": 0.8319310048076407,
+      "learning_rate": 9.773967820724101e-05,
+      "loss": 1.3939,
+      "step": 11331
+    },
+    {
+      "epoch": 0.5220435804118487,
+      "grad_norm": 1.4616525278407964,
+      "learning_rate": 9.772476115802082e-05,
+      "loss": 1.32,
+      "step": 11332
+    },
+    {
+      "epoch": 0.5220896485004837,
+      "grad_norm": 0.9892948865231795,
+      "learning_rate": 9.770984415945487e-05,
+      "loss": 1.2801,
+      "step": 11333
+    },
+    {
+      "epoch": 0.5221357165891187,
+      "grad_norm": 1.4874093409788025,
+      "learning_rate": 9.769492721187537e-05,
+      "loss": 1.2224,
+      "step": 11334
+    },
+    {
+      "epoch": 0.5221817846777537,
+      "grad_norm": 0.988291546560416,
+      "learning_rate": 9.76800103156143e-05,
+      "loss": 1.294,
+      "step": 11335
+    },
+    {
+      "epoch": 0.5222278527663887,
+      "grad_norm": 1.1812018991311177,
+      "learning_rate": 9.76650934710039e-05,
+      "loss": 1.2016,
+      "step": 11336
+    },
+    {
+      "epoch": 0.5222739208550238,
+      "grad_norm": 3.5082841785860746,
+      "learning_rate": 9.76501766783761e-05,
+      "loss": 1.4485,
+      "step": 11337
+    },
+    {
+      "epoch": 0.5223199889436587,
+      "grad_norm": 1.6077620509447708,
+      "learning_rate": 9.763525993806318e-05,
+      "loss": 1.384,
+      "step": 11338
+    },
+    {
+      "epoch": 0.5223660570322938,
+      "grad_norm": 1.0350233885674212,
+      "learning_rate": 9.762034325039705e-05,
+      "loss": 1.3041,
+      "step": 11339
+    },
+    {
+      "epoch": 0.5224121251209287,
+      "grad_norm": 1.1521898533790433,
+      "learning_rate": 9.760542661570997e-05,
+      "loss": 1.2385,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5224581932095638,
+      "grad_norm": 1.56140929291576,
+      "learning_rate": 9.759051003433395e-05,
+      "loss": 1.2485,
+      "step": 11341
+    },
+    {
+      "epoch": 0.5225042612981987,
+      "grad_norm": 1.266156684009039,
+      "learning_rate": 9.757559350660104e-05,
+      "loss": 1.5875,
+      "step": 11342
+    },
+    {
+      "epoch": 0.5225503293868338,
+      "grad_norm": 1.0429150129675004,
+      "learning_rate": 9.756067703284345e-05,
+      "loss": 1.347,
+      "step": 11343
+    },
+    {
+      "epoch": 0.5225963974754687,
+      "grad_norm": 3.475912256884291,
+      "learning_rate": 9.754576061339314e-05,
+      "loss": 1.4075,
+      "step": 11344
+    },
+    {
+      "epoch": 0.5226424655641038,
+      "grad_norm": 4.535518503295349,
+      "learning_rate": 9.753084424858234e-05,
+      "loss": 1.1711,
+      "step": 11345
+    },
+    {
+      "epoch": 0.5226885336527387,
+      "grad_norm": 1.3511052727824533,
+      "learning_rate": 9.751592793874297e-05,
+      "loss": 1.3044,
+      "step": 11346
+    },
+    {
+      "epoch": 0.5227346017413738,
+      "grad_norm": 1.0987251085294942,
+      "learning_rate": 9.750101168420729e-05,
+      "loss": 1.3092,
+      "step": 11347
+    },
+    {
+      "epoch": 0.5227806698300087,
+      "grad_norm": 1.473245020550243,
+      "learning_rate": 9.748609548530721e-05,
+      "loss": 1.4092,
+      "step": 11348
+    },
+    {
+      "epoch": 0.5228267379186438,
+      "grad_norm": 0.8178811456780752,
+      "learning_rate": 9.747117934237499e-05,
+      "loss": 1.6071,
+      "step": 11349
+    },
+    {
+      "epoch": 0.5228728060072788,
+      "grad_norm": 2.748108393934779,
+      "learning_rate": 9.745626325574261e-05,
+      "loss": 1.3448,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5229188740959138,
+      "grad_norm": 1.3014668083674283,
+      "learning_rate": 9.744134722574213e-05,
+      "loss": 1.3507,
+      "step": 11351
+    },
+    {
+      "epoch": 0.5229649421845488,
+      "grad_norm": 1.4314664596926054,
+      "learning_rate": 9.742643125270569e-05,
+      "loss": 1.4616,
+      "step": 11352
+    },
+    {
+      "epoch": 0.5230110102731838,
+      "grad_norm": 1.439296610976063,
+      "learning_rate": 9.741151533696533e-05,
+      "loss": 1.3782,
+      "step": 11353
+    },
+    {
+      "epoch": 0.5230570783618188,
+      "grad_norm": 1.154186213562114,
+      "learning_rate": 9.739659947885316e-05,
+      "loss": 1.1274,
+      "step": 11354
+    },
+    {
+      "epoch": 0.5231031464504537,
+      "grad_norm": 1.2129918665509207,
+      "learning_rate": 9.73816836787012e-05,
+      "loss": 1.1499,
+      "step": 11355
+    },
+    {
+      "epoch": 0.5231492145390888,
+      "grad_norm": 1.8075550590664045,
+      "learning_rate": 9.736676793684162e-05,
+      "loss": 1.2059,
+      "step": 11356
+    },
+    {
+      "epoch": 0.5231952826277237,
+      "grad_norm": 1.4554717627109428,
+      "learning_rate": 9.735185225360637e-05,
+      "loss": 1.1412,
+      "step": 11357
+    },
+    {
+      "epoch": 0.5232413507163588,
+      "grad_norm": 1.0642351571508386,
+      "learning_rate": 9.733693662932766e-05,
+      "loss": 1.0881,
+      "step": 11358
+    },
+    {
+      "epoch": 0.5232874188049937,
+      "grad_norm": 1.9105516416210957,
+      "learning_rate": 9.732202106433748e-05,
+      "loss": 1.2522,
+      "step": 11359
+    },
+    {
+      "epoch": 0.5233334868936288,
+      "grad_norm": 1.0425325420605316,
+      "learning_rate": 9.730710555896787e-05,
+      "loss": 1.569,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5233795549822637,
+      "grad_norm": 1.7021000637465902,
+      "learning_rate": 9.729219011355098e-05,
+      "loss": 1.2329,
+      "step": 11361
+    },
+    {
+      "epoch": 0.5234256230708988,
+      "grad_norm": 0.9976943339738149,
+      "learning_rate": 9.727727472841879e-05,
+      "loss": 1.3848,
+      "step": 11362
+    },
+    {
+      "epoch": 0.5234716911595337,
+      "grad_norm": 1.817698788542918,
+      "learning_rate": 9.726235940390344e-05,
+      "loss": 1.2195,
+      "step": 11363
+    },
+    {
+      "epoch": 0.5235177592481688,
+      "grad_norm": 1.3569502203806996,
+      "learning_rate": 9.724744414033693e-05,
+      "loss": 1.2258,
+      "step": 11364
+    },
+    {
+      "epoch": 0.5235638273368038,
+      "grad_norm": 1.9738137194137066,
+      "learning_rate": 9.72325289380514e-05,
+      "loss": 1.3101,
+      "step": 11365
+    },
+    {
+      "epoch": 0.5236098954254388,
+      "grad_norm": 1.6360109965829754,
+      "learning_rate": 9.721761379737884e-05,
+      "loss": 1.5811,
+      "step": 11366
+    },
+    {
+      "epoch": 0.5236559635140738,
+      "grad_norm": 2.0238534480178036,
+      "learning_rate": 9.720269871865135e-05,
+      "loss": 1.2888,
+      "step": 11367
+    },
+    {
+      "epoch": 0.5237020316027088,
+      "grad_norm": 1.1187736252805296,
+      "learning_rate": 9.718778370220098e-05,
+      "loss": 1.1625,
+      "step": 11368
+    },
+    {
+      "epoch": 0.5237480996913438,
+      "grad_norm": 2.4134198788428862,
+      "learning_rate": 9.717286874835977e-05,
+      "loss": 1.2766,
+      "step": 11369
+    },
+    {
+      "epoch": 0.5237941677799788,
+      "grad_norm": 1.0196950085002738,
+      "learning_rate": 9.71579538574598e-05,
+      "loss": 1.5513,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5238402358686138,
+      "grad_norm": 1.9452060453591113,
+      "learning_rate": 9.71430390298331e-05,
+      "loss": 1.5337,
+      "step": 11371
+    },
+    {
+      "epoch": 0.5238863039572488,
+      "grad_norm": 1.1480264053548555,
+      "learning_rate": 9.712812426581176e-05,
+      "loss": 1.4747,
+      "step": 11372
+    },
+    {
+      "epoch": 0.5239323720458838,
+      "grad_norm": 1.6948822669977825,
+      "learning_rate": 9.71132095657278e-05,
+      "loss": 1.4217,
+      "step": 11373
+    },
+    {
+      "epoch": 0.5239784401345188,
+      "grad_norm": 1.432731558527566,
+      "learning_rate": 9.709829492991328e-05,
+      "loss": 1.4795,
+      "step": 11374
+    },
+    {
+      "epoch": 0.5240245082231538,
+      "grad_norm": 1.3466200797196164,
+      "learning_rate": 9.708338035870023e-05,
+      "loss": 1.3365,
+      "step": 11375
+    },
+    {
+      "epoch": 0.5240705763117888,
+      "grad_norm": 1.9934402552786237,
+      "learning_rate": 9.706846585242074e-05,
+      "loss": 1.3207,
+      "step": 11376
+    },
+    {
+      "epoch": 0.5241166444004238,
+      "grad_norm": 3.3060949928696166,
+      "learning_rate": 9.705355141140683e-05,
+      "loss": 1.1951,
+      "step": 11377
+    },
+    {
+      "epoch": 0.5241627124890589,
+      "grad_norm": 1.373371950198766,
+      "learning_rate": 9.703863703599052e-05,
+      "loss": 1.1211,
+      "step": 11378
+    },
+    {
+      "epoch": 0.5242087805776938,
+      "grad_norm": 1.223206882607817,
+      "learning_rate": 9.702372272650391e-05,
+      "loss": 1.3366,
+      "step": 11379
+    },
+    {
+      "epoch": 0.5242548486663289,
+      "grad_norm": 1.3663651205537206,
+      "learning_rate": 9.700880848327899e-05,
+      "loss": 1.2712,
+      "step": 11380
+    },
+    {
+      "epoch": 0.5243009167549638,
+      "grad_norm": 1.6366359092698934,
+      "learning_rate": 9.699389430664783e-05,
+      "loss": 1.0475,
+      "step": 11381
+    },
+    {
+      "epoch": 0.5243469848435989,
+      "grad_norm": 1.8215753911659105,
+      "learning_rate": 9.697898019694244e-05,
+      "loss": 1.2397,
+      "step": 11382
+    },
+    {
+      "epoch": 0.5243930529322338,
+      "grad_norm": 0.8292250397155876,
+      "learning_rate": 9.69640661544949e-05,
+      "loss": 1.5564,
+      "step": 11383
+    },
+    {
+      "epoch": 0.5244391210208689,
+      "grad_norm": 1.3138272725778029,
+      "learning_rate": 9.69491521796372e-05,
+      "loss": 1.5397,
+      "step": 11384
+    },
+    {
+      "epoch": 0.5244851891095038,
+      "grad_norm": 2.0239537605460067,
+      "learning_rate": 9.693423827270143e-05,
+      "loss": 1.2074,
+      "step": 11385
+    },
+    {
+      "epoch": 0.5245312571981389,
+      "grad_norm": 1.1329564298357142,
+      "learning_rate": 9.691932443401958e-05,
+      "loss": 1.4316,
+      "step": 11386
+    },
+    {
+      "epoch": 0.5245773252867738,
+      "grad_norm": 0.8254481996728954,
+      "learning_rate": 9.690441066392366e-05,
+      "loss": 1.5576,
+      "step": 11387
+    },
+    {
+      "epoch": 0.5246233933754089,
+      "grad_norm": 1.223596842998935,
+      "learning_rate": 9.688949696274575e-05,
+      "loss": 1.0921,
+      "step": 11388
+    },
+    {
+      "epoch": 0.5246694614640438,
+      "grad_norm": 1.3069842054591645,
+      "learning_rate": 9.687458333081783e-05,
+      "loss": 1.4919,
+      "step": 11389
+    },
+    {
+      "epoch": 0.5247155295526789,
+      "grad_norm": 1.1833277784837999,
+      "learning_rate": 9.6859669768472e-05,
+      "loss": 1.3425,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5247615976413139,
+      "grad_norm": 0.7786360826163623,
+      "learning_rate": 9.684475627604019e-05,
+      "loss": 1.52,
+      "step": 11391
+    },
+    {
+      "epoch": 0.5248076657299489,
+      "grad_norm": 1.1674557696408336,
+      "learning_rate": 9.682984285385452e-05,
+      "loss": 1.2434,
+      "step": 11392
+    },
+    {
+      "epoch": 0.5248537338185839,
+      "grad_norm": 1.4668614963709887,
+      "learning_rate": 9.681492950224691e-05,
+      "loss": 1.2671,
+      "step": 11393
+    },
+    {
+      "epoch": 0.5248998019072189,
+      "grad_norm": 2.2250164265382057,
+      "learning_rate": 9.680001622154948e-05,
+      "loss": 1.1505,
+      "step": 11394
+    },
+    {
+      "epoch": 0.5249458699958539,
+      "grad_norm": 1.1959515041308695,
+      "learning_rate": 9.678510301209419e-05,
+      "loss": 1.0239,
+      "step": 11395
+    },
+    {
+      "epoch": 0.5249919380844889,
+      "grad_norm": 1.2411012565534854,
+      "learning_rate": 9.677018987421304e-05,
+      "loss": 1.5106,
+      "step": 11396
+    },
+    {
+      "epoch": 0.5250380061731239,
+      "grad_norm": 1.3774787795981225,
+      "learning_rate": 9.675527680823812e-05,
+      "loss": 1.6191,
+      "step": 11397
+    },
+    {
+      "epoch": 0.5250840742617588,
+      "grad_norm": 1.1780497570128168,
+      "learning_rate": 9.674036381450135e-05,
+      "loss": 1.346,
+      "step": 11398
+    },
+    {
+      "epoch": 0.5251301423503939,
+      "grad_norm": 1.1915702036513418,
+      "learning_rate": 9.672545089333483e-05,
+      "loss": 0.8444,
+      "step": 11399
+    },
+    {
+      "epoch": 0.5251762104390288,
+      "grad_norm": 0.9098703296383519,
+      "learning_rate": 9.671053804507051e-05,
+      "loss": 1.5111,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5252222785276639,
+      "grad_norm": 1.174885421998239,
+      "learning_rate": 9.669562527004044e-05,
+      "loss": 1.266,
+      "step": 11401
+    },
+    {
+      "epoch": 0.5252683466162988,
+      "grad_norm": 2.0898789174988233,
+      "learning_rate": 9.668071256857658e-05,
+      "loss": 1.1977,
+      "step": 11402
+    },
+    {
+      "epoch": 0.5253144147049339,
+      "grad_norm": 1.1008349867831304,
+      "learning_rate": 9.666579994101098e-05,
+      "loss": 1.5524,
+      "step": 11403
+    },
+    {
+      "epoch": 0.525360482793569,
+      "grad_norm": 1.0927822734610828,
+      "learning_rate": 9.665088738767563e-05,
+      "loss": 1.1745,
+      "step": 11404
+    },
+    {
+      "epoch": 0.5254065508822039,
+      "grad_norm": 1.2971732821992752,
+      "learning_rate": 9.663597490890252e-05,
+      "loss": 1.3762,
+      "step": 11405
+    },
+    {
+      "epoch": 0.525452618970839,
+      "grad_norm": 0.8791971647418182,
+      "learning_rate": 9.662106250502367e-05,
+      "loss": 1.4966,
+      "step": 11406
+    },
+    {
+      "epoch": 0.5254986870594739,
+      "grad_norm": 9.388930405489495,
+      "learning_rate": 9.660615017637106e-05,
+      "loss": 1.2161,
+      "step": 11407
+    },
+    {
+      "epoch": 0.5255447551481089,
+      "grad_norm": 1.083966635953691,
+      "learning_rate": 9.659123792327671e-05,
+      "loss": 1.4778,
+      "step": 11408
+    },
+    {
+      "epoch": 0.5255908232367439,
+      "grad_norm": 2.1235017699729615,
+      "learning_rate": 9.657632574607258e-05,
+      "loss": 1.6062,
+      "step": 11409
+    },
+    {
+      "epoch": 0.5256368913253789,
+      "grad_norm": 1.8181974114288892,
+      "learning_rate": 9.656141364509071e-05,
+      "loss": 1.0994,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5256829594140139,
+      "grad_norm": 1.1440240296752202,
+      "learning_rate": 9.654650162066305e-05,
+      "loss": 1.4604,
+      "step": 11411
+    },
+    {
+      "epoch": 0.5257290275026489,
+      "grad_norm": 1.2370178382706833,
+      "learning_rate": 9.653158967312164e-05,
+      "loss": 1.6653,
+      "step": 11412
+    },
+    {
+      "epoch": 0.5257750955912839,
+      "grad_norm": 1.4455447976278266,
+      "learning_rate": 9.651667780279845e-05,
+      "loss": 1.3132,
+      "step": 11413
+    },
+    {
+      "epoch": 0.5258211636799189,
+      "grad_norm": 1.204628137324034,
+      "learning_rate": 9.650176601002539e-05,
+      "loss": 1.0743,
+      "step": 11414
+    },
+    {
+      "epoch": 0.5258672317685539,
+      "grad_norm": 8.533449489461221,
+      "learning_rate": 9.648685429513456e-05,
+      "loss": 1.3393,
+      "step": 11415
+    },
+    {
+      "epoch": 0.5259132998571889,
+      "grad_norm": 1.117389989169131,
+      "learning_rate": 9.647194265845788e-05,
+      "loss": 1.5905,
+      "step": 11416
+    },
+    {
+      "epoch": 0.5259593679458239,
+      "grad_norm": 0.9872800139736703,
+      "learning_rate": 9.645703110032735e-05,
+      "loss": 1.1519,
+      "step": 11417
+    },
+    {
+      "epoch": 0.5260054360344589,
+      "grad_norm": 1.2335659702668031,
+      "learning_rate": 9.644211962107494e-05,
+      "loss": 1.1867,
+      "step": 11418
+    },
+    {
+      "epoch": 0.526051504123094,
+      "grad_norm": 1.1722050279055638,
+      "learning_rate": 9.642720822103265e-05,
+      "loss": 1.5409,
+      "step": 11419
+    },
+    {
+      "epoch": 0.5260975722117289,
+      "grad_norm": 2.025909456294796,
+      "learning_rate": 9.641229690053242e-05,
+      "loss": 0.9919,
+      "step": 11420
+    },
+    {
+      "epoch": 0.526143640300364,
+      "grad_norm": 1.0420213586356792,
+      "learning_rate": 9.639738565990628e-05,
+      "loss": 1.6262,
+      "step": 11421
+    },
+    {
+      "epoch": 0.5261897083889989,
+      "grad_norm": 2.4208969892555627,
+      "learning_rate": 9.63824744994862e-05,
+      "loss": 1.2321,
+      "step": 11422
+    },
+    {
+      "epoch": 0.526235776477634,
+      "grad_norm": 1.3277216354141261,
+      "learning_rate": 9.636756341960403e-05,
+      "loss": 1.4833,
+      "step": 11423
+    },
+    {
+      "epoch": 0.5262818445662689,
+      "grad_norm": 1.0986622599315543,
+      "learning_rate": 9.635265242059192e-05,
+      "loss": 1.4539,
+      "step": 11424
+    },
+    {
+      "epoch": 0.526327912654904,
+      "grad_norm": 1.2795606058418438,
+      "learning_rate": 9.633774150278167e-05,
+      "loss": 1.2844,
+      "step": 11425
+    },
+    {
+      "epoch": 0.5263739807435389,
+      "grad_norm": 0.9578948791651783,
+      "learning_rate": 9.63228306665054e-05,
+      "loss": 1.3013,
+      "step": 11426
+    },
+    {
+      "epoch": 0.526420048832174,
+      "grad_norm": 1.0327538138977352,
+      "learning_rate": 9.630791991209493e-05,
+      "loss": 1.5502,
+      "step": 11427
+    },
+    {
+      "epoch": 0.5264661169208089,
+      "grad_norm": 0.8852938103858337,
+      "learning_rate": 9.629300923988235e-05,
+      "loss": 1.5343,
+      "step": 11428
+    },
+    {
+      "epoch": 0.526512185009444,
+      "grad_norm": 2.4859655650296686,
+      "learning_rate": 9.627809865019953e-05,
+      "loss": 1.4548,
+      "step": 11429
+    },
+    {
+      "epoch": 0.5265582530980789,
+      "grad_norm": 1.3431523439144644,
+      "learning_rate": 9.626318814337849e-05,
+      "loss": 1.3539,
+      "step": 11430
+    },
+    {
+      "epoch": 0.526604321186714,
+      "grad_norm": 1.5940114656851434,
+      "learning_rate": 9.624827771975112e-05,
+      "loss": 1.454,
+      "step": 11431
+    },
+    {
+      "epoch": 0.526650389275349,
+      "grad_norm": 1.155357560680655,
+      "learning_rate": 9.623336737964946e-05,
+      "loss": 1.211,
+      "step": 11432
+    },
+    {
+      "epoch": 0.526696457363984,
+      "grad_norm": 1.00577285089757,
+      "learning_rate": 9.621845712340542e-05,
+      "loss": 1.2326,
+      "step": 11433
+    },
+    {
+      "epoch": 0.526742525452619,
+      "grad_norm": 0.8289204898382551,
+      "learning_rate": 9.620354695135088e-05,
+      "loss": 1.5518,
+      "step": 11434
+    },
+    {
+      "epoch": 0.526788593541254,
+      "grad_norm": 1.0478857269461699,
+      "learning_rate": 9.618863686381794e-05,
+      "loss": 1.4143,
+      "step": 11435
+    },
+    {
+      "epoch": 0.526834661629889,
+      "grad_norm": 1.2214719698814385,
+      "learning_rate": 9.617372686113839e-05,
+      "loss": 1.0559,
+      "step": 11436
+    },
+    {
+      "epoch": 0.526880729718524,
+      "grad_norm": 0.9475083336630984,
+      "learning_rate": 9.615881694364432e-05,
+      "loss": 1.4792,
+      "step": 11437
+    },
+    {
+      "epoch": 0.526926797807159,
+      "grad_norm": 1.0581139687885226,
+      "learning_rate": 9.614390711166754e-05,
+      "loss": 1.4257,
+      "step": 11438
+    },
+    {
+      "epoch": 0.526972865895794,
+      "grad_norm": 0.9355691393196599,
+      "learning_rate": 9.612899736554012e-05,
+      "loss": 1.6033,
+      "step": 11439
+    },
+    {
+      "epoch": 0.527018933984429,
+      "grad_norm": 1.0725464349514102,
+      "learning_rate": 9.611408770559387e-05,
+      "loss": 1.5631,
+      "step": 11440
+    },
+    {
+      "epoch": 0.527065002073064,
+      "grad_norm": 1.936590707407087,
+      "learning_rate": 9.609917813216083e-05,
+      "loss": 1.7453,
+      "step": 11441
+    },
+    {
+      "epoch": 0.527111070161699,
+      "grad_norm": 1.8479046225379425,
+      "learning_rate": 9.608426864557293e-05,
+      "loss": 1.1398,
+      "step": 11442
+    },
+    {
+      "epoch": 0.527157138250334,
+      "grad_norm": 1.0097098357499203,
+      "learning_rate": 9.606935924616201e-05,
+      "loss": 1.3945,
+      "step": 11443
+    },
+    {
+      "epoch": 0.527203206338969,
+      "grad_norm": 1.065436035492015,
+      "learning_rate": 9.605444993426014e-05,
+      "loss": 1.1461,
+      "step": 11444
+    },
+    {
+      "epoch": 0.527249274427604,
+      "grad_norm": 1.1401311719397522,
+      "learning_rate": 9.603954071019909e-05,
+      "loss": 1.313,
+      "step": 11445
+    },
+    {
+      "epoch": 0.527295342516239,
+      "grad_norm": 1.1667376736557917,
+      "learning_rate": 9.602463157431096e-05,
+      "loss": 1.2576,
+      "step": 11446
+    },
+    {
+      "epoch": 0.527341410604874,
+      "grad_norm": 1.1808378578546557,
+      "learning_rate": 9.600972252692751e-05,
+      "loss": 1.3002,
+      "step": 11447
+    },
+    {
+      "epoch": 0.527387478693509,
+      "grad_norm": 1.4144736947125596,
+      "learning_rate": 9.599481356838083e-05,
+      "loss": 1.1021,
+      "step": 11448
+    },
+    {
+      "epoch": 0.527433546782144,
+      "grad_norm": 1.8823393697109847,
+      "learning_rate": 9.597990469900268e-05,
+      "loss": 1.3488,
+      "step": 11449
+    },
+    {
+      "epoch": 0.527479614870779,
+      "grad_norm": 1.1542667678944267,
+      "learning_rate": 9.596499591912513e-05,
+      "loss": 1.265,
+      "step": 11450
+    },
+    {
+      "epoch": 0.527525682959414,
+      "grad_norm": 0.8758563991980967,
+      "learning_rate": 9.595008722908001e-05,
+      "loss": 1.4975,
+      "step": 11451
+    },
+    {
+      "epoch": 0.527571751048049,
+      "grad_norm": 3.134424161745272,
+      "learning_rate": 9.593517862919922e-05,
+      "loss": 1.1986,
+      "step": 11452
+    },
+    {
+      "epoch": 0.527617819136684,
+      "grad_norm": 3.920585993152053,
+      "learning_rate": 9.592027011981474e-05,
+      "loss": 1.276,
+      "step": 11453
+    },
+    {
+      "epoch": 0.527663887225319,
+      "grad_norm": 1.0558204487957752,
+      "learning_rate": 9.59053617012584e-05,
+      "loss": 1.5422,
+      "step": 11454
+    },
+    {
+      "epoch": 0.527709955313954,
+      "grad_norm": 1.2632013171957766,
+      "learning_rate": 9.589045337386223e-05,
+      "loss": 1.2826,
+      "step": 11455
+    },
+    {
+      "epoch": 0.527756023402589,
+      "grad_norm": 1.6437063799290026,
+      "learning_rate": 9.5875545137958e-05,
+      "loss": 1.3896,
+      "step": 11456
+    },
+    {
+      "epoch": 0.527802091491224,
+      "grad_norm": 2.8851861519345308,
+      "learning_rate": 9.586063699387776e-05,
+      "loss": 0.9746,
+      "step": 11457
+    },
+    {
+      "epoch": 0.527848159579859,
+      "grad_norm": 1.0871214122499686,
+      "learning_rate": 9.584572894195325e-05,
+      "loss": 1.242,
+      "step": 11458
+    },
+    {
+      "epoch": 0.527894227668494,
+      "grad_norm": 1.4086134383816122,
+      "learning_rate": 9.583082098251657e-05,
+      "loss": 1.0157,
+      "step": 11459
+    },
+    {
+      "epoch": 0.5279402957571291,
+      "grad_norm": 0.9443014844547262,
+      "learning_rate": 9.581591311589946e-05,
+      "loss": 1.6539,
+      "step": 11460
+    },
+    {
+      "epoch": 0.527986363845764,
+      "grad_norm": 1.4364682992678652,
+      "learning_rate": 9.580100534243385e-05,
+      "loss": 1.2985,
+      "step": 11461
+    },
+    {
+      "epoch": 0.5280324319343991,
+      "grad_norm": 3.4817530876333245,
+      "learning_rate": 9.578609766245169e-05,
+      "loss": 1.3431,
+      "step": 11462
+    },
+    {
+      "epoch": 0.528078500023034,
+      "grad_norm": 1.7856332252585563,
+      "learning_rate": 9.57711900762848e-05,
+      "loss": 1.3828,
+      "step": 11463
+    },
+    {
+      "epoch": 0.5281245681116691,
+      "grad_norm": 0.8160676555362798,
+      "learning_rate": 9.575628258426516e-05,
+      "loss": 1.5028,
+      "step": 11464
+    },
+    {
+      "epoch": 0.528170636200304,
+      "grad_norm": 1.883404629043889,
+      "learning_rate": 9.574137518672457e-05,
+      "loss": 1.1439,
+      "step": 11465
+    },
+    {
+      "epoch": 0.5282167042889391,
+      "grad_norm": 1.7192396470154554,
+      "learning_rate": 9.572646788399499e-05,
+      "loss": 1.1395,
+      "step": 11466
+    },
+    {
+      "epoch": 0.528262772377574,
+      "grad_norm": 4.555511039396264,
+      "learning_rate": 9.571156067640823e-05,
+      "loss": 1.1422,
+      "step": 11467
+    },
+    {
+      "epoch": 0.5283088404662091,
+      "grad_norm": 1.0774957507871536,
+      "learning_rate": 9.569665356429628e-05,
+      "loss": 1.5883,
+      "step": 11468
+    },
+    {
+      "epoch": 0.528354908554844,
+      "grad_norm": 2.6432029751828767,
+      "learning_rate": 9.568174654799094e-05,
+      "loss": 1.3397,
+      "step": 11469
+    },
+    {
+      "epoch": 0.5284009766434791,
+      "grad_norm": 1.8843169487213216,
+      "learning_rate": 9.566683962782408e-05,
+      "loss": 1.4596,
+      "step": 11470
+    },
+    {
+      "epoch": 0.528447044732114,
+      "grad_norm": 1.509795563277276,
+      "learning_rate": 9.565193280412764e-05,
+      "loss": 1.4345,
+      "step": 11471
+    },
+    {
+      "epoch": 0.5284931128207491,
+      "grad_norm": 25.920082267060476,
+      "learning_rate": 9.563702607723342e-05,
+      "loss": 1.226,
+      "step": 11472
+    },
+    {
+      "epoch": 0.5285391809093841,
+      "grad_norm": 1.429164700976811,
+      "learning_rate": 9.562211944747338e-05,
+      "loss": 1.469,
+      "step": 11473
+    },
+    {
+      "epoch": 0.5285852489980191,
+      "grad_norm": 1.1664458922033687,
+      "learning_rate": 9.560721291517931e-05,
+      "loss": 1.5341,
+      "step": 11474
+    },
+    {
+      "epoch": 0.5286313170866541,
+      "grad_norm": 2.214743140131127,
+      "learning_rate": 9.559230648068313e-05,
+      "loss": 1.1409,
+      "step": 11475
+    },
+    {
+      "epoch": 0.5286773851752891,
+      "grad_norm": 2.5473033233173696,
+      "learning_rate": 9.557740014431667e-05,
+      "loss": 1.2711,
+      "step": 11476
+    },
+    {
+      "epoch": 0.5287234532639241,
+      "grad_norm": 2.259503681314892,
+      "learning_rate": 9.556249390641184e-05,
+      "loss": 1.1338,
+      "step": 11477
+    },
+    {
+      "epoch": 0.5287695213525591,
+      "grad_norm": 1.8039295638234158,
+      "learning_rate": 9.554758776730049e-05,
+      "loss": 1.5018,
+      "step": 11478
+    },
+    {
+      "epoch": 0.5288155894411941,
+      "grad_norm": 9.077388028644858,
+      "learning_rate": 9.55326817273144e-05,
+      "loss": 1.3558,
+      "step": 11479
+    },
+    {
+      "epoch": 0.528861657529829,
+      "grad_norm": 1.736713382814658,
+      "learning_rate": 9.551777578678555e-05,
+      "loss": 1.4043,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5289077256184641,
+      "grad_norm": 1.616510996867352,
+      "learning_rate": 9.55028699460457e-05,
+      "loss": 1.4212,
+      "step": 11481
+    },
+    {
+      "epoch": 0.528953793707099,
+      "grad_norm": 1.9531748027965594,
+      "learning_rate": 9.548796420542677e-05,
+      "loss": 1.467,
+      "step": 11482
+    },
+    {
+      "epoch": 0.5289998617957341,
+      "grad_norm": 3.286258077992849,
+      "learning_rate": 9.547305856526054e-05,
+      "loss": 1.1874,
+      "step": 11483
+    },
+    {
+      "epoch": 0.529045929884369,
+      "grad_norm": 2.5706220010447214,
+      "learning_rate": 9.545815302587894e-05,
+      "loss": 1.3484,
+      "step": 11484
+    },
+    {
+      "epoch": 0.5290919979730041,
+      "grad_norm": 4.326011754450391,
+      "learning_rate": 9.544324758761376e-05,
+      "loss": 1.3565,
+      "step": 11485
+    },
+    {
+      "epoch": 0.5291380660616392,
+      "grad_norm": 1.8608292557383033,
+      "learning_rate": 9.542834225079686e-05,
+      "loss": 1.5065,
+      "step": 11486
+    },
+    {
+      "epoch": 0.5291841341502741,
+      "grad_norm": 4.970303303308014,
+      "learning_rate": 9.541343701576008e-05,
+      "loss": 1.4752,
+      "step": 11487
+    },
+    {
+      "epoch": 0.5292302022389092,
+      "grad_norm": 4.661686924267525,
+      "learning_rate": 9.539853188283526e-05,
+      "loss": 1.3477,
+      "step": 11488
+    },
+    {
+      "epoch": 0.5292762703275441,
+      "grad_norm": 2.7116540545471413,
+      "learning_rate": 9.538362685235424e-05,
+      "loss": 1.3438,
+      "step": 11489
+    },
+    {
+      "epoch": 0.5293223384161792,
+      "grad_norm": 2.30195361884438,
+      "learning_rate": 9.536872192464883e-05,
+      "loss": 1.326,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5293684065048141,
+      "grad_norm": 3.0611283039177133,
+      "learning_rate": 9.535381710005092e-05,
+      "loss": 1.5674,
+      "step": 11491
+    },
+    {
+      "epoch": 0.5294144745934491,
+      "grad_norm": 2.331782251792762,
+      "learning_rate": 9.533891237889227e-05,
+      "loss": 1.1575,
+      "step": 11492
+    },
+    {
+      "epoch": 0.5294605426820841,
+      "grad_norm": 1.4485798610237552,
+      "learning_rate": 9.532400776150479e-05,
+      "loss": 1.3596,
+      "step": 11493
+    },
+    {
+      "epoch": 0.5295066107707191,
+      "grad_norm": 2.1436777417597224,
+      "learning_rate": 9.530910324822021e-05,
+      "loss": 1.4411,
+      "step": 11494
+    },
+    {
+      "epoch": 0.5295526788593541,
+      "grad_norm": 2.396356492765842,
+      "learning_rate": 9.529419883937044e-05,
+      "loss": 1.1223,
+      "step": 11495
+    },
+    {
+      "epoch": 0.5295987469479891,
+      "grad_norm": 1.6426144882057372,
+      "learning_rate": 9.527929453528727e-05,
+      "loss": 1.1186,
+      "step": 11496
+    },
+    {
+      "epoch": 0.5296448150366241,
+      "grad_norm": 1.152705451508126,
+      "learning_rate": 9.526439033630248e-05,
+      "loss": 1.31,
+      "step": 11497
+    },
+    {
+      "epoch": 0.5296908831252591,
+      "grad_norm": 3.269896017563574,
+      "learning_rate": 9.524948624274794e-05,
+      "loss": 1.1772,
+      "step": 11498
+    },
+    {
+      "epoch": 0.5297369512138941,
+      "grad_norm": 2.4462982164611984,
+      "learning_rate": 9.523458225495544e-05,
+      "loss": 1.3795,
+      "step": 11499
+    },
+    {
+      "epoch": 0.5297830193025291,
+      "grad_norm": 0.9459895716795057,
+      "learning_rate": 9.521967837325678e-05,
+      "loss": 1.334,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5298290873911642,
+      "grad_norm": 1.4229558253586079,
+      "learning_rate": 9.52047745979838e-05,
+      "loss": 1.1411,
+      "step": 11501
+    },
+    {
+      "epoch": 0.5298751554797991,
+      "grad_norm": 1.2379010162405524,
+      "learning_rate": 9.51898709294683e-05,
+      "loss": 1.4299,
+      "step": 11502
+    },
+    {
+      "epoch": 0.5299212235684342,
+      "grad_norm": 1.279866150178876,
+      "learning_rate": 9.517496736804204e-05,
+      "loss": 1.7102,
+      "step": 11503
+    },
+    {
+      "epoch": 0.5299672916570691,
+      "grad_norm": 3.0806358267843295,
+      "learning_rate": 9.51600639140369e-05,
+      "loss": 1.2861,
+      "step": 11504
+    },
+    {
+      "epoch": 0.5300133597457042,
+      "grad_norm": 1.805127995013149,
+      "learning_rate": 9.514516056778463e-05,
+      "loss": 1.9305,
+      "step": 11505
+    },
+    {
+      "epoch": 0.5300594278343391,
+      "grad_norm": 1.2646690736513486,
+      "learning_rate": 9.513025732961701e-05,
+      "loss": 1.4511,
+      "step": 11506
+    },
+    {
+      "epoch": 0.5301054959229742,
+      "grad_norm": 2.1101459675787857,
+      "learning_rate": 9.511535419986589e-05,
+      "loss": 1.3701,
+      "step": 11507
+    },
+    {
+      "epoch": 0.5301515640116091,
+      "grad_norm": 4.5891126633817,
+      "learning_rate": 9.510045117886301e-05,
+      "loss": 1.456,
+      "step": 11508
+    },
+    {
+      "epoch": 0.5301976321002442,
+      "grad_norm": 1.9485275754435016,
+      "learning_rate": 9.50855482669402e-05,
+      "loss": 1.3122,
+      "step": 11509
+    },
+    {
+      "epoch": 0.5302437001888791,
+      "grad_norm": 1.168354289481421,
+      "learning_rate": 9.507064546442922e-05,
+      "loss": 1.2108,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5302897682775142,
+      "grad_norm": 2.7911570568296034,
+      "learning_rate": 9.505574277166188e-05,
+      "loss": 1.5197,
+      "step": 11511
+    },
+    {
+      "epoch": 0.5303358363661491,
+      "grad_norm": 1.5975315765382723,
+      "learning_rate": 9.504084018896992e-05,
+      "loss": 1.1994,
+      "step": 11512
+    },
+    {
+      "epoch": 0.5303819044547842,
+      "grad_norm": 2.464190472000926,
+      "learning_rate": 9.502593771668519e-05,
+      "loss": 1.3565,
+      "step": 11513
+    },
+    {
+      "epoch": 0.5304279725434192,
+      "grad_norm": 1.2163108311146686,
+      "learning_rate": 9.501103535513942e-05,
+      "loss": 1.4673,
+      "step": 11514
+    },
+    {
+      "epoch": 0.5304740406320542,
+      "grad_norm": 2.1392852927238803,
+      "learning_rate": 9.499613310466437e-05,
+      "loss": 1.0022,
+      "step": 11515
+    },
+    {
+      "epoch": 0.5305201087206892,
+      "grad_norm": 8.24998063513023,
+      "learning_rate": 9.498123096559186e-05,
+      "loss": 1.0928,
+      "step": 11516
+    },
+    {
+      "epoch": 0.5305661768093242,
+      "grad_norm": 1.6178093093095745,
+      "learning_rate": 9.496632893825363e-05,
+      "loss": 1.2504,
+      "step": 11517
+    },
+    {
+      "epoch": 0.5306122448979592,
+      "grad_norm": 3.774861229225861,
+      "learning_rate": 9.495142702298145e-05,
+      "loss": 1.3391,
+      "step": 11518
+    },
+    {
+      "epoch": 0.5306583129865942,
+      "grad_norm": 2.288187313674821,
+      "learning_rate": 9.493652522010708e-05,
+      "loss": 1.4205,
+      "step": 11519
+    },
+    {
+      "epoch": 0.5307043810752292,
+      "grad_norm": 10.242098680059915,
+      "learning_rate": 9.492162352996232e-05,
+      "loss": 1.3045,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5307504491638642,
+      "grad_norm": 2.1890258786344963,
+      "learning_rate": 9.490672195287888e-05,
+      "loss": 1.304,
+      "step": 11521
+    },
+    {
+      "epoch": 0.5307965172524992,
+      "grad_norm": 1.8135913052646682,
+      "learning_rate": 9.489182048918856e-05,
+      "loss": 1.2139,
+      "step": 11522
+    },
+    {
+      "epoch": 0.5308425853411342,
+      "grad_norm": 2.527643593219692,
+      "learning_rate": 9.487691913922312e-05,
+      "loss": 1.3125,
+      "step": 11523
+    },
+    {
+      "epoch": 0.5308886534297692,
+      "grad_norm": 0.9289972524976661,
+      "learning_rate": 9.486201790331424e-05,
+      "loss": 1.7032,
+      "step": 11524
+    },
+    {
+      "epoch": 0.5309347215184042,
+      "grad_norm": 1.0657351425021677,
+      "learning_rate": 9.484711678179377e-05,
+      "loss": 1.4996,
+      "step": 11525
+    },
+    {
+      "epoch": 0.5309807896070392,
+      "grad_norm": 1.5406353888999058,
+      "learning_rate": 9.483221577499336e-05,
+      "loss": 1.3356,
+      "step": 11526
+    },
+    {
+      "epoch": 0.5310268576956743,
+      "grad_norm": 2.1971832776415043,
+      "learning_rate": 9.481731488324484e-05,
+      "loss": 1.3567,
+      "step": 11527
+    },
+    {
+      "epoch": 0.5310729257843092,
+      "grad_norm": 2.604696328535365,
+      "learning_rate": 9.48024141068799e-05,
+      "loss": 1.4312,
+      "step": 11528
+    },
+    {
+      "epoch": 0.5311189938729443,
+      "grad_norm": 1.273871681684133,
+      "learning_rate": 9.478751344623031e-05,
+      "loss": 1.4238,
+      "step": 11529
+    },
+    {
+      "epoch": 0.5311650619615792,
+      "grad_norm": 7.212599033625544,
+      "learning_rate": 9.477261290162779e-05,
+      "loss": 1.3752,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5312111300502143,
+      "grad_norm": 1.3267736242136678,
+      "learning_rate": 9.475771247340409e-05,
+      "loss": 1.6167,
+      "step": 11531
+    },
+    {
+      "epoch": 0.5312571981388492,
+      "grad_norm": 1.806392696989968,
+      "learning_rate": 9.474281216189096e-05,
+      "loss": 1.2536,
+      "step": 11532
+    },
+    {
+      "epoch": 0.5313032662274842,
+      "grad_norm": 1.763138756603292,
+      "learning_rate": 9.472791196742003e-05,
+      "loss": 1.3601,
+      "step": 11533
+    },
+    {
+      "epoch": 0.5313493343161192,
+      "grad_norm": 3.1596655982320656,
+      "learning_rate": 9.471301189032317e-05,
+      "loss": 1.4169,
+      "step": 11534
+    },
+    {
+      "epoch": 0.5313954024047542,
+      "grad_norm": 1.3210016834408054,
+      "learning_rate": 9.469811193093196e-05,
+      "loss": 1.3203,
+      "step": 11535
+    },
+    {
+      "epoch": 0.5314414704933892,
+      "grad_norm": 1.582601591754497,
+      "learning_rate": 9.468321208957828e-05,
+      "loss": 1.2441,
+      "step": 11536
+    },
+    {
+      "epoch": 0.5314875385820242,
+      "grad_norm": 1.821607812631824,
+      "learning_rate": 9.466831236659369e-05,
+      "loss": 1.4632,
+      "step": 11537
+    },
+    {
+      "epoch": 0.5315336066706592,
+      "grad_norm": 5.856723825666248,
+      "learning_rate": 9.465341276231005e-05,
+      "loss": 1.2395,
+      "step": 11538
+    },
+    {
+      "epoch": 0.5315796747592942,
+      "grad_norm": 3.0549720090029457,
+      "learning_rate": 9.463851327705896e-05,
+      "loss": 1.5548,
+      "step": 11539
+    },
+    {
+      "epoch": 0.5316257428479293,
+      "grad_norm": 2.03995586053443,
+      "learning_rate": 9.46236139111722e-05,
+      "loss": 1.472,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5316718109365642,
+      "grad_norm": 1.8249495908059667,
+      "learning_rate": 9.460871466498149e-05,
+      "loss": 0.9814,
+      "step": 11541
+    },
+    {
+      "epoch": 0.5317178790251993,
+      "grad_norm": 1.6023799228850404,
+      "learning_rate": 9.459381553881844e-05,
+      "loss": 1.1547,
+      "step": 11542
+    },
+    {
+      "epoch": 0.5317639471138342,
+      "grad_norm": 0.9856582314091292,
+      "learning_rate": 9.457891653301488e-05,
+      "loss": 1.6205,
+      "step": 11543
+    },
+    {
+      "epoch": 0.5318100152024693,
+      "grad_norm": 1.636647824904356,
+      "learning_rate": 9.456401764790237e-05,
+      "loss": 1.2935,
+      "step": 11544
+    },
+    {
+      "epoch": 0.5318560832911042,
+      "grad_norm": 1.5921843107700873,
+      "learning_rate": 9.454911888381277e-05,
+      "loss": 1.0016,
+      "step": 11545
+    },
+    {
+      "epoch": 0.5319021513797393,
+      "grad_norm": 1.6827447422729997,
+      "learning_rate": 9.453422024107761e-05,
+      "loss": 1.4653,
+      "step": 11546
+    },
+    {
+      "epoch": 0.5319482194683742,
+      "grad_norm": 0.7886286778174599,
+      "learning_rate": 9.451932172002874e-05,
+      "loss": 1.6761,
+      "step": 11547
+    },
+    {
+      "epoch": 0.5319942875570093,
+      "grad_norm": 1.058293717332798,
+      "learning_rate": 9.450442332099771e-05,
+      "loss": 1.4792,
+      "step": 11548
+    },
+    {
+      "epoch": 0.5320403556456442,
+      "grad_norm": 2.7123704888303957,
+      "learning_rate": 9.448952504431634e-05,
+      "loss": 1.406,
+      "step": 11549
+    },
+    {
+      "epoch": 0.5320864237342793,
+      "grad_norm": 1.2188590112898159,
+      "learning_rate": 9.447462689031618e-05,
+      "loss": 1.3161,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5321324918229142,
+      "grad_norm": 1.4013046496085748,
+      "learning_rate": 9.445972885932904e-05,
+      "loss": 1.5655,
+      "step": 11551
+    },
+    {
+      "epoch": 0.5321785599115493,
+      "grad_norm": 1.0862511602890148,
+      "learning_rate": 9.444483095168652e-05,
+      "loss": 1.3837,
+      "step": 11552
+    },
+    {
+      "epoch": 0.5322246280001842,
+      "grad_norm": 1.8797020444128394,
+      "learning_rate": 9.442993316772025e-05,
+      "loss": 1.4186,
+      "step": 11553
+    },
+    {
+      "epoch": 0.5322706960888193,
+      "grad_norm": 1.02720240636564,
+      "learning_rate": 9.441503550776205e-05,
+      "loss": 1.3512,
+      "step": 11554
+    },
+    {
+      "epoch": 0.5323167641774543,
+      "grad_norm": 0.8609311253054887,
+      "learning_rate": 9.440013797214344e-05,
+      "loss": 1.3723,
+      "step": 11555
+    },
+    {
+      "epoch": 0.5323628322660893,
+      "grad_norm": 1.7975208026149543,
+      "learning_rate": 9.438524056119623e-05,
+      "loss": 1.4165,
+      "step": 11556
+    },
+    {
+      "epoch": 0.5324089003547243,
+      "grad_norm": 1.2662675351373016,
+      "learning_rate": 9.437034327525194e-05,
+      "loss": 1.3382,
+      "step": 11557
+    },
+    {
+      "epoch": 0.5324549684433593,
+      "grad_norm": 0.8068525471187602,
+      "learning_rate": 9.435544611464237e-05,
+      "loss": 1.4686,
+      "step": 11558
+    },
+    {
+      "epoch": 0.5325010365319943,
+      "grad_norm": 1.2794048066851356,
+      "learning_rate": 9.434054907969905e-05,
+      "loss": 1.2691,
+      "step": 11559
+    },
+    {
+      "epoch": 0.5325471046206293,
+      "grad_norm": 1.3048235270807473,
+      "learning_rate": 9.432565217075375e-05,
+      "loss": 1.264,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5325931727092643,
+      "grad_norm": 1.0229772369898757,
+      "learning_rate": 9.431075538813808e-05,
+      "loss": 1.5999,
+      "step": 11561
+    },
+    {
+      "epoch": 0.5326392407978993,
+      "grad_norm": 1.5957649051133993,
+      "learning_rate": 9.429585873218365e-05,
+      "loss": 1.3536,
+      "step": 11562
+    },
+    {
+      "epoch": 0.5326853088865343,
+      "grad_norm": 1.8270374294002987,
+      "learning_rate": 9.428096220322217e-05,
+      "loss": 1.4511,
+      "step": 11563
+    },
+    {
+      "epoch": 0.5327313769751693,
+      "grad_norm": 1.1997254252419693,
+      "learning_rate": 9.426606580158523e-05,
+      "loss": 1.1854,
+      "step": 11564
+    },
+    {
+      "epoch": 0.5327774450638043,
+      "grad_norm": 2.4091608464857943,
+      "learning_rate": 9.425116952760454e-05,
+      "loss": 1.4695,
+      "step": 11565
+    },
+    {
+      "epoch": 0.5328235131524393,
+      "grad_norm": 1.3358347748758848,
+      "learning_rate": 9.423627338161166e-05,
+      "loss": 1.3626,
+      "step": 11566
+    },
+    {
+      "epoch": 0.5328695812410743,
+      "grad_norm": 2.577502566878929,
+      "learning_rate": 9.42213773639383e-05,
+      "loss": 1.4398,
+      "step": 11567
+    },
+    {
+      "epoch": 0.5329156493297094,
+      "grad_norm": 2.4335405490988955,
+      "learning_rate": 9.420648147491605e-05,
+      "loss": 1.3777,
+      "step": 11568
+    },
+    {
+      "epoch": 0.5329617174183443,
+      "grad_norm": 1.6544184222174754,
+      "learning_rate": 9.419158571487659e-05,
+      "loss": 1.2952,
+      "step": 11569
+    },
+    {
+      "epoch": 0.5330077855069794,
+      "grad_norm": 1.8646945546235,
+      "learning_rate": 9.41766900841515e-05,
+      "loss": 1.1288,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5330538535956143,
+      "grad_norm": 2.0570919288816962,
+      "learning_rate": 9.41617945830724e-05,
+      "loss": 1.4571,
+      "step": 11571
+    },
+    {
+      "epoch": 0.5330999216842494,
+      "grad_norm": 1.4300794953377658,
+      "learning_rate": 9.414689921197096e-05,
+      "loss": 1.5596,
+      "step": 11572
+    },
+    {
+      "epoch": 0.5331459897728843,
+      "grad_norm": 5.1047025932899315,
+      "learning_rate": 9.413200397117875e-05,
+      "loss": 1.1141,
+      "step": 11573
+    },
+    {
+      "epoch": 0.5331920578615194,
+      "grad_norm": 1.540759235909657,
+      "learning_rate": 9.411710886102743e-05,
+      "loss": 1.2556,
+      "step": 11574
+    },
+    {
+      "epoch": 0.5332381259501543,
+      "grad_norm": 3.0949348979880305,
+      "learning_rate": 9.410221388184857e-05,
+      "loss": 1.0776,
+      "step": 11575
+    },
+    {
+      "epoch": 0.5332841940387893,
+      "grad_norm": 5.073094068634337,
+      "learning_rate": 9.408731903397382e-05,
+      "loss": 1.4203,
+      "step": 11576
+    },
+    {
+      "epoch": 0.5333302621274243,
+      "grad_norm": 2.678601534583723,
+      "learning_rate": 9.407242431773476e-05,
+      "loss": 1.257,
+      "step": 11577
+    },
+    {
+      "epoch": 0.5333763302160593,
+      "grad_norm": 1.7450403551035016,
+      "learning_rate": 9.405752973346304e-05,
+      "loss": 1.5023,
+      "step": 11578
+    },
+    {
+      "epoch": 0.5334223983046943,
+      "grad_norm": 1.2438986056504049,
+      "learning_rate": 9.404263528149022e-05,
+      "loss": 1.456,
+      "step": 11579
+    },
+    {
+      "epoch": 0.5334684663933293,
+      "grad_norm": 0.8942095787576653,
+      "learning_rate": 9.402774096214788e-05,
+      "loss": 1.5163,
+      "step": 11580
+    },
+    {
+      "epoch": 0.5335145344819644,
+      "grad_norm": 2.7944118028626006,
+      "learning_rate": 9.401284677576768e-05,
+      "loss": 1.3171,
+      "step": 11581
+    },
+    {
+      "epoch": 0.5335606025705993,
+      "grad_norm": 1.249558753430973,
+      "learning_rate": 9.399795272268117e-05,
+      "loss": 1.4769,
+      "step": 11582
+    },
+    {
+      "epoch": 0.5336066706592344,
+      "grad_norm": 2.56366137085162,
+      "learning_rate": 9.398305880321995e-05,
+      "loss": 1.2809,
+      "step": 11583
+    },
+    {
+      "epoch": 0.5336527387478693,
+      "grad_norm": 1.2227880556276751,
+      "learning_rate": 9.39681650177156e-05,
+      "loss": 1.5421,
+      "step": 11584
+    },
+    {
+      "epoch": 0.5336988068365044,
+      "grad_norm": 0.9002320826328721,
+      "learning_rate": 9.395327136649975e-05,
+      "loss": 1.4483,
+      "step": 11585
+    },
+    {
+      "epoch": 0.5337448749251393,
+      "grad_norm": 2.27286706807954,
+      "learning_rate": 9.39383778499039e-05,
+      "loss": 1.6071,
+      "step": 11586
+    },
+    {
+      "epoch": 0.5337909430137744,
+      "grad_norm": 1.2936313169534162,
+      "learning_rate": 9.39234844682597e-05,
+      "loss": 1.2216,
+      "step": 11587
+    },
+    {
+      "epoch": 0.5338370111024093,
+      "grad_norm": 3.0207454369547087,
+      "learning_rate": 9.390859122189872e-05,
+      "loss": 1.3442,
+      "step": 11588
+    },
+    {
+      "epoch": 0.5338830791910444,
+      "grad_norm": 1.5058582504380054,
+      "learning_rate": 9.389369811115247e-05,
+      "loss": 1.3432,
+      "step": 11589
+    },
+    {
+      "epoch": 0.5339291472796793,
+      "grad_norm": 2.8699642334418463,
+      "learning_rate": 9.387880513635258e-05,
+      "loss": 1.0958,
+      "step": 11590
+    },
+    {
+      "epoch": 0.5339752153683144,
+      "grad_norm": 1.7549208513870453,
+      "learning_rate": 9.386391229783057e-05,
+      "loss": 1.7435,
+      "step": 11591
+    },
+    {
+      "epoch": 0.5340212834569493,
+      "grad_norm": 1.5639982426428882,
+      "learning_rate": 9.384901959591807e-05,
+      "loss": 1.2315,
+      "step": 11592
+    },
+    {
+      "epoch": 0.5340673515455844,
+      "grad_norm": 5.137441192734479,
+      "learning_rate": 9.383412703094659e-05,
+      "loss": 1.3173,
+      "step": 11593
+    },
+    {
+      "epoch": 0.5341134196342193,
+      "grad_norm": 0.8719542488246855,
+      "learning_rate": 9.38192346032477e-05,
+      "loss": 1.1699,
+      "step": 11594
+    },
+    {
+      "epoch": 0.5341594877228544,
+      "grad_norm": 1.181927809524169,
+      "learning_rate": 9.380434231315293e-05,
+      "loss": 1.5914,
+      "step": 11595
+    },
+    {
+      "epoch": 0.5342055558114894,
+      "grad_norm": 2.634532903107933,
+      "learning_rate": 9.37894501609939e-05,
+      "loss": 1.2997,
+      "step": 11596
+    },
+    {
+      "epoch": 0.5342516239001244,
+      "grad_norm": 1.227886639413389,
+      "learning_rate": 9.37745581471021e-05,
+      "loss": 1.4744,
+      "step": 11597
+    },
+    {
+      "epoch": 0.5342976919887594,
+      "grad_norm": 1.4452182098393491,
+      "learning_rate": 9.375966627180905e-05,
+      "loss": 1.5677,
+      "step": 11598
+    },
+    {
+      "epoch": 0.5343437600773944,
+      "grad_norm": 2.3447476560494662,
+      "learning_rate": 9.374477453544637e-05,
+      "loss": 1.3851,
+      "step": 11599
+    },
+    {
+      "epoch": 0.5343898281660294,
+      "grad_norm": 3.9595262261870543,
+      "learning_rate": 9.372988293834555e-05,
+      "loss": 1.2432,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5344358962546644,
+      "grad_norm": 4.818832710235684,
+      "learning_rate": 9.371499148083814e-05,
+      "loss": 1.3134,
+      "step": 11601
+    },
+    {
+      "epoch": 0.5344819643432994,
+      "grad_norm": 1.6554288968220385,
+      "learning_rate": 9.370010016325565e-05,
+      "loss": 1.4255,
+      "step": 11602
+    },
+    {
+      "epoch": 0.5345280324319344,
+      "grad_norm": 132.93899642127658,
+      "learning_rate": 9.368520898592964e-05,
+      "loss": 1.2977,
+      "step": 11603
+    },
+    {
+      "epoch": 0.5345741005205694,
+      "grad_norm": 2.7065474160852965,
+      "learning_rate": 9.367031794919162e-05,
+      "loss": 1.1258,
+      "step": 11604
+    },
+    {
+      "epoch": 0.5346201686092044,
+      "grad_norm": 2.6760283990120746,
+      "learning_rate": 9.365542705337313e-05,
+      "loss": 1.3145,
+      "step": 11605
+    },
+    {
+      "epoch": 0.5346662366978394,
+      "grad_norm": 2.5819510935985464,
+      "learning_rate": 9.364053629880568e-05,
+      "loss": 1.6655,
+      "step": 11606
+    },
+    {
+      "epoch": 0.5347123047864744,
+      "grad_norm": 3.4211591564636232,
+      "learning_rate": 9.362564568582077e-05,
+      "loss": 1.2891,
+      "step": 11607
+    },
+    {
+      "epoch": 0.5347583728751094,
+      "grad_norm": 4.0392408228167715,
+      "learning_rate": 9.361075521474993e-05,
+      "loss": 1.7276,
+      "step": 11608
+    },
+    {
+      "epoch": 0.5348044409637445,
+      "grad_norm": 11.067459072941462,
+      "learning_rate": 9.359586488592467e-05,
+      "loss": 1.7529,
+      "step": 11609
+    },
+    {
+      "epoch": 0.5348505090523794,
+      "grad_norm": 3.7407760623225497,
+      "learning_rate": 9.358097469967652e-05,
+      "loss": 1.629,
+      "step": 11610
+    },
+    {
+      "epoch": 0.5348965771410145,
+      "grad_norm": 2.4952043911898376,
+      "learning_rate": 9.356608465633693e-05,
+      "loss": 1.6642,
+      "step": 11611
+    },
+    {
+      "epoch": 0.5349426452296494,
+      "grad_norm": 1.9251979812519808,
+      "learning_rate": 9.355119475623745e-05,
+      "loss": 1.6363,
+      "step": 11612
+    },
+    {
+      "epoch": 0.5349887133182845,
+      "grad_norm": 2.639071178120629,
+      "learning_rate": 9.353630499970955e-05,
+      "loss": 1.5847,
+      "step": 11613
+    },
+    {
+      "epoch": 0.5350347814069194,
+      "grad_norm": 3.830544582480521,
+      "learning_rate": 9.352141538708476e-05,
+      "loss": 1.4557,
+      "step": 11614
+    },
+    {
+      "epoch": 0.5350808494955545,
+      "grad_norm": 2.4061350469031355,
+      "learning_rate": 9.350652591869454e-05,
+      "loss": 1.6068,
+      "step": 11615
+    },
+    {
+      "epoch": 0.5351269175841894,
+      "grad_norm": 2.369295397703836,
+      "learning_rate": 9.349163659487037e-05,
+      "loss": 1.2317,
+      "step": 11616
+    },
+    {
+      "epoch": 0.5351729856728245,
+      "grad_norm": 3.70800070668108,
+      "learning_rate": 9.347674741594376e-05,
+      "loss": 1.2454,
+      "step": 11617
+    },
+    {
+      "epoch": 0.5352190537614594,
+      "grad_norm": 3.8290192260550215,
+      "learning_rate": 9.346185838224619e-05,
+      "loss": 1.423,
+      "step": 11618
+    },
+    {
+      "epoch": 0.5352651218500944,
+      "grad_norm": 6.529175365915309,
+      "learning_rate": 9.344696949410912e-05,
+      "loss": 1.2347,
+      "step": 11619
+    },
+    {
+      "epoch": 0.5353111899387294,
+      "grad_norm": 1.2237805180151142,
+      "learning_rate": 9.343208075186404e-05,
+      "loss": 1.542,
+      "step": 11620
+    },
+    {
+      "epoch": 0.5353572580273644,
+      "grad_norm": 2.999126951040956,
+      "learning_rate": 9.341719215584244e-05,
+      "loss": 1.5283,
+      "step": 11621
+    },
+    {
+      "epoch": 0.5354033261159995,
+      "grad_norm": 4.067627590492185,
+      "learning_rate": 9.340230370637573e-05,
+      "loss": 1.6573,
+      "step": 11622
+    },
+    {
+      "epoch": 0.5354493942046344,
+      "grad_norm": 2.5000593841610366,
+      "learning_rate": 9.338741540379546e-05,
+      "loss": 1.1009,
+      "step": 11623
+    },
+    {
+      "epoch": 0.5354954622932695,
+      "grad_norm": 8.417899503255414,
+      "learning_rate": 9.337252724843306e-05,
+      "loss": 1.4396,
+      "step": 11624
+    },
+    {
+      "epoch": 0.5355415303819044,
+      "grad_norm": 12.038614142689305,
+      "learning_rate": 9.33576392406199e-05,
+      "loss": 1.3124,
+      "step": 11625
+    },
+    {
+      "epoch": 0.5355875984705395,
+      "grad_norm": 3.9552273686613373,
+      "learning_rate": 9.334275138068758e-05,
+      "loss": 1.3361,
+      "step": 11626
+    },
+    {
+      "epoch": 0.5356336665591744,
+      "grad_norm": 4.070870136408656,
+      "learning_rate": 9.332786366896746e-05,
+      "loss": 1.5016,
+      "step": 11627
+    },
+    {
+      "epoch": 0.5356797346478095,
+      "grad_norm": 1.7858579368047691,
+      "learning_rate": 9.331297610579103e-05,
+      "loss": 1.2819,
+      "step": 11628
+    },
+    {
+      "epoch": 0.5357258027364444,
+      "grad_norm": 1.6547077123841094,
+      "learning_rate": 9.32980886914897e-05,
+      "loss": 1.6988,
+      "step": 11629
+    },
+    {
+      "epoch": 0.5357718708250795,
+      "grad_norm": 1.8119393402225574,
+      "learning_rate": 9.328320142639495e-05,
+      "loss": 1.3387,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5358179389137144,
+      "grad_norm": 3.953949497434051,
+      "learning_rate": 9.32683143108382e-05,
+      "loss": 1.6775,
+      "step": 11631
+    },
+    {
+      "epoch": 0.5358640070023495,
+      "grad_norm": 2.056571869821988,
+      "learning_rate": 9.32534273451509e-05,
+      "loss": 1.297,
+      "step": 11632
+    },
+    {
+      "epoch": 0.5359100750909844,
+      "grad_norm": 1.5218706848402457,
+      "learning_rate": 9.323854052966448e-05,
+      "loss": 1.5485,
+      "step": 11633
+    },
+    {
+      "epoch": 0.5359561431796195,
+      "grad_norm": 2.1316558316113925,
+      "learning_rate": 9.32236538647103e-05,
+      "loss": 1.4407,
+      "step": 11634
+    },
+    {
+      "epoch": 0.5360022112682544,
+      "grad_norm": 2.7142940175525765,
+      "learning_rate": 9.320876735061993e-05,
+      "loss": 1.4577,
+      "step": 11635
+    },
+    {
+      "epoch": 0.5360482793568895,
+      "grad_norm": 2.121260281537178,
+      "learning_rate": 9.319388098772464e-05,
+      "loss": 1.7393,
+      "step": 11636
+    },
+    {
+      "epoch": 0.5360943474455245,
+      "grad_norm": 1.4597938238608281,
+      "learning_rate": 9.317899477635598e-05,
+      "loss": 1.2605,
+      "step": 11637
+    },
+    {
+      "epoch": 0.5361404155341595,
+      "grad_norm": 1.5528690056704597,
+      "learning_rate": 9.316410871684525e-05,
+      "loss": 1.4204,
+      "step": 11638
+    },
+    {
+      "epoch": 0.5361864836227945,
+      "grad_norm": 2.704026892274474,
+      "learning_rate": 9.314922280952395e-05,
+      "loss": 1.3747,
+      "step": 11639
+    },
+    {
+      "epoch": 0.5362325517114295,
+      "grad_norm": 0.9964076458886306,
+      "learning_rate": 9.313433705472343e-05,
+      "loss": 1.551,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5362786198000645,
+      "grad_norm": 1.8092136718352452,
+      "learning_rate": 9.311945145277516e-05,
+      "loss": 1.149,
+      "step": 11641
+    },
+    {
+      "epoch": 0.5363246878886995,
+      "grad_norm": 2.4415366739782165,
+      "learning_rate": 9.310456600401053e-05,
+      "loss": 1.4297,
+      "step": 11642
+    },
+    {
+      "epoch": 0.5363707559773345,
+      "grad_norm": 1.0729711957163652,
+      "learning_rate": 9.308968070876085e-05,
+      "loss": 1.6215,
+      "step": 11643
+    },
+    {
+      "epoch": 0.5364168240659695,
+      "grad_norm": 2.3421738246113994,
+      "learning_rate": 9.307479556735764e-05,
+      "loss": 1.4113,
+      "step": 11644
+    },
+    {
+      "epoch": 0.5364628921546045,
+      "grad_norm": 1.5953460686939431,
+      "learning_rate": 9.305991058013215e-05,
+      "loss": 1.3429,
+      "step": 11645
+    },
+    {
+      "epoch": 0.5365089602432395,
+      "grad_norm": 1.8483750706987225,
+      "learning_rate": 9.304502574741593e-05,
+      "loss": 1.3096,
+      "step": 11646
+    },
+    {
+      "epoch": 0.5365550283318745,
+      "grad_norm": 2.3654378510066008,
+      "learning_rate": 9.303014106954023e-05,
+      "loss": 1.3362,
+      "step": 11647
+    },
+    {
+      "epoch": 0.5366010964205095,
+      "grad_norm": 2.3077266377676975,
+      "learning_rate": 9.301525654683654e-05,
+      "loss": 1.2933,
+      "step": 11648
+    },
+    {
+      "epoch": 0.5366471645091445,
+      "grad_norm": 1.3853525602455663,
+      "learning_rate": 9.300037217963613e-05,
+      "loss": 1.5191,
+      "step": 11649
+    },
+    {
+      "epoch": 0.5366932325977796,
+      "grad_norm": 1.6421411063736975,
+      "learning_rate": 9.298548796827049e-05,
+      "loss": 1.2393,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5367393006864145,
+      "grad_norm": 1.6230770585461274,
+      "learning_rate": 9.297060391307092e-05,
+      "loss": 1.5045,
+      "step": 11651
+    },
+    {
+      "epoch": 0.5367853687750496,
+      "grad_norm": 2.4804861283972905,
+      "learning_rate": 9.295572001436876e-05,
+      "loss": 1.5363,
+      "step": 11652
+    },
+    {
+      "epoch": 0.5368314368636845,
+      "grad_norm": 1.9215675323323569,
+      "learning_rate": 9.294083627249547e-05,
+      "loss": 1.2205,
+      "step": 11653
+    },
+    {
+      "epoch": 0.5368775049523196,
+      "grad_norm": 1.4146400640531684,
+      "learning_rate": 9.29259526877823e-05,
+      "loss": 1.1687,
+      "step": 11654
+    },
+    {
+      "epoch": 0.5369235730409545,
+      "grad_norm": 20.890952613074365,
+      "learning_rate": 9.291106926056073e-05,
+      "loss": 1.5938,
+      "step": 11655
+    },
+    {
+      "epoch": 0.5369696411295896,
+      "grad_norm": 1.8592409124477811,
+      "learning_rate": 9.289618599116197e-05,
+      "loss": 1.5205,
+      "step": 11656
+    },
+    {
+      "epoch": 0.5370157092182245,
+      "grad_norm": 1.2270503528964765,
+      "learning_rate": 9.288130287991752e-05,
+      "loss": 1.6968,
+      "step": 11657
+    },
+    {
+      "epoch": 0.5370617773068596,
+      "grad_norm": 1.4637969042666708,
+      "learning_rate": 9.286641992715859e-05,
+      "loss": 1.1802,
+      "step": 11658
+    },
+    {
+      "epoch": 0.5371078453954945,
+      "grad_norm": 2.1776074456890897,
+      "learning_rate": 9.285153713321666e-05,
+      "loss": 1.8628,
+      "step": 11659
+    },
+    {
+      "epoch": 0.5371539134841296,
+      "grad_norm": 2.8269491256194788,
+      "learning_rate": 9.283665449842296e-05,
+      "loss": 1.4961,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5371999815727645,
+      "grad_norm": 2.545854882904222,
+      "learning_rate": 9.282177202310886e-05,
+      "loss": 1.2951,
+      "step": 11661
+    },
+    {
+      "epoch": 0.5372460496613995,
+      "grad_norm": 2.1089067193391586,
+      "learning_rate": 9.280688970760573e-05,
+      "loss": 1.314,
+      "step": 11662
+    },
+    {
+      "epoch": 0.5372921177500346,
+      "grad_norm": 1.1216927477287835,
+      "learning_rate": 9.279200755224483e-05,
+      "loss": 1.4445,
+      "step": 11663
+    },
+    {
+      "epoch": 0.5373381858386695,
+      "grad_norm": 1.7704300790205167,
+      "learning_rate": 9.277712555735754e-05,
+      "loss": 1.4237,
+      "step": 11664
+    },
+    {
+      "epoch": 0.5373842539273046,
+      "grad_norm": 3.612199785170032,
+      "learning_rate": 9.276224372327513e-05,
+      "loss": 1.4667,
+      "step": 11665
+    },
+    {
+      "epoch": 0.5374303220159395,
+      "grad_norm": 0.9899963247552164,
+      "learning_rate": 9.2747362050329e-05,
+      "loss": 1.3822,
+      "step": 11666
+    },
+    {
+      "epoch": 0.5374763901045746,
+      "grad_norm": 1.2802045737244592,
+      "learning_rate": 9.273248053885039e-05,
+      "loss": 1.5415,
+      "step": 11667
+    },
+    {
+      "epoch": 0.5375224581932095,
+      "grad_norm": 2.7549011371994325,
+      "learning_rate": 9.271759918917068e-05,
+      "loss": 1.5088,
+      "step": 11668
+    },
+    {
+      "epoch": 0.5375685262818446,
+      "grad_norm": 1.7236938892711677,
+      "learning_rate": 9.270271800162107e-05,
+      "loss": 1.5034,
+      "step": 11669
+    },
+    {
+      "epoch": 0.5376145943704795,
+      "grad_norm": 2.4118234569808714,
+      "learning_rate": 9.2687836976533e-05,
+      "loss": 1.3342,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5376606624591146,
+      "grad_norm": 1.2297432910522752,
+      "learning_rate": 9.267295611423769e-05,
+      "loss": 1.745,
+      "step": 11671
+    },
+    {
+      "epoch": 0.5377067305477495,
+      "grad_norm": 2.6000393279332012,
+      "learning_rate": 9.265807541506642e-05,
+      "loss": 1.3671,
+      "step": 11672
+    },
+    {
+      "epoch": 0.5377527986363846,
+      "grad_norm": 1.7417985104856675,
+      "learning_rate": 9.264319487935053e-05,
+      "loss": 1.2314,
+      "step": 11673
+    },
+    {
+      "epoch": 0.5377988667250195,
+      "grad_norm": 2.28055008333208,
+      "learning_rate": 9.262831450742128e-05,
+      "loss": 1.3876,
+      "step": 11674
+    },
+    {
+      "epoch": 0.5378449348136546,
+      "grad_norm": 2.424861379369418,
+      "learning_rate": 9.261343429960998e-05,
+      "loss": 1.0769,
+      "step": 11675
+    },
+    {
+      "epoch": 0.5378910029022896,
+      "grad_norm": 3.050865823461511,
+      "learning_rate": 9.259855425624788e-05,
+      "loss": 1.3123,
+      "step": 11676
+    },
+    {
+      "epoch": 0.5379370709909246,
+      "grad_norm": 1.7624633696327185,
+      "learning_rate": 9.25836743776663e-05,
+      "loss": 1.2328,
+      "step": 11677
+    },
+    {
+      "epoch": 0.5379831390795596,
+      "grad_norm": 2.1657087268466624,
+      "learning_rate": 9.256879466419648e-05,
+      "loss": 1.2464,
+      "step": 11678
+    },
+    {
+      "epoch": 0.5380292071681946,
+      "grad_norm": 1.1133081756663077,
+      "learning_rate": 9.255391511616972e-05,
+      "loss": 1.2811,
+      "step": 11679
+    },
+    {
+      "epoch": 0.5380752752568296,
+      "grad_norm": 1.2730344605229953,
+      "learning_rate": 9.253903573391728e-05,
+      "loss": 1.0734,
+      "step": 11680
+    },
+    {
+      "epoch": 0.5381213433454646,
+      "grad_norm": 1.0030080396758698,
+      "learning_rate": 9.25241565177704e-05,
+      "loss": 1.583,
+      "step": 11681
+    },
+    {
+      "epoch": 0.5381674114340996,
+      "grad_norm": 1.3865864755564226,
+      "learning_rate": 9.250927746806035e-05,
+      "loss": 1.6834,
+      "step": 11682
+    },
+    {
+      "epoch": 0.5382134795227346,
+      "grad_norm": 1.6610394853135277,
+      "learning_rate": 9.249439858511839e-05,
+      "loss": 1.2041,
+      "step": 11683
+    },
+    {
+      "epoch": 0.5382595476113696,
+      "grad_norm": 3.0710985035366583,
+      "learning_rate": 9.24795198692758e-05,
+      "loss": 1.5419,
+      "step": 11684
+    },
+    {
+      "epoch": 0.5383056157000046,
+      "grad_norm": 1.2940375122344305,
+      "learning_rate": 9.246464132086378e-05,
+      "loss": 1.1012,
+      "step": 11685
+    },
+    {
+      "epoch": 0.5383516837886396,
+      "grad_norm": 1.532876688153164,
+      "learning_rate": 9.244976294021362e-05,
+      "loss": 1.1637,
+      "step": 11686
+    },
+    {
+      "epoch": 0.5383977518772746,
+      "grad_norm": 0.9787846507791678,
+      "learning_rate": 9.243488472765652e-05,
+      "loss": 1.4585,
+      "step": 11687
+    },
+    {
+      "epoch": 0.5384438199659096,
+      "grad_norm": 1.8751090812789188,
+      "learning_rate": 9.242000668352377e-05,
+      "loss": 1.3873,
+      "step": 11688
+    },
+    {
+      "epoch": 0.5384898880545446,
+      "grad_norm": 1.4208201831320737,
+      "learning_rate": 9.240512880814657e-05,
+      "loss": 1.4862,
+      "step": 11689
+    },
+    {
+      "epoch": 0.5385359561431796,
+      "grad_norm": 2.329081829227414,
+      "learning_rate": 9.239025110185613e-05,
+      "loss": 1.1724,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5385820242318147,
+      "grad_norm": 1.270458706787049,
+      "learning_rate": 9.237537356498374e-05,
+      "loss": 1.6254,
+      "step": 11691
+    },
+    {
+      "epoch": 0.5386280923204496,
+      "grad_norm": 3.287846538115525,
+      "learning_rate": 9.236049619786054e-05,
+      "loss": 1.5902,
+      "step": 11692
+    },
+    {
+      "epoch": 0.5386741604090847,
+      "grad_norm": 1.7783663471835676,
+      "learning_rate": 9.234561900081782e-05,
+      "loss": 1.246,
+      "step": 11693
+    },
+    {
+      "epoch": 0.5387202284977196,
+      "grad_norm": 5.858334797519375,
+      "learning_rate": 9.233074197418675e-05,
+      "loss": 1.5145,
+      "step": 11694
+    },
+    {
+      "epoch": 0.5387662965863547,
+      "grad_norm": 1.1223807202696197,
+      "learning_rate": 9.231586511829859e-05,
+      "loss": 1.6213,
+      "step": 11695
+    },
+    {
+      "epoch": 0.5388123646749896,
+      "grad_norm": 1.2447273673713652,
+      "learning_rate": 9.230098843348448e-05,
+      "loss": 1.5375,
+      "step": 11696
+    },
+    {
+      "epoch": 0.5388584327636247,
+      "grad_norm": 1.3523292670259763,
+      "learning_rate": 9.22861119200757e-05,
+      "loss": 1.3689,
+      "step": 11697
+    },
+    {
+      "epoch": 0.5389045008522596,
+      "grad_norm": 2.018511480315925,
+      "learning_rate": 9.22712355784034e-05,
+      "loss": 1.5419,
+      "step": 11698
+    },
+    {
+      "epoch": 0.5389505689408947,
+      "grad_norm": 1.9888183205622034,
+      "learning_rate": 9.225635940879875e-05,
+      "loss": 1.2075,
+      "step": 11699
+    },
+    {
+      "epoch": 0.5389966370295296,
+      "grad_norm": 1.225960743035796,
+      "learning_rate": 9.224148341159304e-05,
+      "loss": 1.6022,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5390427051181647,
+      "grad_norm": 2.0183486465473,
+      "learning_rate": 9.222660758711734e-05,
+      "loss": 1.2101,
+      "step": 11701
+    },
+    {
+      "epoch": 0.5390887732067996,
+      "grad_norm": 1.8211948340961592,
+      "learning_rate": 9.221173193570294e-05,
+      "loss": 1.7007,
+      "step": 11702
+    },
+    {
+      "epoch": 0.5391348412954347,
+      "grad_norm": 1.0478260753901454,
+      "learning_rate": 9.219685645768095e-05,
+      "loss": 1.3619,
+      "step": 11703
+    },
+    {
+      "epoch": 0.5391809093840697,
+      "grad_norm": 10.367734440059742,
+      "learning_rate": 9.218198115338258e-05,
+      "loss": 1.6442,
+      "step": 11704
+    },
+    {
+      "epoch": 0.5392269774727046,
+      "grad_norm": 1.2293197504340203,
+      "learning_rate": 9.216710602313899e-05,
+      "loss": 1.4075,
+      "step": 11705
+    },
+    {
+      "epoch": 0.5392730455613397,
+      "grad_norm": 6.33375739646645,
+      "learning_rate": 9.215223106728135e-05,
+      "loss": 1.2916,
+      "step": 11706
+    },
+    {
+      "epoch": 0.5393191136499746,
+      "grad_norm": 1.0449251933327868,
+      "learning_rate": 9.213735628614086e-05,
+      "loss": 1.3228,
+      "step": 11707
+    },
+    {
+      "epoch": 0.5393651817386097,
+      "grad_norm": 1.379843530688071,
+      "learning_rate": 9.212248168004861e-05,
+      "loss": 1.4615,
+      "step": 11708
+    },
+    {
+      "epoch": 0.5394112498272446,
+      "grad_norm": 2.4542011553502214,
+      "learning_rate": 9.210760724933582e-05,
+      "loss": 1.5857,
+      "step": 11709
+    },
+    {
+      "epoch": 0.5394573179158797,
+      "grad_norm": 2.1415266119456655,
+      "learning_rate": 9.20927329943336e-05,
+      "loss": 1.288,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5395033860045146,
+      "grad_norm": 1.8541697004866067,
+      "learning_rate": 9.207785891537316e-05,
+      "loss": 1.2791,
+      "step": 11711
+    },
+    {
+      "epoch": 0.5395494540931497,
+      "grad_norm": 1.247300516790598,
+      "learning_rate": 9.206298501278557e-05,
+      "loss": 1.4416,
+      "step": 11712
+    },
+    {
+      "epoch": 0.5395955221817846,
+      "grad_norm": 7.711064290521077,
+      "learning_rate": 9.204811128690205e-05,
+      "loss": 1.2715,
+      "step": 11713
+    },
+    {
+      "epoch": 0.5396415902704197,
+      "grad_norm": 1.1685097413255454,
+      "learning_rate": 9.203323773805365e-05,
+      "loss": 1.5374,
+      "step": 11714
+    },
+    {
+      "epoch": 0.5396876583590546,
+      "grad_norm": 3.435935045217426,
+      "learning_rate": 9.20183643665716e-05,
+      "loss": 1.3975,
+      "step": 11715
+    },
+    {
+      "epoch": 0.5397337264476897,
+      "grad_norm": 1.6997709126829994,
+      "learning_rate": 9.200349117278697e-05,
+      "loss": 1.2078,
+      "step": 11716
+    },
+    {
+      "epoch": 0.5397797945363247,
+      "grad_norm": 1.8137267244106323,
+      "learning_rate": 9.198861815703085e-05,
+      "loss": 1.6833,
+      "step": 11717
+    },
+    {
+      "epoch": 0.5398258626249597,
+      "grad_norm": 2.513409160075112,
+      "learning_rate": 9.197374531963447e-05,
+      "loss": 1.4491,
+      "step": 11718
+    },
+    {
+      "epoch": 0.5398719307135947,
+      "grad_norm": 0.84772765026208,
+      "learning_rate": 9.195887266092885e-05,
+      "loss": 1.6168,
+      "step": 11719
+    },
+    {
+      "epoch": 0.5399179988022297,
+      "grad_norm": 0.9529198778478198,
+      "learning_rate": 9.194400018124517e-05,
+      "loss": 1.3622,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5399640668908647,
+      "grad_norm": 1.3080155970417053,
+      "learning_rate": 9.192912788091448e-05,
+      "loss": 1.5581,
+      "step": 11721
+    },
+    {
+      "epoch": 0.5400101349794997,
+      "grad_norm": 1.4290874695536415,
+      "learning_rate": 9.191425576026796e-05,
+      "loss": 1.4758,
+      "step": 11722
+    },
+    {
+      "epoch": 0.5400562030681347,
+      "grad_norm": 1.114202539811552,
+      "learning_rate": 9.189938381963663e-05,
+      "loss": 1.118,
+      "step": 11723
+    },
+    {
+      "epoch": 0.5401022711567697,
+      "grad_norm": 0.8195601022426819,
+      "learning_rate": 9.188451205935167e-05,
+      "loss": 1.5272,
+      "step": 11724
+    },
+    {
+      "epoch": 0.5401483392454047,
+      "grad_norm": 1.0531429579206384,
+      "learning_rate": 9.186964047974412e-05,
+      "loss": 1.2901,
+      "step": 11725
+    },
+    {
+      "epoch": 0.5401944073340397,
+      "grad_norm": 0.9340619407494489,
+      "learning_rate": 9.185476908114506e-05,
+      "loss": 1.245,
+      "step": 11726
+    },
+    {
+      "epoch": 0.5402404754226747,
+      "grad_norm": 1.117652367984404,
+      "learning_rate": 9.183989786388564e-05,
+      "loss": 1.6389,
+      "step": 11727
+    },
+    {
+      "epoch": 0.5402865435113097,
+      "grad_norm": 1.0617133022845056,
+      "learning_rate": 9.182502682829685e-05,
+      "loss": 1.4875,
+      "step": 11728
+    },
+    {
+      "epoch": 0.5403326115999447,
+      "grad_norm": 1.732412799137513,
+      "learning_rate": 9.181015597470988e-05,
+      "loss": 1.5209,
+      "step": 11729
+    },
+    {
+      "epoch": 0.5403786796885797,
+      "grad_norm": 1.3188591543355714,
+      "learning_rate": 9.17952853034557e-05,
+      "loss": 1.4594,
+      "step": 11730
+    },
+    {
+      "epoch": 0.5404247477772147,
+      "grad_norm": 1.553616147440485,
+      "learning_rate": 9.178041481486545e-05,
+      "loss": 1.6464,
+      "step": 11731
+    },
+    {
+      "epoch": 0.5404708158658498,
+      "grad_norm": 1.6986582987496346,
+      "learning_rate": 9.176554450927015e-05,
+      "loss": 1.2242,
+      "step": 11732
+    },
+    {
+      "epoch": 0.5405168839544847,
+      "grad_norm": 1.0257442838618311,
+      "learning_rate": 9.17506743870009e-05,
+      "loss": 1.3026,
+      "step": 11733
+    },
+    {
+      "epoch": 0.5405629520431198,
+      "grad_norm": 2.591700523193511,
+      "learning_rate": 9.173580444838876e-05,
+      "loss": 1.6735,
+      "step": 11734
+    },
+    {
+      "epoch": 0.5406090201317547,
+      "grad_norm": 0.9734319106920846,
+      "learning_rate": 9.172093469376471e-05,
+      "loss": 1.2052,
+      "step": 11735
+    },
+    {
+      "epoch": 0.5406550882203898,
+      "grad_norm": 0.9928136180905125,
+      "learning_rate": 9.170606512345992e-05,
+      "loss": 1.4522,
+      "step": 11736
+    },
+    {
+      "epoch": 0.5407011563090247,
+      "grad_norm": 1.8647319707560694,
+      "learning_rate": 9.169119573780529e-05,
+      "loss": 1.3797,
+      "step": 11737
+    },
+    {
+      "epoch": 0.5407472243976598,
+      "grad_norm": 1.0725106052559914,
+      "learning_rate": 9.167632653713199e-05,
+      "loss": 1.3776,
+      "step": 11738
+    },
+    {
+      "epoch": 0.5407932924862947,
+      "grad_norm": 1.0050498952621971,
+      "learning_rate": 9.166145752177099e-05,
+      "loss": 1.459,
+      "step": 11739
+    },
+    {
+      "epoch": 0.5408393605749298,
+      "grad_norm": 0.9500004234910842,
+      "learning_rate": 9.164658869205335e-05,
+      "loss": 1.0876,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5408854286635647,
+      "grad_norm": 1.2575392362073972,
+      "learning_rate": 9.163172004831008e-05,
+      "loss": 1.558,
+      "step": 11741
+    },
+    {
+      "epoch": 0.5409314967521998,
+      "grad_norm": 2.6722213628597404,
+      "learning_rate": 9.161685159087223e-05,
+      "loss": 1.4054,
+      "step": 11742
+    },
+    {
+      "epoch": 0.5409775648408347,
+      "grad_norm": 1.1586798492620352,
+      "learning_rate": 9.160198332007084e-05,
+      "loss": 1.3242,
+      "step": 11743
+    },
+    {
+      "epoch": 0.5410236329294698,
+      "grad_norm": 0.9569825065735521,
+      "learning_rate": 9.158711523623682e-05,
+      "loss": 1.3111,
+      "step": 11744
+    },
+    {
+      "epoch": 0.5410697010181048,
+      "grad_norm": 0.9061415034801757,
+      "learning_rate": 9.157224733970132e-05,
+      "loss": 1.5538,
+      "step": 11745
+    },
+    {
+      "epoch": 0.5411157691067398,
+      "grad_norm": 0.8626854055975514,
+      "learning_rate": 9.15573796307952e-05,
+      "loss": 1.4083,
+      "step": 11746
+    },
+    {
+      "epoch": 0.5411618371953748,
+      "grad_norm": 1.0833627621419073,
+      "learning_rate": 9.154251210984963e-05,
+      "loss": 1.188,
+      "step": 11747
+    },
+    {
+      "epoch": 0.5412079052840097,
+      "grad_norm": 0.8948608114129192,
+      "learning_rate": 9.152764477719545e-05,
+      "loss": 1.2909,
+      "step": 11748
+    },
+    {
+      "epoch": 0.5412539733726448,
+      "grad_norm": 2.27061213047437,
+      "learning_rate": 9.151277763316381e-05,
+      "loss": 1.3641,
+      "step": 11749
+    },
+    {
+      "epoch": 0.5413000414612797,
+      "grad_norm": 1.2607836713068652,
+      "learning_rate": 9.149791067808555e-05,
+      "loss": 1.4304,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5413461095499148,
+      "grad_norm": 1.1059989326587205,
+      "learning_rate": 9.148304391229178e-05,
+      "loss": 1.3684,
+      "step": 11751
+    },
+    {
+      "epoch": 0.5413921776385497,
+      "grad_norm": 1.6392010826713723,
+      "learning_rate": 9.146817733611346e-05,
+      "loss": 1.2113,
+      "step": 11752
+    },
+    {
+      "epoch": 0.5414382457271848,
+      "grad_norm": 1.3536587639465074,
+      "learning_rate": 9.145331094988146e-05,
+      "loss": 1.2671,
+      "step": 11753
+    },
+    {
+      "epoch": 0.5414843138158197,
+      "grad_norm": 1.2977845408511353,
+      "learning_rate": 9.143844475392691e-05,
+      "loss": 1.2193,
+      "step": 11754
+    },
+    {
+      "epoch": 0.5415303819044548,
+      "grad_norm": 0.8632562651877824,
+      "learning_rate": 9.142357874858066e-05,
+      "loss": 1.139,
+      "step": 11755
+    },
+    {
+      "epoch": 0.5415764499930897,
+      "grad_norm": 0.9340252226578216,
+      "learning_rate": 9.140871293417376e-05,
+      "loss": 1.3851,
+      "step": 11756
+    },
+    {
+      "epoch": 0.5416225180817248,
+      "grad_norm": 1.0972305230940587,
+      "learning_rate": 9.139384731103709e-05,
+      "loss": 1.5412,
+      "step": 11757
+    },
+    {
+      "epoch": 0.5416685861703598,
+      "grad_norm": 1.033347664828138,
+      "learning_rate": 9.137898187950171e-05,
+      "loss": 1.465,
+      "step": 11758
+    },
+    {
+      "epoch": 0.5417146542589948,
+      "grad_norm": 1.3181258273357002,
+      "learning_rate": 9.136411663989845e-05,
+      "loss": 1.3704,
+      "step": 11759
+    },
+    {
+      "epoch": 0.5417607223476298,
+      "grad_norm": 1.4175110253460352,
+      "learning_rate": 9.134925159255839e-05,
+      "loss": 1.3905,
+      "step": 11760
+    },
+    {
+      "epoch": 0.5418067904362648,
+      "grad_norm": 1.0535202042321636,
+      "learning_rate": 9.13343867378124e-05,
+      "loss": 1.4275,
+      "step": 11761
+    },
+    {
+      "epoch": 0.5418528585248998,
+      "grad_norm": 1.6473959608869004,
+      "learning_rate": 9.13195220759914e-05,
+      "loss": 1.7028,
+      "step": 11762
+    },
+    {
+      "epoch": 0.5418989266135348,
+      "grad_norm": 4.629003623742191,
+      "learning_rate": 9.130465760742639e-05,
+      "loss": 1.1495,
+      "step": 11763
+    },
+    {
+      "epoch": 0.5419449947021698,
+      "grad_norm": 1.4189810286542976,
+      "learning_rate": 9.128979333244821e-05,
+      "loss": 1.0592,
+      "step": 11764
+    },
+    {
+      "epoch": 0.5419910627908048,
+      "grad_norm": 1.6063916825968347,
+      "learning_rate": 9.127492925138792e-05,
+      "loss": 1.0955,
+      "step": 11765
+    },
+    {
+      "epoch": 0.5420371308794398,
+      "grad_norm": 2.8544679408046503,
+      "learning_rate": 9.12600653645763e-05,
+      "loss": 1.2254,
+      "step": 11766
+    },
+    {
+      "epoch": 0.5420831989680748,
+      "grad_norm": 1.6394710975778148,
+      "learning_rate": 9.124520167234442e-05,
+      "loss": 0.8986,
+      "step": 11767
+    },
+    {
+      "epoch": 0.5421292670567098,
+      "grad_norm": 2.304650972636839,
+      "learning_rate": 9.123033817502302e-05,
+      "loss": 1.0661,
+      "step": 11768
+    },
+    {
+      "epoch": 0.5421753351453448,
+      "grad_norm": 1.1625411062483033,
+      "learning_rate": 9.12154748729432e-05,
+      "loss": 1.5461,
+      "step": 11769
+    },
+    {
+      "epoch": 0.5422214032339798,
+      "grad_norm": 0.7808197005431522,
+      "learning_rate": 9.120061176643574e-05,
+      "loss": 1.6855,
+      "step": 11770
+    },
+    {
+      "epoch": 0.5422674713226149,
+      "grad_norm": 0.8703232099222176,
+      "learning_rate": 9.118574885583153e-05,
+      "loss": 1.3852,
+      "step": 11771
+    },
+    {
+      "epoch": 0.5423135394112498,
+      "grad_norm": 0.6724742629083758,
+      "learning_rate": 9.117088614146154e-05,
+      "loss": 1.4451,
+      "step": 11772
+    },
+    {
+      "epoch": 0.5423596074998849,
+      "grad_norm": 0.9723221761226404,
+      "learning_rate": 9.115602362365661e-05,
+      "loss": 0.8916,
+      "step": 11773
+    },
+    {
+      "epoch": 0.5424056755885198,
+      "grad_norm": 1.7752701620465654,
+      "learning_rate": 9.114116130274768e-05,
+      "loss": 1.1644,
+      "step": 11774
+    },
+    {
+      "epoch": 0.5424517436771549,
+      "grad_norm": 1.172446974326252,
+      "learning_rate": 9.112629917906558e-05,
+      "loss": 1.3926,
+      "step": 11775
+    },
+    {
+      "epoch": 0.5424978117657898,
+      "grad_norm": 1.0487348592775607,
+      "learning_rate": 9.111143725294123e-05,
+      "loss": 1.3853,
+      "step": 11776
+    },
+    {
+      "epoch": 0.5425438798544249,
+      "grad_norm": 1.116440521561286,
+      "learning_rate": 9.109657552470546e-05,
+      "loss": 1.4815,
+      "step": 11777
+    },
+    {
+      "epoch": 0.5425899479430598,
+      "grad_norm": 0.9645352016347574,
+      "learning_rate": 9.10817139946892e-05,
+      "loss": 1.0541,
+      "step": 11778
+    },
+    {
+      "epoch": 0.5426360160316949,
+      "grad_norm": 2.5349968586182174,
+      "learning_rate": 9.106685266322328e-05,
+      "loss": 1.2118,
+      "step": 11779
+    },
+    {
+      "epoch": 0.5426820841203298,
+      "grad_norm": 1.1238997715778414,
+      "learning_rate": 9.105199153063857e-05,
+      "loss": 1.4337,
+      "step": 11780
+    },
+    {
+      "epoch": 0.5427281522089649,
+      "grad_norm": 1.5940201599791324,
+      "learning_rate": 9.103713059726593e-05,
+      "loss": 0.9803,
+      "step": 11781
+    },
+    {
+      "epoch": 0.5427742202975998,
+      "grad_norm": 0.7132877989491758,
+      "learning_rate": 9.102226986343619e-05,
+      "loss": 1.1995,
+      "step": 11782
+    },
+    {
+      "epoch": 0.5428202883862349,
+      "grad_norm": 0.8739778366613736,
+      "learning_rate": 9.100740932948023e-05,
+      "loss": 1.4122,
+      "step": 11783
+    },
+    {
+      "epoch": 0.5428663564748698,
+      "grad_norm": 1.0724056579470649,
+      "learning_rate": 9.099254899572887e-05,
+      "loss": 1.4808,
+      "step": 11784
+    },
+    {
+      "epoch": 0.5429124245635049,
+      "grad_norm": 0.8527229476162929,
+      "learning_rate": 9.097768886251299e-05,
+      "loss": 1.7657,
+      "step": 11785
+    },
+    {
+      "epoch": 0.5429584926521399,
+      "grad_norm": 0.8479712078424686,
+      "learning_rate": 9.096282893016336e-05,
+      "loss": 1.4401,
+      "step": 11786
+    },
+    {
+      "epoch": 0.5430045607407749,
+      "grad_norm": 1.8527492520900224,
+      "learning_rate": 9.094796919901089e-05,
+      "loss": 1.5193,
+      "step": 11787
+    },
+    {
+      "epoch": 0.5430506288294099,
+      "grad_norm": 1.250543717538726,
+      "learning_rate": 9.093310966938634e-05,
+      "loss": 1.4222,
+      "step": 11788
+    },
+    {
+      "epoch": 0.5430966969180449,
+      "grad_norm": 1.0231508205747875,
+      "learning_rate": 9.091825034162056e-05,
+      "loss": 1.2754,
+      "step": 11789
+    },
+    {
+      "epoch": 0.5431427650066799,
+      "grad_norm": 0.9273447538306677,
+      "learning_rate": 9.090339121604436e-05,
+      "loss": 1.4162,
+      "step": 11790
+    },
+    {
+      "epoch": 0.5431888330953148,
+      "grad_norm": 1.2100687831581756,
+      "learning_rate": 9.088853229298856e-05,
+      "loss": 1.0219,
+      "step": 11791
+    },
+    {
+      "epoch": 0.5432349011839499,
+      "grad_norm": 1.3161782587622575,
+      "learning_rate": 9.087367357278398e-05,
+      "loss": 1.2132,
+      "step": 11792
+    },
+    {
+      "epoch": 0.5432809692725848,
+      "grad_norm": 0.9095681367446603,
+      "learning_rate": 9.085881505576138e-05,
+      "loss": 1.3917,
+      "step": 11793
+    },
+    {
+      "epoch": 0.5433270373612199,
+      "grad_norm": 1.7901196701453628,
+      "learning_rate": 9.084395674225161e-05,
+      "loss": 1.1264,
+      "step": 11794
+    },
+    {
+      "epoch": 0.5433731054498548,
+      "grad_norm": 1.232585121299249,
+      "learning_rate": 9.082909863258544e-05,
+      "loss": 1.1383,
+      "step": 11795
+    },
+    {
+      "epoch": 0.5434191735384899,
+      "grad_norm": 0.9986521879024295,
+      "learning_rate": 9.081424072709367e-05,
+      "loss": 1.451,
+      "step": 11796
+    },
+    {
+      "epoch": 0.5434652416271248,
+      "grad_norm": 0.7959860436716647,
+      "learning_rate": 9.079938302610707e-05,
+      "loss": 1.2156,
+      "step": 11797
+    },
+    {
+      "epoch": 0.5435113097157599,
+      "grad_norm": 1.0476038267367886,
+      "learning_rate": 9.078452552995644e-05,
+      "loss": 1.3865,
+      "step": 11798
+    },
+    {
+      "epoch": 0.543557377804395,
+      "grad_norm": 1.131164534212315,
+      "learning_rate": 9.076966823897255e-05,
+      "loss": 1.358,
+      "step": 11799
+    },
+    {
+      "epoch": 0.5436034458930299,
+      "grad_norm": 1.2200563872910681,
+      "learning_rate": 9.075481115348616e-05,
+      "loss": 1.3413,
+      "step": 11800
+    },
+    {
+      "epoch": 0.543649513981665,
+      "grad_norm": 1.3308296056327562,
+      "learning_rate": 9.073995427382805e-05,
+      "loss": 1.1456,
+      "step": 11801
+    },
+    {
+      "epoch": 0.5436955820702999,
+      "grad_norm": 1.020691368780016,
+      "learning_rate": 9.072509760032898e-05,
+      "loss": 1.4086,
+      "step": 11802
+    },
+    {
+      "epoch": 0.5437416501589349,
+      "grad_norm": 1.394508803741389,
+      "learning_rate": 9.071024113331972e-05,
+      "loss": 1.3597,
+      "step": 11803
+    },
+    {
+      "epoch": 0.5437877182475699,
+      "grad_norm": 1.4314368693288961,
+      "learning_rate": 9.069538487313099e-05,
+      "loss": 1.2657,
+      "step": 11804
+    },
+    {
+      "epoch": 0.5438337863362049,
+      "grad_norm": 0.9878049685349356,
+      "learning_rate": 9.068052882009358e-05,
+      "loss": 1.1654,
+      "step": 11805
+    },
+    {
+      "epoch": 0.5438798544248399,
+      "grad_norm": 0.9788700274528244,
+      "learning_rate": 9.066567297453821e-05,
+      "loss": 1.3816,
+      "step": 11806
+    },
+    {
+      "epoch": 0.5439259225134749,
+      "grad_norm": 0.7941822058136017,
+      "learning_rate": 9.065081733679564e-05,
+      "loss": 1.1474,
+      "step": 11807
+    },
+    {
+      "epoch": 0.5439719906021099,
+      "grad_norm": 1.2228677483938029,
+      "learning_rate": 9.063596190719657e-05,
+      "loss": 0.9761,
+      "step": 11808
+    },
+    {
+      "epoch": 0.5440180586907449,
+      "grad_norm": 1.2334744706421052,
+      "learning_rate": 9.062110668607176e-05,
+      "loss": 1.4179,
+      "step": 11809
+    },
+    {
+      "epoch": 0.5440641267793799,
+      "grad_norm": 1.851179469354622,
+      "learning_rate": 9.060625167375194e-05,
+      "loss": 1.4252,
+      "step": 11810
+    },
+    {
+      "epoch": 0.5441101948680149,
+      "grad_norm": 1.173297602789486,
+      "learning_rate": 9.05913968705678e-05,
+      "loss": 1.1705,
+      "step": 11811
+    },
+    {
+      "epoch": 0.54415626295665,
+      "grad_norm": 1.3713282491927363,
+      "learning_rate": 9.057654227685009e-05,
+      "loss": 1.2378,
+      "step": 11812
+    },
+    {
+      "epoch": 0.5442023310452849,
+      "grad_norm": 98.1856556900129,
+      "learning_rate": 9.05616878929295e-05,
+      "loss": 1.3045,
+      "step": 11813
+    },
+    {
+      "epoch": 0.54424839913392,
+      "grad_norm": 0.9805432312057613,
+      "learning_rate": 9.054683371913675e-05,
+      "loss": 1.5442,
+      "step": 11814
+    },
+    {
+      "epoch": 0.5442944672225549,
+      "grad_norm": 1.4646541335155407,
+      "learning_rate": 9.053197975580253e-05,
+      "loss": 1.285,
+      "step": 11815
+    },
+    {
+      "epoch": 0.54434053531119,
+      "grad_norm": 1.4112245925091933,
+      "learning_rate": 9.051712600325757e-05,
+      "loss": 1.5085,
+      "step": 11816
+    },
+    {
+      "epoch": 0.5443866033998249,
+      "grad_norm": 2.4186614547195213,
+      "learning_rate": 9.050227246183252e-05,
+      "loss": 1.4659,
+      "step": 11817
+    },
+    {
+      "epoch": 0.54443267148846,
+      "grad_norm": 4.669863137367536,
+      "learning_rate": 9.04874191318581e-05,
+      "loss": 1.4815,
+      "step": 11818
+    },
+    {
+      "epoch": 0.5444787395770949,
+      "grad_norm": 1.6999381363335797,
+      "learning_rate": 9.047256601366498e-05,
+      "loss": 1.7983,
+      "step": 11819
+    },
+    {
+      "epoch": 0.54452480766573,
+      "grad_norm": 2.3859390183588878,
+      "learning_rate": 9.045771310758382e-05,
+      "loss": 1.5501,
+      "step": 11820
+    },
+    {
+      "epoch": 0.5445708757543649,
+      "grad_norm": 1.6967179205993834,
+      "learning_rate": 9.044286041394535e-05,
+      "loss": 1.8474,
+      "step": 11821
+    },
+    {
+      "epoch": 0.544616943843,
+      "grad_norm": 1.119031789333899,
+      "learning_rate": 9.042800793308017e-05,
+      "loss": 1.2918,
+      "step": 11822
+    },
+    {
+      "epoch": 0.5446630119316349,
+      "grad_norm": 1.7513798032034322,
+      "learning_rate": 9.041315566531902e-05,
+      "loss": 1.3769,
+      "step": 11823
+    },
+    {
+      "epoch": 0.54470908002027,
+      "grad_norm": 2.220564195242546,
+      "learning_rate": 9.039830361099249e-05,
+      "loss": 1.4278,
+      "step": 11824
+    },
+    {
+      "epoch": 0.5447551481089049,
+      "grad_norm": 1.3697329810971726,
+      "learning_rate": 9.038345177043128e-05,
+      "loss": 1.3653,
+      "step": 11825
+    },
+    {
+      "epoch": 0.54480121619754,
+      "grad_norm": 1.0844900437990976,
+      "learning_rate": 9.036860014396604e-05,
+      "loss": 0.9615,
+      "step": 11826
+    },
+    {
+      "epoch": 0.544847284286175,
+      "grad_norm": 1.7202306188442231,
+      "learning_rate": 9.035374873192738e-05,
+      "loss": 1.572,
+      "step": 11827
+    },
+    {
+      "epoch": 0.54489335237481,
+      "grad_norm": 1.0598960831808082,
+      "learning_rate": 9.033889753464599e-05,
+      "loss": 1.5269,
+      "step": 11828
+    },
+    {
+      "epoch": 0.544939420463445,
+      "grad_norm": 1.1657441410384672,
+      "learning_rate": 9.032404655245244e-05,
+      "loss": 1.2918,
+      "step": 11829
+    },
+    {
+      "epoch": 0.54498548855208,
+      "grad_norm": 1.1078484415758814,
+      "learning_rate": 9.030919578567744e-05,
+      "loss": 1.6797,
+      "step": 11830
+    },
+    {
+      "epoch": 0.545031556640715,
+      "grad_norm": 1.151895694569022,
+      "learning_rate": 9.029434523465156e-05,
+      "loss": 1.4877,
+      "step": 11831
+    },
+    {
+      "epoch": 0.54507762472935,
+      "grad_norm": 2.752500140675328,
+      "learning_rate": 9.027949489970547e-05,
+      "loss": 1.3286,
+      "step": 11832
+    },
+    {
+      "epoch": 0.545123692817985,
+      "grad_norm": 2.915678458936296,
+      "learning_rate": 9.026464478116972e-05,
+      "loss": 1.1482,
+      "step": 11833
+    },
+    {
+      "epoch": 0.54516976090662,
+      "grad_norm": 1.2085972076026408,
+      "learning_rate": 9.024979487937499e-05,
+      "loss": 1.7236,
+      "step": 11834
+    },
+    {
+      "epoch": 0.545215828995255,
+      "grad_norm": 1.1409940759599355,
+      "learning_rate": 9.023494519465188e-05,
+      "loss": 1.4838,
+      "step": 11835
+    },
+    {
+      "epoch": 0.54526189708389,
+      "grad_norm": 1.634458897804687,
+      "learning_rate": 9.02200957273309e-05,
+      "loss": 1.1211,
+      "step": 11836
+    },
+    {
+      "epoch": 0.545307965172525,
+      "grad_norm": 0.7468708716695346,
+      "learning_rate": 9.020524647774278e-05,
+      "loss": 1.5096,
+      "step": 11837
+    },
+    {
+      "epoch": 0.5453540332611599,
+      "grad_norm": 1.2657858648739269,
+      "learning_rate": 9.019039744621803e-05,
+      "loss": 1.3806,
+      "step": 11838
+    },
+    {
+      "epoch": 0.545400101349795,
+      "grad_norm": 1.630898810749243,
+      "learning_rate": 9.017554863308725e-05,
+      "loss": 1.5869,
+      "step": 11839
+    },
+    {
+      "epoch": 0.54544616943843,
+      "grad_norm": 1.3710765250753534,
+      "learning_rate": 9.016070003868104e-05,
+      "loss": 1.1028,
+      "step": 11840
+    },
+    {
+      "epoch": 0.545492237527065,
+      "grad_norm": 2.900959662186275,
+      "learning_rate": 9.014585166333e-05,
+      "loss": 1.3954,
+      "step": 11841
+    },
+    {
+      "epoch": 0.5455383056157,
+      "grad_norm": 1.8336985030538826,
+      "learning_rate": 9.013100350736464e-05,
+      "loss": 1.5578,
+      "step": 11842
+    },
+    {
+      "epoch": 0.545584373704335,
+      "grad_norm": 1.2185731431195224,
+      "learning_rate": 9.01161555711156e-05,
+      "loss": 1.1997,
+      "step": 11843
+    },
+    {
+      "epoch": 0.54563044179297,
+      "grad_norm": 1.128563908737987,
+      "learning_rate": 9.010130785491342e-05,
+      "loss": 1.5159,
+      "step": 11844
+    },
+    {
+      "epoch": 0.545676509881605,
+      "grad_norm": 0.948651127237892,
+      "learning_rate": 9.008646035908858e-05,
+      "loss": 1.4172,
+      "step": 11845
+    },
+    {
+      "epoch": 0.54572257797024,
+      "grad_norm": 0.8643222212793386,
+      "learning_rate": 9.007161308397179e-05,
+      "loss": 1.1899,
+      "step": 11846
+    },
+    {
+      "epoch": 0.545768646058875,
+      "grad_norm": 0.9741668094596517,
+      "learning_rate": 9.005676602989343e-05,
+      "loss": 1.3564,
+      "step": 11847
+    },
+    {
+      "epoch": 0.54581471414751,
+      "grad_norm": 0.8785347338965334,
+      "learning_rate": 9.004191919718419e-05,
+      "loss": 1.4854,
+      "step": 11848
+    },
+    {
+      "epoch": 0.545860782236145,
+      "grad_norm": 1.31445092846886,
+      "learning_rate": 9.002707258617448e-05,
+      "loss": 1.4228,
+      "step": 11849
+    },
+    {
+      "epoch": 0.54590685032478,
+      "grad_norm": 1.0953738656768142,
+      "learning_rate": 9.001222619719493e-05,
+      "loss": 1.4399,
+      "step": 11850
+    },
+    {
+      "epoch": 0.545952918413415,
+      "grad_norm": 1.179803253068292,
+      "learning_rate": 8.999738003057603e-05,
+      "loss": 1.253,
+      "step": 11851
+    },
+    {
+      "epoch": 0.54599898650205,
+      "grad_norm": 1.1006330352632003,
+      "learning_rate": 8.998253408664834e-05,
+      "loss": 1.7427,
+      "step": 11852
+    },
+    {
+      "epoch": 0.5460450545906851,
+      "grad_norm": 2.053629802901146,
+      "learning_rate": 8.996768836574237e-05,
+      "loss": 1.2868,
+      "step": 11853
+    },
+    {
+      "epoch": 0.54609112267932,
+      "grad_norm": 0.7890776302314413,
+      "learning_rate": 8.995284286818854e-05,
+      "loss": 1.474,
+      "step": 11854
+    },
+    {
+      "epoch": 0.5461371907679551,
+      "grad_norm": 1.0589501496647729,
+      "learning_rate": 8.993799759431751e-05,
+      "loss": 1.6226,
+      "step": 11855
+    },
+    {
+      "epoch": 0.54618325885659,
+      "grad_norm": 0.7769333707230595,
+      "learning_rate": 8.992315254445965e-05,
+      "loss": 1.7702,
+      "step": 11856
+    },
+    {
+      "epoch": 0.5462293269452251,
+      "grad_norm": 2.9557676514180478,
+      "learning_rate": 8.990830771894559e-05,
+      "loss": 1.5014,
+      "step": 11857
+    },
+    {
+      "epoch": 0.54627539503386,
+      "grad_norm": 0.875065394152828,
+      "learning_rate": 8.989346311810567e-05,
+      "loss": 1.5911,
+      "step": 11858
+    },
+    {
+      "epoch": 0.5463214631224951,
+      "grad_norm": 1.289646276119124,
+      "learning_rate": 8.987861874227055e-05,
+      "loss": 1.5072,
+      "step": 11859
+    },
+    {
+      "epoch": 0.54636753121113,
+      "grad_norm": 1.0611358738409165,
+      "learning_rate": 8.986377459177056e-05,
+      "loss": 1.2422,
+      "step": 11860
+    },
+    {
+      "epoch": 0.5464135992997651,
+      "grad_norm": 2.4149671891026907,
+      "learning_rate": 8.984893066693631e-05,
+      "loss": 1.5311,
+      "step": 11861
+    },
+    {
+      "epoch": 0.5464596673884,
+      "grad_norm": 4.907146754534857,
+      "learning_rate": 8.983408696809819e-05,
+      "loss": 1.1526,
+      "step": 11862
+    },
+    {
+      "epoch": 0.5465057354770351,
+      "grad_norm": 1.4899806790503194,
+      "learning_rate": 8.981924349558666e-05,
+      "loss": 1.3876,
+      "step": 11863
+    },
+    {
+      "epoch": 0.54655180356567,
+      "grad_norm": 0.7925319481014788,
+      "learning_rate": 8.980440024973225e-05,
+      "loss": 1.5222,
+      "step": 11864
+    },
+    {
+      "epoch": 0.5465978716543051,
+      "grad_norm": 3.074939873570702,
+      "learning_rate": 8.978955723086535e-05,
+      "loss": 1.3088,
+      "step": 11865
+    },
+    {
+      "epoch": 0.54664393974294,
+      "grad_norm": 2.3370527029319357,
+      "learning_rate": 8.977471443931652e-05,
+      "loss": 1.2377,
+      "step": 11866
+    },
+    {
+      "epoch": 0.5466900078315751,
+      "grad_norm": 1.5951613579679187,
+      "learning_rate": 8.975987187541605e-05,
+      "loss": 1.2567,
+      "step": 11867
+    },
+    {
+      "epoch": 0.5467360759202101,
+      "grad_norm": 1.2362031537454234,
+      "learning_rate": 8.974502953949456e-05,
+      "loss": 1.1685,
+      "step": 11868
+    },
+    {
+      "epoch": 0.5467821440088451,
+      "grad_norm": 1.0532671320494527,
+      "learning_rate": 8.973018743188233e-05,
+      "loss": 1.652,
+      "step": 11869
+    },
+    {
+      "epoch": 0.5468282120974801,
+      "grad_norm": 1.2700313474553429,
+      "learning_rate": 8.971534555290995e-05,
+      "loss": 1.2499,
+      "step": 11870
+    },
+    {
+      "epoch": 0.5468742801861151,
+      "grad_norm": 1.0106885338238998,
+      "learning_rate": 8.970050390290773e-05,
+      "loss": 1.3159,
+      "step": 11871
+    },
+    {
+      "epoch": 0.5469203482747501,
+      "grad_norm": 3.3443880499579195,
+      "learning_rate": 8.968566248220612e-05,
+      "loss": 1.4965,
+      "step": 11872
+    },
+    {
+      "epoch": 0.546966416363385,
+      "grad_norm": 1.0836142096668786,
+      "learning_rate": 8.967082129113555e-05,
+      "loss": 1.4767,
+      "step": 11873
+    },
+    {
+      "epoch": 0.5470124844520201,
+      "grad_norm": 1.163527484330239,
+      "learning_rate": 8.965598033002643e-05,
+      "loss": 1.502,
+      "step": 11874
+    },
+    {
+      "epoch": 0.547058552540655,
+      "grad_norm": 0.9512184008205999,
+      "learning_rate": 8.964113959920919e-05,
+      "loss": 1.651,
+      "step": 11875
+    },
+    {
+      "epoch": 0.5471046206292901,
+      "grad_norm": 2.1884943486073083,
+      "learning_rate": 8.962629909901419e-05,
+      "loss": 1.1046,
+      "step": 11876
+    },
+    {
+      "epoch": 0.547150688717925,
+      "grad_norm": 0.7776968515622402,
+      "learning_rate": 8.961145882977187e-05,
+      "loss": 1.2305,
+      "step": 11877
+    },
+    {
+      "epoch": 0.5471967568065601,
+      "grad_norm": 0.9704623108697699,
+      "learning_rate": 8.959661879181258e-05,
+      "loss": 1.3783,
+      "step": 11878
+    },
+    {
+      "epoch": 0.547242824895195,
+      "grad_norm": 0.7744171178706791,
+      "learning_rate": 8.95817789854668e-05,
+      "loss": 1.5738,
+      "step": 11879
+    },
+    {
+      "epoch": 0.5472888929838301,
+      "grad_norm": 1.5448886543161091,
+      "learning_rate": 8.956693941106481e-05,
+      "loss": 1.5743,
+      "step": 11880
+    },
+    {
+      "epoch": 0.5473349610724652,
+      "grad_norm": 17.09412632046919,
+      "learning_rate": 8.955210006893701e-05,
+      "loss": 1.1541,
+      "step": 11881
+    },
+    {
+      "epoch": 0.5473810291611001,
+      "grad_norm": 0.9186722160819165,
+      "learning_rate": 8.953726095941381e-05,
+      "loss": 1.2922,
+      "step": 11882
+    },
+    {
+      "epoch": 0.5474270972497352,
+      "grad_norm": 1.269809366474395,
+      "learning_rate": 8.952242208282553e-05,
+      "loss": 1.2928,
+      "step": 11883
+    },
+    {
+      "epoch": 0.5474731653383701,
+      "grad_norm": 1.4640871970403937,
+      "learning_rate": 8.950758343950257e-05,
+      "loss": 1.6124,
+      "step": 11884
+    },
+    {
+      "epoch": 0.5475192334270051,
+      "grad_norm": 2.115494325668049,
+      "learning_rate": 8.949274502977528e-05,
+      "loss": 1.4273,
+      "step": 11885
+    },
+    {
+      "epoch": 0.5475653015156401,
+      "grad_norm": 1.7924182381678466,
+      "learning_rate": 8.9477906853974e-05,
+      "loss": 1.1441,
+      "step": 11886
+    },
+    {
+      "epoch": 0.5476113696042751,
+      "grad_norm": 1.6650892253086427,
+      "learning_rate": 8.946306891242907e-05,
+      "loss": 1.2856,
+      "step": 11887
+    },
+    {
+      "epoch": 0.5476574376929101,
+      "grad_norm": 4.008728676762909,
+      "learning_rate": 8.944823120547087e-05,
+      "loss": 1.319,
+      "step": 11888
+    },
+    {
+      "epoch": 0.5477035057815451,
+      "grad_norm": 1.616585195811668,
+      "learning_rate": 8.94333937334297e-05,
+      "loss": 1.5633,
+      "step": 11889
+    },
+    {
+      "epoch": 0.5477495738701801,
+      "grad_norm": 1.6003538032908202,
+      "learning_rate": 8.941855649663589e-05,
+      "loss": 1.6058,
+      "step": 11890
+    },
+    {
+      "epoch": 0.5477956419588151,
+      "grad_norm": 2.6421654134957184,
+      "learning_rate": 8.94037194954198e-05,
+      "loss": 1.2586,
+      "step": 11891
+    },
+    {
+      "epoch": 0.5478417100474501,
+      "grad_norm": 1.9699264415081332,
+      "learning_rate": 8.93888827301117e-05,
+      "loss": 1.4088,
+      "step": 11892
+    },
+    {
+      "epoch": 0.5478877781360851,
+      "grad_norm": 1.258184023464178,
+      "learning_rate": 8.937404620104195e-05,
+      "loss": 1.2456,
+      "step": 11893
+    },
+    {
+      "epoch": 0.5479338462247202,
+      "grad_norm": 1.0324570991662758,
+      "learning_rate": 8.935920990854082e-05,
+      "loss": 1.4894,
+      "step": 11894
+    },
+    {
+      "epoch": 0.5479799143133551,
+      "grad_norm": 1.2690382644045786,
+      "learning_rate": 8.934437385293865e-05,
+      "loss": 1.546,
+      "step": 11895
+    },
+    {
+      "epoch": 0.5480259824019902,
+      "grad_norm": 2.580464698469881,
+      "learning_rate": 8.932953803456572e-05,
+      "loss": 1.4663,
+      "step": 11896
+    },
+    {
+      "epoch": 0.5480720504906251,
+      "grad_norm": 0.9774894731313233,
+      "learning_rate": 8.931470245375234e-05,
+      "loss": 1.5971,
+      "step": 11897
+    },
+    {
+      "epoch": 0.5481181185792602,
+      "grad_norm": 1.2124468358465486,
+      "learning_rate": 8.92998671108288e-05,
+      "loss": 1.503,
+      "step": 11898
+    },
+    {
+      "epoch": 0.5481641866678951,
+      "grad_norm": 2.9611239599177757,
+      "learning_rate": 8.928503200612534e-05,
+      "loss": 1.2496,
+      "step": 11899
+    },
+    {
+      "epoch": 0.5482102547565302,
+      "grad_norm": 1.0480010779716489,
+      "learning_rate": 8.92701971399723e-05,
+      "loss": 1.5529,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5482563228451651,
+      "grad_norm": 1.377002584262804,
+      "learning_rate": 8.92553625126999e-05,
+      "loss": 1.3166,
+      "step": 11901
+    },
+    {
+      "epoch": 0.5483023909338002,
+      "grad_norm": 2.2124430073592025,
+      "learning_rate": 8.924052812463844e-05,
+      "loss": 1.5691,
+      "step": 11902
+    },
+    {
+      "epoch": 0.5483484590224351,
+      "grad_norm": 1.0962318366169919,
+      "learning_rate": 8.922569397611816e-05,
+      "loss": 1.419,
+      "step": 11903
+    },
+    {
+      "epoch": 0.5483945271110702,
+      "grad_norm": 1.4645137465377411,
+      "learning_rate": 8.921086006746936e-05,
+      "loss": 1.8126,
+      "step": 11904
+    },
+    {
+      "epoch": 0.5484405951997051,
+      "grad_norm": 1.1165678030877255,
+      "learning_rate": 8.919602639902223e-05,
+      "loss": 1.6417,
+      "step": 11905
+    },
+    {
+      "epoch": 0.5484866632883402,
+      "grad_norm": 1.3971683338578034,
+      "learning_rate": 8.918119297110707e-05,
+      "loss": 1.2368,
+      "step": 11906
+    },
+    {
+      "epoch": 0.5485327313769752,
+      "grad_norm": 1.5041221180124358,
+      "learning_rate": 8.91663597840541e-05,
+      "loss": 1.0153,
+      "step": 11907
+    },
+    {
+      "epoch": 0.5485787994656102,
+      "grad_norm": 1.318620470615105,
+      "learning_rate": 8.915152683819353e-05,
+      "loss": 1.363,
+      "step": 11908
+    },
+    {
+      "epoch": 0.5486248675542452,
+      "grad_norm": 1.3143458899704104,
+      "learning_rate": 8.913669413385564e-05,
+      "loss": 1.228,
+      "step": 11909
+    },
+    {
+      "epoch": 0.5486709356428802,
+      "grad_norm": 1.7717061525570739,
+      "learning_rate": 8.91218616713706e-05,
+      "loss": 1.2621,
+      "step": 11910
+    },
+    {
+      "epoch": 0.5487170037315152,
+      "grad_norm": 1.2399461642072669,
+      "learning_rate": 8.910702945106869e-05,
+      "loss": 1.377,
+      "step": 11911
+    },
+    {
+      "epoch": 0.5487630718201502,
+      "grad_norm": 1.234613518436009,
+      "learning_rate": 8.909219747328006e-05,
+      "loss": 1.3132,
+      "step": 11912
+    },
+    {
+      "epoch": 0.5488091399087852,
+      "grad_norm": 1.2736922801617991,
+      "learning_rate": 8.907736573833498e-05,
+      "loss": 1.5418,
+      "step": 11913
+    },
+    {
+      "epoch": 0.5488552079974202,
+      "grad_norm": 1.3398582934931005,
+      "learning_rate": 8.90625342465636e-05,
+      "loss": 1.2913,
+      "step": 11914
+    },
+    {
+      "epoch": 0.5489012760860552,
+      "grad_norm": 1.1335380806360773,
+      "learning_rate": 8.904770299829615e-05,
+      "loss": 1.3936,
+      "step": 11915
+    },
+    {
+      "epoch": 0.5489473441746902,
+      "grad_norm": 1.5616032579307524,
+      "learning_rate": 8.90328719938628e-05,
+      "loss": 1.585,
+      "step": 11916
+    },
+    {
+      "epoch": 0.5489934122633252,
+      "grad_norm": 0.9987675570336794,
+      "learning_rate": 8.901804123359378e-05,
+      "loss": 1.0874,
+      "step": 11917
+    },
+    {
+      "epoch": 0.5490394803519602,
+      "grad_norm": 0.9660373369257507,
+      "learning_rate": 8.900321071781923e-05,
+      "loss": 1.2472,
+      "step": 11918
+    },
+    {
+      "epoch": 0.5490855484405952,
+      "grad_norm": 2.87091690159696,
+      "learning_rate": 8.898838044686933e-05,
+      "loss": 1.3742,
+      "step": 11919
+    },
+    {
+      "epoch": 0.5491316165292301,
+      "grad_norm": 0.9180556818418245,
+      "learning_rate": 8.897355042107427e-05,
+      "loss": 1.4956,
+      "step": 11920
+    },
+    {
+      "epoch": 0.5491776846178652,
+      "grad_norm": 1.7139738707548045,
+      "learning_rate": 8.895872064076418e-05,
+      "loss": 1.2233,
+      "step": 11921
+    },
+    {
+      "epoch": 0.5492237527065003,
+      "grad_norm": 0.8890636082314379,
+      "learning_rate": 8.894389110626928e-05,
+      "loss": 1.4917,
+      "step": 11922
+    },
+    {
+      "epoch": 0.5492698207951352,
+      "grad_norm": 1.6668013446514662,
+      "learning_rate": 8.892906181791964e-05,
+      "loss": 1.4652,
+      "step": 11923
+    },
+    {
+      "epoch": 0.5493158888837703,
+      "grad_norm": 1.0622073483498269,
+      "learning_rate": 8.891423277604549e-05,
+      "loss": 1.3344,
+      "step": 11924
+    },
+    {
+      "epoch": 0.5493619569724052,
+      "grad_norm": 1.6813400672662022,
+      "learning_rate": 8.889940398097693e-05,
+      "loss": 1.2334,
+      "step": 11925
+    },
+    {
+      "epoch": 0.5494080250610403,
+      "grad_norm": 1.180251279970634,
+      "learning_rate": 8.888457543304412e-05,
+      "loss": 1.3762,
+      "step": 11926
+    },
+    {
+      "epoch": 0.5494540931496752,
+      "grad_norm": 1.6921214100580408,
+      "learning_rate": 8.886974713257718e-05,
+      "loss": 1.0943,
+      "step": 11927
+    },
+    {
+      "epoch": 0.5495001612383102,
+      "grad_norm": 0.7938592538156554,
+      "learning_rate": 8.88549190799062e-05,
+      "loss": 1.2932,
+      "step": 11928
+    },
+    {
+      "epoch": 0.5495462293269452,
+      "grad_norm": 1.3624580029813778,
+      "learning_rate": 8.884009127536137e-05,
+      "loss": 1.2912,
+      "step": 11929
+    },
+    {
+      "epoch": 0.5495922974155802,
+      "grad_norm": 1.2180337345647685,
+      "learning_rate": 8.882526371927276e-05,
+      "loss": 1.2338,
+      "step": 11930
+    },
+    {
+      "epoch": 0.5496383655042152,
+      "grad_norm": 1.114656665172745,
+      "learning_rate": 8.88104364119705e-05,
+      "loss": 1.2746,
+      "step": 11931
+    },
+    {
+      "epoch": 0.5496844335928502,
+      "grad_norm": 0.9289628335473005,
+      "learning_rate": 8.879560935378467e-05,
+      "loss": 1.5207,
+      "step": 11932
+    },
+    {
+      "epoch": 0.5497305016814852,
+      "grad_norm": 0.9008698951296904,
+      "learning_rate": 8.87807825450454e-05,
+      "loss": 1.2663,
+      "step": 11933
+    },
+    {
+      "epoch": 0.5497765697701202,
+      "grad_norm": 1.7610017260431883,
+      "learning_rate": 8.876595598608277e-05,
+      "loss": 1.4453,
+      "step": 11934
+    },
+    {
+      "epoch": 0.5498226378587553,
+      "grad_norm": 1.6182798169975838,
+      "learning_rate": 8.875112967722687e-05,
+      "loss": 1.2787,
+      "step": 11935
+    },
+    {
+      "epoch": 0.5498687059473902,
+      "grad_norm": 1.936400630003218,
+      "learning_rate": 8.873630361880779e-05,
+      "loss": 1.2627,
+      "step": 11936
+    },
+    {
+      "epoch": 0.5499147740360253,
+      "grad_norm": 1.9364301806638262,
+      "learning_rate": 8.872147781115557e-05,
+      "loss": 1.1993,
+      "step": 11937
+    },
+    {
+      "epoch": 0.5499608421246602,
+      "grad_norm": 5.717325316838684,
+      "learning_rate": 8.870665225460033e-05,
+      "loss": 1.2347,
+      "step": 11938
+    },
+    {
+      "epoch": 0.5500069102132953,
+      "grad_norm": 4.440964615173271,
+      "learning_rate": 8.86918269494721e-05,
+      "loss": 1.4113,
+      "step": 11939
+    },
+    {
+      "epoch": 0.5500529783019302,
+      "grad_norm": 0.9122922639915353,
+      "learning_rate": 8.867700189610097e-05,
+      "loss": 1.4687,
+      "step": 11940
+    },
+    {
+      "epoch": 0.5500990463905653,
+      "grad_norm": 1.1673112130271484,
+      "learning_rate": 8.866217709481696e-05,
+      "loss": 1.2433,
+      "step": 11941
+    },
+    {
+      "epoch": 0.5501451144792002,
+      "grad_norm": 1.9337888251896673,
+      "learning_rate": 8.864735254595015e-05,
+      "loss": 1.5951,
+      "step": 11942
+    },
+    {
+      "epoch": 0.5501911825678353,
+      "grad_norm": 0.9316610410666936,
+      "learning_rate": 8.863252824983054e-05,
+      "loss": 1.5215,
+      "step": 11943
+    },
+    {
+      "epoch": 0.5502372506564702,
+      "grad_norm": 1.171902948051788,
+      "learning_rate": 8.861770420678824e-05,
+      "loss": 1.3146,
+      "step": 11944
+    },
+    {
+      "epoch": 0.5502833187451053,
+      "grad_norm": 2.2995567362366924,
+      "learning_rate": 8.860288041715326e-05,
+      "loss": 1.6441,
+      "step": 11945
+    },
+    {
+      "epoch": 0.5503293868337402,
+      "grad_norm": 1.2600695571149965,
+      "learning_rate": 8.858805688125552e-05,
+      "loss": 1.1709,
+      "step": 11946
+    },
+    {
+      "epoch": 0.5503754549223753,
+      "grad_norm": 3.2636467403687854,
+      "learning_rate": 8.857323359942521e-05,
+      "loss": 1.271,
+      "step": 11947
+    },
+    {
+      "epoch": 0.5504215230110103,
+      "grad_norm": 1.1523967532492556,
+      "learning_rate": 8.855841057199219e-05,
+      "loss": 1.2524,
+      "step": 11948
+    },
+    {
+      "epoch": 0.5504675910996453,
+      "grad_norm": 1.2222753161802993,
+      "learning_rate": 8.854358779928659e-05,
+      "loss": 1.2363,
+      "step": 11949
+    },
+    {
+      "epoch": 0.5505136591882803,
+      "grad_norm": 0.9677739893333467,
+      "learning_rate": 8.852876528163834e-05,
+      "loss": 1.6437,
+      "step": 11950
+    },
+    {
+      "epoch": 0.5505597272769153,
+      "grad_norm": 2.9753297459747956,
+      "learning_rate": 8.851394301937748e-05,
+      "loss": 1.472,
+      "step": 11951
+    },
+    {
+      "epoch": 0.5506057953655503,
+      "grad_norm": 1.272943681361081,
+      "learning_rate": 8.849912101283396e-05,
+      "loss": 1.2706,
+      "step": 11952
+    },
+    {
+      "epoch": 0.5506518634541853,
+      "grad_norm": 1.0036797993423063,
+      "learning_rate": 8.848429926233781e-05,
+      "loss": 1.3453,
+      "step": 11953
+    },
+    {
+      "epoch": 0.5506979315428203,
+      "grad_norm": 1.058507630428817,
+      "learning_rate": 8.846947776821903e-05,
+      "loss": 1.1648,
+      "step": 11954
+    },
+    {
+      "epoch": 0.5507439996314553,
+      "grad_norm": 2.455795050626941,
+      "learning_rate": 8.845465653080747e-05,
+      "loss": 1.2254,
+      "step": 11955
+    },
+    {
+      "epoch": 0.5507900677200903,
+      "grad_norm": 1.7446419809163412,
+      "learning_rate": 8.843983555043326e-05,
+      "loss": 1.4533,
+      "step": 11956
+    },
+    {
+      "epoch": 0.5508361358087253,
+      "grad_norm": 1.3133826675228735,
+      "learning_rate": 8.842501482742623e-05,
+      "loss": 1.4852,
+      "step": 11957
+    },
+    {
+      "epoch": 0.5508822038973603,
+      "grad_norm": 2.011452341061443,
+      "learning_rate": 8.841019436211646e-05,
+      "loss": 1.3678,
+      "step": 11958
+    },
+    {
+      "epoch": 0.5509282719859953,
+      "grad_norm": 1.3302675361487226,
+      "learning_rate": 8.839537415483378e-05,
+      "loss": 1.5229,
+      "step": 11959
+    },
+    {
+      "epoch": 0.5509743400746303,
+      "grad_norm": 1.4356361641304622,
+      "learning_rate": 8.838055420590825e-05,
+      "loss": 1.4184,
+      "step": 11960
+    },
+    {
+      "epoch": 0.5510204081632653,
+      "grad_norm": 1.1392832111278879,
+      "learning_rate": 8.83657345156697e-05,
+      "loss": 1.4322,
+      "step": 11961
+    },
+    {
+      "epoch": 0.5510664762519003,
+      "grad_norm": 1.3665867832562655,
+      "learning_rate": 8.835091508444819e-05,
+      "loss": 0.9484,
+      "step": 11962
+    },
+    {
+      "epoch": 0.5511125443405354,
+      "grad_norm": 1.1861142640488092,
+      "learning_rate": 8.833609591257355e-05,
+      "loss": 1.3328,
+      "step": 11963
+    },
+    {
+      "epoch": 0.5511586124291703,
+      "grad_norm": 0.7748927969916382,
+      "learning_rate": 8.83212770003757e-05,
+      "loss": 1.5251,
+      "step": 11964
+    },
+    {
+      "epoch": 0.5512046805178054,
+      "grad_norm": 1.1278998035476469,
+      "learning_rate": 8.830645834818466e-05,
+      "loss": 1.4849,
+      "step": 11965
+    },
+    {
+      "epoch": 0.5512507486064403,
+      "grad_norm": 1.2312789695846782,
+      "learning_rate": 8.829163995633021e-05,
+      "loss": 1.3225,
+      "step": 11966
+    },
+    {
+      "epoch": 0.5512968166950754,
+      "grad_norm": 0.9812884679557714,
+      "learning_rate": 8.827682182514238e-05,
+      "loss": 1.4953,
+      "step": 11967
+    },
+    {
+      "epoch": 0.5513428847837103,
+      "grad_norm": 1.034857954522184,
+      "learning_rate": 8.826200395495095e-05,
+      "loss": 1.2243,
+      "step": 11968
+    },
+    {
+      "epoch": 0.5513889528723454,
+      "grad_norm": 1.6241681177345615,
+      "learning_rate": 8.824718634608594e-05,
+      "loss": 1.5251,
+      "step": 11969
+    },
+    {
+      "epoch": 0.5514350209609803,
+      "grad_norm": 1.6458076256732075,
+      "learning_rate": 8.823236899887711e-05,
+      "loss": 1.435,
+      "step": 11970
+    },
+    {
+      "epoch": 0.5514810890496153,
+      "grad_norm": 1.0088662728852436,
+      "learning_rate": 8.821755191365447e-05,
+      "loss": 1.0468,
+      "step": 11971
+    },
+    {
+      "epoch": 0.5515271571382503,
+      "grad_norm": 1.1627519237082586,
+      "learning_rate": 8.820273509074781e-05,
+      "loss": 1.163,
+      "step": 11972
+    },
+    {
+      "epoch": 0.5515732252268853,
+      "grad_norm": 1.6163608853883848,
+      "learning_rate": 8.818791853048701e-05,
+      "loss": 1.1158,
+      "step": 11973
+    },
+    {
+      "epoch": 0.5516192933155203,
+      "grad_norm": 1.5551528659955647,
+      "learning_rate": 8.817310223320196e-05,
+      "loss": 1.5897,
+      "step": 11974
+    },
+    {
+      "epoch": 0.5516653614041553,
+      "grad_norm": 2.040403762807953,
+      "learning_rate": 8.81582861992225e-05,
+      "loss": 1.35,
+      "step": 11975
+    },
+    {
+      "epoch": 0.5517114294927904,
+      "grad_norm": 2.7372979832816,
+      "learning_rate": 8.814347042887852e-05,
+      "loss": 1.2781,
+      "step": 11976
+    },
+    {
+      "epoch": 0.5517574975814253,
+      "grad_norm": 0.9093684299403284,
+      "learning_rate": 8.812865492249978e-05,
+      "loss": 1.6085,
+      "step": 11977
+    },
+    {
+      "epoch": 0.5518035656700604,
+      "grad_norm": 3.6453814940329567,
+      "learning_rate": 8.811383968041626e-05,
+      "loss": 1.4539,
+      "step": 11978
+    },
+    {
+      "epoch": 0.5518496337586953,
+      "grad_norm": 0.8392151408439139,
+      "learning_rate": 8.809902470295765e-05,
+      "loss": 1.5619,
+      "step": 11979
+    },
+    {
+      "epoch": 0.5518957018473304,
+      "grad_norm": 1.181065985683156,
+      "learning_rate": 8.808420999045393e-05,
+      "loss": 1.3632,
+      "step": 11980
+    },
+    {
+      "epoch": 0.5519417699359653,
+      "grad_norm": 1.471602307399071,
+      "learning_rate": 8.806939554323482e-05,
+      "loss": 1.6615,
+      "step": 11981
+    },
+    {
+      "epoch": 0.5519878380246004,
+      "grad_norm": 1.0191789402709668,
+      "learning_rate": 8.805458136163012e-05,
+      "loss": 1.5457,
+      "step": 11982
+    },
+    {
+      "epoch": 0.5520339061132353,
+      "grad_norm": 1.2870234643616487,
+      "learning_rate": 8.803976744596972e-05,
+      "loss": 1.3851,
+      "step": 11983
+    },
+    {
+      "epoch": 0.5520799742018704,
+      "grad_norm": 1.11387771720687,
+      "learning_rate": 8.802495379658335e-05,
+      "loss": 1.387,
+      "step": 11984
+    },
+    {
+      "epoch": 0.5521260422905053,
+      "grad_norm": 1.5000623201281358,
+      "learning_rate": 8.801014041380089e-05,
+      "loss": 1.1218,
+      "step": 11985
+    },
+    {
+      "epoch": 0.5521721103791404,
+      "grad_norm": 1.3621658702313217,
+      "learning_rate": 8.799532729795207e-05,
+      "loss": 1.4792,
+      "step": 11986
+    },
+    {
+      "epoch": 0.5522181784677753,
+      "grad_norm": 2.4721546010333353,
+      "learning_rate": 8.798051444936673e-05,
+      "loss": 1.2667,
+      "step": 11987
+    },
+    {
+      "epoch": 0.5522642465564104,
+      "grad_norm": 1.0638305482753923,
+      "learning_rate": 8.796570186837459e-05,
+      "loss": 1.2862,
+      "step": 11988
+    },
+    {
+      "epoch": 0.5523103146450454,
+      "grad_norm": 3.7177240020375417,
+      "learning_rate": 8.79508895553055e-05,
+      "loss": 1.2475,
+      "step": 11989
+    },
+    {
+      "epoch": 0.5523563827336804,
+      "grad_norm": 1.0692328042497645,
+      "learning_rate": 8.793607751048919e-05,
+      "loss": 1.1912,
+      "step": 11990
+    },
+    {
+      "epoch": 0.5524024508223154,
+      "grad_norm": 2.0272072543173927,
+      "learning_rate": 8.792126573425542e-05,
+      "loss": 1.0205,
+      "step": 11991
+    },
+    {
+      "epoch": 0.5524485189109504,
+      "grad_norm": 4.739119638994194,
+      "learning_rate": 8.790645422693396e-05,
+      "loss": 1.4575,
+      "step": 11992
+    },
+    {
+      "epoch": 0.5524945869995854,
+      "grad_norm": 2.3039206683712106,
+      "learning_rate": 8.789164298885455e-05,
+      "loss": 1.4726,
+      "step": 11993
+    },
+    {
+      "epoch": 0.5525406550882204,
+      "grad_norm": 1.2600589523586707,
+      "learning_rate": 8.787683202034696e-05,
+      "loss": 1.362,
+      "step": 11994
+    },
+    {
+      "epoch": 0.5525867231768554,
+      "grad_norm": 2.03616664475525,
+      "learning_rate": 8.786202132174089e-05,
+      "loss": 1.0414,
+      "step": 11995
+    },
+    {
+      "epoch": 0.5526327912654904,
+      "grad_norm": 1.0478474344184703,
+      "learning_rate": 8.784721089336614e-05,
+      "loss": 1.4637,
+      "step": 11996
+    },
+    {
+      "epoch": 0.5526788593541254,
+      "grad_norm": 1.6602172921593035,
+      "learning_rate": 8.783240073555238e-05,
+      "loss": 1.2969,
+      "step": 11997
+    },
+    {
+      "epoch": 0.5527249274427604,
+      "grad_norm": 0.9138655530934214,
+      "learning_rate": 8.781759084862937e-05,
+      "loss": 1.4875,
+      "step": 11998
+    },
+    {
+      "epoch": 0.5527709955313954,
+      "grad_norm": 1.6967367664340849,
+      "learning_rate": 8.780278123292682e-05,
+      "loss": 1.5378,
+      "step": 11999
+    },
+    {
+      "epoch": 0.5528170636200304,
+      "grad_norm": 1.2922097905057548,
+      "learning_rate": 8.77879718887744e-05,
+      "loss": 1.2732,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5528631317086654,
+      "grad_norm": 1.6871483949862538,
+      "learning_rate": 8.777316281650187e-05,
+      "loss": 1.5828,
+      "step": 12001
+    },
+    {
+      "epoch": 0.5529091997973004,
+      "grad_norm": 0.9182853117373786,
+      "learning_rate": 8.775835401643888e-05,
+      "loss": 1.4803,
+      "step": 12002
+    },
+    {
+      "epoch": 0.5529552678859354,
+      "grad_norm": 1.2342695602727676,
+      "learning_rate": 8.774354548891518e-05,
+      "loss": 1.2331,
+      "step": 12003
+    },
+    {
+      "epoch": 0.5530013359745705,
+      "grad_norm": 1.0740248310839473,
+      "learning_rate": 8.77287372342604e-05,
+      "loss": 1.2995,
+      "step": 12004
+    },
+    {
+      "epoch": 0.5530474040632054,
+      "grad_norm": 1.5287383428863501,
+      "learning_rate": 8.771392925280427e-05,
+      "loss": 1.419,
+      "step": 12005
+    },
+    {
+      "epoch": 0.5530934721518405,
+      "grad_norm": 1.5630814175759724,
+      "learning_rate": 8.76991215448764e-05,
+      "loss": 1.4035,
+      "step": 12006
+    },
+    {
+      "epoch": 0.5531395402404754,
+      "grad_norm": 3.802053001683808,
+      "learning_rate": 8.768431411080655e-05,
+      "loss": 1.4683,
+      "step": 12007
+    },
+    {
+      "epoch": 0.5531856083291105,
+      "grad_norm": 1.452616217426781,
+      "learning_rate": 8.76695069509243e-05,
+      "loss": 1.1444,
+      "step": 12008
+    },
+    {
+      "epoch": 0.5532316764177454,
+      "grad_norm": 1.3491022921586415,
+      "learning_rate": 8.765470006555933e-05,
+      "loss": 1.3458,
+      "step": 12009
+    },
+    {
+      "epoch": 0.5532777445063805,
+      "grad_norm": 4.175129189332462,
+      "learning_rate": 8.763989345504131e-05,
+      "loss": 1.3513,
+      "step": 12010
+    },
+    {
+      "epoch": 0.5533238125950154,
+      "grad_norm": 0.9935440788624413,
+      "learning_rate": 8.762508711969986e-05,
+      "loss": 1.2673,
+      "step": 12011
+    },
+    {
+      "epoch": 0.5533698806836505,
+      "grad_norm": 0.6757169558708799,
+      "learning_rate": 8.761028105986463e-05,
+      "loss": 1.4328,
+      "step": 12012
+    },
+    {
+      "epoch": 0.5534159487722854,
+      "grad_norm": 1.7904536032766043,
+      "learning_rate": 8.759547527586523e-05,
+      "loss": 1.3896,
+      "step": 12013
+    },
+    {
+      "epoch": 0.5534620168609204,
+      "grad_norm": 1.5919001832518367,
+      "learning_rate": 8.758066976803133e-05,
+      "loss": 1.1291,
+      "step": 12014
+    },
+    {
+      "epoch": 0.5535080849495554,
+      "grad_norm": 1.4486063028485494,
+      "learning_rate": 8.756586453669251e-05,
+      "loss": 1.4346,
+      "step": 12015
+    },
+    {
+      "epoch": 0.5535541530381904,
+      "grad_norm": 1.1640185772544598,
+      "learning_rate": 8.755105958217841e-05,
+      "loss": 1.4225,
+      "step": 12016
+    },
+    {
+      "epoch": 0.5536002211268255,
+      "grad_norm": 1.9529254115214765,
+      "learning_rate": 8.753625490481862e-05,
+      "loss": 1.464,
+      "step": 12017
+    },
+    {
+      "epoch": 0.5536462892154604,
+      "grad_norm": 1.6130165789501827,
+      "learning_rate": 8.752145050494272e-05,
+      "loss": 1.1896,
+      "step": 12018
+    },
+    {
+      "epoch": 0.5536923573040955,
+      "grad_norm": 1.4400179988307726,
+      "learning_rate": 8.750664638288035e-05,
+      "loss": 1.5431,
+      "step": 12019
+    },
+    {
+      "epoch": 0.5537384253927304,
+      "grad_norm": 1.3496629370864928,
+      "learning_rate": 8.749184253896106e-05,
+      "loss": 1.0999,
+      "step": 12020
+    },
+    {
+      "epoch": 0.5537844934813655,
+      "grad_norm": 0.7901829786859121,
+      "learning_rate": 8.747703897351448e-05,
+      "loss": 1.4315,
+      "step": 12021
+    },
+    {
+      "epoch": 0.5538305615700004,
+      "grad_norm": 1.2125401735491088,
+      "learning_rate": 8.746223568687012e-05,
+      "loss": 1.2768,
+      "step": 12022
+    },
+    {
+      "epoch": 0.5538766296586355,
+      "grad_norm": 1.9754167001557221,
+      "learning_rate": 8.744743267935761e-05,
+      "loss": 1.3443,
+      "step": 12023
+    },
+    {
+      "epoch": 0.5539226977472704,
+      "grad_norm": 2.3596422805320754,
+      "learning_rate": 8.743262995130646e-05,
+      "loss": 1.0626,
+      "step": 12024
+    },
+    {
+      "epoch": 0.5539687658359055,
+      "grad_norm": 1.7877804608599195,
+      "learning_rate": 8.741782750304629e-05,
+      "loss": 1.1122,
+      "step": 12025
+    },
+    {
+      "epoch": 0.5540148339245404,
+      "grad_norm": 2.800399962836771,
+      "learning_rate": 8.740302533490661e-05,
+      "loss": 1.1606,
+      "step": 12026
+    },
+    {
+      "epoch": 0.5540609020131755,
+      "grad_norm": 1.6347010735563736,
+      "learning_rate": 8.738822344721696e-05,
+      "loss": 1.279,
+      "step": 12027
+    },
+    {
+      "epoch": 0.5541069701018104,
+      "grad_norm": 1.3699709704348333,
+      "learning_rate": 8.73734218403069e-05,
+      "loss": 1.5039,
+      "step": 12028
+    },
+    {
+      "epoch": 0.5541530381904455,
+      "grad_norm": 1.5077315873012642,
+      "learning_rate": 8.735862051450593e-05,
+      "loss": 1.3343,
+      "step": 12029
+    },
+    {
+      "epoch": 0.5541991062790805,
+      "grad_norm": 1.443149644231017,
+      "learning_rate": 8.734381947014363e-05,
+      "loss": 1.0991,
+      "step": 12030
+    },
+    {
+      "epoch": 0.5542451743677155,
+      "grad_norm": 0.7436787389970108,
+      "learning_rate": 8.732901870754946e-05,
+      "loss": 1.5561,
+      "step": 12031
+    },
+    {
+      "epoch": 0.5542912424563505,
+      "grad_norm": 1.3072088137994764,
+      "learning_rate": 8.731421822705299e-05,
+      "loss": 1.3935,
+      "step": 12032
+    },
+    {
+      "epoch": 0.5543373105449855,
+      "grad_norm": 2.239042714210015,
+      "learning_rate": 8.729941802898367e-05,
+      "loss": 1.2442,
+      "step": 12033
+    },
+    {
+      "epoch": 0.5543833786336205,
+      "grad_norm": 1.064015496469021,
+      "learning_rate": 8.728461811367106e-05,
+      "loss": 1.153,
+      "step": 12034
+    },
+    {
+      "epoch": 0.5544294467222555,
+      "grad_norm": 1.5599187240080348,
+      "learning_rate": 8.72698184814446e-05,
+      "loss": 1.3391,
+      "step": 12035
+    },
+    {
+      "epoch": 0.5544755148108905,
+      "grad_norm": 1.3004292081128406,
+      "learning_rate": 8.725501913263382e-05,
+      "loss": 1.4578,
+      "step": 12036
+    },
+    {
+      "epoch": 0.5545215828995255,
+      "grad_norm": 8.508043504577323,
+      "learning_rate": 8.72402200675682e-05,
+      "loss": 1.5407,
+      "step": 12037
+    },
+    {
+      "epoch": 0.5545676509881605,
+      "grad_norm": 2.1381545763375858,
+      "learning_rate": 8.722542128657716e-05,
+      "loss": 1.2183,
+      "step": 12038
+    },
+    {
+      "epoch": 0.5546137190767955,
+      "grad_norm": 1.6910270775910274,
+      "learning_rate": 8.721062278999026e-05,
+      "loss": 1.0756,
+      "step": 12039
+    },
+    {
+      "epoch": 0.5546597871654305,
+      "grad_norm": 1.5242874301330094,
+      "learning_rate": 8.719582457813687e-05,
+      "loss": 1.2793,
+      "step": 12040
+    },
+    {
+      "epoch": 0.5547058552540655,
+      "grad_norm": 1.1730479285213786,
+      "learning_rate": 8.718102665134653e-05,
+      "loss": 1.4478,
+      "step": 12041
+    },
+    {
+      "epoch": 0.5547519233427005,
+      "grad_norm": 1.0960408899602894,
+      "learning_rate": 8.716622900994862e-05,
+      "loss": 1.4769,
+      "step": 12042
+    },
+    {
+      "epoch": 0.5547979914313356,
+      "grad_norm": 1.899923729835973,
+      "learning_rate": 8.715143165427263e-05,
+      "loss": 1.1967,
+      "step": 12043
+    },
+    {
+      "epoch": 0.5548440595199705,
+      "grad_norm": 1.8046539392703733,
+      "learning_rate": 8.713663458464797e-05,
+      "loss": 1.2376,
+      "step": 12044
+    },
+    {
+      "epoch": 0.5548901276086056,
+      "grad_norm": 0.8802186283292295,
+      "learning_rate": 8.712183780140411e-05,
+      "loss": 1.3053,
+      "step": 12045
+    },
+    {
+      "epoch": 0.5549361956972405,
+      "grad_norm": 2.108449206637332,
+      "learning_rate": 8.710704130487047e-05,
+      "loss": 1.4085,
+      "step": 12046
+    },
+    {
+      "epoch": 0.5549822637858756,
+      "grad_norm": 1.0951898965767863,
+      "learning_rate": 8.709224509537638e-05,
+      "loss": 1.2728,
+      "step": 12047
+    },
+    {
+      "epoch": 0.5550283318745105,
+      "grad_norm": 4.344630766162411,
+      "learning_rate": 8.707744917325138e-05,
+      "loss": 1.5598,
+      "step": 12048
+    },
+    {
+      "epoch": 0.5550743999631456,
+      "grad_norm": 1.306609449513669,
+      "learning_rate": 8.706265353882477e-05,
+      "loss": 1.3834,
+      "step": 12049
+    },
+    {
+      "epoch": 0.5551204680517805,
+      "grad_norm": 3.4722802808458533,
+      "learning_rate": 8.704785819242604e-05,
+      "loss": 1.2491,
+      "step": 12050
+    },
+    {
+      "epoch": 0.5551665361404156,
+      "grad_norm": 1.470294932495579,
+      "learning_rate": 8.70330631343845e-05,
+      "loss": 1.3325,
+      "step": 12051
+    },
+    {
+      "epoch": 0.5552126042290505,
+      "grad_norm": 0.997784131439425,
+      "learning_rate": 8.701826836502959e-05,
+      "loss": 1.3265,
+      "step": 12052
+    },
+    {
+      "epoch": 0.5552586723176856,
+      "grad_norm": 1.63423571765148,
+      "learning_rate": 8.700347388469068e-05,
+      "loss": 1.5233,
+      "step": 12053
+    },
+    {
+      "epoch": 0.5553047404063205,
+      "grad_norm": 1.8355468892574842,
+      "learning_rate": 8.698867969369715e-05,
+      "loss": 1.4932,
+      "step": 12054
+    },
+    {
+      "epoch": 0.5553508084949556,
+      "grad_norm": 1.5934043588805316,
+      "learning_rate": 8.697388579237838e-05,
+      "loss": 1.2727,
+      "step": 12055
+    },
+    {
+      "epoch": 0.5553968765835905,
+      "grad_norm": 8.39972433358346,
+      "learning_rate": 8.695909218106364e-05,
+      "loss": 1.2446,
+      "step": 12056
+    },
+    {
+      "epoch": 0.5554429446722255,
+      "grad_norm": 1.3124053896535683,
+      "learning_rate": 8.694429886008243e-05,
+      "loss": 1.2838,
+      "step": 12057
+    },
+    {
+      "epoch": 0.5554890127608606,
+      "grad_norm": 1.9790444550912731,
+      "learning_rate": 8.692950582976393e-05,
+      "loss": 1.2355,
+      "step": 12058
+    },
+    {
+      "epoch": 0.5555350808494955,
+      "grad_norm": 2.7960678560463563,
+      "learning_rate": 8.691471309043767e-05,
+      "loss": 1.6376,
+      "step": 12059
+    },
+    {
+      "epoch": 0.5555811489381306,
+      "grad_norm": 1.2121417570011173,
+      "learning_rate": 8.68999206424328e-05,
+      "loss": 1.3521,
+      "step": 12060
+    },
+    {
+      "epoch": 0.5556272170267655,
+      "grad_norm": 1.2124288477045881,
+      "learning_rate": 8.68851284860788e-05,
+      "loss": 1.4227,
+      "step": 12061
+    },
+    {
+      "epoch": 0.5556732851154006,
+      "grad_norm": 1.5280311333947723,
+      "learning_rate": 8.687033662170486e-05,
+      "loss": 1.4183,
+      "step": 12062
+    },
+    {
+      "epoch": 0.5557193532040355,
+      "grad_norm": 1.2172877428904212,
+      "learning_rate": 8.685554504964043e-05,
+      "loss": 1.5001,
+      "step": 12063
+    },
+    {
+      "epoch": 0.5557654212926706,
+      "grad_norm": 1.3291032524984299,
+      "learning_rate": 8.684075377021473e-05,
+      "loss": 1.0409,
+      "step": 12064
+    },
+    {
+      "epoch": 0.5558114893813055,
+      "grad_norm": 1.658184291007993,
+      "learning_rate": 8.682596278375705e-05,
+      "loss": 1.3264,
+      "step": 12065
+    },
+    {
+      "epoch": 0.5558575574699406,
+      "grad_norm": 1.9811664721136384,
+      "learning_rate": 8.681117209059678e-05,
+      "loss": 1.0854,
+      "step": 12066
+    },
+    {
+      "epoch": 0.5559036255585755,
+      "grad_norm": 1.169482483454106,
+      "learning_rate": 8.679638169106307e-05,
+      "loss": 1.4271,
+      "step": 12067
+    },
+    {
+      "epoch": 0.5559496936472106,
+      "grad_norm": 1.261675542222043,
+      "learning_rate": 8.678159158548537e-05,
+      "loss": 1.4612,
+      "step": 12068
+    },
+    {
+      "epoch": 0.5559957617358455,
+      "grad_norm": 1.2762431679299877,
+      "learning_rate": 8.676680177419279e-05,
+      "loss": 1.3237,
+      "step": 12069
+    },
+    {
+      "epoch": 0.5560418298244806,
+      "grad_norm": 1.1986616640070182,
+      "learning_rate": 8.675201225751474e-05,
+      "loss": 1.3454,
+      "step": 12070
+    },
+    {
+      "epoch": 0.5560878979131156,
+      "grad_norm": 1.0296496530957604,
+      "learning_rate": 8.673722303578038e-05,
+      "loss": 1.1425,
+      "step": 12071
+    },
+    {
+      "epoch": 0.5561339660017506,
+      "grad_norm": 1.3444547303112488,
+      "learning_rate": 8.672243410931905e-05,
+      "loss": 1.6114,
+      "step": 12072
+    },
+    {
+      "epoch": 0.5561800340903856,
+      "grad_norm": 4.104446307990613,
+      "learning_rate": 8.670764547845994e-05,
+      "loss": 1.1988,
+      "step": 12073
+    },
+    {
+      "epoch": 0.5562261021790206,
+      "grad_norm": 2.1924276204865896,
+      "learning_rate": 8.66928571435323e-05,
+      "loss": 1.3489,
+      "step": 12074
+    },
+    {
+      "epoch": 0.5562721702676556,
+      "grad_norm": 1.4784753792972865,
+      "learning_rate": 8.667806910486538e-05,
+      "loss": 1.2824,
+      "step": 12075
+    },
+    {
+      "epoch": 0.5563182383562906,
+      "grad_norm": 1.1969404832048247,
+      "learning_rate": 8.66632813627884e-05,
+      "loss": 1.0208,
+      "step": 12076
+    },
+    {
+      "epoch": 0.5563643064449256,
+      "grad_norm": 1.0763670626650002,
+      "learning_rate": 8.664849391763063e-05,
+      "loss": 1.5641,
+      "step": 12077
+    },
+    {
+      "epoch": 0.5564103745335606,
+      "grad_norm": 1.1228171446792181,
+      "learning_rate": 8.66337067697212e-05,
+      "loss": 1.2624,
+      "step": 12078
+    },
+    {
+      "epoch": 0.5564564426221956,
+      "grad_norm": 1.62063908219795,
+      "learning_rate": 8.661891991938942e-05,
+      "loss": 1.3797,
+      "step": 12079
+    },
+    {
+      "epoch": 0.5565025107108306,
+      "grad_norm": 1.4303613520483582,
+      "learning_rate": 8.660413336696439e-05,
+      "loss": 1.3489,
+      "step": 12080
+    },
+    {
+      "epoch": 0.5565485787994656,
+      "grad_norm": 1.0849982249778818,
+      "learning_rate": 8.658934711277542e-05,
+      "loss": 1.4029,
+      "step": 12081
+    },
+    {
+      "epoch": 0.5565946468881006,
+      "grad_norm": 1.4546136599133752,
+      "learning_rate": 8.657456115715163e-05,
+      "loss": 1.1924,
+      "step": 12082
+    },
+    {
+      "epoch": 0.5566407149767356,
+      "grad_norm": 1.394642921902655,
+      "learning_rate": 8.655977550042219e-05,
+      "loss": 1.4494,
+      "step": 12083
+    },
+    {
+      "epoch": 0.5566867830653707,
+      "grad_norm": 2.7226832482008945,
+      "learning_rate": 8.654499014291632e-05,
+      "loss": 1.3467,
+      "step": 12084
+    },
+    {
+      "epoch": 0.5567328511540056,
+      "grad_norm": 2.7195923958019246,
+      "learning_rate": 8.653020508496316e-05,
+      "loss": 1.5551,
+      "step": 12085
+    },
+    {
+      "epoch": 0.5567789192426407,
+      "grad_norm": 1.220247032595566,
+      "learning_rate": 8.651542032689191e-05,
+      "loss": 1.3717,
+      "step": 12086
+    },
+    {
+      "epoch": 0.5568249873312756,
+      "grad_norm": 2.4364125463658133,
+      "learning_rate": 8.650063586903168e-05,
+      "loss": 1.4167,
+      "step": 12087
+    },
+    {
+      "epoch": 0.5568710554199107,
+      "grad_norm": 1.1923396086344433,
+      "learning_rate": 8.648585171171167e-05,
+      "loss": 1.2839,
+      "step": 12088
+    },
+    {
+      "epoch": 0.5569171235085456,
+      "grad_norm": 1.1226500694366233,
+      "learning_rate": 8.647106785526094e-05,
+      "loss": 1.1882,
+      "step": 12089
+    },
+    {
+      "epoch": 0.5569631915971807,
+      "grad_norm": 1.276992809145151,
+      "learning_rate": 8.645628430000877e-05,
+      "loss": 1.2074,
+      "step": 12090
+    },
+    {
+      "epoch": 0.5570092596858156,
+      "grad_norm": 1.557678040587173,
+      "learning_rate": 8.64415010462842e-05,
+      "loss": 1.217,
+      "step": 12091
+    },
+    {
+      "epoch": 0.5570553277744507,
+      "grad_norm": 1.8013875684949243,
+      "learning_rate": 8.642671809441632e-05,
+      "loss": 1.3936,
+      "step": 12092
+    },
+    {
+      "epoch": 0.5571013958630856,
+      "grad_norm": 4.3760823547809595,
+      "learning_rate": 8.641193544473431e-05,
+      "loss": 1.2867,
+      "step": 12093
+    },
+    {
+      "epoch": 0.5571474639517207,
+      "grad_norm": 2.101802032003909,
+      "learning_rate": 8.639715309756722e-05,
+      "loss": 1.2829,
+      "step": 12094
+    },
+    {
+      "epoch": 0.5571935320403556,
+      "grad_norm": 1.5140726646930325,
+      "learning_rate": 8.638237105324424e-05,
+      "loss": 1.5838,
+      "step": 12095
+    },
+    {
+      "epoch": 0.5572396001289907,
+      "grad_norm": 1.7609441212004517,
+      "learning_rate": 8.63675893120944e-05,
+      "loss": 1.3971,
+      "step": 12096
+    },
+    {
+      "epoch": 0.5572856682176256,
+      "grad_norm": 1.2405215820033082,
+      "learning_rate": 8.635280787444681e-05,
+      "loss": 1.5553,
+      "step": 12097
+    },
+    {
+      "epoch": 0.5573317363062606,
+      "grad_norm": 0.9985894155792882,
+      "learning_rate": 8.633802674063056e-05,
+      "loss": 1.4875,
+      "step": 12098
+    },
+    {
+      "epoch": 0.5573778043948957,
+      "grad_norm": 1.8285825393673765,
+      "learning_rate": 8.632324591097471e-05,
+      "loss": 1.5368,
+      "step": 12099
+    },
+    {
+      "epoch": 0.5574238724835306,
+      "grad_norm": 1.3722846037297005,
+      "learning_rate": 8.630846538580836e-05,
+      "loss": 1.32,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5574699405721657,
+      "grad_norm": 0.8346466864625266,
+      "learning_rate": 8.629368516546052e-05,
+      "loss": 1.557,
+      "step": 12101
+    },
+    {
+      "epoch": 0.5575160086608006,
+      "grad_norm": 1.0124978225129078,
+      "learning_rate": 8.627890525026031e-05,
+      "loss": 1.3404,
+      "step": 12102
+    },
+    {
+      "epoch": 0.5575620767494357,
+      "grad_norm": 1.4357557060033264,
+      "learning_rate": 8.626412564053672e-05,
+      "loss": 1.5092,
+      "step": 12103
+    },
+    {
+      "epoch": 0.5576081448380706,
+      "grad_norm": 1.1636612995246407,
+      "learning_rate": 8.624934633661884e-05,
+      "loss": 1.3609,
+      "step": 12104
+    },
+    {
+      "epoch": 0.5576542129267057,
+      "grad_norm": 1.480238116483167,
+      "learning_rate": 8.623456733883568e-05,
+      "loss": 1.4587,
+      "step": 12105
+    },
+    {
+      "epoch": 0.5577002810153406,
+      "grad_norm": 1.1141260205687855,
+      "learning_rate": 8.621978864751628e-05,
+      "loss": 1.4254,
+      "step": 12106
+    },
+    {
+      "epoch": 0.5577463491039757,
+      "grad_norm": 1.5213236772122238,
+      "learning_rate": 8.620501026298964e-05,
+      "loss": 1.4722,
+      "step": 12107
+    },
+    {
+      "epoch": 0.5577924171926106,
+      "grad_norm": 1.6419225315134562,
+      "learning_rate": 8.61902321855848e-05,
+      "loss": 1.2443,
+      "step": 12108
+    },
+    {
+      "epoch": 0.5578384852812457,
+      "grad_norm": 1.3301461917904844,
+      "learning_rate": 8.617545441563078e-05,
+      "loss": 1.2956,
+      "step": 12109
+    },
+    {
+      "epoch": 0.5578845533698806,
+      "grad_norm": 1.3454439044382898,
+      "learning_rate": 8.616067695345654e-05,
+      "loss": 1.3606,
+      "step": 12110
+    },
+    {
+      "epoch": 0.5579306214585157,
+      "grad_norm": 1.6374174208995877,
+      "learning_rate": 8.614589979939113e-05,
+      "loss": 1.6559,
+      "step": 12111
+    },
+    {
+      "epoch": 0.5579766895471507,
+      "grad_norm": 1.0368540317750699,
+      "learning_rate": 8.613112295376347e-05,
+      "loss": 1.2542,
+      "step": 12112
+    },
+    {
+      "epoch": 0.5580227576357857,
+      "grad_norm": 2.515562433181572,
+      "learning_rate": 8.611634641690261e-05,
+      "loss": 1.1794,
+      "step": 12113
+    },
+    {
+      "epoch": 0.5580688257244207,
+      "grad_norm": 1.6247723916661583,
+      "learning_rate": 8.610157018913745e-05,
+      "loss": 1.0842,
+      "step": 12114
+    },
+    {
+      "epoch": 0.5581148938130557,
+      "grad_norm": 1.3794907074835538,
+      "learning_rate": 8.608679427079704e-05,
+      "loss": 1.3438,
+      "step": 12115
+    },
+    {
+      "epoch": 0.5581609619016907,
+      "grad_norm": 2.6220637593695573,
+      "learning_rate": 8.607201866221026e-05,
+      "loss": 1.4485,
+      "step": 12116
+    },
+    {
+      "epoch": 0.5582070299903257,
+      "grad_norm": 2.132822813120038,
+      "learning_rate": 8.605724336370615e-05,
+      "loss": 1.516,
+      "step": 12117
+    },
+    {
+      "epoch": 0.5582530980789607,
+      "grad_norm": 1.1301829907039593,
+      "learning_rate": 8.604246837561359e-05,
+      "loss": 1.228,
+      "step": 12118
+    },
+    {
+      "epoch": 0.5582991661675957,
+      "grad_norm": 2.6840612417499417,
+      "learning_rate": 8.602769369826152e-05,
+      "loss": 1.1756,
+      "step": 12119
+    },
+    {
+      "epoch": 0.5583452342562307,
+      "grad_norm": 1.8980889194081074,
+      "learning_rate": 8.601291933197893e-05,
+      "loss": 1.377,
+      "step": 12120
+    },
+    {
+      "epoch": 0.5583913023448657,
+      "grad_norm": 1.028577823781846,
+      "learning_rate": 8.599814527709466e-05,
+      "loss": 1.4295,
+      "step": 12121
+    },
+    {
+      "epoch": 0.5584373704335007,
+      "grad_norm": 1.3558303954927704,
+      "learning_rate": 8.598337153393772e-05,
+      "loss": 1.2278,
+      "step": 12122
+    },
+    {
+      "epoch": 0.5584834385221357,
+      "grad_norm": 1.0165351342028153,
+      "learning_rate": 8.596859810283695e-05,
+      "loss": 1.526,
+      "step": 12123
+    },
+    {
+      "epoch": 0.5585295066107707,
+      "grad_norm": 1.1284026261263247,
+      "learning_rate": 8.595382498412131e-05,
+      "loss": 1.1921,
+      "step": 12124
+    },
+    {
+      "epoch": 0.5585755746994058,
+      "grad_norm": 0.9147741604755935,
+      "learning_rate": 8.593905217811965e-05,
+      "loss": 1.3035,
+      "step": 12125
+    },
+    {
+      "epoch": 0.5586216427880407,
+      "grad_norm": 1.2866400362047086,
+      "learning_rate": 8.592427968516091e-05,
+      "loss": 1.5226,
+      "step": 12126
+    },
+    {
+      "epoch": 0.5586677108766758,
+      "grad_norm": 1.9953537135317654,
+      "learning_rate": 8.590950750557394e-05,
+      "loss": 1.1114,
+      "step": 12127
+    },
+    {
+      "epoch": 0.5587137789653107,
+      "grad_norm": 1.9442118320451784,
+      "learning_rate": 8.58947356396876e-05,
+      "loss": 1.1291,
+      "step": 12128
+    },
+    {
+      "epoch": 0.5587598470539458,
+      "grad_norm": 0.8501981576994078,
+      "learning_rate": 8.587996408783082e-05,
+      "loss": 1.2685,
+      "step": 12129
+    },
+    {
+      "epoch": 0.5588059151425807,
+      "grad_norm": 1.0889033392108816,
+      "learning_rate": 8.58651928503324e-05,
+      "loss": 1.2399,
+      "step": 12130
+    },
+    {
+      "epoch": 0.5588519832312158,
+      "grad_norm": 2.134086804907824,
+      "learning_rate": 8.585042192752124e-05,
+      "loss": 1.5212,
+      "step": 12131
+    },
+    {
+      "epoch": 0.5588980513198507,
+      "grad_norm": 0.9204689186737233,
+      "learning_rate": 8.583565131972616e-05,
+      "loss": 1.5839,
+      "step": 12132
+    },
+    {
+      "epoch": 0.5589441194084858,
+      "grad_norm": 1.5483516964725592,
+      "learning_rate": 8.582088102727605e-05,
+      "loss": 1.2713,
+      "step": 12133
+    },
+    {
+      "epoch": 0.5589901874971207,
+      "grad_norm": 1.269956340828244,
+      "learning_rate": 8.580611105049967e-05,
+      "loss": 1.4922,
+      "step": 12134
+    },
+    {
+      "epoch": 0.5590362555857558,
+      "grad_norm": 1.353241189056946,
+      "learning_rate": 8.579134138972592e-05,
+      "loss": 1.2212,
+      "step": 12135
+    },
+    {
+      "epoch": 0.5590823236743907,
+      "grad_norm": 1.1302719280573588,
+      "learning_rate": 8.577657204528359e-05,
+      "loss": 1.2455,
+      "step": 12136
+    },
+    {
+      "epoch": 0.5591283917630258,
+      "grad_norm": 0.8782280240853845,
+      "learning_rate": 8.576180301750147e-05,
+      "loss": 1.4974,
+      "step": 12137
+    },
+    {
+      "epoch": 0.5591744598516608,
+      "grad_norm": 1.3590500457000956,
+      "learning_rate": 8.574703430670843e-05,
+      "loss": 1.1453,
+      "step": 12138
+    },
+    {
+      "epoch": 0.5592205279402958,
+      "grad_norm": 0.8801464794004048,
+      "learning_rate": 8.57322659132332e-05,
+      "loss": 1.4625,
+      "step": 12139
+    },
+    {
+      "epoch": 0.5592665960289308,
+      "grad_norm": 1.2028790204393425,
+      "learning_rate": 8.571749783740463e-05,
+      "loss": 1.5096,
+      "step": 12140
+    },
+    {
+      "epoch": 0.5593126641175657,
+      "grad_norm": 2.511900418084835,
+      "learning_rate": 8.570273007955145e-05,
+      "loss": 1.2743,
+      "step": 12141
+    },
+    {
+      "epoch": 0.5593587322062008,
+      "grad_norm": 2.054489663887731,
+      "learning_rate": 8.568796264000251e-05,
+      "loss": 1.4643,
+      "step": 12142
+    },
+    {
+      "epoch": 0.5594048002948357,
+      "grad_norm": 1.8323257800186508,
+      "learning_rate": 8.567319551908651e-05,
+      "loss": 1.0942,
+      "step": 12143
+    },
+    {
+      "epoch": 0.5594508683834708,
+      "grad_norm": 1.369893305274981,
+      "learning_rate": 8.565842871713227e-05,
+      "loss": 1.3804,
+      "step": 12144
+    },
+    {
+      "epoch": 0.5594969364721057,
+      "grad_norm": 1.3788702826466648,
+      "learning_rate": 8.564366223446854e-05,
+      "loss": 1.2497,
+      "step": 12145
+    },
+    {
+      "epoch": 0.5595430045607408,
+      "grad_norm": 0.9341146955934831,
+      "learning_rate": 8.5628896071424e-05,
+      "loss": 1.4817,
+      "step": 12146
+    },
+    {
+      "epoch": 0.5595890726493757,
+      "grad_norm": 1.1845295026259681,
+      "learning_rate": 8.561413022832751e-05,
+      "loss": 1.4463,
+      "step": 12147
+    },
+    {
+      "epoch": 0.5596351407380108,
+      "grad_norm": 2.9463565692280937,
+      "learning_rate": 8.559936470550766e-05,
+      "loss": 1.4143,
+      "step": 12148
+    },
+    {
+      "epoch": 0.5596812088266457,
+      "grad_norm": 0.8608313107514977,
+      "learning_rate": 8.558459950329333e-05,
+      "loss": 1.5044,
+      "step": 12149
+    },
+    {
+      "epoch": 0.5597272769152808,
+      "grad_norm": 1.3227629202578814,
+      "learning_rate": 8.556983462201315e-05,
+      "loss": 1.3855,
+      "step": 12150
+    },
+    {
+      "epoch": 0.5597733450039157,
+      "grad_norm": 1.7130796917166253,
+      "learning_rate": 8.555507006199587e-05,
+      "loss": 1.5965,
+      "step": 12151
+    },
+    {
+      "epoch": 0.5598194130925508,
+      "grad_norm": 2.630352249179765,
+      "learning_rate": 8.554030582357017e-05,
+      "loss": 1.469,
+      "step": 12152
+    },
+    {
+      "epoch": 0.5598654811811858,
+      "grad_norm": 1.052187950859482,
+      "learning_rate": 8.552554190706479e-05,
+      "loss": 1.6279,
+      "step": 12153
+    },
+    {
+      "epoch": 0.5599115492698208,
+      "grad_norm": 1.0705844472447164,
+      "learning_rate": 8.551077831280838e-05,
+      "loss": 1.341,
+      "step": 12154
+    },
+    {
+      "epoch": 0.5599576173584558,
+      "grad_norm": 0.7426678918042804,
+      "learning_rate": 8.549601504112967e-05,
+      "loss": 1.5121,
+      "step": 12155
+    },
+    {
+      "epoch": 0.5600036854470908,
+      "grad_norm": 2.2391630379343463,
+      "learning_rate": 8.548125209235734e-05,
+      "loss": 1.2178,
+      "step": 12156
+    },
+    {
+      "epoch": 0.5600497535357258,
+      "grad_norm": 1.2113926111367472,
+      "learning_rate": 8.546648946681996e-05,
+      "loss": 1.2551,
+      "step": 12157
+    },
+    {
+      "epoch": 0.5600958216243608,
+      "grad_norm": 0.9071660726006798,
+      "learning_rate": 8.545172716484637e-05,
+      "loss": 1.3395,
+      "step": 12158
+    },
+    {
+      "epoch": 0.5601418897129958,
+      "grad_norm": 1.7137656587526027,
+      "learning_rate": 8.543696518676503e-05,
+      "loss": 1.2534,
+      "step": 12159
+    },
+    {
+      "epoch": 0.5601879578016308,
+      "grad_norm": 0.9354648879245884,
+      "learning_rate": 8.542220353290477e-05,
+      "loss": 1.4923,
+      "step": 12160
+    },
+    {
+      "epoch": 0.5602340258902658,
+      "grad_norm": 0.9123232439991846,
+      "learning_rate": 8.540744220359409e-05,
+      "loss": 1.3017,
+      "step": 12161
+    },
+    {
+      "epoch": 0.5602800939789008,
+      "grad_norm": 8.160125861390584,
+      "learning_rate": 8.539268119916173e-05,
+      "loss": 1.2664,
+      "step": 12162
+    },
+    {
+      "epoch": 0.5603261620675358,
+      "grad_norm": 1.066604741216553,
+      "learning_rate": 8.537792051993626e-05,
+      "loss": 1.5306,
+      "step": 12163
+    },
+    {
+      "epoch": 0.5603722301561708,
+      "grad_norm": 2.172550096599808,
+      "learning_rate": 8.536316016624632e-05,
+      "loss": 1.3298,
+      "step": 12164
+    },
+    {
+      "epoch": 0.5604182982448058,
+      "grad_norm": 1.5700208231232715,
+      "learning_rate": 8.534840013842056e-05,
+      "loss": 0.9519,
+      "step": 12165
+    },
+    {
+      "epoch": 0.5604643663334409,
+      "grad_norm": 2.1480138879868074,
+      "learning_rate": 8.533364043678747e-05,
+      "loss": 1.1451,
+      "step": 12166
+    },
+    {
+      "epoch": 0.5605104344220758,
+      "grad_norm": 1.1816455415132654,
+      "learning_rate": 8.53188810616758e-05,
+      "loss": 1.3084,
+      "step": 12167
+    },
+    {
+      "epoch": 0.5605565025107109,
+      "grad_norm": 1.5268618621638557,
+      "learning_rate": 8.530412201341402e-05,
+      "loss": 1.0926,
+      "step": 12168
+    },
+    {
+      "epoch": 0.5606025705993458,
+      "grad_norm": 1.1946535562756067,
+      "learning_rate": 8.528936329233082e-05,
+      "loss": 1.1117,
+      "step": 12169
+    },
+    {
+      "epoch": 0.5606486386879809,
+      "grad_norm": 1.4154105913034776,
+      "learning_rate": 8.527460489875465e-05,
+      "loss": 1.3339,
+      "step": 12170
+    },
+    {
+      "epoch": 0.5606947067766158,
+      "grad_norm": 3.9447770819849577,
+      "learning_rate": 8.525984683301423e-05,
+      "loss": 1.3395,
+      "step": 12171
+    },
+    {
+      "epoch": 0.5607407748652509,
+      "grad_norm": 1.7262953512520176,
+      "learning_rate": 8.524508909543796e-05,
+      "loss": 1.4665,
+      "step": 12172
+    },
+    {
+      "epoch": 0.5607868429538858,
+      "grad_norm": 1.1759390421398959,
+      "learning_rate": 8.523033168635457e-05,
+      "loss": 1.2005,
+      "step": 12173
+    },
+    {
+      "epoch": 0.5608329110425209,
+      "grad_norm": 2.8755395226275122,
+      "learning_rate": 8.52155746060925e-05,
+      "loss": 1.3639,
+      "step": 12174
+    },
+    {
+      "epoch": 0.5608789791311558,
+      "grad_norm": 1.310536601095752,
+      "learning_rate": 8.520081785498026e-05,
+      "loss": 1.3018,
+      "step": 12175
+    },
+    {
+      "epoch": 0.5609250472197909,
+      "grad_norm": 1.1950967313947773,
+      "learning_rate": 8.51860614333465e-05,
+      "loss": 1.3258,
+      "step": 12176
+    },
+    {
+      "epoch": 0.5609711153084258,
+      "grad_norm": 1.3172944424761162,
+      "learning_rate": 8.517130534151961e-05,
+      "loss": 1.2146,
+      "step": 12177
+    },
+    {
+      "epoch": 0.5610171833970609,
+      "grad_norm": 0.9281794139197652,
+      "learning_rate": 8.515654957982826e-05,
+      "loss": 1.2831,
+      "step": 12178
+    },
+    {
+      "epoch": 0.5610632514856959,
+      "grad_norm": 1.0891610137110743,
+      "learning_rate": 8.51417941486008e-05,
+      "loss": 1.2194,
+      "step": 12179
+    },
+    {
+      "epoch": 0.5611093195743309,
+      "grad_norm": 1.3059904430016933,
+      "learning_rate": 8.51270390481659e-05,
+      "loss": 1.1676,
+      "step": 12180
+    },
+    {
+      "epoch": 0.5611553876629659,
+      "grad_norm": 1.0182017011619682,
+      "learning_rate": 8.511228427885189e-05,
+      "loss": 0.9662,
+      "step": 12181
+    },
+    {
+      "epoch": 0.5612014557516009,
+      "grad_norm": 1.8785948688869643,
+      "learning_rate": 8.509752984098744e-05,
+      "loss": 1.3137,
+      "step": 12182
+    },
+    {
+      "epoch": 0.5612475238402359,
+      "grad_norm": 1.44214934499016,
+      "learning_rate": 8.508277573490089e-05,
+      "loss": 1.3414,
+      "step": 12183
+    },
+    {
+      "epoch": 0.5612935919288708,
+      "grad_norm": 1.0892557250377577,
+      "learning_rate": 8.506802196092075e-05,
+      "loss": 1.1797,
+      "step": 12184
+    },
+    {
+      "epoch": 0.5613396600175059,
+      "grad_norm": 1.379040394563239,
+      "learning_rate": 8.50532685193755e-05,
+      "loss": 1.3959,
+      "step": 12185
+    },
+    {
+      "epoch": 0.5613857281061408,
+      "grad_norm": 2.8864585235764997,
+      "learning_rate": 8.50385154105936e-05,
+      "loss": 1.4503,
+      "step": 12186
+    },
+    {
+      "epoch": 0.5614317961947759,
+      "grad_norm": 1.2934753915970287,
+      "learning_rate": 8.50237626349035e-05,
+      "loss": 1.3724,
+      "step": 12187
+    },
+    {
+      "epoch": 0.5614778642834108,
+      "grad_norm": 5.456077075362509,
+      "learning_rate": 8.500901019263361e-05,
+      "loss": 1.2347,
+      "step": 12188
+    },
+    {
+      "epoch": 0.5615239323720459,
+      "grad_norm": 1.056526557105973,
+      "learning_rate": 8.499425808411249e-05,
+      "loss": 1.264,
+      "step": 12189
+    },
+    {
+      "epoch": 0.5615700004606808,
+      "grad_norm": 1.641496909246085,
+      "learning_rate": 8.49795063096684e-05,
+      "loss": 1.4658,
+      "step": 12190
+    },
+    {
+      "epoch": 0.5616160685493159,
+      "grad_norm": 1.3877462509538596,
+      "learning_rate": 8.496475486962992e-05,
+      "loss": 1.3482,
+      "step": 12191
+    },
+    {
+      "epoch": 0.5616621366379508,
+      "grad_norm": 1.025129450381095,
+      "learning_rate": 8.495000376432536e-05,
+      "loss": 1.4075,
+      "step": 12192
+    },
+    {
+      "epoch": 0.5617082047265859,
+      "grad_norm": 2.525980651314234,
+      "learning_rate": 8.493525299408315e-05,
+      "loss": 1.3051,
+      "step": 12193
+    },
+    {
+      "epoch": 0.561754272815221,
+      "grad_norm": 2.066414992174816,
+      "learning_rate": 8.492050255923172e-05,
+      "loss": 1.6989,
+      "step": 12194
+    },
+    {
+      "epoch": 0.5618003409038559,
+      "grad_norm": 14.410671376063556,
+      "learning_rate": 8.490575246009942e-05,
+      "loss": 1.3539,
+      "step": 12195
+    },
+    {
+      "epoch": 0.5618464089924909,
+      "grad_norm": 1.3947126723972372,
+      "learning_rate": 8.489100269701469e-05,
+      "loss": 1.6161,
+      "step": 12196
+    },
+    {
+      "epoch": 0.5618924770811259,
+      "grad_norm": 3.5112280486295324,
+      "learning_rate": 8.487625327030586e-05,
+      "loss": 1.3872,
+      "step": 12197
+    },
+    {
+      "epoch": 0.5619385451697609,
+      "grad_norm": 1.7297392409430143,
+      "learning_rate": 8.486150418030132e-05,
+      "loss": 1.3245,
+      "step": 12198
+    },
+    {
+      "epoch": 0.5619846132583959,
+      "grad_norm": 2.622233157353531,
+      "learning_rate": 8.484675542732942e-05,
+      "loss": 1.3715,
+      "step": 12199
+    },
+    {
+      "epoch": 0.5620306813470309,
+      "grad_norm": 1.709426930727549,
+      "learning_rate": 8.483200701171856e-05,
+      "loss": 1.329,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5620767494356659,
+      "grad_norm": 2.338930294588052,
+      "learning_rate": 8.481725893379705e-05,
+      "loss": 1.1104,
+      "step": 12201
+    },
+    {
+      "epoch": 0.5621228175243009,
+      "grad_norm": 1.9725451102065497,
+      "learning_rate": 8.480251119389322e-05,
+      "loss": 1.5659,
+      "step": 12202
+    },
+    {
+      "epoch": 0.5621688856129359,
+      "grad_norm": 2.2422957168073094,
+      "learning_rate": 8.478776379233542e-05,
+      "loss": 1.6474,
+      "step": 12203
+    },
+    {
+      "epoch": 0.5622149537015709,
+      "grad_norm": 1.8117985887861985,
+      "learning_rate": 8.477301672945198e-05,
+      "loss": 1.1739,
+      "step": 12204
+    },
+    {
+      "epoch": 0.5622610217902059,
+      "grad_norm": 2.5733531627079995,
+      "learning_rate": 8.475827000557122e-05,
+      "loss": 1.296,
+      "step": 12205
+    },
+    {
+      "epoch": 0.5623070898788409,
+      "grad_norm": 2.3205274691058544,
+      "learning_rate": 8.474352362102143e-05,
+      "loss": 1.3575,
+      "step": 12206
+    },
+    {
+      "epoch": 0.562353157967476,
+      "grad_norm": 1.0533091667409211,
+      "learning_rate": 8.472877757613096e-05,
+      "loss": 1.1064,
+      "step": 12207
+    },
+    {
+      "epoch": 0.5623992260561109,
+      "grad_norm": 1.1068062182397482,
+      "learning_rate": 8.471403187122804e-05,
+      "loss": 1.5486,
+      "step": 12208
+    },
+    {
+      "epoch": 0.562445294144746,
+      "grad_norm": 1.4111100432470942,
+      "learning_rate": 8.469928650664101e-05,
+      "loss": 1.3637,
+      "step": 12209
+    },
+    {
+      "epoch": 0.5624913622333809,
+      "grad_norm": 1.2648834985335988,
+      "learning_rate": 8.468454148269813e-05,
+      "loss": 0.9441,
+      "step": 12210
+    },
+    {
+      "epoch": 0.562537430322016,
+      "grad_norm": 1.286060334758966,
+      "learning_rate": 8.466979679972765e-05,
+      "loss": 1.3171,
+      "step": 12211
+    },
+    {
+      "epoch": 0.5625834984106509,
+      "grad_norm": 0.9646380870441059,
+      "learning_rate": 8.46550524580579e-05,
+      "loss": 1.0572,
+      "step": 12212
+    },
+    {
+      "epoch": 0.562629566499286,
+      "grad_norm": 2.7531922451687327,
+      "learning_rate": 8.464030845801705e-05,
+      "loss": 1.2624,
+      "step": 12213
+    },
+    {
+      "epoch": 0.5626756345879209,
+      "grad_norm": 1.1877758770395495,
+      "learning_rate": 8.462556479993342e-05,
+      "loss": 1.5707,
+      "step": 12214
+    },
+    {
+      "epoch": 0.562721702676556,
+      "grad_norm": 2.942140667808909,
+      "learning_rate": 8.461082148413521e-05,
+      "loss": 1.2013,
+      "step": 12215
+    },
+    {
+      "epoch": 0.5627677707651909,
+      "grad_norm": 6.9170873937818,
+      "learning_rate": 8.45960785109507e-05,
+      "loss": 1.3604,
+      "step": 12216
+    },
+    {
+      "epoch": 0.562813838853826,
+      "grad_norm": 1.2491403436604795,
+      "learning_rate": 8.458133588070806e-05,
+      "loss": 1.3173,
+      "step": 12217
+    },
+    {
+      "epoch": 0.5628599069424609,
+      "grad_norm": 3.0508209319401174,
+      "learning_rate": 8.456659359373558e-05,
+      "loss": 1.2317,
+      "step": 12218
+    },
+    {
+      "epoch": 0.562905975031096,
+      "grad_norm": 1.4185187751127886,
+      "learning_rate": 8.455185165036141e-05,
+      "loss": 1.1439,
+      "step": 12219
+    },
+    {
+      "epoch": 0.562952043119731,
+      "grad_norm": 1.7749491560279764,
+      "learning_rate": 8.453711005091377e-05,
+      "loss": 1.1688,
+      "step": 12220
+    },
+    {
+      "epoch": 0.562998111208366,
+      "grad_norm": 1.6584920421673583,
+      "learning_rate": 8.452236879572088e-05,
+      "loss": 1.3982,
+      "step": 12221
+    },
+    {
+      "epoch": 0.563044179297001,
+      "grad_norm": 1.8369285153963777,
+      "learning_rate": 8.450762788511087e-05,
+      "loss": 1.2617,
+      "step": 12222
+    },
+    {
+      "epoch": 0.563090247385636,
+      "grad_norm": 1.9602293058523934,
+      "learning_rate": 8.4492887319412e-05,
+      "loss": 1.4215,
+      "step": 12223
+    },
+    {
+      "epoch": 0.563136315474271,
+      "grad_norm": 1.2592821675728711,
+      "learning_rate": 8.447814709895239e-05,
+      "loss": 1.521,
+      "step": 12224
+    },
+    {
+      "epoch": 0.563182383562906,
+      "grad_norm": 0.8344873883808684,
+      "learning_rate": 8.446340722406023e-05,
+      "loss": 1.3094,
+      "step": 12225
+    },
+    {
+      "epoch": 0.563228451651541,
+      "grad_norm": 2.3811816586298673,
+      "learning_rate": 8.444866769506364e-05,
+      "loss": 1.357,
+      "step": 12226
+    },
+    {
+      "epoch": 0.563274519740176,
+      "grad_norm": 1.2853933209542636,
+      "learning_rate": 8.443392851229083e-05,
+      "loss": 1.5552,
+      "step": 12227
+    },
+    {
+      "epoch": 0.563320587828811,
+      "grad_norm": 1.339742599354049,
+      "learning_rate": 8.441918967606991e-05,
+      "loss": 1.3447,
+      "step": 12228
+    },
+    {
+      "epoch": 0.563366655917446,
+      "grad_norm": 1.4123846792264811,
+      "learning_rate": 8.4404451186729e-05,
+      "loss": 1.1702,
+      "step": 12229
+    },
+    {
+      "epoch": 0.563412724006081,
+      "grad_norm": 1.490842210769235,
+      "learning_rate": 8.438971304459626e-05,
+      "loss": 1.3035,
+      "step": 12230
+    },
+    {
+      "epoch": 0.5634587920947159,
+      "grad_norm": 1.1917231326345876,
+      "learning_rate": 8.437497524999978e-05,
+      "loss": 1.451,
+      "step": 12231
+    },
+    {
+      "epoch": 0.563504860183351,
+      "grad_norm": 1.3880892500852224,
+      "learning_rate": 8.436023780326769e-05,
+      "loss": 1.3497,
+      "step": 12232
+    },
+    {
+      "epoch": 0.5635509282719859,
+      "grad_norm": 1.251587615633366,
+      "learning_rate": 8.434550070472808e-05,
+      "loss": 1.2874,
+      "step": 12233
+    },
+    {
+      "epoch": 0.563596996360621,
+      "grad_norm": 2.895806182796619,
+      "learning_rate": 8.433076395470906e-05,
+      "loss": 1.2077,
+      "step": 12234
+    },
+    {
+      "epoch": 0.563643064449256,
+      "grad_norm": 1.6118145566528241,
+      "learning_rate": 8.431602755353871e-05,
+      "loss": 1.1924,
+      "step": 12235
+    },
+    {
+      "epoch": 0.563689132537891,
+      "grad_norm": 1.1976336694390135,
+      "learning_rate": 8.430129150154513e-05,
+      "loss": 1.3517,
+      "step": 12236
+    },
+    {
+      "epoch": 0.563735200626526,
+      "grad_norm": 0.9348154304835318,
+      "learning_rate": 8.428655579905638e-05,
+      "loss": 1.1287,
+      "step": 12237
+    },
+    {
+      "epoch": 0.563781268715161,
+      "grad_norm": 1.1320120835932301,
+      "learning_rate": 8.42718204464005e-05,
+      "loss": 1.3242,
+      "step": 12238
+    },
+    {
+      "epoch": 0.563827336803796,
+      "grad_norm": 5.675673102773622,
+      "learning_rate": 8.425708544390559e-05,
+      "loss": 1.0897,
+      "step": 12239
+    },
+    {
+      "epoch": 0.563873404892431,
+      "grad_norm": 0.9493884039824539,
+      "learning_rate": 8.424235079189966e-05,
+      "loss": 1.3786,
+      "step": 12240
+    },
+    {
+      "epoch": 0.563919472981066,
+      "grad_norm": 1.2305620610813346,
+      "learning_rate": 8.42276164907108e-05,
+      "loss": 1.1331,
+      "step": 12241
+    },
+    {
+      "epoch": 0.563965541069701,
+      "grad_norm": 1.1814691308578873,
+      "learning_rate": 8.421288254066697e-05,
+      "loss": 1.663,
+      "step": 12242
+    },
+    {
+      "epoch": 0.564011609158336,
+      "grad_norm": 1.0034298730470608,
+      "learning_rate": 8.419814894209628e-05,
+      "loss": 1.2979,
+      "step": 12243
+    },
+    {
+      "epoch": 0.564057677246971,
+      "grad_norm": 0.8458163001065672,
+      "learning_rate": 8.418341569532669e-05,
+      "loss": 1.5078,
+      "step": 12244
+    },
+    {
+      "epoch": 0.564103745335606,
+      "grad_norm": 1.3753749762336462,
+      "learning_rate": 8.416868280068624e-05,
+      "loss": 1.4888,
+      "step": 12245
+    },
+    {
+      "epoch": 0.564149813424241,
+      "grad_norm": 0.883174598611018,
+      "learning_rate": 8.415395025850294e-05,
+      "loss": 1.4534,
+      "step": 12246
+    },
+    {
+      "epoch": 0.564195881512876,
+      "grad_norm": 1.601320058949749,
+      "learning_rate": 8.413921806910471e-05,
+      "loss": 1.2288,
+      "step": 12247
+    },
+    {
+      "epoch": 0.5642419496015111,
+      "grad_norm": 2.5374116758616747,
+      "learning_rate": 8.412448623281966e-05,
+      "loss": 1.282,
+      "step": 12248
+    },
+    {
+      "epoch": 0.564288017690146,
+      "grad_norm": 2.0948008667291167,
+      "learning_rate": 8.410975474997566e-05,
+      "loss": 1.0731,
+      "step": 12249
+    },
+    {
+      "epoch": 0.5643340857787811,
+      "grad_norm": 1.2975539317703388,
+      "learning_rate": 8.409502362090074e-05,
+      "loss": 1.1239,
+      "step": 12250
+    },
+    {
+      "epoch": 0.564380153867416,
+      "grad_norm": 0.9919248763049456,
+      "learning_rate": 8.408029284592284e-05,
+      "loss": 1.3863,
+      "step": 12251
+    },
+    {
+      "epoch": 0.5644262219560511,
+      "grad_norm": 1.346866851139995,
+      "learning_rate": 8.406556242536994e-05,
+      "loss": 1.2103,
+      "step": 12252
+    },
+    {
+      "epoch": 0.564472290044686,
+      "grad_norm": 1.4292574852451978,
+      "learning_rate": 8.405083235956994e-05,
+      "loss": 1.1759,
+      "step": 12253
+    },
+    {
+      "epoch": 0.5645183581333211,
+      "grad_norm": 0.8471454566423774,
+      "learning_rate": 8.403610264885085e-05,
+      "loss": 1.5238,
+      "step": 12254
+    },
+    {
+      "epoch": 0.564564426221956,
+      "grad_norm": 1.046925571670545,
+      "learning_rate": 8.402137329354057e-05,
+      "loss": 1.2773,
+      "step": 12255
+    },
+    {
+      "epoch": 0.5646104943105911,
+      "grad_norm": 4.7267203259969595,
+      "learning_rate": 8.400664429396697e-05,
+      "loss": 1.0633,
+      "step": 12256
+    },
+    {
+      "epoch": 0.564656562399226,
+      "grad_norm": 1.0879990045868124,
+      "learning_rate": 8.399191565045806e-05,
+      "loss": 1.5592,
+      "step": 12257
+    },
+    {
+      "epoch": 0.5647026304878611,
+      "grad_norm": 2.071006996627912,
+      "learning_rate": 8.397718736334162e-05,
+      "loss": 1.3484,
+      "step": 12258
+    },
+    {
+      "epoch": 0.564748698576496,
+      "grad_norm": 1.0421506991153995,
+      "learning_rate": 8.396245943294571e-05,
+      "loss": 1.5341,
+      "step": 12259
+    },
+    {
+      "epoch": 0.5647947666651311,
+      "grad_norm": 1.3393287063535366,
+      "learning_rate": 8.394773185959809e-05,
+      "loss": 1.1515,
+      "step": 12260
+    },
+    {
+      "epoch": 0.5648408347537661,
+      "grad_norm": 3.8903659137538815,
+      "learning_rate": 8.393300464362673e-05,
+      "loss": 1.1939,
+      "step": 12261
+    },
+    {
+      "epoch": 0.5648869028424011,
+      "grad_norm": 1.6186580771784569,
+      "learning_rate": 8.391827778535946e-05,
+      "loss": 1.4603,
+      "step": 12262
+    },
+    {
+      "epoch": 0.5649329709310361,
+      "grad_norm": 1.2933166700439214,
+      "learning_rate": 8.390355128512417e-05,
+      "loss": 1.2931,
+      "step": 12263
+    },
+    {
+      "epoch": 0.5649790390196711,
+      "grad_norm": 0.9193813265518019,
+      "learning_rate": 8.388882514324874e-05,
+      "loss": 1.3613,
+      "step": 12264
+    },
+    {
+      "epoch": 0.5650251071083061,
+      "grad_norm": 1.568408695144395,
+      "learning_rate": 8.387409936006093e-05,
+      "loss": 1.3041,
+      "step": 12265
+    },
+    {
+      "epoch": 0.5650711751969411,
+      "grad_norm": 1.1443118513733126,
+      "learning_rate": 8.38593739358887e-05,
+      "loss": 1.4457,
+      "step": 12266
+    },
+    {
+      "epoch": 0.5651172432855761,
+      "grad_norm": 0.8563154808951656,
+      "learning_rate": 8.384464887105977e-05,
+      "loss": 1.1867,
+      "step": 12267
+    },
+    {
+      "epoch": 0.565163311374211,
+      "grad_norm": 3.6618403884987547,
+      "learning_rate": 8.38299241659021e-05,
+      "loss": 1.2691,
+      "step": 12268
+    },
+    {
+      "epoch": 0.5652093794628461,
+      "grad_norm": 1.998746694557228,
+      "learning_rate": 8.381519982074339e-05,
+      "loss": 1.3829,
+      "step": 12269
+    },
+    {
+      "epoch": 0.565255447551481,
+      "grad_norm": 1.3444624507549674,
+      "learning_rate": 8.380047583591156e-05,
+      "loss": 1.3847,
+      "step": 12270
+    },
+    {
+      "epoch": 0.5653015156401161,
+      "grad_norm": 1.2514679804596498,
+      "learning_rate": 8.37857522117343e-05,
+      "loss": 1.413,
+      "step": 12271
+    },
+    {
+      "epoch": 0.565347583728751,
+      "grad_norm": 2.030876269615959,
+      "learning_rate": 8.377102894853952e-05,
+      "loss": 1.5145,
+      "step": 12272
+    },
+    {
+      "epoch": 0.5653936518173861,
+      "grad_norm": 1.458505157038733,
+      "learning_rate": 8.375630604665494e-05,
+      "loss": 1.0747,
+      "step": 12273
+    },
+    {
+      "epoch": 0.5654397199060212,
+      "grad_norm": 1.3795388268412447,
+      "learning_rate": 8.374158350640831e-05,
+      "loss": 1.3695,
+      "step": 12274
+    },
+    {
+      "epoch": 0.5654857879946561,
+      "grad_norm": 1.8537118667014307,
+      "learning_rate": 8.372686132812752e-05,
+      "loss": 1.5835,
+      "step": 12275
+    },
+    {
+      "epoch": 0.5655318560832912,
+      "grad_norm": 0.9516499587212632,
+      "learning_rate": 8.371213951214018e-05,
+      "loss": 1.4243,
+      "step": 12276
+    },
+    {
+      "epoch": 0.5655779241719261,
+      "grad_norm": 1.5133415051539716,
+      "learning_rate": 8.369741805877418e-05,
+      "loss": 1.2493,
+      "step": 12277
+    },
+    {
+      "epoch": 0.5656239922605611,
+      "grad_norm": 1.8024329077201184,
+      "learning_rate": 8.368269696835718e-05,
+      "loss": 1.4469,
+      "step": 12278
+    },
+    {
+      "epoch": 0.5656700603491961,
+      "grad_norm": 1.5132376262494556,
+      "learning_rate": 8.3667976241217e-05,
+      "loss": 1.5785,
+      "step": 12279
+    },
+    {
+      "epoch": 0.5657161284378311,
+      "grad_norm": 1.2295377078338425,
+      "learning_rate": 8.365325587768127e-05,
+      "loss": 1.6992,
+      "step": 12280
+    },
+    {
+      "epoch": 0.5657621965264661,
+      "grad_norm": 0.9380450051749124,
+      "learning_rate": 8.363853587807783e-05,
+      "loss": 1.3287,
+      "step": 12281
+    },
+    {
+      "epoch": 0.5658082646151011,
+      "grad_norm": 1.3431038004054778,
+      "learning_rate": 8.362381624273426e-05,
+      "loss": 1.2605,
+      "step": 12282
+    },
+    {
+      "epoch": 0.5658543327037361,
+      "grad_norm": 1.4519392982807593,
+      "learning_rate": 8.360909697197841e-05,
+      "loss": 1.3693,
+      "step": 12283
+    },
+    {
+      "epoch": 0.5659004007923711,
+      "grad_norm": 2.815710531225836,
+      "learning_rate": 8.35943780661379e-05,
+      "loss": 1.4035,
+      "step": 12284
+    },
+    {
+      "epoch": 0.5659464688810061,
+      "grad_norm": 1.7377253047582013,
+      "learning_rate": 8.357965952554042e-05,
+      "loss": 1.366,
+      "step": 12285
+    },
+    {
+      "epoch": 0.5659925369696411,
+      "grad_norm": 1.2888889959456538,
+      "learning_rate": 8.356494135051367e-05,
+      "loss": 1.503,
+      "step": 12286
+    },
+    {
+      "epoch": 0.5660386050582761,
+      "grad_norm": 1.0446262882854482,
+      "learning_rate": 8.355022354138528e-05,
+      "loss": 1.3715,
+      "step": 12287
+    },
+    {
+      "epoch": 0.5660846731469111,
+      "grad_norm": 0.7746583846085481,
+      "learning_rate": 8.353550609848303e-05,
+      "loss": 1.2801,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5661307412355462,
+      "grad_norm": 1.1591479893067056,
+      "learning_rate": 8.352078902213443e-05,
+      "loss": 1.4888,
+      "step": 12289
+    },
+    {
+      "epoch": 0.5661768093241811,
+      "grad_norm": 9.501714976178945,
+      "learning_rate": 8.350607231266728e-05,
+      "loss": 1.2418,
+      "step": 12290
+    },
+    {
+      "epoch": 0.5662228774128162,
+      "grad_norm": 1.9550257893895207,
+      "learning_rate": 8.349135597040908e-05,
+      "loss": 1.1195,
+      "step": 12291
+    },
+    {
+      "epoch": 0.5662689455014511,
+      "grad_norm": 2.9205767985775464,
+      "learning_rate": 8.34766399956876e-05,
+      "loss": 1.3483,
+      "step": 12292
+    },
+    {
+      "epoch": 0.5663150135900862,
+      "grad_norm": 1.5134986267628492,
+      "learning_rate": 8.346192438883038e-05,
+      "loss": 1.3951,
+      "step": 12293
+    },
+    {
+      "epoch": 0.5663610816787211,
+      "grad_norm": 1.6438578396065098,
+      "learning_rate": 8.344720915016503e-05,
+      "loss": 1.3232,
+      "step": 12294
+    },
+    {
+      "epoch": 0.5664071497673562,
+      "grad_norm": 1.4881073403732183,
+      "learning_rate": 8.34324942800192e-05,
+      "loss": 1.2571,
+      "step": 12295
+    },
+    {
+      "epoch": 0.5664532178559911,
+      "grad_norm": 1.257674736979128,
+      "learning_rate": 8.341777977872047e-05,
+      "loss": 1.1896,
+      "step": 12296
+    },
+    {
+      "epoch": 0.5664992859446262,
+      "grad_norm": 0.9996429660011614,
+      "learning_rate": 8.340306564659646e-05,
+      "loss": 1.6478,
+      "step": 12297
+    },
+    {
+      "epoch": 0.5665453540332611,
+      "grad_norm": 4.314373891320858,
+      "learning_rate": 8.338835188397471e-05,
+      "loss": 1.1085,
+      "step": 12298
+    },
+    {
+      "epoch": 0.5665914221218962,
+      "grad_norm": 1.0330427651220535,
+      "learning_rate": 8.337363849118285e-05,
+      "loss": 1.0477,
+      "step": 12299
+    },
+    {
+      "epoch": 0.5666374902105311,
+      "grad_norm": 1.4769961668194045,
+      "learning_rate": 8.335892546854838e-05,
+      "loss": 1.1895,
+      "step": 12300
+    },
+    {
+      "epoch": 0.5666835582991662,
+      "grad_norm": 0.8724824361104329,
+      "learning_rate": 8.334421281639895e-05,
+      "loss": 1.3857,
+      "step": 12301
+    },
+    {
+      "epoch": 0.5667296263878012,
+      "grad_norm": 0.9465681516457444,
+      "learning_rate": 8.332950053506205e-05,
+      "loss": 1.3386,
+      "step": 12302
+    },
+    {
+      "epoch": 0.5667756944764362,
+      "grad_norm": 0.9526008756337161,
+      "learning_rate": 8.331478862486522e-05,
+      "loss": 1.4665,
+      "step": 12303
+    },
+    {
+      "epoch": 0.5668217625650712,
+      "grad_norm": 1.3910157719940297,
+      "learning_rate": 8.330007708613604e-05,
+      "loss": 1.5378,
+      "step": 12304
+    },
+    {
+      "epoch": 0.5668678306537062,
+      "grad_norm": 1.3771354064392267,
+      "learning_rate": 8.328536591920198e-05,
+      "loss": 1.3664,
+      "step": 12305
+    },
+    {
+      "epoch": 0.5669138987423412,
+      "grad_norm": 1.234327053110017,
+      "learning_rate": 8.327065512439059e-05,
+      "loss": 1.2516,
+      "step": 12306
+    },
+    {
+      "epoch": 0.5669599668309762,
+      "grad_norm": 0.9794010201729998,
+      "learning_rate": 8.325594470202939e-05,
+      "loss": 1.3383,
+      "step": 12307
+    },
+    {
+      "epoch": 0.5670060349196112,
+      "grad_norm": 1.0386242672623267,
+      "learning_rate": 8.324123465244588e-05,
+      "loss": 1.439,
+      "step": 12308
+    },
+    {
+      "epoch": 0.5670521030082462,
+      "grad_norm": 1.0348079895783346,
+      "learning_rate": 8.322652497596752e-05,
+      "loss": 1.245,
+      "step": 12309
+    },
+    {
+      "epoch": 0.5670981710968812,
+      "grad_norm": 1.4302723430702107,
+      "learning_rate": 8.321181567292184e-05,
+      "loss": 1.6996,
+      "step": 12310
+    },
+    {
+      "epoch": 0.5671442391855162,
+      "grad_norm": 6.734981888496114,
+      "learning_rate": 8.31971067436363e-05,
+      "loss": 1.124,
+      "step": 12311
+    },
+    {
+      "epoch": 0.5671903072741512,
+      "grad_norm": 1.4353587843041629,
+      "learning_rate": 8.318239818843836e-05,
+      "loss": 1.1244,
+      "step": 12312
+    },
+    {
+      "epoch": 0.5672363753627861,
+      "grad_norm": 1.1139493042090114,
+      "learning_rate": 8.316769000765549e-05,
+      "loss": 1.5351,
+      "step": 12313
+    },
+    {
+      "epoch": 0.5672824434514212,
+      "grad_norm": 1.5127393914377396,
+      "learning_rate": 8.315298220161513e-05,
+      "loss": 1.3439,
+      "step": 12314
+    },
+    {
+      "epoch": 0.5673285115400563,
+      "grad_norm": 1.0223605324012603,
+      "learning_rate": 8.313827477064476e-05,
+      "loss": 1.2329,
+      "step": 12315
+    },
+    {
+      "epoch": 0.5673745796286912,
+      "grad_norm": 1.4719170706496483,
+      "learning_rate": 8.312356771507175e-05,
+      "loss": 1.1868,
+      "step": 12316
+    },
+    {
+      "epoch": 0.5674206477173263,
+      "grad_norm": 0.9750735080762596,
+      "learning_rate": 8.310886103522362e-05,
+      "loss": 1.1594,
+      "step": 12317
+    },
+    {
+      "epoch": 0.5674667158059612,
+      "grad_norm": 1.1451888197155433,
+      "learning_rate": 8.30941547314277e-05,
+      "loss": 1.2209,
+      "step": 12318
+    },
+    {
+      "epoch": 0.5675127838945963,
+      "grad_norm": 1.2641617671783234,
+      "learning_rate": 8.307944880401145e-05,
+      "loss": 1.3128,
+      "step": 12319
+    },
+    {
+      "epoch": 0.5675588519832312,
+      "grad_norm": 0.8376632215455716,
+      "learning_rate": 8.306474325330226e-05,
+      "loss": 1.3023,
+      "step": 12320
+    },
+    {
+      "epoch": 0.5676049200718662,
+      "grad_norm": 1.1040508202389376,
+      "learning_rate": 8.305003807962752e-05,
+      "loss": 1.1137,
+      "step": 12321
+    },
+    {
+      "epoch": 0.5676509881605012,
+      "grad_norm": 1.3417994252145982,
+      "learning_rate": 8.303533328331464e-05,
+      "loss": 1.5529,
+      "step": 12322
+    },
+    {
+      "epoch": 0.5676970562491362,
+      "grad_norm": 1.724226731735638,
+      "learning_rate": 8.302062886469093e-05,
+      "loss": 1.4961,
+      "step": 12323
+    },
+    {
+      "epoch": 0.5677431243377712,
+      "grad_norm": 1.6038666576145966,
+      "learning_rate": 8.300592482408385e-05,
+      "loss": 1.2035,
+      "step": 12324
+    },
+    {
+      "epoch": 0.5677891924264062,
+      "grad_norm": 1.341157926119182,
+      "learning_rate": 8.299122116182069e-05,
+      "loss": 1.2702,
+      "step": 12325
+    },
+    {
+      "epoch": 0.5678352605150412,
+      "grad_norm": 0.9944912217376054,
+      "learning_rate": 8.297651787822884e-05,
+      "loss": 1.6168,
+      "step": 12326
+    },
+    {
+      "epoch": 0.5678813286036762,
+      "grad_norm": 6.835371914976813,
+      "learning_rate": 8.296181497363561e-05,
+      "loss": 1.4621,
+      "step": 12327
+    },
+    {
+      "epoch": 0.5679273966923112,
+      "grad_norm": 0.7512226300012753,
+      "learning_rate": 8.294711244836838e-05,
+      "loss": 1.1947,
+      "step": 12328
+    },
+    {
+      "epoch": 0.5679734647809462,
+      "grad_norm": 1.1070933149399491,
+      "learning_rate": 8.293241030275444e-05,
+      "loss": 1.5905,
+      "step": 12329
+    },
+    {
+      "epoch": 0.5680195328695813,
+      "grad_norm": 1.3821429211889051,
+      "learning_rate": 8.291770853712112e-05,
+      "loss": 1.3162,
+      "step": 12330
+    },
+    {
+      "epoch": 0.5680656009582162,
+      "grad_norm": 1.2100486680761013,
+      "learning_rate": 8.290300715179575e-05,
+      "loss": 1.143,
+      "step": 12331
+    },
+    {
+      "epoch": 0.5681116690468513,
+      "grad_norm": 1.843728686638158,
+      "learning_rate": 8.288830614710557e-05,
+      "loss": 1.3283,
+      "step": 12332
+    },
+    {
+      "epoch": 0.5681577371354862,
+      "grad_norm": 1.304382864442343,
+      "learning_rate": 8.287360552337796e-05,
+      "loss": 1.3729,
+      "step": 12333
+    },
+    {
+      "epoch": 0.5682038052241213,
+      "grad_norm": 1.4528691791404647,
+      "learning_rate": 8.285890528094011e-05,
+      "loss": 1.3327,
+      "step": 12334
+    },
+    {
+      "epoch": 0.5682498733127562,
+      "grad_norm": 2.5814691289775005,
+      "learning_rate": 8.284420542011939e-05,
+      "loss": 1.141,
+      "step": 12335
+    },
+    {
+      "epoch": 0.5682959414013913,
+      "grad_norm": 1.5158506729668295,
+      "learning_rate": 8.282950594124296e-05,
+      "loss": 1.1797,
+      "step": 12336
+    },
+    {
+      "epoch": 0.5683420094900262,
+      "grad_norm": 1.4210780072360072,
+      "learning_rate": 8.28148068446382e-05,
+      "loss": 1.2472,
+      "step": 12337
+    },
+    {
+      "epoch": 0.5683880775786613,
+      "grad_norm": 7.250036210158566,
+      "learning_rate": 8.280010813063228e-05,
+      "loss": 1.2231,
+      "step": 12338
+    },
+    {
+      "epoch": 0.5684341456672962,
+      "grad_norm": 1.8890163897475558,
+      "learning_rate": 8.278540979955245e-05,
+      "loss": 1.4899,
+      "step": 12339
+    },
+    {
+      "epoch": 0.5684802137559313,
+      "grad_norm": 1.072492748731796,
+      "learning_rate": 8.277071185172597e-05,
+      "loss": 1.3372,
+      "step": 12340
+    },
+    {
+      "epoch": 0.5685262818445662,
+      "grad_norm": 1.6275650655796579,
+      "learning_rate": 8.275601428748003e-05,
+      "loss": 1.123,
+      "step": 12341
+    },
+    {
+      "epoch": 0.5685723499332013,
+      "grad_norm": 0.9377221066066349,
+      "learning_rate": 8.274131710714188e-05,
+      "loss": 1.6378,
+      "step": 12342
+    },
+    {
+      "epoch": 0.5686184180218363,
+      "grad_norm": 1.1111646465618221,
+      "learning_rate": 8.27266203110387e-05,
+      "loss": 1.403,
+      "step": 12343
+    },
+    {
+      "epoch": 0.5686644861104713,
+      "grad_norm": 1.228908306696692,
+      "learning_rate": 8.271192389949771e-05,
+      "loss": 1.4449,
+      "step": 12344
+    },
+    {
+      "epoch": 0.5687105541991063,
+      "grad_norm": 2.855924906454042,
+      "learning_rate": 8.269722787284608e-05,
+      "loss": 1.2682,
+      "step": 12345
+    },
+    {
+      "epoch": 0.5687566222877413,
+      "grad_norm": 2.3607111965994183,
+      "learning_rate": 8.268253223141102e-05,
+      "loss": 0.979,
+      "step": 12346
+    },
+    {
+      "epoch": 0.5688026903763763,
+      "grad_norm": 1.8140030382513948,
+      "learning_rate": 8.26678369755197e-05,
+      "loss": 1.3093,
+      "step": 12347
+    },
+    {
+      "epoch": 0.5688487584650113,
+      "grad_norm": 1.7634991476686992,
+      "learning_rate": 8.265314210549924e-05,
+      "loss": 1.1181,
+      "step": 12348
+    },
+    {
+      "epoch": 0.5688948265536463,
+      "grad_norm": 1.2240453137871325,
+      "learning_rate": 8.263844762167684e-05,
+      "loss": 1.41,
+      "step": 12349
+    },
+    {
+      "epoch": 0.5689408946422813,
+      "grad_norm": 1.784041202503339,
+      "learning_rate": 8.262375352437962e-05,
+      "loss": 1.2694,
+      "step": 12350
+    },
+    {
+      "epoch": 0.5689869627309163,
+      "grad_norm": 1.5875707184825412,
+      "learning_rate": 8.260905981393476e-05,
+      "loss": 1.444,
+      "step": 12351
+    },
+    {
+      "epoch": 0.5690330308195513,
+      "grad_norm": 1.5288141446751493,
+      "learning_rate": 8.259436649066934e-05,
+      "loss": 1.22,
+      "step": 12352
+    },
+    {
+      "epoch": 0.5690790989081863,
+      "grad_norm": 2.326201421456273,
+      "learning_rate": 8.257967355491053e-05,
+      "loss": 1.1085,
+      "step": 12353
+    },
+    {
+      "epoch": 0.5691251669968213,
+      "grad_norm": 1.6915113423971293,
+      "learning_rate": 8.25649810069854e-05,
+      "loss": 1.384,
+      "step": 12354
+    },
+    {
+      "epoch": 0.5691712350854563,
+      "grad_norm": 2.0445982624819266,
+      "learning_rate": 8.255028884722109e-05,
+      "loss": 1.3285,
+      "step": 12355
+    },
+    {
+      "epoch": 0.5692173031740914,
+      "grad_norm": 1.9260165873874722,
+      "learning_rate": 8.25355970759447e-05,
+      "loss": 1.1372,
+      "step": 12356
+    },
+    {
+      "epoch": 0.5692633712627263,
+      "grad_norm": 1.7934376530647151,
+      "learning_rate": 8.252090569348323e-05,
+      "loss": 1.2971,
+      "step": 12357
+    },
+    {
+      "epoch": 0.5693094393513614,
+      "grad_norm": 1.787426055654165,
+      "learning_rate": 8.250621470016388e-05,
+      "loss": 1.3747,
+      "step": 12358
+    },
+    {
+      "epoch": 0.5693555074399963,
+      "grad_norm": 1.133309156935856,
+      "learning_rate": 8.249152409631362e-05,
+      "loss": 0.8984,
+      "step": 12359
+    },
+    {
+      "epoch": 0.5694015755286314,
+      "grad_norm": 1.294268801950533,
+      "learning_rate": 8.247683388225959e-05,
+      "loss": 1.6159,
+      "step": 12360
+    },
+    {
+      "epoch": 0.5694476436172663,
+      "grad_norm": 1.2695217860580723,
+      "learning_rate": 8.246214405832878e-05,
+      "loss": 1.7042,
+      "step": 12361
+    },
+    {
+      "epoch": 0.5694937117059014,
+      "grad_norm": 1.7973349089871717,
+      "learning_rate": 8.244745462484828e-05,
+      "loss": 1.2819,
+      "step": 12362
+    },
+    {
+      "epoch": 0.5695397797945363,
+      "grad_norm": 1.9847117755295245,
+      "learning_rate": 8.24327655821451e-05,
+      "loss": 1.4322,
+      "step": 12363
+    },
+    {
+      "epoch": 0.5695858478831713,
+      "grad_norm": 1.965071234802995,
+      "learning_rate": 8.241807693054628e-05,
+      "loss": 1.2712,
+      "step": 12364
+    },
+    {
+      "epoch": 0.5696319159718063,
+      "grad_norm": 3.8225068123343147,
+      "learning_rate": 8.240338867037887e-05,
+      "loss": 1.0995,
+      "step": 12365
+    },
+    {
+      "epoch": 0.5696779840604413,
+      "grad_norm": 1.7655405796279005,
+      "learning_rate": 8.238870080196977e-05,
+      "loss": 1.1443,
+      "step": 12366
+    },
+    {
+      "epoch": 0.5697240521490763,
+      "grad_norm": 2.2580735471165934,
+      "learning_rate": 8.237401332564611e-05,
+      "loss": 1.3347,
+      "step": 12367
+    },
+    {
+      "epoch": 0.5697701202377113,
+      "grad_norm": 1.0710152709682783,
+      "learning_rate": 8.235932624173475e-05,
+      "loss": 1.133,
+      "step": 12368
+    },
+    {
+      "epoch": 0.5698161883263463,
+      "grad_norm": 2.025420857374866,
+      "learning_rate": 8.234463955056281e-05,
+      "loss": 1.3957,
+      "step": 12369
+    },
+    {
+      "epoch": 0.5698622564149813,
+      "grad_norm": 2.0324856682667343,
+      "learning_rate": 8.232995325245713e-05,
+      "loss": 1.1274,
+      "step": 12370
+    },
+    {
+      "epoch": 0.5699083245036164,
+      "grad_norm": 1.0461447671241693,
+      "learning_rate": 8.231526734774483e-05,
+      "loss": 1.3952,
+      "step": 12371
+    },
+    {
+      "epoch": 0.5699543925922513,
+      "grad_norm": 1.099314930668819,
+      "learning_rate": 8.23005818367527e-05,
+      "loss": 1.2407,
+      "step": 12372
+    },
+    {
+      "epoch": 0.5700004606808864,
+      "grad_norm": 2.372103333793192,
+      "learning_rate": 8.228589671980781e-05,
+      "loss": 1.234,
+      "step": 12373
+    },
+    {
+      "epoch": 0.5700465287695213,
+      "grad_norm": 1.2138461883464773,
+      "learning_rate": 8.227121199723708e-05,
+      "loss": 1.4769,
+      "step": 12374
+    },
+    {
+      "epoch": 0.5700925968581564,
+      "grad_norm": 3.125067785414022,
+      "learning_rate": 8.225652766936734e-05,
+      "loss": 1.1724,
+      "step": 12375
+    },
+    {
+      "epoch": 0.5701386649467913,
+      "grad_norm": 3.808053947503697,
+      "learning_rate": 8.224184373652567e-05,
+      "loss": 1.1187,
+      "step": 12376
+    },
+    {
+      "epoch": 0.5701847330354264,
+      "grad_norm": 1.1243072759952606,
+      "learning_rate": 8.222716019903883e-05,
+      "loss": 1.6588,
+      "step": 12377
+    },
+    {
+      "epoch": 0.5702308011240613,
+      "grad_norm": 9.642762936122397,
+      "learning_rate": 8.221247705723386e-05,
+      "loss": 1.3121,
+      "step": 12378
+    },
+    {
+      "epoch": 0.5702768692126964,
+      "grad_norm": 2.8094724549146726,
+      "learning_rate": 8.219779431143753e-05,
+      "loss": 1.149,
+      "step": 12379
+    },
+    {
+      "epoch": 0.5703229373013313,
+      "grad_norm": 1.913869742765219,
+      "learning_rate": 8.218311196197684e-05,
+      "loss": 1.3422,
+      "step": 12380
+    },
+    {
+      "epoch": 0.5703690053899664,
+      "grad_norm": 4.982684789656869,
+      "learning_rate": 8.216843000917854e-05,
+      "loss": 1.4189,
+      "step": 12381
+    },
+    {
+      "epoch": 0.5704150734786013,
+      "grad_norm": 1.1810921765460114,
+      "learning_rate": 8.215374845336965e-05,
+      "loss": 1.518,
+      "step": 12382
+    },
+    {
+      "epoch": 0.5704611415672364,
+      "grad_norm": 1.370074173884551,
+      "learning_rate": 8.213906729487693e-05,
+      "loss": 1.4323,
+      "step": 12383
+    },
+    {
+      "epoch": 0.5705072096558714,
+      "grad_norm": 9.488766932958539,
+      "learning_rate": 8.212438653402723e-05,
+      "loss": 1.4126,
+      "step": 12384
+    },
+    {
+      "epoch": 0.5705532777445064,
+      "grad_norm": 1.1009634931169665,
+      "learning_rate": 8.210970617114743e-05,
+      "loss": 1.6414,
+      "step": 12385
+    },
+    {
+      "epoch": 0.5705993458331414,
+      "grad_norm": 1.3553178758284454,
+      "learning_rate": 8.209502620656431e-05,
+      "loss": 1.5751,
+      "step": 12386
+    },
+    {
+      "epoch": 0.5706454139217764,
+      "grad_norm": 1.45922703519302,
+      "learning_rate": 8.208034664060478e-05,
+      "loss": 1.3373,
+      "step": 12387
+    },
+    {
+      "epoch": 0.5706914820104114,
+      "grad_norm": 1.201015013859376,
+      "learning_rate": 8.206566747359555e-05,
+      "loss": 1.4934,
+      "step": 12388
+    },
+    {
+      "epoch": 0.5707375500990464,
+      "grad_norm": 3.4880615192962328,
+      "learning_rate": 8.205098870586356e-05,
+      "loss": 1.3986,
+      "step": 12389
+    },
+    {
+      "epoch": 0.5707836181876814,
+      "grad_norm": 0.8593898415172713,
+      "learning_rate": 8.203631033773546e-05,
+      "loss": 1.4503,
+      "step": 12390
+    },
+    {
+      "epoch": 0.5708296862763164,
+      "grad_norm": 1.6337394997674168,
+      "learning_rate": 8.202163236953817e-05,
+      "loss": 1.2436,
+      "step": 12391
+    },
+    {
+      "epoch": 0.5708757543649514,
+      "grad_norm": 1.1747923435989338,
+      "learning_rate": 8.200695480159838e-05,
+      "loss": 1.4603,
+      "step": 12392
+    },
+    {
+      "epoch": 0.5709218224535864,
+      "grad_norm": 1.4032684368235016,
+      "learning_rate": 8.199227763424288e-05,
+      "loss": 1.2761,
+      "step": 12393
+    },
+    {
+      "epoch": 0.5709678905422214,
+      "grad_norm": 0.9852945434679133,
+      "learning_rate": 8.197760086779846e-05,
+      "loss": 1.3146,
+      "step": 12394
+    },
+    {
+      "epoch": 0.5710139586308564,
+      "grad_norm": 2.4962061619843046,
+      "learning_rate": 8.196292450259183e-05,
+      "loss": 1.533,
+      "step": 12395
+    },
+    {
+      "epoch": 0.5710600267194914,
+      "grad_norm": 3.0569721207532714,
+      "learning_rate": 8.194824853894978e-05,
+      "loss": 1.2498,
+      "step": 12396
+    },
+    {
+      "epoch": 0.5711060948081265,
+      "grad_norm": 1.0886318672714337,
+      "learning_rate": 8.193357297719902e-05,
+      "loss": 1.3661,
+      "step": 12397
+    },
+    {
+      "epoch": 0.5711521628967614,
+      "grad_norm": 13.714153570381235,
+      "learning_rate": 8.191889781766628e-05,
+      "loss": 1.5239,
+      "step": 12398
+    },
+    {
+      "epoch": 0.5711982309853965,
+      "grad_norm": 1.9595443541075652,
+      "learning_rate": 8.190422306067824e-05,
+      "loss": 1.4085,
+      "step": 12399
+    },
+    {
+      "epoch": 0.5712442990740314,
+      "grad_norm": 1.9887761342994406,
+      "learning_rate": 8.188954870656172e-05,
+      "loss": 1.3851,
+      "step": 12400
+    },
+    {
+      "epoch": 0.5712903671626665,
+      "grad_norm": 1.2302459454414876,
+      "learning_rate": 8.187487475564326e-05,
+      "loss": 1.4033,
+      "step": 12401
+    },
+    {
+      "epoch": 0.5713364352513014,
+      "grad_norm": 1.8234109538173642,
+      "learning_rate": 8.186020120824972e-05,
+      "loss": 1.4679,
+      "step": 12402
+    },
+    {
+      "epoch": 0.5713825033399365,
+      "grad_norm": 2.677837457330607,
+      "learning_rate": 8.184552806470767e-05,
+      "loss": 1.354,
+      "step": 12403
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 2.1710007912865654,
+      "learning_rate": 8.183085532534378e-05,
+      "loss": 1.5312,
+      "step": 12404
+    },
+    {
+      "epoch": 0.5714746395172065,
+      "grad_norm": 1.1235874509392814,
+      "learning_rate": 8.181618299048478e-05,
+      "loss": 1.2629,
+      "step": 12405
+    },
+    {
+      "epoch": 0.5715207076058414,
+      "grad_norm": 0.8783678481577011,
+      "learning_rate": 8.180151106045726e-05,
+      "loss": 1.6504,
+      "step": 12406
+    },
+    {
+      "epoch": 0.5715667756944764,
+      "grad_norm": 1.2106625272815261,
+      "learning_rate": 8.17868395355879e-05,
+      "loss": 1.2122,
+      "step": 12407
+    },
+    {
+      "epoch": 0.5716128437831114,
+      "grad_norm": 1.8203358186564382,
+      "learning_rate": 8.177216841620332e-05,
+      "loss": 1.2061,
+      "step": 12408
+    },
+    {
+      "epoch": 0.5716589118717464,
+      "grad_norm": 1.0521638271135194,
+      "learning_rate": 8.175749770263018e-05,
+      "loss": 1.1649,
+      "step": 12409
+    },
+    {
+      "epoch": 0.5717049799603815,
+      "grad_norm": 0.9210586643138254,
+      "learning_rate": 8.174282739519506e-05,
+      "loss": 1.5605,
+      "step": 12410
+    },
+    {
+      "epoch": 0.5717510480490164,
+      "grad_norm": 0.8087289247851046,
+      "learning_rate": 8.172815749422457e-05,
+      "loss": 1.4101,
+      "step": 12411
+    },
+    {
+      "epoch": 0.5717971161376515,
+      "grad_norm": 2.0969325806521395,
+      "learning_rate": 8.171348800004537e-05,
+      "loss": 1.2204,
+      "step": 12412
+    },
+    {
+      "epoch": 0.5718431842262864,
+      "grad_norm": 0.9851608500054939,
+      "learning_rate": 8.169881891298395e-05,
+      "loss": 1.3555,
+      "step": 12413
+    },
+    {
+      "epoch": 0.5718892523149215,
+      "grad_norm": 1.409060765448586,
+      "learning_rate": 8.168415023336698e-05,
+      "loss": 1.167,
+      "step": 12414
+    },
+    {
+      "epoch": 0.5719353204035564,
+      "grad_norm": 1.1322691013947435,
+      "learning_rate": 8.166948196152098e-05,
+      "loss": 1.3034,
+      "step": 12415
+    },
+    {
+      "epoch": 0.5719813884921915,
+      "grad_norm": 1.448032062139682,
+      "learning_rate": 8.165481409777255e-05,
+      "loss": 1.4914,
+      "step": 12416
+    },
+    {
+      "epoch": 0.5720274565808264,
+      "grad_norm": 0.7604838304376694,
+      "learning_rate": 8.164014664244822e-05,
+      "loss": 1.4034,
+      "step": 12417
+    },
+    {
+      "epoch": 0.5720735246694615,
+      "grad_norm": 3.354883682410306,
+      "learning_rate": 8.162547959587455e-05,
+      "loss": 1.3083,
+      "step": 12418
+    },
+    {
+      "epoch": 0.5721195927580964,
+      "grad_norm": 1.2897678549243625,
+      "learning_rate": 8.161081295837806e-05,
+      "loss": 1.502,
+      "step": 12419
+    },
+    {
+      "epoch": 0.5721656608467315,
+      "grad_norm": 1.1591534932576943,
+      "learning_rate": 8.159614673028532e-05,
+      "loss": 1.355,
+      "step": 12420
+    },
+    {
+      "epoch": 0.5722117289353664,
+      "grad_norm": 1.3010016386086285,
+      "learning_rate": 8.15814809119228e-05,
+      "loss": 1.2441,
+      "step": 12421
+    },
+    {
+      "epoch": 0.5722577970240015,
+      "grad_norm": 1.6500219590974277,
+      "learning_rate": 8.156681550361701e-05,
+      "loss": 1.5617,
+      "step": 12422
+    },
+    {
+      "epoch": 0.5723038651126364,
+      "grad_norm": 0.7123028994432667,
+      "learning_rate": 8.155215050569449e-05,
+      "loss": 1.4302,
+      "step": 12423
+    },
+    {
+      "epoch": 0.5723499332012715,
+      "grad_norm": 2.787560793124655,
+      "learning_rate": 8.153748591848168e-05,
+      "loss": 1.4644,
+      "step": 12424
+    },
+    {
+      "epoch": 0.5723960012899065,
+      "grad_norm": 3.6904147585955545,
+      "learning_rate": 8.152282174230513e-05,
+      "loss": 1.0559,
+      "step": 12425
+    },
+    {
+      "epoch": 0.5724420693785415,
+      "grad_norm": 0.9845581882095855,
+      "learning_rate": 8.150815797749123e-05,
+      "loss": 1.3016,
+      "step": 12426
+    },
+    {
+      "epoch": 0.5724881374671765,
+      "grad_norm": 1.011018772529775,
+      "learning_rate": 8.14934946243665e-05,
+      "loss": 1.6379,
+      "step": 12427
+    },
+    {
+      "epoch": 0.5725342055558115,
+      "grad_norm": 1.9834957453694442,
+      "learning_rate": 8.147883168325739e-05,
+      "loss": 1.3356,
+      "step": 12428
+    },
+    {
+      "epoch": 0.5725802736444465,
+      "grad_norm": 0.8213208657706754,
+      "learning_rate": 8.146416915449033e-05,
+      "loss": 1.1666,
+      "step": 12429
+    },
+    {
+      "epoch": 0.5726263417330815,
+      "grad_norm": 1.4447350923257687,
+      "learning_rate": 8.144950703839177e-05,
+      "loss": 1.2068,
+      "step": 12430
+    },
+    {
+      "epoch": 0.5726724098217165,
+      "grad_norm": 1.0133877288974082,
+      "learning_rate": 8.143484533528812e-05,
+      "loss": 1.505,
+      "step": 12431
+    },
+    {
+      "epoch": 0.5727184779103515,
+      "grad_norm": 0.9188069013539675,
+      "learning_rate": 8.142018404550581e-05,
+      "loss": 1.3171,
+      "step": 12432
+    },
+    {
+      "epoch": 0.5727645459989865,
+      "grad_norm": 0.8796234995126593,
+      "learning_rate": 8.140552316937123e-05,
+      "loss": 1.4059,
+      "step": 12433
+    },
+    {
+      "epoch": 0.5728106140876215,
+      "grad_norm": 0.9208913854886785,
+      "learning_rate": 8.139086270721081e-05,
+      "loss": 1.594,
+      "step": 12434
+    },
+    {
+      "epoch": 0.5728566821762565,
+      "grad_norm": 1.2650259299015372,
+      "learning_rate": 8.13762026593509e-05,
+      "loss": 1.1698,
+      "step": 12435
+    },
+    {
+      "epoch": 0.5729027502648915,
+      "grad_norm": 1.4418753162534186,
+      "learning_rate": 8.136154302611795e-05,
+      "loss": 1.3568,
+      "step": 12436
+    },
+    {
+      "epoch": 0.5729488183535265,
+      "grad_norm": 1.074374066901835,
+      "learning_rate": 8.134688380783824e-05,
+      "loss": 1.3696,
+      "step": 12437
+    },
+    {
+      "epoch": 0.5729948864421616,
+      "grad_norm": 1.045350920132124,
+      "learning_rate": 8.133222500483822e-05,
+      "loss": 1.1749,
+      "step": 12438
+    },
+    {
+      "epoch": 0.5730409545307965,
+      "grad_norm": 1.1705560671094222,
+      "learning_rate": 8.131756661744419e-05,
+      "loss": 1.4946,
+      "step": 12439
+    },
+    {
+      "epoch": 0.5730870226194316,
+      "grad_norm": 0.6873436845829912,
+      "learning_rate": 8.130290864598247e-05,
+      "loss": 1.6168,
+      "step": 12440
+    },
+    {
+      "epoch": 0.5731330907080665,
+      "grad_norm": 2.0637322927015123,
+      "learning_rate": 8.128825109077947e-05,
+      "loss": 1.0704,
+      "step": 12441
+    },
+    {
+      "epoch": 0.5731791587967016,
+      "grad_norm": 0.8397279300649836,
+      "learning_rate": 8.127359395216145e-05,
+      "loss": 1.359,
+      "step": 12442
+    },
+    {
+      "epoch": 0.5732252268853365,
+      "grad_norm": 0.9277133787206447,
+      "learning_rate": 8.125893723045478e-05,
+      "loss": 1.1451,
+      "step": 12443
+    },
+    {
+      "epoch": 0.5732712949739716,
+      "grad_norm": 0.7750559330548388,
+      "learning_rate": 8.124428092598572e-05,
+      "loss": 1.5085,
+      "step": 12444
+    },
+    {
+      "epoch": 0.5733173630626065,
+      "grad_norm": 0.877102920778876,
+      "learning_rate": 8.12296250390806e-05,
+      "loss": 1.3834,
+      "step": 12445
+    },
+    {
+      "epoch": 0.5733634311512416,
+      "grad_norm": 1.0195549119695133,
+      "learning_rate": 8.121496957006567e-05,
+      "loss": 1.277,
+      "step": 12446
+    },
+    {
+      "epoch": 0.5734094992398765,
+      "grad_norm": 0.882645915539558,
+      "learning_rate": 8.120031451926725e-05,
+      "loss": 1.1301,
+      "step": 12447
+    },
+    {
+      "epoch": 0.5734555673285116,
+      "grad_norm": 1.0725956768746538,
+      "learning_rate": 8.11856598870116e-05,
+      "loss": 1.2674,
+      "step": 12448
+    },
+    {
+      "epoch": 0.5735016354171465,
+      "grad_norm": 0.6923161601457265,
+      "learning_rate": 8.117100567362494e-05,
+      "loss": 1.5258,
+      "step": 12449
+    },
+    {
+      "epoch": 0.5735477035057815,
+      "grad_norm": 2.0138957566104225,
+      "learning_rate": 8.11563518794336e-05,
+      "loss": 1.4592,
+      "step": 12450
+    },
+    {
+      "epoch": 0.5735937715944166,
+      "grad_norm": 1.0129456253039324,
+      "learning_rate": 8.114169850476374e-05,
+      "loss": 1.4964,
+      "step": 12451
+    },
+    {
+      "epoch": 0.5736398396830515,
+      "grad_norm": 2.225335271827939,
+      "learning_rate": 8.112704554994164e-05,
+      "loss": 1.232,
+      "step": 12452
+    },
+    {
+      "epoch": 0.5736859077716866,
+      "grad_norm": 1.0262746374685252,
+      "learning_rate": 8.111239301529349e-05,
+      "loss": 1.3403,
+      "step": 12453
+    },
+    {
+      "epoch": 0.5737319758603215,
+      "grad_norm": 0.8762058365683801,
+      "learning_rate": 8.109774090114557e-05,
+      "loss": 1.5066,
+      "step": 12454
+    },
+    {
+      "epoch": 0.5737780439489566,
+      "grad_norm": 1.0766702297196928,
+      "learning_rate": 8.1083089207824e-05,
+      "loss": 1.4586,
+      "step": 12455
+    },
+    {
+      "epoch": 0.5738241120375915,
+      "grad_norm": 4.86302582529242,
+      "learning_rate": 8.106843793565503e-05,
+      "loss": 1.3215,
+      "step": 12456
+    },
+    {
+      "epoch": 0.5738701801262266,
+      "grad_norm": 0.897104275959643,
+      "learning_rate": 8.105378708496486e-05,
+      "loss": 1.5848,
+      "step": 12457
+    },
+    {
+      "epoch": 0.5739162482148615,
+      "grad_norm": 1.4913003545794479,
+      "learning_rate": 8.103913665607957e-05,
+      "loss": 1.3362,
+      "step": 12458
+    },
+    {
+      "epoch": 0.5739623163034966,
+      "grad_norm": 0.9702000223037006,
+      "learning_rate": 8.102448664932543e-05,
+      "loss": 1.0942,
+      "step": 12459
+    },
+    {
+      "epoch": 0.5740083843921315,
+      "grad_norm": 0.9118394776026858,
+      "learning_rate": 8.100983706502853e-05,
+      "loss": 1.3217,
+      "step": 12460
+    },
+    {
+      "epoch": 0.5740544524807666,
+      "grad_norm": 13.99241628496714,
+      "learning_rate": 8.099518790351507e-05,
+      "loss": 1.12,
+      "step": 12461
+    },
+    {
+      "epoch": 0.5741005205694015,
+      "grad_norm": 0.8344333163137283,
+      "learning_rate": 8.098053916511116e-05,
+      "loss": 1.4382,
+      "step": 12462
+    },
+    {
+      "epoch": 0.5741465886580366,
+      "grad_norm": 0.8986756321690529,
+      "learning_rate": 8.096589085014294e-05,
+      "loss": 1.3961,
+      "step": 12463
+    },
+    {
+      "epoch": 0.5741926567466715,
+      "grad_norm": 1.2588070929996795,
+      "learning_rate": 8.09512429589365e-05,
+      "loss": 1.2156,
+      "step": 12464
+    },
+    {
+      "epoch": 0.5742387248353066,
+      "grad_norm": 0.8849858551811357,
+      "learning_rate": 8.093659549181802e-05,
+      "loss": 1.4687,
+      "step": 12465
+    },
+    {
+      "epoch": 0.5742847929239416,
+      "grad_norm": 2.704754171169206,
+      "learning_rate": 8.092194844911355e-05,
+      "loss": 1.4774,
+      "step": 12466
+    },
+    {
+      "epoch": 0.5743308610125766,
+      "grad_norm": 24.946185622200883,
+      "learning_rate": 8.090730183114913e-05,
+      "loss": 1.1743,
+      "step": 12467
+    },
+    {
+      "epoch": 0.5743769291012116,
+      "grad_norm": 1.4705797253694046,
+      "learning_rate": 8.089265563825096e-05,
+      "loss": 1.2584,
+      "step": 12468
+    },
+    {
+      "epoch": 0.5744229971898466,
+      "grad_norm": 0.6194529476823395,
+      "learning_rate": 8.087800987074499e-05,
+      "loss": 1.247,
+      "step": 12469
+    },
+    {
+      "epoch": 0.5744690652784816,
+      "grad_norm": 1.652395512413796,
+      "learning_rate": 8.086336452895741e-05,
+      "loss": 1.5146,
+      "step": 12470
+    },
+    {
+      "epoch": 0.5745151333671166,
+      "grad_norm": 1.1441011685381004,
+      "learning_rate": 8.084871961321413e-05,
+      "loss": 1.2919,
+      "step": 12471
+    },
+    {
+      "epoch": 0.5745612014557516,
+      "grad_norm": 1.1888941175986578,
+      "learning_rate": 8.083407512384133e-05,
+      "loss": 1.3793,
+      "step": 12472
+    },
+    {
+      "epoch": 0.5746072695443866,
+      "grad_norm": 1.044128525455616,
+      "learning_rate": 8.081943106116495e-05,
+      "loss": 1.3974,
+      "step": 12473
+    },
+    {
+      "epoch": 0.5746533376330216,
+      "grad_norm": 4.12915426496496,
+      "learning_rate": 8.080478742551107e-05,
+      "loss": 1.3667,
+      "step": 12474
+    },
+    {
+      "epoch": 0.5746994057216566,
+      "grad_norm": 1.0744779229322792,
+      "learning_rate": 8.079014421720573e-05,
+      "loss": 1.3644,
+      "step": 12475
+    },
+    {
+      "epoch": 0.5747454738102916,
+      "grad_norm": 1.1753432457319506,
+      "learning_rate": 8.07755014365748e-05,
+      "loss": 1.4864,
+      "step": 12476
+    },
+    {
+      "epoch": 0.5747915418989266,
+      "grad_norm": 1.578167663608265,
+      "learning_rate": 8.076085908394444e-05,
+      "loss": 1.1125,
+      "step": 12477
+    },
+    {
+      "epoch": 0.5748376099875616,
+      "grad_norm": 1.6953377572413617,
+      "learning_rate": 8.07462171596405e-05,
+      "loss": 1.2219,
+      "step": 12478
+    },
+    {
+      "epoch": 0.5748836780761967,
+      "grad_norm": 1.0259779258132038,
+      "learning_rate": 8.073157566398909e-05,
+      "loss": 1.2773,
+      "step": 12479
+    },
+    {
+      "epoch": 0.5749297461648316,
+      "grad_norm": 0.920430201794464,
+      "learning_rate": 8.071693459731605e-05,
+      "loss": 1.3702,
+      "step": 12480
+    },
+    {
+      "epoch": 0.5749758142534667,
+      "grad_norm": 1.336461215838847,
+      "learning_rate": 8.070229395994744e-05,
+      "loss": 1.3525,
+      "step": 12481
+    },
+    {
+      "epoch": 0.5750218823421016,
+      "grad_norm": 4.776659265280963,
+      "learning_rate": 8.068765375220913e-05,
+      "loss": 1.2746,
+      "step": 12482
+    },
+    {
+      "epoch": 0.5750679504307367,
+      "grad_norm": 2.0150808312669457,
+      "learning_rate": 8.067301397442713e-05,
+      "loss": 1.4486,
+      "step": 12483
+    },
+    {
+      "epoch": 0.5751140185193716,
+      "grad_norm": 0.9269345030997967,
+      "learning_rate": 8.065837462692734e-05,
+      "loss": 1.6758,
+      "step": 12484
+    },
+    {
+      "epoch": 0.5751600866080067,
+      "grad_norm": 0.9884582472086284,
+      "learning_rate": 8.064373571003561e-05,
+      "loss": 1.2573,
+      "step": 12485
+    },
+    {
+      "epoch": 0.5752061546966416,
+      "grad_norm": 0.9096806266562211,
+      "learning_rate": 8.0629097224078e-05,
+      "loss": 1.4546,
+      "step": 12486
+    },
+    {
+      "epoch": 0.5752522227852767,
+      "grad_norm": 0.9739395110649921,
+      "learning_rate": 8.061445916938026e-05,
+      "loss": 1.2259,
+      "step": 12487
+    },
+    {
+      "epoch": 0.5752982908739116,
+      "grad_norm": 0.8838656723736203,
+      "learning_rate": 8.059982154626841e-05,
+      "loss": 1.4885,
+      "step": 12488
+    },
+    {
+      "epoch": 0.5753443589625467,
+      "grad_norm": 1.4561572806202563,
+      "learning_rate": 8.058518435506819e-05,
+      "loss": 1.3443,
+      "step": 12489
+    },
+    {
+      "epoch": 0.5753904270511816,
+      "grad_norm": 3.4181543987180216,
+      "learning_rate": 8.057054759610562e-05,
+      "loss": 1.283,
+      "step": 12490
+    },
+    {
+      "epoch": 0.5754364951398167,
+      "grad_norm": 1.4878924107367637,
+      "learning_rate": 8.055591126970645e-05,
+      "loss": 1.3216,
+      "step": 12491
+    },
+    {
+      "epoch": 0.5754825632284517,
+      "grad_norm": 1.4096897464068558,
+      "learning_rate": 8.054127537619662e-05,
+      "loss": 1.4325,
+      "step": 12492
+    },
+    {
+      "epoch": 0.5755286313170866,
+      "grad_norm": 1.1081807478610204,
+      "learning_rate": 8.052663991590192e-05,
+      "loss": 1.1881,
+      "step": 12493
+    },
+    {
+      "epoch": 0.5755746994057217,
+      "grad_norm": 2.125111997354679,
+      "learning_rate": 8.051200488914816e-05,
+      "loss": 1.3917,
+      "step": 12494
+    },
+    {
+      "epoch": 0.5756207674943566,
+      "grad_norm": 1.476512490553539,
+      "learning_rate": 8.049737029626124e-05,
+      "loss": 1.4398,
+      "step": 12495
+    },
+    {
+      "epoch": 0.5756668355829917,
+      "grad_norm": 1.126330292066929,
+      "learning_rate": 8.048273613756687e-05,
+      "loss": 1.4601,
+      "step": 12496
+    },
+    {
+      "epoch": 0.5757129036716266,
+      "grad_norm": 1.1528440650042135,
+      "learning_rate": 8.046810241339096e-05,
+      "loss": 1.4523,
+      "step": 12497
+    },
+    {
+      "epoch": 0.5757589717602617,
+      "grad_norm": 1.0410411892844469,
+      "learning_rate": 8.045346912405922e-05,
+      "loss": 1.5927,
+      "step": 12498
+    },
+    {
+      "epoch": 0.5758050398488966,
+      "grad_norm": 0.9328575969888174,
+      "learning_rate": 8.043883626989753e-05,
+      "loss": 1.3038,
+      "step": 12499
+    },
+    {
+      "epoch": 0.5758511079375317,
+      "grad_norm": 1.196217519266265,
+      "learning_rate": 8.042420385123155e-05,
+      "loss": 1.0358,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5758971760261666,
+      "grad_norm": 0.8313162893763568,
+      "learning_rate": 8.040957186838718e-05,
+      "loss": 1.4028,
+      "step": 12501
+    },
+    {
+      "epoch": 0.5759432441148017,
+      "grad_norm": 1.4585375017341349,
+      "learning_rate": 8.039494032169008e-05,
+      "loss": 1.6751,
+      "step": 12502
+    },
+    {
+      "epoch": 0.5759893122034366,
+      "grad_norm": 1.0365175487480496,
+      "learning_rate": 8.038030921146597e-05,
+      "loss": 1.4507,
+      "step": 12503
+    },
+    {
+      "epoch": 0.5760353802920717,
+      "grad_norm": 5.351471858881662,
+      "learning_rate": 8.036567853804068e-05,
+      "loss": 1.235,
+      "step": 12504
+    },
+    {
+      "epoch": 0.5760814483807067,
+      "grad_norm": 0.7987966511469692,
+      "learning_rate": 8.035104830173986e-05,
+      "loss": 1.4239,
+      "step": 12505
+    },
+    {
+      "epoch": 0.5761275164693417,
+      "grad_norm": 0.7304840561806848,
+      "learning_rate": 8.033641850288928e-05,
+      "loss": 1.5259,
+      "step": 12506
+    },
+    {
+      "epoch": 0.5761735845579767,
+      "grad_norm": 3.3319528575609954,
+      "learning_rate": 8.032178914181462e-05,
+      "loss": 1.4704,
+      "step": 12507
+    },
+    {
+      "epoch": 0.5762196526466117,
+      "grad_norm": 1.2134803794542057,
+      "learning_rate": 8.030716021884159e-05,
+      "loss": 1.5723,
+      "step": 12508
+    },
+    {
+      "epoch": 0.5762657207352467,
+      "grad_norm": 1.7394582245673855,
+      "learning_rate": 8.029253173429586e-05,
+      "loss": 1.3039,
+      "step": 12509
+    },
+    {
+      "epoch": 0.5763117888238817,
+      "grad_norm": 1.2789670880004922,
+      "learning_rate": 8.027790368850314e-05,
+      "loss": 1.3952,
+      "step": 12510
+    },
+    {
+      "epoch": 0.5763578569125167,
+      "grad_norm": 1.5965161469442353,
+      "learning_rate": 8.026327608178908e-05,
+      "loss": 1.3178,
+      "step": 12511
+    },
+    {
+      "epoch": 0.5764039250011517,
+      "grad_norm": 0.9770213215116929,
+      "learning_rate": 8.02486489144793e-05,
+      "loss": 1.6601,
+      "step": 12512
+    },
+    {
+      "epoch": 0.5764499930897867,
+      "grad_norm": 2.3029634088334943,
+      "learning_rate": 8.023402218689951e-05,
+      "loss": 1.439,
+      "step": 12513
+    },
+    {
+      "epoch": 0.5764960611784217,
+      "grad_norm": 5.021821478613329,
+      "learning_rate": 8.021939589937532e-05,
+      "loss": 1.3934,
+      "step": 12514
+    },
+    {
+      "epoch": 0.5765421292670567,
+      "grad_norm": 1.559567364714133,
+      "learning_rate": 8.020477005223237e-05,
+      "loss": 1.1207,
+      "step": 12515
+    },
+    {
+      "epoch": 0.5765881973556917,
+      "grad_norm": 0.9552682466933552,
+      "learning_rate": 8.019014464579627e-05,
+      "loss": 1.3055,
+      "step": 12516
+    },
+    {
+      "epoch": 0.5766342654443267,
+      "grad_norm": 1.3757016363890109,
+      "learning_rate": 8.017551968039263e-05,
+      "loss": 1.1975,
+      "step": 12517
+    },
+    {
+      "epoch": 0.5766803335329617,
+      "grad_norm": 6.442906027719921,
+      "learning_rate": 8.016089515634706e-05,
+      "loss": 1.0817,
+      "step": 12518
+    },
+    {
+      "epoch": 0.5767264016215967,
+      "grad_norm": 1.2451570678178456,
+      "learning_rate": 8.014627107398513e-05,
+      "loss": 1.4439,
+      "step": 12519
+    },
+    {
+      "epoch": 0.5767724697102318,
+      "grad_norm": 1.0147197078133012,
+      "learning_rate": 8.013164743363243e-05,
+      "loss": 1.6687,
+      "step": 12520
+    },
+    {
+      "epoch": 0.5768185377988667,
+      "grad_norm": 2.182874200231815,
+      "learning_rate": 8.011702423561455e-05,
+      "loss": 1.4392,
+      "step": 12521
+    },
+    {
+      "epoch": 0.5768646058875018,
+      "grad_norm": 1.026955001743617,
+      "learning_rate": 8.010240148025704e-05,
+      "loss": 1.4094,
+      "step": 12522
+    },
+    {
+      "epoch": 0.5769106739761367,
+      "grad_norm": 1.2430835937854463,
+      "learning_rate": 8.008777916788541e-05,
+      "loss": 1.2836,
+      "step": 12523
+    },
+    {
+      "epoch": 0.5769567420647718,
+      "grad_norm": 1.1877624601780141,
+      "learning_rate": 8.007315729882527e-05,
+      "loss": 1.3512,
+      "step": 12524
+    },
+    {
+      "epoch": 0.5770028101534067,
+      "grad_norm": 1.2872625805904252,
+      "learning_rate": 8.00585358734021e-05,
+      "loss": 1.7002,
+      "step": 12525
+    },
+    {
+      "epoch": 0.5770488782420418,
+      "grad_norm": 2.9532139520870624,
+      "learning_rate": 8.004391489194144e-05,
+      "loss": 1.086,
+      "step": 12526
+    },
+    {
+      "epoch": 0.5770949463306767,
+      "grad_norm": 1.6108285654044578,
+      "learning_rate": 8.002929435476878e-05,
+      "loss": 1.2492,
+      "step": 12527
+    },
+    {
+      "epoch": 0.5771410144193118,
+      "grad_norm": 1.8851959113685433,
+      "learning_rate": 8.001467426220966e-05,
+      "loss": 0.9646,
+      "step": 12528
+    },
+    {
+      "epoch": 0.5771870825079467,
+      "grad_norm": 0.8378090616603635,
+      "learning_rate": 8.000005461458954e-05,
+      "loss": 1.4223,
+      "step": 12529
+    },
+    {
+      "epoch": 0.5772331505965818,
+      "grad_norm": 1.3720929811289024,
+      "learning_rate": 7.998543541223392e-05,
+      "loss": 1.0976,
+      "step": 12530
+    },
+    {
+      "epoch": 0.5772792186852167,
+      "grad_norm": 1.6689847202590231,
+      "learning_rate": 7.997081665546827e-05,
+      "loss": 1.1803,
+      "step": 12531
+    },
+    {
+      "epoch": 0.5773252867738518,
+      "grad_norm": 1.6027651411743107,
+      "learning_rate": 7.995619834461801e-05,
+      "loss": 1.3881,
+      "step": 12532
+    },
+    {
+      "epoch": 0.5773713548624868,
+      "grad_norm": 1.0636675074554764,
+      "learning_rate": 7.994158048000866e-05,
+      "loss": 1.358,
+      "step": 12533
+    },
+    {
+      "epoch": 0.5774174229511218,
+      "grad_norm": 1.9836522638045024,
+      "learning_rate": 7.99269630619656e-05,
+      "loss": 1.525,
+      "step": 12534
+    },
+    {
+      "epoch": 0.5774634910397568,
+      "grad_norm": 1.038624365055852,
+      "learning_rate": 7.991234609081432e-05,
+      "loss": 1.3543,
+      "step": 12535
+    },
+    {
+      "epoch": 0.5775095591283917,
+      "grad_norm": 1.1389860667284086,
+      "learning_rate": 7.98977295668802e-05,
+      "loss": 1.1051,
+      "step": 12536
+    },
+    {
+      "epoch": 0.5775556272170268,
+      "grad_norm": 5.14624720828951,
+      "learning_rate": 7.988311349048866e-05,
+      "loss": 1.2108,
+      "step": 12537
+    },
+    {
+      "epoch": 0.5776016953056617,
+      "grad_norm": 1.4447605361762168,
+      "learning_rate": 7.98684978619651e-05,
+      "loss": 1.2357,
+      "step": 12538
+    },
+    {
+      "epoch": 0.5776477633942968,
+      "grad_norm": 3.691635973450358,
+      "learning_rate": 7.985388268163493e-05,
+      "loss": 1.3247,
+      "step": 12539
+    },
+    {
+      "epoch": 0.5776938314829317,
+      "grad_norm": 2.557587445295601,
+      "learning_rate": 7.983926794982351e-05,
+      "loss": 1.1772,
+      "step": 12540
+    },
+    {
+      "epoch": 0.5777398995715668,
+      "grad_norm": 0.8914060571965612,
+      "learning_rate": 7.982465366685623e-05,
+      "loss": 1.2346,
+      "step": 12541
+    },
+    {
+      "epoch": 0.5777859676602017,
+      "grad_norm": 1.1050039617188963,
+      "learning_rate": 7.981003983305844e-05,
+      "loss": 1.3479,
+      "step": 12542
+    },
+    {
+      "epoch": 0.5778320357488368,
+      "grad_norm": 1.38919827780766,
+      "learning_rate": 7.979542644875548e-05,
+      "loss": 1.4269,
+      "step": 12543
+    },
+    {
+      "epoch": 0.5778781038374717,
+      "grad_norm": 1.5751035969220675,
+      "learning_rate": 7.978081351427271e-05,
+      "loss": 1.3426,
+      "step": 12544
+    },
+    {
+      "epoch": 0.5779241719261068,
+      "grad_norm": 1.5469714758777435,
+      "learning_rate": 7.976620102993546e-05,
+      "loss": 1.2566,
+      "step": 12545
+    },
+    {
+      "epoch": 0.5779702400147418,
+      "grad_norm": 1.101560740807484,
+      "learning_rate": 7.975158899606907e-05,
+      "loss": 1.6393,
+      "step": 12546
+    },
+    {
+      "epoch": 0.5780163081033768,
+      "grad_norm": 1.1803998828187523,
+      "learning_rate": 7.97369774129988e-05,
+      "loss": 1.3458,
+      "step": 12547
+    },
+    {
+      "epoch": 0.5780623761920118,
+      "grad_norm": 1.3444702533482265,
+      "learning_rate": 7.972236628105001e-05,
+      "loss": 1.5171,
+      "step": 12548
+    },
+    {
+      "epoch": 0.5781084442806468,
+      "grad_norm": 1.1198498381305666,
+      "learning_rate": 7.970775560054798e-05,
+      "loss": 1.346,
+      "step": 12549
+    },
+    {
+      "epoch": 0.5781545123692818,
+      "grad_norm": 1.4782000084529818,
+      "learning_rate": 7.969314537181793e-05,
+      "loss": 1.4751,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5782005804579168,
+      "grad_norm": 0.8640209853817278,
+      "learning_rate": 7.967853559518522e-05,
+      "loss": 1.4616,
+      "step": 12551
+    },
+    {
+      "epoch": 0.5782466485465518,
+      "grad_norm": 8.847525755122513,
+      "learning_rate": 7.966392627097505e-05,
+      "loss": 1.4707,
+      "step": 12552
+    },
+    {
+      "epoch": 0.5782927166351868,
+      "grad_norm": 0.8450559665694678,
+      "learning_rate": 7.964931739951271e-05,
+      "loss": 1.4678,
+      "step": 12553
+    },
+    {
+      "epoch": 0.5783387847238218,
+      "grad_norm": 1.1790985138502161,
+      "learning_rate": 7.963470898112339e-05,
+      "loss": 1.3398,
+      "step": 12554
+    },
+    {
+      "epoch": 0.5783848528124568,
+      "grad_norm": 0.9966511684812079,
+      "learning_rate": 7.96201010161324e-05,
+      "loss": 1.4702,
+      "step": 12555
+    },
+    {
+      "epoch": 0.5784309209010918,
+      "grad_norm": 2.1619716739392416,
+      "learning_rate": 7.960549350486488e-05,
+      "loss": 1.4867,
+      "step": 12556
+    },
+    {
+      "epoch": 0.5784769889897268,
+      "grad_norm": 0.7966961225095459,
+      "learning_rate": 7.959088644764611e-05,
+      "loss": 1.3292,
+      "step": 12557
+    },
+    {
+      "epoch": 0.5785230570783618,
+      "grad_norm": 1.2591744116572783,
+      "learning_rate": 7.957627984480125e-05,
+      "loss": 1.3446,
+      "step": 12558
+    },
+    {
+      "epoch": 0.5785691251669968,
+      "grad_norm": 2.2483334791611043,
+      "learning_rate": 7.956167369665548e-05,
+      "loss": 1.5194,
+      "step": 12559
+    },
+    {
+      "epoch": 0.5786151932556318,
+      "grad_norm": 1.1876566526589574,
+      "learning_rate": 7.954706800353403e-05,
+      "loss": 1.6342,
+      "step": 12560
+    },
+    {
+      "epoch": 0.5786612613442669,
+      "grad_norm": 1.3787613591626138,
+      "learning_rate": 7.9532462765762e-05,
+      "loss": 1.5383,
+      "step": 12561
+    },
+    {
+      "epoch": 0.5787073294329018,
+      "grad_norm": 1.4654361979998272,
+      "learning_rate": 7.951785798366463e-05,
+      "loss": 1.2338,
+      "step": 12562
+    },
+    {
+      "epoch": 0.5787533975215369,
+      "grad_norm": 1.6527430343481062,
+      "learning_rate": 7.950325365756701e-05,
+      "loss": 1.5242,
+      "step": 12563
+    },
+    {
+      "epoch": 0.5787994656101718,
+      "grad_norm": 1.1086403719269446,
+      "learning_rate": 7.948864978779434e-05,
+      "loss": 1.3273,
+      "step": 12564
+    },
+    {
+      "epoch": 0.5788455336988069,
+      "grad_norm": 4.677811588255119,
+      "learning_rate": 7.947404637467166e-05,
+      "loss": 1.317,
+      "step": 12565
+    },
+    {
+      "epoch": 0.5788916017874418,
+      "grad_norm": 0.8865340089063214,
+      "learning_rate": 7.945944341852419e-05,
+      "loss": 1.4015,
+      "step": 12566
+    },
+    {
+      "epoch": 0.5789376698760769,
+      "grad_norm": 1.3294825166647124,
+      "learning_rate": 7.9444840919677e-05,
+      "loss": 1.2975,
+      "step": 12567
+    },
+    {
+      "epoch": 0.5789837379647118,
+      "grad_norm": 0.7959354913694966,
+      "learning_rate": 7.943023887845512e-05,
+      "loss": 1.468,
+      "step": 12568
+    },
+    {
+      "epoch": 0.5790298060533469,
+      "grad_norm": 2.0948761733690824,
+      "learning_rate": 7.941563729518376e-05,
+      "loss": 1.3942,
+      "step": 12569
+    },
+    {
+      "epoch": 0.5790758741419818,
+      "grad_norm": 2.3730497835429323,
+      "learning_rate": 7.940103617018789e-05,
+      "loss": 1.0655,
+      "step": 12570
+    },
+    {
+      "epoch": 0.5791219422306169,
+      "grad_norm": 0.9889942241229227,
+      "learning_rate": 7.938643550379267e-05,
+      "loss": 1.4492,
+      "step": 12571
+    },
+    {
+      "epoch": 0.5791680103192518,
+      "grad_norm": 1.2827118255447898,
+      "learning_rate": 7.93718352963231e-05,
+      "loss": 0.8542,
+      "step": 12572
+    },
+    {
+      "epoch": 0.5792140784078869,
+      "grad_norm": 1.0089590645252609,
+      "learning_rate": 7.935723554810426e-05,
+      "loss": 1.2937,
+      "step": 12573
+    },
+    {
+      "epoch": 0.5792601464965219,
+      "grad_norm": 2.0758152787315876,
+      "learning_rate": 7.934263625946116e-05,
+      "loss": 1.3177,
+      "step": 12574
+    },
+    {
+      "epoch": 0.5793062145851569,
+      "grad_norm": 0.9920538991794583,
+      "learning_rate": 7.932803743071885e-05,
+      "loss": 1.0371,
+      "step": 12575
+    },
+    {
+      "epoch": 0.5793522826737919,
+      "grad_norm": 0.906380962487213,
+      "learning_rate": 7.931343906220238e-05,
+      "loss": 1.2687,
+      "step": 12576
+    },
+    {
+      "epoch": 0.5793983507624269,
+      "grad_norm": 1.2065007441117357,
+      "learning_rate": 7.929884115423666e-05,
+      "loss": 1.217,
+      "step": 12577
+    },
+    {
+      "epoch": 0.5794444188510619,
+      "grad_norm": 1.061929371453385,
+      "learning_rate": 7.92842437071468e-05,
+      "loss": 1.2735,
+      "step": 12578
+    },
+    {
+      "epoch": 0.5794904869396968,
+      "grad_norm": 0.8446757229439221,
+      "learning_rate": 7.926964672125767e-05,
+      "loss": 1.3208,
+      "step": 12579
+    },
+    {
+      "epoch": 0.5795365550283319,
+      "grad_norm": 1.0568036679868962,
+      "learning_rate": 7.925505019689437e-05,
+      "loss": 1.3453,
+      "step": 12580
+    },
+    {
+      "epoch": 0.5795826231169668,
+      "grad_norm": 1.3913151248655982,
+      "learning_rate": 7.924045413438176e-05,
+      "loss": 1.4027,
+      "step": 12581
+    },
+    {
+      "epoch": 0.5796286912056019,
+      "grad_norm": 1.0956057495399063,
+      "learning_rate": 7.92258585340449e-05,
+      "loss": 1.1603,
+      "step": 12582
+    },
+    {
+      "epoch": 0.5796747592942368,
+      "grad_norm": 1.1234453036264638,
+      "learning_rate": 7.921126339620861e-05,
+      "loss": 1.6107,
+      "step": 12583
+    },
+    {
+      "epoch": 0.5797208273828719,
+      "grad_norm": 1.8001674498259719,
+      "learning_rate": 7.919666872119794e-05,
+      "loss": 1.4464,
+      "step": 12584
+    },
+    {
+      "epoch": 0.5797668954715068,
+      "grad_norm": 0.9846090236184045,
+      "learning_rate": 7.91820745093378e-05,
+      "loss": 1.2172,
+      "step": 12585
+    },
+    {
+      "epoch": 0.5798129635601419,
+      "grad_norm": 1.6944045503619274,
+      "learning_rate": 7.9167480760953e-05,
+      "loss": 1.4041,
+      "step": 12586
+    },
+    {
+      "epoch": 0.579859031648777,
+      "grad_norm": 0.8898172959558059,
+      "learning_rate": 7.91528874763686e-05,
+      "loss": 1.3856,
+      "step": 12587
+    },
+    {
+      "epoch": 0.5799050997374119,
+      "grad_norm": 1.433929838435523,
+      "learning_rate": 7.913829465590935e-05,
+      "loss": 1.1695,
+      "step": 12588
+    },
+    {
+      "epoch": 0.5799511678260469,
+      "grad_norm": 1.3825826914010335,
+      "learning_rate": 7.912370229990025e-05,
+      "loss": 1.0742,
+      "step": 12589
+    },
+    {
+      "epoch": 0.5799972359146819,
+      "grad_norm": 1.603155590591908,
+      "learning_rate": 7.910911040866606e-05,
+      "loss": 1.5565,
+      "step": 12590
+    },
+    {
+      "epoch": 0.5800433040033169,
+      "grad_norm": 1.3250710584718608,
+      "learning_rate": 7.909451898253178e-05,
+      "loss": 1.1643,
+      "step": 12591
+    },
+    {
+      "epoch": 0.5800893720919519,
+      "grad_norm": 1.2289446391954164,
+      "learning_rate": 7.907992802182211e-05,
+      "loss": 1.5385,
+      "step": 12592
+    },
+    {
+      "epoch": 0.5801354401805869,
+      "grad_norm": 0.9554085879697177,
+      "learning_rate": 7.906533752686205e-05,
+      "loss": 1.2666,
+      "step": 12593
+    },
+    {
+      "epoch": 0.5801815082692219,
+      "grad_norm": 0.8851809988935465,
+      "learning_rate": 7.905074749797633e-05,
+      "loss": 1.4776,
+      "step": 12594
+    },
+    {
+      "epoch": 0.5802275763578569,
+      "grad_norm": 1.1358289038269183,
+      "learning_rate": 7.903615793548978e-05,
+      "loss": 1.4291,
+      "step": 12595
+    },
+    {
+      "epoch": 0.5802736444464919,
+      "grad_norm": 11.821159231375043,
+      "learning_rate": 7.902156883972725e-05,
+      "loss": 1.0393,
+      "step": 12596
+    },
+    {
+      "epoch": 0.5803197125351269,
+      "grad_norm": 1.8469567265757973,
+      "learning_rate": 7.900698021101347e-05,
+      "loss": 1.3924,
+      "step": 12597
+    },
+    {
+      "epoch": 0.5803657806237619,
+      "grad_norm": 0.761056446540655,
+      "learning_rate": 7.899239204967334e-05,
+      "loss": 1.3231,
+      "step": 12598
+    },
+    {
+      "epoch": 0.5804118487123969,
+      "grad_norm": 1.0540577549330603,
+      "learning_rate": 7.897780435603153e-05,
+      "loss": 1.4134,
+      "step": 12599
+    },
+    {
+      "epoch": 0.5804579168010319,
+      "grad_norm": 0.7148391674893564,
+      "learning_rate": 7.896321713041292e-05,
+      "loss": 1.3391,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5805039848896669,
+      "grad_norm": 0.9791628134853352,
+      "learning_rate": 7.894863037314214e-05,
+      "loss": 1.3248,
+      "step": 12601
+    },
+    {
+      "epoch": 0.580550052978302,
+      "grad_norm": 2.5035619964253786,
+      "learning_rate": 7.893404408454408e-05,
+      "loss": 1.2665,
+      "step": 12602
+    },
+    {
+      "epoch": 0.5805961210669369,
+      "grad_norm": 2.644878460044675,
+      "learning_rate": 7.891945826494338e-05,
+      "loss": 1.4251,
+      "step": 12603
+    },
+    {
+      "epoch": 0.580642189155572,
+      "grad_norm": 1.2467754094737435,
+      "learning_rate": 7.890487291466479e-05,
+      "loss": 1.3966,
+      "step": 12604
+    },
+    {
+      "epoch": 0.5806882572442069,
+      "grad_norm": 1.1953667259387428,
+      "learning_rate": 7.889028803403304e-05,
+      "loss": 1.3515,
+      "step": 12605
+    },
+    {
+      "epoch": 0.580734325332842,
+      "grad_norm": 1.0181720348792445,
+      "learning_rate": 7.887570362337282e-05,
+      "loss": 1.0759,
+      "step": 12606
+    },
+    {
+      "epoch": 0.5807803934214769,
+      "grad_norm": 1.529504096849134,
+      "learning_rate": 7.886111968300884e-05,
+      "loss": 1.3053,
+      "step": 12607
+    },
+    {
+      "epoch": 0.580826461510112,
+      "grad_norm": 1.8957147546361455,
+      "learning_rate": 7.884653621326579e-05,
+      "loss": 1.4298,
+      "step": 12608
+    },
+    {
+      "epoch": 0.5808725295987469,
+      "grad_norm": 1.217924638949267,
+      "learning_rate": 7.883195321446834e-05,
+      "loss": 1.1538,
+      "step": 12609
+    },
+    {
+      "epoch": 0.580918597687382,
+      "grad_norm": 0.9981529985868516,
+      "learning_rate": 7.881737068694113e-05,
+      "loss": 1.1542,
+      "step": 12610
+    },
+    {
+      "epoch": 0.5809646657760169,
+      "grad_norm": 1.0597108186203743,
+      "learning_rate": 7.880278863100891e-05,
+      "loss": 1.242,
+      "step": 12611
+    },
+    {
+      "epoch": 0.581010733864652,
+      "grad_norm": 0.9822122263685856,
+      "learning_rate": 7.878820704699621e-05,
+      "loss": 1.4809,
+      "step": 12612
+    },
+    {
+      "epoch": 0.5810568019532869,
+      "grad_norm": 1.2743666662231121,
+      "learning_rate": 7.877362593522771e-05,
+      "loss": 1.4011,
+      "step": 12613
+    },
+    {
+      "epoch": 0.581102870041922,
+      "grad_norm": 1.4113525780375742,
+      "learning_rate": 7.875904529602805e-05,
+      "loss": 1.2634,
+      "step": 12614
+    },
+    {
+      "epoch": 0.581148938130557,
+      "grad_norm": 1.520312131013396,
+      "learning_rate": 7.87444651297218e-05,
+      "loss": 1.0602,
+      "step": 12615
+    },
+    {
+      "epoch": 0.581195006219192,
+      "grad_norm": 1.9074916187844957,
+      "learning_rate": 7.872988543663361e-05,
+      "loss": 1.0408,
+      "step": 12616
+    },
+    {
+      "epoch": 0.581241074307827,
+      "grad_norm": 1.090831955597546,
+      "learning_rate": 7.871530621708803e-05,
+      "loss": 1.5231,
+      "step": 12617
+    },
+    {
+      "epoch": 0.581287142396462,
+      "grad_norm": 1.4947043377370313,
+      "learning_rate": 7.870072747140968e-05,
+      "loss": 1.2286,
+      "step": 12618
+    },
+    {
+      "epoch": 0.581333210485097,
+      "grad_norm": 1.4870696154016065,
+      "learning_rate": 7.86861491999231e-05,
+      "loss": 1.3194,
+      "step": 12619
+    },
+    {
+      "epoch": 0.581379278573732,
+      "grad_norm": 1.3672206783275735,
+      "learning_rate": 7.867157140295287e-05,
+      "loss": 1.5075,
+      "step": 12620
+    },
+    {
+      "epoch": 0.581425346662367,
+      "grad_norm": 1.1430900019672934,
+      "learning_rate": 7.865699408082354e-05,
+      "loss": 1.2906,
+      "step": 12621
+    },
+    {
+      "epoch": 0.581471414751002,
+      "grad_norm": 1.1667908550574442,
+      "learning_rate": 7.864241723385963e-05,
+      "loss": 1.3958,
+      "step": 12622
+    },
+    {
+      "epoch": 0.581517482839637,
+      "grad_norm": 0.9790001666193123,
+      "learning_rate": 7.862784086238568e-05,
+      "loss": 1.5096,
+      "step": 12623
+    },
+    {
+      "epoch": 0.5815635509282719,
+      "grad_norm": 0.8995228012236287,
+      "learning_rate": 7.861326496672621e-05,
+      "loss": 1.1735,
+      "step": 12624
+    },
+    {
+      "epoch": 0.581609619016907,
+      "grad_norm": 1.8117151087240275,
+      "learning_rate": 7.859868954720573e-05,
+      "loss": 1.1848,
+      "step": 12625
+    },
+    {
+      "epoch": 0.5816556871055419,
+      "grad_norm": 1.2297784831256153,
+      "learning_rate": 7.858411460414871e-05,
+      "loss": 1.0908,
+      "step": 12626
+    },
+    {
+      "epoch": 0.581701755194177,
+      "grad_norm": 1.6782254985248992,
+      "learning_rate": 7.85695401378797e-05,
+      "loss": 1.3381,
+      "step": 12627
+    },
+    {
+      "epoch": 0.581747823282812,
+      "grad_norm": 1.1432396535261302,
+      "learning_rate": 7.855496614872308e-05,
+      "loss": 1.1476,
+      "step": 12628
+    },
+    {
+      "epoch": 0.581793891371447,
+      "grad_norm": 1.505942059592879,
+      "learning_rate": 7.854039263700341e-05,
+      "loss": 1.3854,
+      "step": 12629
+    },
+    {
+      "epoch": 0.581839959460082,
+      "grad_norm": 1.8661185882668272,
+      "learning_rate": 7.85258196030451e-05,
+      "loss": 1.3257,
+      "step": 12630
+    },
+    {
+      "epoch": 0.581886027548717,
+      "grad_norm": 0.9881365402014493,
+      "learning_rate": 7.851124704717258e-05,
+      "loss": 1.3072,
+      "step": 12631
+    },
+    {
+      "epoch": 0.581932095637352,
+      "grad_norm": 1.352002263162576,
+      "learning_rate": 7.849667496971032e-05,
+      "loss": 1.0349,
+      "step": 12632
+    },
+    {
+      "epoch": 0.581978163725987,
+      "grad_norm": 1.8043107192232126,
+      "learning_rate": 7.848210337098268e-05,
+      "loss": 1.2009,
+      "step": 12633
+    },
+    {
+      "epoch": 0.582024231814622,
+      "grad_norm": 0.9276478708323517,
+      "learning_rate": 7.846753225131416e-05,
+      "loss": 1.1867,
+      "step": 12634
+    },
+    {
+      "epoch": 0.582070299903257,
+      "grad_norm": 2.3266444625734617,
+      "learning_rate": 7.845296161102907e-05,
+      "loss": 1.3067,
+      "step": 12635
+    },
+    {
+      "epoch": 0.582116367991892,
+      "grad_norm": 1.3027929847004749,
+      "learning_rate": 7.843839145045188e-05,
+      "loss": 1.5631,
+      "step": 12636
+    },
+    {
+      "epoch": 0.582162436080527,
+      "grad_norm": 1.2612459120436215,
+      "learning_rate": 7.842382176990691e-05,
+      "loss": 1.2859,
+      "step": 12637
+    },
+    {
+      "epoch": 0.582208504169162,
+      "grad_norm": 1.582629383581813,
+      "learning_rate": 7.840925256971859e-05,
+      "loss": 1.3143,
+      "step": 12638
+    },
+    {
+      "epoch": 0.582254572257797,
+      "grad_norm": 0.8852560786424788,
+      "learning_rate": 7.839468385021121e-05,
+      "loss": 1.2202,
+      "step": 12639
+    },
+    {
+      "epoch": 0.582300640346432,
+      "grad_norm": 1.1234484128518456,
+      "learning_rate": 7.838011561170915e-05,
+      "loss": 1.3912,
+      "step": 12640
+    },
+    {
+      "epoch": 0.5823467084350671,
+      "grad_norm": 1.3545756211982078,
+      "learning_rate": 7.836554785453676e-05,
+      "loss": 1.5812,
+      "step": 12641
+    },
+    {
+      "epoch": 0.582392776523702,
+      "grad_norm": 1.7940552126695783,
+      "learning_rate": 7.835098057901833e-05,
+      "loss": 1.1684,
+      "step": 12642
+    },
+    {
+      "epoch": 0.5824388446123371,
+      "grad_norm": 1.227715500694917,
+      "learning_rate": 7.833641378547823e-05,
+      "loss": 1.3465,
+      "step": 12643
+    },
+    {
+      "epoch": 0.582484912700972,
+      "grad_norm": 1.8353059508913248,
+      "learning_rate": 7.83218474742407e-05,
+      "loss": 1.1037,
+      "step": 12644
+    },
+    {
+      "epoch": 0.5825309807896071,
+      "grad_norm": 1.1424427352669697,
+      "learning_rate": 7.830728164563008e-05,
+      "loss": 1.3574,
+      "step": 12645
+    },
+    {
+      "epoch": 0.582577048878242,
+      "grad_norm": 2.2349637761288528,
+      "learning_rate": 7.829271629997062e-05,
+      "loss": 1.2958,
+      "step": 12646
+    },
+    {
+      "epoch": 0.5826231169668771,
+      "grad_norm": 0.9503531897133694,
+      "learning_rate": 7.827815143758664e-05,
+      "loss": 1.2585,
+      "step": 12647
+    },
+    {
+      "epoch": 0.582669185055512,
+      "grad_norm": 1.634139643376419,
+      "learning_rate": 7.826358705880236e-05,
+      "loss": 1.2181,
+      "step": 12648
+    },
+    {
+      "epoch": 0.5827152531441471,
+      "grad_norm": 1.5735223423943165,
+      "learning_rate": 7.824902316394205e-05,
+      "loss": 1.222,
+      "step": 12649
+    },
+    {
+      "epoch": 0.582761321232782,
+      "grad_norm": 2.193617975281137,
+      "learning_rate": 7.823445975332995e-05,
+      "loss": 1.2122,
+      "step": 12650
+    },
+    {
+      "epoch": 0.5828073893214171,
+      "grad_norm": 2.246120133429024,
+      "learning_rate": 7.821989682729028e-05,
+      "loss": 1.3486,
+      "step": 12651
+    },
+    {
+      "epoch": 0.582853457410052,
+      "grad_norm": 1.2368990870628842,
+      "learning_rate": 7.820533438614726e-05,
+      "loss": 1.2701,
+      "step": 12652
+    },
+    {
+      "epoch": 0.5828995254986871,
+      "grad_norm": 1.0067680183915817,
+      "learning_rate": 7.819077243022508e-05,
+      "loss": 1.1636,
+      "step": 12653
+    },
+    {
+      "epoch": 0.582945593587322,
+      "grad_norm": 1.7460520416382344,
+      "learning_rate": 7.817621095984798e-05,
+      "loss": 1.2627,
+      "step": 12654
+    },
+    {
+      "epoch": 0.5829916616759571,
+      "grad_norm": 0.9453123642522238,
+      "learning_rate": 7.81616499753401e-05,
+      "loss": 1.2535,
+      "step": 12655
+    },
+    {
+      "epoch": 0.5830377297645921,
+      "grad_norm": 14.809609952619251,
+      "learning_rate": 7.814708947702567e-05,
+      "loss": 1.4058,
+      "step": 12656
+    },
+    {
+      "epoch": 0.5830837978532271,
+      "grad_norm": 1.1233441875110954,
+      "learning_rate": 7.81325294652288e-05,
+      "loss": 1.0886,
+      "step": 12657
+    },
+    {
+      "epoch": 0.5831298659418621,
+      "grad_norm": 1.5994405745141786,
+      "learning_rate": 7.811796994027369e-05,
+      "loss": 1.0255,
+      "step": 12658
+    },
+    {
+      "epoch": 0.5831759340304971,
+      "grad_norm": 0.9373427161577258,
+      "learning_rate": 7.810341090248445e-05,
+      "loss": 1.2638,
+      "step": 12659
+    },
+    {
+      "epoch": 0.5832220021191321,
+      "grad_norm": 0.8371233770957726,
+      "learning_rate": 7.80888523521852e-05,
+      "loss": 1.3631,
+      "step": 12660
+    },
+    {
+      "epoch": 0.583268070207767,
+      "grad_norm": 1.907970605615188,
+      "learning_rate": 7.807429428970011e-05,
+      "loss": 0.9492,
+      "step": 12661
+    },
+    {
+      "epoch": 0.5833141382964021,
+      "grad_norm": 0.9203459718456918,
+      "learning_rate": 7.805973671535325e-05,
+      "loss": 1.512,
+      "step": 12662
+    },
+    {
+      "epoch": 0.583360206385037,
+      "grad_norm": 1.2733299993427876,
+      "learning_rate": 7.804517962946873e-05,
+      "loss": 1.1955,
+      "step": 12663
+    },
+    {
+      "epoch": 0.5834062744736721,
+      "grad_norm": 1.2336856513343233,
+      "learning_rate": 7.803062303237065e-05,
+      "loss": 1.2187,
+      "step": 12664
+    },
+    {
+      "epoch": 0.583452342562307,
+      "grad_norm": 1.000144306930514,
+      "learning_rate": 7.801606692438307e-05,
+      "loss": 1.4593,
+      "step": 12665
+    },
+    {
+      "epoch": 0.5834984106509421,
+      "grad_norm": 1.3225800250631754,
+      "learning_rate": 7.800151130583006e-05,
+      "loss": 1.1687,
+      "step": 12666
+    },
+    {
+      "epoch": 0.583544478739577,
+      "grad_norm": 0.9695640366805525,
+      "learning_rate": 7.798695617703571e-05,
+      "loss": 1.2636,
+      "step": 12667
+    },
+    {
+      "epoch": 0.5835905468282121,
+      "grad_norm": 2.4922620474549277,
+      "learning_rate": 7.797240153832405e-05,
+      "loss": 1.2688,
+      "step": 12668
+    },
+    {
+      "epoch": 0.5836366149168472,
+      "grad_norm": 1.0195730182711333,
+      "learning_rate": 7.795784739001903e-05,
+      "loss": 1.4134,
+      "step": 12669
+    },
+    {
+      "epoch": 0.5836826830054821,
+      "grad_norm": 1.2413262975377166,
+      "learning_rate": 7.79432937324448e-05,
+      "loss": 1.3948,
+      "step": 12670
+    },
+    {
+      "epoch": 0.5837287510941171,
+      "grad_norm": 1.172984595795549,
+      "learning_rate": 7.792874056592528e-05,
+      "loss": 1.5775,
+      "step": 12671
+    },
+    {
+      "epoch": 0.5837748191827521,
+      "grad_norm": 2.2116942993800697,
+      "learning_rate": 7.791418789078454e-05,
+      "loss": 1.2181,
+      "step": 12672
+    },
+    {
+      "epoch": 0.5838208872713871,
+      "grad_norm": 1.48878360424377,
+      "learning_rate": 7.789963570734651e-05,
+      "loss": 1.3269,
+      "step": 12673
+    },
+    {
+      "epoch": 0.5838669553600221,
+      "grad_norm": 1.104327309035751,
+      "learning_rate": 7.788508401593523e-05,
+      "loss": 1.047,
+      "step": 12674
+    },
+    {
+      "epoch": 0.5839130234486571,
+      "grad_norm": 1.0065290191613248,
+      "learning_rate": 7.78705328168746e-05,
+      "loss": 1.4757,
+      "step": 12675
+    },
+    {
+      "epoch": 0.5839590915372921,
+      "grad_norm": 1.1775647162018708,
+      "learning_rate": 7.785598211048863e-05,
+      "loss": 1.4802,
+      "step": 12676
+    },
+    {
+      "epoch": 0.5840051596259271,
+      "grad_norm": 1.0505929870072952,
+      "learning_rate": 7.78414318971013e-05,
+      "loss": 1.1805,
+      "step": 12677
+    },
+    {
+      "epoch": 0.5840512277145621,
+      "grad_norm": 1.0007406050477932,
+      "learning_rate": 7.78268821770364e-05,
+      "loss": 1.282,
+      "step": 12678
+    },
+    {
+      "epoch": 0.5840972958031971,
+      "grad_norm": 0.8245562097094887,
+      "learning_rate": 7.781233295061802e-05,
+      "loss": 1.3342,
+      "step": 12679
+    },
+    {
+      "epoch": 0.5841433638918321,
+      "grad_norm": 1.0403368073366033,
+      "learning_rate": 7.779778421816995e-05,
+      "loss": 1.0228,
+      "step": 12680
+    },
+    {
+      "epoch": 0.5841894319804671,
+      "grad_norm": 0.860974576206569,
+      "learning_rate": 7.77832359800162e-05,
+      "loss": 1.4368,
+      "step": 12681
+    },
+    {
+      "epoch": 0.5842355000691022,
+      "grad_norm": 0.8008089374314685,
+      "learning_rate": 7.776868823648055e-05,
+      "loss": 1.4529,
+      "step": 12682
+    },
+    {
+      "epoch": 0.5842815681577371,
+      "grad_norm": 0.97024201128623,
+      "learning_rate": 7.775414098788698e-05,
+      "loss": 1.5243,
+      "step": 12683
+    },
+    {
+      "epoch": 0.5843276362463722,
+      "grad_norm": 2.2273008759958572,
+      "learning_rate": 7.773959423455927e-05,
+      "loss": 1.0601,
+      "step": 12684
+    },
+    {
+      "epoch": 0.5843737043350071,
+      "grad_norm": 0.8496750519528343,
+      "learning_rate": 7.772504797682139e-05,
+      "loss": 1.3178,
+      "step": 12685
+    },
+    {
+      "epoch": 0.5844197724236422,
+      "grad_norm": 3.3666386932471912,
+      "learning_rate": 7.771050221499711e-05,
+      "loss": 1.1262,
+      "step": 12686
+    },
+    {
+      "epoch": 0.5844658405122771,
+      "grad_norm": 1.184749023705712,
+      "learning_rate": 7.769595694941024e-05,
+      "loss": 1.3591,
+      "step": 12687
+    },
+    {
+      "epoch": 0.5845119086009122,
+      "grad_norm": 1.7199230119258275,
+      "learning_rate": 7.768141218038469e-05,
+      "loss": 1.2406,
+      "step": 12688
+    },
+    {
+      "epoch": 0.5845579766895471,
+      "grad_norm": 1.2825781672182346,
+      "learning_rate": 7.766686790824417e-05,
+      "loss": 1.5301,
+      "step": 12689
+    },
+    {
+      "epoch": 0.5846040447781822,
+      "grad_norm": 1.0116920108146599,
+      "learning_rate": 7.765232413331263e-05,
+      "loss": 1.2592,
+      "step": 12690
+    },
+    {
+      "epoch": 0.5846501128668171,
+      "grad_norm": 2.7939246559179676,
+      "learning_rate": 7.763778085591369e-05,
+      "loss": 1.2704,
+      "step": 12691
+    },
+    {
+      "epoch": 0.5846961809554522,
+      "grad_norm": 1.7485272821630933,
+      "learning_rate": 7.762323807637128e-05,
+      "loss": 0.9865,
+      "step": 12692
+    },
+    {
+      "epoch": 0.5847422490440871,
+      "grad_norm": 5.776956945629147,
+      "learning_rate": 7.760869579500905e-05,
+      "loss": 1.6604,
+      "step": 12693
+    },
+    {
+      "epoch": 0.5847883171327222,
+      "grad_norm": 1.0467568964864329,
+      "learning_rate": 7.759415401215088e-05,
+      "loss": 1.2333,
+      "step": 12694
+    },
+    {
+      "epoch": 0.5848343852213571,
+      "grad_norm": 0.9577002322846241,
+      "learning_rate": 7.757961272812042e-05,
+      "loss": 1.2731,
+      "step": 12695
+    },
+    {
+      "epoch": 0.5848804533099922,
+      "grad_norm": 2.1183258388792594,
+      "learning_rate": 7.756507194324141e-05,
+      "loss": 1.2998,
+      "step": 12696
+    },
+    {
+      "epoch": 0.5849265213986272,
+      "grad_norm": 1.0719309142172104,
+      "learning_rate": 7.755053165783767e-05,
+      "loss": 1.3541,
+      "step": 12697
+    },
+    {
+      "epoch": 0.5849725894872622,
+      "grad_norm": 0.9476844534077116,
+      "learning_rate": 7.753599187223277e-05,
+      "loss": 1.1485,
+      "step": 12698
+    },
+    {
+      "epoch": 0.5850186575758972,
+      "grad_norm": 1.4166541736877818,
+      "learning_rate": 7.752145258675057e-05,
+      "loss": 1.0807,
+      "step": 12699
+    },
+    {
+      "epoch": 0.5850647256645322,
+      "grad_norm": 0.9310192154857525,
+      "learning_rate": 7.750691380171462e-05,
+      "loss": 1.4305,
+      "step": 12700
+    },
+    {
+      "epoch": 0.5851107937531672,
+      "grad_norm": 1.2524701549398918,
+      "learning_rate": 7.749237551744873e-05,
+      "loss": 1.4142,
+      "step": 12701
+    },
+    {
+      "epoch": 0.5851568618418022,
+      "grad_norm": 1.157404567843879,
+      "learning_rate": 7.747783773427644e-05,
+      "loss": 1.3609,
+      "step": 12702
+    },
+    {
+      "epoch": 0.5852029299304372,
+      "grad_norm": 1.1969165978416079,
+      "learning_rate": 7.746330045252155e-05,
+      "loss": 1.5762,
+      "step": 12703
+    },
+    {
+      "epoch": 0.5852489980190722,
+      "grad_norm": 1.416596515638548,
+      "learning_rate": 7.744876367250759e-05,
+      "loss": 1.3525,
+      "step": 12704
+    },
+    {
+      "epoch": 0.5852950661077072,
+      "grad_norm": 0.8467714495324649,
+      "learning_rate": 7.743422739455824e-05,
+      "loss": 1.3617,
+      "step": 12705
+    },
+    {
+      "epoch": 0.5853411341963421,
+      "grad_norm": 1.0377444983946567,
+      "learning_rate": 7.741969161899714e-05,
+      "loss": 1.7034,
+      "step": 12706
+    },
+    {
+      "epoch": 0.5853872022849772,
+      "grad_norm": 0.9000293953425528,
+      "learning_rate": 7.740515634614787e-05,
+      "loss": 1.1108,
+      "step": 12707
+    },
+    {
+      "epoch": 0.5854332703736121,
+      "grad_norm": 1.3757576141930477,
+      "learning_rate": 7.739062157633408e-05,
+      "loss": 1.1735,
+      "step": 12708
+    },
+    {
+      "epoch": 0.5854793384622472,
+      "grad_norm": 1.1166860771795677,
+      "learning_rate": 7.737608730987927e-05,
+      "loss": 1.5082,
+      "step": 12709
+    },
+    {
+      "epoch": 0.5855254065508823,
+      "grad_norm": 2.0237694692285464,
+      "learning_rate": 7.736155354710717e-05,
+      "loss": 0.9035,
+      "step": 12710
+    },
+    {
+      "epoch": 0.5855714746395172,
+      "grad_norm": 7.177310685474377,
+      "learning_rate": 7.734702028834117e-05,
+      "loss": 1.3157,
+      "step": 12711
+    },
+    {
+      "epoch": 0.5856175427281523,
+      "grad_norm": 1.822211595385456,
+      "learning_rate": 7.733248753390502e-05,
+      "loss": 1.0873,
+      "step": 12712
+    },
+    {
+      "epoch": 0.5856636108167872,
+      "grad_norm": 1.4174428520852658,
+      "learning_rate": 7.731795528412213e-05,
+      "loss": 1.2073,
+      "step": 12713
+    },
+    {
+      "epoch": 0.5857096789054222,
+      "grad_norm": 1.4474062399930354,
+      "learning_rate": 7.730342353931606e-05,
+      "loss": 1.0005,
+      "step": 12714
+    },
+    {
+      "epoch": 0.5857557469940572,
+      "grad_norm": 2.55459794236209,
+      "learning_rate": 7.728889229981036e-05,
+      "loss": 1.282,
+      "step": 12715
+    },
+    {
+      "epoch": 0.5858018150826922,
+      "grad_norm": 4.132918149620938,
+      "learning_rate": 7.727436156592852e-05,
+      "loss": 1.454,
+      "step": 12716
+    },
+    {
+      "epoch": 0.5858478831713272,
+      "grad_norm": 1.05514993447451,
+      "learning_rate": 7.725983133799406e-05,
+      "loss": 1.2748,
+      "step": 12717
+    },
+    {
+      "epoch": 0.5858939512599622,
+      "grad_norm": 1.5568741483314663,
+      "learning_rate": 7.724530161633046e-05,
+      "loss": 1.1054,
+      "step": 12718
+    },
+    {
+      "epoch": 0.5859400193485972,
+      "grad_norm": 1.2966442464994041,
+      "learning_rate": 7.723077240126121e-05,
+      "loss": 1.0071,
+      "step": 12719
+    },
+    {
+      "epoch": 0.5859860874372322,
+      "grad_norm": 3.3387133089095173,
+      "learning_rate": 7.721624369310977e-05,
+      "loss": 1.5115,
+      "step": 12720
+    },
+    {
+      "epoch": 0.5860321555258672,
+      "grad_norm": 1.6085712682415603,
+      "learning_rate": 7.720171549219961e-05,
+      "loss": 1.5131,
+      "step": 12721
+    },
+    {
+      "epoch": 0.5860782236145022,
+      "grad_norm": 3.651928318280505,
+      "learning_rate": 7.718718779885416e-05,
+      "loss": 1.4303,
+      "step": 12722
+    },
+    {
+      "epoch": 0.5861242917031373,
+      "grad_norm": 0.9685670069256723,
+      "learning_rate": 7.717266061339684e-05,
+      "loss": 1.2794,
+      "step": 12723
+    },
+    {
+      "epoch": 0.5861703597917722,
+      "grad_norm": 0.94348500340816,
+      "learning_rate": 7.71581339361511e-05,
+      "loss": 1.1898,
+      "step": 12724
+    },
+    {
+      "epoch": 0.5862164278804073,
+      "grad_norm": 2.46743150254277,
+      "learning_rate": 7.714360776744035e-05,
+      "loss": 1.5879,
+      "step": 12725
+    },
+    {
+      "epoch": 0.5862624959690422,
+      "grad_norm": 0.8680720695422941,
+      "learning_rate": 7.712908210758798e-05,
+      "loss": 1.0081,
+      "step": 12726
+    },
+    {
+      "epoch": 0.5863085640576773,
+      "grad_norm": 0.9753273668792285,
+      "learning_rate": 7.711455695691737e-05,
+      "loss": 1.2722,
+      "step": 12727
+    },
+    {
+      "epoch": 0.5863546321463122,
+      "grad_norm": 0.8818867184222114,
+      "learning_rate": 7.710003231575194e-05,
+      "loss": 1.1424,
+      "step": 12728
+    },
+    {
+      "epoch": 0.5864007002349473,
+      "grad_norm": 1.525065233997188,
+      "learning_rate": 7.708550818441498e-05,
+      "loss": 0.9754,
+      "step": 12729
+    },
+    {
+      "epoch": 0.5864467683235822,
+      "grad_norm": 0.8665348608581841,
+      "learning_rate": 7.707098456322991e-05,
+      "loss": 1.3892,
+      "step": 12730
+    },
+    {
+      "epoch": 0.5864928364122173,
+      "grad_norm": 0.9749282684535884,
+      "learning_rate": 7.705646145252008e-05,
+      "loss": 1.5141,
+      "step": 12731
+    },
+    {
+      "epoch": 0.5865389045008522,
+      "grad_norm": 0.8893675681702781,
+      "learning_rate": 7.704193885260876e-05,
+      "loss": 1.2796,
+      "step": 12732
+    },
+    {
+      "epoch": 0.5865849725894873,
+      "grad_norm": 1.4939436254559455,
+      "learning_rate": 7.702741676381934e-05,
+      "loss": 1.2879,
+      "step": 12733
+    },
+    {
+      "epoch": 0.5866310406781222,
+      "grad_norm": 1.6375108891969241,
+      "learning_rate": 7.701289518647506e-05,
+      "loss": 1.4069,
+      "step": 12734
+    },
+    {
+      "epoch": 0.5866771087667573,
+      "grad_norm": 1.154449467328008,
+      "learning_rate": 7.699837412089928e-05,
+      "loss": 1.2532,
+      "step": 12735
+    },
+    {
+      "epoch": 0.5867231768553922,
+      "grad_norm": 1.2676715897658826,
+      "learning_rate": 7.698385356741524e-05,
+      "loss": 1.429,
+      "step": 12736
+    },
+    {
+      "epoch": 0.5867692449440273,
+      "grad_norm": 1.3384178407050857,
+      "learning_rate": 7.696933352634626e-05,
+      "loss": 1.1586,
+      "step": 12737
+    },
+    {
+      "epoch": 0.5868153130326623,
+      "grad_norm": 0.8149936663543096,
+      "learning_rate": 7.695481399801557e-05,
+      "loss": 1.3104,
+      "step": 12738
+    },
+    {
+      "epoch": 0.5868613811212973,
+      "grad_norm": 1.618078842982936,
+      "learning_rate": 7.694029498274643e-05,
+      "loss": 1.5005,
+      "step": 12739
+    },
+    {
+      "epoch": 0.5869074492099323,
+      "grad_norm": 0.9276609811565183,
+      "learning_rate": 7.692577648086208e-05,
+      "loss": 1.1792,
+      "step": 12740
+    },
+    {
+      "epoch": 0.5869535172985673,
+      "grad_norm": 1.7722061636473525,
+      "learning_rate": 7.691125849268576e-05,
+      "loss": 1.3733,
+      "step": 12741
+    },
+    {
+      "epoch": 0.5869995853872023,
+      "grad_norm": 1.683759481810189,
+      "learning_rate": 7.689674101854068e-05,
+      "loss": 1.285,
+      "step": 12742
+    },
+    {
+      "epoch": 0.5870456534758373,
+      "grad_norm": 1.5760958275581858,
+      "learning_rate": 7.688222405875003e-05,
+      "loss": 0.967,
+      "step": 12743
+    },
+    {
+      "epoch": 0.5870917215644723,
+      "grad_norm": 0.9678274530281733,
+      "learning_rate": 7.686770761363705e-05,
+      "loss": 1.2335,
+      "step": 12744
+    },
+    {
+      "epoch": 0.5871377896531073,
+      "grad_norm": 1.2358501702831122,
+      "learning_rate": 7.685319168352488e-05,
+      "loss": 1.2347,
+      "step": 12745
+    },
+    {
+      "epoch": 0.5871838577417423,
+      "grad_norm": 1.7138824222042726,
+      "learning_rate": 7.683867626873671e-05,
+      "loss": 1.2277,
+      "step": 12746
+    },
+    {
+      "epoch": 0.5872299258303773,
+      "grad_norm": 0.8501120104932286,
+      "learning_rate": 7.68241613695957e-05,
+      "loss": 1.418,
+      "step": 12747
+    },
+    {
+      "epoch": 0.5872759939190123,
+      "grad_norm": 0.7810106741416013,
+      "learning_rate": 7.680964698642501e-05,
+      "loss": 1.4075,
+      "step": 12748
+    },
+    {
+      "epoch": 0.5873220620076472,
+      "grad_norm": 2.4956947238395,
+      "learning_rate": 7.679513311954777e-05,
+      "loss": 1.4593,
+      "step": 12749
+    },
+    {
+      "epoch": 0.5873681300962823,
+      "grad_norm": 0.894134612176068,
+      "learning_rate": 7.678061976928707e-05,
+      "loss": 1.4173,
+      "step": 12750
+    },
+    {
+      "epoch": 0.5874141981849174,
+      "grad_norm": 1.0451370159462336,
+      "learning_rate": 7.676610693596611e-05,
+      "loss": 1.4787,
+      "step": 12751
+    },
+    {
+      "epoch": 0.5874602662735523,
+      "grad_norm": 1.263484266640038,
+      "learning_rate": 7.67515946199079e-05,
+      "loss": 1.1304,
+      "step": 12752
+    },
+    {
+      "epoch": 0.5875063343621874,
+      "grad_norm": 1.302066076749168,
+      "learning_rate": 7.673708282143559e-05,
+      "loss": 1.081,
+      "step": 12753
+    },
+    {
+      "epoch": 0.5875524024508223,
+      "grad_norm": 1.7075571503721167,
+      "learning_rate": 7.672257154087222e-05,
+      "loss": 1.0044,
+      "step": 12754
+    },
+    {
+      "epoch": 0.5875984705394574,
+      "grad_norm": 3.0096081757568136,
+      "learning_rate": 7.670806077854092e-05,
+      "loss": 1.566,
+      "step": 12755
+    },
+    {
+      "epoch": 0.5876445386280923,
+      "grad_norm": 0.9320129958388962,
+      "learning_rate": 7.669355053476467e-05,
+      "loss": 1.2347,
+      "step": 12756
+    },
+    {
+      "epoch": 0.5876906067167273,
+      "grad_norm": 0.9275062987434632,
+      "learning_rate": 7.667904080986658e-05,
+      "loss": 1.5232,
+      "step": 12757
+    },
+    {
+      "epoch": 0.5877366748053623,
+      "grad_norm": 1.609798273456736,
+      "learning_rate": 7.666453160416967e-05,
+      "loss": 1.1961,
+      "step": 12758
+    },
+    {
+      "epoch": 0.5877827428939973,
+      "grad_norm": 1.3897931966016985,
+      "learning_rate": 7.665002291799693e-05,
+      "loss": 1.0769,
+      "step": 12759
+    },
+    {
+      "epoch": 0.5878288109826323,
+      "grad_norm": 23.56432839698702,
+      "learning_rate": 7.663551475167141e-05,
+      "loss": 1.0603,
+      "step": 12760
+    },
+    {
+      "epoch": 0.5878748790712673,
+      "grad_norm": 1.5966391672347682,
+      "learning_rate": 7.662100710551607e-05,
+      "loss": 1.2707,
+      "step": 12761
+    },
+    {
+      "epoch": 0.5879209471599023,
+      "grad_norm": 0.922552515227778,
+      "learning_rate": 7.660649997985395e-05,
+      "loss": 1.4945,
+      "step": 12762
+    },
+    {
+      "epoch": 0.5879670152485373,
+      "grad_norm": 0.7936407907219376,
+      "learning_rate": 7.659199337500797e-05,
+      "loss": 1.4437,
+      "step": 12763
+    },
+    {
+      "epoch": 0.5880130833371724,
+      "grad_norm": 1.1669038687714897,
+      "learning_rate": 7.657748729130114e-05,
+      "loss": 1.2448,
+      "step": 12764
+    },
+    {
+      "epoch": 0.5880591514258073,
+      "grad_norm": 3.2141694652093107,
+      "learning_rate": 7.656298172905638e-05,
+      "loss": 1.3081,
+      "step": 12765
+    },
+    {
+      "epoch": 0.5881052195144424,
+      "grad_norm": 1.0548678913833232,
+      "learning_rate": 7.654847668859666e-05,
+      "loss": 1.6549,
+      "step": 12766
+    },
+    {
+      "epoch": 0.5881512876030773,
+      "grad_norm": 1.0154178569995655,
+      "learning_rate": 7.653397217024487e-05,
+      "loss": 1.486,
+      "step": 12767
+    },
+    {
+      "epoch": 0.5881973556917124,
+      "grad_norm": 2.5352368448794174,
+      "learning_rate": 7.651946817432398e-05,
+      "loss": 1.5418,
+      "step": 12768
+    },
+    {
+      "epoch": 0.5882434237803473,
+      "grad_norm": 0.7859248984413847,
+      "learning_rate": 7.650496470115689e-05,
+      "loss": 1.558,
+      "step": 12769
+    },
+    {
+      "epoch": 0.5882894918689824,
+      "grad_norm": 1.2118144939059459,
+      "learning_rate": 7.649046175106643e-05,
+      "loss": 1.2884,
+      "step": 12770
+    },
+    {
+      "epoch": 0.5883355599576173,
+      "grad_norm": 1.3120399322453984,
+      "learning_rate": 7.647595932437556e-05,
+      "loss": 1.2435,
+      "step": 12771
+    },
+    {
+      "epoch": 0.5883816280462524,
+      "grad_norm": 1.8537425066223256,
+      "learning_rate": 7.646145742140709e-05,
+      "loss": 1.2051,
+      "step": 12772
+    },
+    {
+      "epoch": 0.5884276961348873,
+      "grad_norm": 1.113724603009303,
+      "learning_rate": 7.644695604248394e-05,
+      "loss": 1.4521,
+      "step": 12773
+    },
+    {
+      "epoch": 0.5884737642235224,
+      "grad_norm": 0.7508859564960019,
+      "learning_rate": 7.643245518792891e-05,
+      "loss": 1.5316,
+      "step": 12774
+    },
+    {
+      "epoch": 0.5885198323121573,
+      "grad_norm": 1.0524245458468602,
+      "learning_rate": 7.641795485806486e-05,
+      "loss": 1.2947,
+      "step": 12775
+    },
+    {
+      "epoch": 0.5885659004007924,
+      "grad_norm": 2.343023085841893,
+      "learning_rate": 7.640345505321461e-05,
+      "loss": 1.3785,
+      "step": 12776
+    },
+    {
+      "epoch": 0.5886119684894274,
+      "grad_norm": 1.2600793208730647,
+      "learning_rate": 7.638895577370099e-05,
+      "loss": 1.4534,
+      "step": 12777
+    },
+    {
+      "epoch": 0.5886580365780624,
+      "grad_norm": 1.5428388132095356,
+      "learning_rate": 7.63744570198468e-05,
+      "loss": 1.3322,
+      "step": 12778
+    },
+    {
+      "epoch": 0.5887041046666974,
+      "grad_norm": 1.0522080984135442,
+      "learning_rate": 7.635995879197476e-05,
+      "loss": 1.4072,
+      "step": 12779
+    },
+    {
+      "epoch": 0.5887501727553324,
+      "grad_norm": 0.9557328650437036,
+      "learning_rate": 7.634546109040776e-05,
+      "loss": 1.4659,
+      "step": 12780
+    },
+    {
+      "epoch": 0.5887962408439674,
+      "grad_norm": 1.0366721338968,
+      "learning_rate": 7.633096391546846e-05,
+      "loss": 1.3218,
+      "step": 12781
+    },
+    {
+      "epoch": 0.5888423089326024,
+      "grad_norm": 1.4115893272018154,
+      "learning_rate": 7.631646726747968e-05,
+      "loss": 1.1413,
+      "step": 12782
+    },
+    {
+      "epoch": 0.5888883770212374,
+      "grad_norm": 1.0624415504128903,
+      "learning_rate": 7.630197114676415e-05,
+      "loss": 1.2391,
+      "step": 12783
+    },
+    {
+      "epoch": 0.5889344451098724,
+      "grad_norm": 1.1977794606806507,
+      "learning_rate": 7.628747555364461e-05,
+      "loss": 1.2281,
+      "step": 12784
+    },
+    {
+      "epoch": 0.5889805131985074,
+      "grad_norm": 1.0158189999086837,
+      "learning_rate": 7.627298048844376e-05,
+      "loss": 1.1225,
+      "step": 12785
+    },
+    {
+      "epoch": 0.5890265812871424,
+      "grad_norm": 1.136617117957974,
+      "learning_rate": 7.625848595148432e-05,
+      "loss": 1.436,
+      "step": 12786
+    },
+    {
+      "epoch": 0.5890726493757774,
+      "grad_norm": 1.2289746779885913,
+      "learning_rate": 7.624399194308901e-05,
+      "loss": 1.2857,
+      "step": 12787
+    },
+    {
+      "epoch": 0.5891187174644124,
+      "grad_norm": 0.8190353953531592,
+      "learning_rate": 7.622949846358043e-05,
+      "loss": 1.3924,
+      "step": 12788
+    },
+    {
+      "epoch": 0.5891647855530474,
+      "grad_norm": 1.350550214664251,
+      "learning_rate": 7.621500551328136e-05,
+      "loss": 1.3701,
+      "step": 12789
+    },
+    {
+      "epoch": 0.5892108536416824,
+      "grad_norm": 0.9689319844039332,
+      "learning_rate": 7.620051309251435e-05,
+      "loss": 1.3599,
+      "step": 12790
+    },
+    {
+      "epoch": 0.5892569217303174,
+      "grad_norm": 1.116456555160551,
+      "learning_rate": 7.618602120160219e-05,
+      "loss": 1.3132,
+      "step": 12791
+    },
+    {
+      "epoch": 0.5893029898189525,
+      "grad_norm": 0.8756540140743913,
+      "learning_rate": 7.617152984086735e-05,
+      "loss": 1.3432,
+      "step": 12792
+    },
+    {
+      "epoch": 0.5893490579075874,
+      "grad_norm": 1.1235784262574784,
+      "learning_rate": 7.615703901063261e-05,
+      "loss": 1.2563,
+      "step": 12793
+    },
+    {
+      "epoch": 0.5893951259962225,
+      "grad_norm": 1.191124019143287,
+      "learning_rate": 7.614254871122045e-05,
+      "loss": 1.3845,
+      "step": 12794
+    },
+    {
+      "epoch": 0.5894411940848574,
+      "grad_norm": 0.8605458327099575,
+      "learning_rate": 7.612805894295359e-05,
+      "loss": 1.5493,
+      "step": 12795
+    },
+    {
+      "epoch": 0.5894872621734925,
+      "grad_norm": 1.263009811152541,
+      "learning_rate": 7.611356970615457e-05,
+      "loss": 0.9184,
+      "step": 12796
+    },
+    {
+      "epoch": 0.5895333302621274,
+      "grad_norm": 1.3428233623705401,
+      "learning_rate": 7.609908100114591e-05,
+      "loss": 1.3965,
+      "step": 12797
+    },
+    {
+      "epoch": 0.5895793983507625,
+      "grad_norm": 1.8146697529990694,
+      "learning_rate": 7.608459282825029e-05,
+      "loss": 1.4946,
+      "step": 12798
+    },
+    {
+      "epoch": 0.5896254664393974,
+      "grad_norm": 0.8668763230154719,
+      "learning_rate": 7.607010518779014e-05,
+      "loss": 1.3065,
+      "step": 12799
+    },
+    {
+      "epoch": 0.5896715345280324,
+      "grad_norm": 1.3045419348188134,
+      "learning_rate": 7.605561808008814e-05,
+      "loss": 1.2859,
+      "step": 12800
+    },
+    {
+      "epoch": 0.5897176026166674,
+      "grad_norm": 0.945107030681302,
+      "learning_rate": 7.604113150546667e-05,
+      "loss": 1.5119,
+      "step": 12801
+    },
+    {
+      "epoch": 0.5897636707053024,
+      "grad_norm": 0.8143995656894908,
+      "learning_rate": 7.602664546424841e-05,
+      "loss": 1.3218,
+      "step": 12802
+    },
+    {
+      "epoch": 0.5898097387939374,
+      "grad_norm": 0.93166332536323,
+      "learning_rate": 7.60121599567557e-05,
+      "loss": 1.34,
+      "step": 12803
+    },
+    {
+      "epoch": 0.5898558068825724,
+      "grad_norm": 1.6949682856356179,
+      "learning_rate": 7.59976749833112e-05,
+      "loss": 1.3761,
+      "step": 12804
+    },
+    {
+      "epoch": 0.5899018749712075,
+      "grad_norm": 1.7456032235013559,
+      "learning_rate": 7.598319054423728e-05,
+      "loss": 1.3703,
+      "step": 12805
+    },
+    {
+      "epoch": 0.5899479430598424,
+      "grad_norm": 1.157502539106094,
+      "learning_rate": 7.596870663985643e-05,
+      "loss": 1.3988,
+      "step": 12806
+    },
+    {
+      "epoch": 0.5899940111484775,
+      "grad_norm": 0.8398195133018597,
+      "learning_rate": 7.595422327049114e-05,
+      "loss": 1.3039,
+      "step": 12807
+    },
+    {
+      "epoch": 0.5900400792371124,
+      "grad_norm": 1.2503847056993174,
+      "learning_rate": 7.593974043646378e-05,
+      "loss": 1.3224,
+      "step": 12808
+    },
+    {
+      "epoch": 0.5900861473257475,
+      "grad_norm": 1.250573333819432,
+      "learning_rate": 7.592525813809692e-05,
+      "loss": 1.4385,
+      "step": 12809
+    },
+    {
+      "epoch": 0.5901322154143824,
+      "grad_norm": 0.9265727502358104,
+      "learning_rate": 7.591077637571285e-05,
+      "loss": 1.2898,
+      "step": 12810
+    },
+    {
+      "epoch": 0.5901782835030175,
+      "grad_norm": 9.630992130508451,
+      "learning_rate": 7.589629514963408e-05,
+      "loss": 1.3549,
+      "step": 12811
+    },
+    {
+      "epoch": 0.5902243515916524,
+      "grad_norm": 1.6821395745605077,
+      "learning_rate": 7.588181446018293e-05,
+      "loss": 1.2413,
+      "step": 12812
+    },
+    {
+      "epoch": 0.5902704196802875,
+      "grad_norm": 1.2807122845882715,
+      "learning_rate": 7.586733430768189e-05,
+      "loss": 1.2122,
+      "step": 12813
+    },
+    {
+      "epoch": 0.5903164877689224,
+      "grad_norm": 5.609136550386813,
+      "learning_rate": 7.585285469245325e-05,
+      "loss": 1.2517,
+      "step": 12814
+    },
+    {
+      "epoch": 0.5903625558575575,
+      "grad_norm": 0.850971938129049,
+      "learning_rate": 7.583837561481937e-05,
+      "loss": 1.596,
+      "step": 12815
+    },
+    {
+      "epoch": 0.5904086239461924,
+      "grad_norm": 0.7843935829291006,
+      "learning_rate": 7.582389707510265e-05,
+      "loss": 1.4294,
+      "step": 12816
+    },
+    {
+      "epoch": 0.5904546920348275,
+      "grad_norm": 1.2185320797219108,
+      "learning_rate": 7.580941907362539e-05,
+      "loss": 1.1657,
+      "step": 12817
+    },
+    {
+      "epoch": 0.5905007601234625,
+      "grad_norm": 1.016279390875634,
+      "learning_rate": 7.579494161070995e-05,
+      "loss": 1.2381,
+      "step": 12818
+    },
+    {
+      "epoch": 0.5905468282120975,
+      "grad_norm": 1.0822231237510975,
+      "learning_rate": 7.578046468667862e-05,
+      "loss": 1.539,
+      "step": 12819
+    },
+    {
+      "epoch": 0.5905928963007325,
+      "grad_norm": 3.5732101657709854,
+      "learning_rate": 7.576598830185373e-05,
+      "loss": 1.6092,
+      "step": 12820
+    },
+    {
+      "epoch": 0.5906389643893675,
+      "grad_norm": 1.4372742157183518,
+      "learning_rate": 7.575151245655752e-05,
+      "loss": 1.0794,
+      "step": 12821
+    },
+    {
+      "epoch": 0.5906850324780025,
+      "grad_norm": 1.2521366458932064,
+      "learning_rate": 7.573703715111238e-05,
+      "loss": 1.415,
+      "step": 12822
+    },
+    {
+      "epoch": 0.5907311005666375,
+      "grad_norm": 0.9400554865369007,
+      "learning_rate": 7.572256238584048e-05,
+      "loss": 1.2188,
+      "step": 12823
+    },
+    {
+      "epoch": 0.5907771686552725,
+      "grad_norm": 1.0382665008973246,
+      "learning_rate": 7.570808816106408e-05,
+      "loss": 1.5844,
+      "step": 12824
+    },
+    {
+      "epoch": 0.5908232367439075,
+      "grad_norm": 1.2766262837822067,
+      "learning_rate": 7.569361447710545e-05,
+      "loss": 1.3666,
+      "step": 12825
+    },
+    {
+      "epoch": 0.5908693048325425,
+      "grad_norm": 1.0261625549789808,
+      "learning_rate": 7.567914133428682e-05,
+      "loss": 1.1693,
+      "step": 12826
+    },
+    {
+      "epoch": 0.5909153729211775,
+      "grad_norm": 1.0109695043703968,
+      "learning_rate": 7.566466873293041e-05,
+      "loss": 1.5027,
+      "step": 12827
+    },
+    {
+      "epoch": 0.5909614410098125,
+      "grad_norm": 1.2741967044795697,
+      "learning_rate": 7.565019667335841e-05,
+      "loss": 1.2487,
+      "step": 12828
+    },
+    {
+      "epoch": 0.5910075090984475,
+      "grad_norm": 0.9656212193997781,
+      "learning_rate": 7.563572515589306e-05,
+      "loss": 1.4128,
+      "step": 12829
+    },
+    {
+      "epoch": 0.5910535771870825,
+      "grad_norm": 1.8244353232603736,
+      "learning_rate": 7.562125418085648e-05,
+      "loss": 1.2038,
+      "step": 12830
+    },
+    {
+      "epoch": 0.5910996452757175,
+      "grad_norm": 1.1173810682672811,
+      "learning_rate": 7.560678374857089e-05,
+      "loss": 1.537,
+      "step": 12831
+    },
+    {
+      "epoch": 0.5911457133643525,
+      "grad_norm": 1.2933477180607666,
+      "learning_rate": 7.559231385935846e-05,
+      "loss": 1.1855,
+      "step": 12832
+    },
+    {
+      "epoch": 0.5911917814529876,
+      "grad_norm": 1.2088925654315374,
+      "learning_rate": 7.557784451354128e-05,
+      "loss": 1.7068,
+      "step": 12833
+    },
+    {
+      "epoch": 0.5912378495416225,
+      "grad_norm": 0.7543344663551503,
+      "learning_rate": 7.556337571144153e-05,
+      "loss": 1.2572,
+      "step": 12834
+    },
+    {
+      "epoch": 0.5912839176302576,
+      "grad_norm": 1.4697819515504622,
+      "learning_rate": 7.554890745338131e-05,
+      "loss": 1.6836,
+      "step": 12835
+    },
+    {
+      "epoch": 0.5913299857188925,
+      "grad_norm": 1.3109931246881796,
+      "learning_rate": 7.553443973968275e-05,
+      "loss": 1.5544,
+      "step": 12836
+    },
+    {
+      "epoch": 0.5913760538075276,
+      "grad_norm": 2.825211147586158,
+      "learning_rate": 7.551997257066791e-05,
+      "loss": 1.3835,
+      "step": 12837
+    },
+    {
+      "epoch": 0.5914221218961625,
+      "grad_norm": 1.6075042579945942,
+      "learning_rate": 7.550550594665895e-05,
+      "loss": 1.2967,
+      "step": 12838
+    },
+    {
+      "epoch": 0.5914681899847976,
+      "grad_norm": 1.3434727027306346,
+      "learning_rate": 7.549103986797787e-05,
+      "loss": 1.3988,
+      "step": 12839
+    },
+    {
+      "epoch": 0.5915142580734325,
+      "grad_norm": 2.2668011970079527,
+      "learning_rate": 7.547657433494678e-05,
+      "loss": 1.2915,
+      "step": 12840
+    },
+    {
+      "epoch": 0.5915603261620676,
+      "grad_norm": 1.0556815577645284,
+      "learning_rate": 7.546210934788772e-05,
+      "loss": 1.5306,
+      "step": 12841
+    },
+    {
+      "epoch": 0.5916063942507025,
+      "grad_norm": 0.7576892502196663,
+      "learning_rate": 7.544764490712271e-05,
+      "loss": 1.1133,
+      "step": 12842
+    },
+    {
+      "epoch": 0.5916524623393375,
+      "grad_norm": 1.1353189474470518,
+      "learning_rate": 7.543318101297378e-05,
+      "loss": 1.5928,
+      "step": 12843
+    },
+    {
+      "epoch": 0.5916985304279725,
+      "grad_norm": 1.2638489024237525,
+      "learning_rate": 7.541871766576296e-05,
+      "loss": 1.0943,
+      "step": 12844
+    },
+    {
+      "epoch": 0.5917445985166075,
+      "grad_norm": 1.2095137834334018,
+      "learning_rate": 7.540425486581226e-05,
+      "loss": 0.9843,
+      "step": 12845
+    },
+    {
+      "epoch": 0.5917906666052426,
+      "grad_norm": 1.3414866771692313,
+      "learning_rate": 7.538979261344362e-05,
+      "loss": 1.3896,
+      "step": 12846
+    },
+    {
+      "epoch": 0.5918367346938775,
+      "grad_norm": 1.3039841982810583,
+      "learning_rate": 7.537533090897908e-05,
+      "loss": 1.459,
+      "step": 12847
+    },
+    {
+      "epoch": 0.5918828027825126,
+      "grad_norm": 1.5226116680747341,
+      "learning_rate": 7.536086975274054e-05,
+      "loss": 1.0302,
+      "step": 12848
+    },
+    {
+      "epoch": 0.5919288708711475,
+      "grad_norm": 1.0138221598155177,
+      "learning_rate": 7.534640914505004e-05,
+      "loss": 1.1595,
+      "step": 12849
+    },
+    {
+      "epoch": 0.5919749389597826,
+      "grad_norm": 3.434020992867892,
+      "learning_rate": 7.533194908622943e-05,
+      "loss": 1.3178,
+      "step": 12850
+    },
+    {
+      "epoch": 0.5920210070484175,
+      "grad_norm": 1.2235090368806016,
+      "learning_rate": 7.531748957660069e-05,
+      "loss": 1.4361,
+      "step": 12851
+    },
+    {
+      "epoch": 0.5920670751370526,
+      "grad_norm": 1.1620643825947066,
+      "learning_rate": 7.530303061648571e-05,
+      "loss": 1.4136,
+      "step": 12852
+    },
+    {
+      "epoch": 0.5921131432256875,
+      "grad_norm": 3.229419636817809,
+      "learning_rate": 7.528857220620642e-05,
+      "loss": 1.29,
+      "step": 12853
+    },
+    {
+      "epoch": 0.5921592113143226,
+      "grad_norm": 1.1463921564985813,
+      "learning_rate": 7.52741143460847e-05,
+      "loss": 1.3505,
+      "step": 12854
+    },
+    {
+      "epoch": 0.5922052794029575,
+      "grad_norm": 1.9276735343405287,
+      "learning_rate": 7.525965703644242e-05,
+      "loss": 1.6154,
+      "step": 12855
+    },
+    {
+      "epoch": 0.5922513474915926,
+      "grad_norm": 1.382513985629884,
+      "learning_rate": 7.524520027760146e-05,
+      "loss": 1.1057,
+      "step": 12856
+    },
+    {
+      "epoch": 0.5922974155802275,
+      "grad_norm": 1.2594859970407521,
+      "learning_rate": 7.523074406988367e-05,
+      "loss": 1.3332,
+      "step": 12857
+    },
+    {
+      "epoch": 0.5923434836688626,
+      "grad_norm": 1.1092678979030925,
+      "learning_rate": 7.521628841361088e-05,
+      "loss": 1.4279,
+      "step": 12858
+    },
+    {
+      "epoch": 0.5923895517574976,
+      "grad_norm": 0.9074954943244617,
+      "learning_rate": 7.520183330910496e-05,
+      "loss": 1.2709,
+      "step": 12859
+    },
+    {
+      "epoch": 0.5924356198461326,
+      "grad_norm": 1.7795534920311915,
+      "learning_rate": 7.518737875668769e-05,
+      "loss": 1.1934,
+      "step": 12860
+    },
+    {
+      "epoch": 0.5924816879347676,
+      "grad_norm": 2.0185377103634887,
+      "learning_rate": 7.517292475668089e-05,
+      "loss": 1.2549,
+      "step": 12861
+    },
+    {
+      "epoch": 0.5925277560234026,
+      "grad_norm": 0.745930877708851,
+      "learning_rate": 7.515847130940633e-05,
+      "loss": 1.1829,
+      "step": 12862
+    },
+    {
+      "epoch": 0.5925738241120376,
+      "grad_norm": 1.1599759508604628,
+      "learning_rate": 7.514401841518584e-05,
+      "loss": 1.0393,
+      "step": 12863
+    },
+    {
+      "epoch": 0.5926198922006726,
+      "grad_norm": 1.044000867028745,
+      "learning_rate": 7.512956607434113e-05,
+      "loss": 1.3118,
+      "step": 12864
+    },
+    {
+      "epoch": 0.5926659602893076,
+      "grad_norm": 1.162377696840393,
+      "learning_rate": 7.511511428719402e-05,
+      "loss": 1.5849,
+      "step": 12865
+    },
+    {
+      "epoch": 0.5927120283779426,
+      "grad_norm": 2.4956412476546617,
+      "learning_rate": 7.51006630540662e-05,
+      "loss": 1.3195,
+      "step": 12866
+    },
+    {
+      "epoch": 0.5927580964665776,
+      "grad_norm": 1.4205626374801885,
+      "learning_rate": 7.508621237527944e-05,
+      "loss": 1.5097,
+      "step": 12867
+    },
+    {
+      "epoch": 0.5928041645552126,
+      "grad_norm": 1.1309603570601612,
+      "learning_rate": 7.507176225115544e-05,
+      "loss": 1.131,
+      "step": 12868
+    },
+    {
+      "epoch": 0.5928502326438476,
+      "grad_norm": 0.8035476579793129,
+      "learning_rate": 7.50573126820159e-05,
+      "loss": 1.1338,
+      "step": 12869
+    },
+    {
+      "epoch": 0.5928963007324826,
+      "grad_norm": 0.8556873720646706,
+      "learning_rate": 7.504286366818253e-05,
+      "loss": 1.5088,
+      "step": 12870
+    },
+    {
+      "epoch": 0.5929423688211176,
+      "grad_norm": 1.2802492688756464,
+      "learning_rate": 7.502841520997701e-05,
+      "loss": 1.2731,
+      "step": 12871
+    },
+    {
+      "epoch": 0.5929884369097527,
+      "grad_norm": 0.9110397796239688,
+      "learning_rate": 7.501396730772102e-05,
+      "loss": 1.4908,
+      "step": 12872
+    },
+    {
+      "epoch": 0.5930345049983876,
+      "grad_norm": 1.4244342429948569,
+      "learning_rate": 7.499951996173619e-05,
+      "loss": 1.2861,
+      "step": 12873
+    },
+    {
+      "epoch": 0.5930805730870227,
+      "grad_norm": 1.0709468021320616,
+      "learning_rate": 7.498507317234419e-05,
+      "loss": 1.3115,
+      "step": 12874
+    },
+    {
+      "epoch": 0.5931266411756576,
+      "grad_norm": 1.6310994843840296,
+      "learning_rate": 7.497062693986663e-05,
+      "loss": 1.1477,
+      "step": 12875
+    },
+    {
+      "epoch": 0.5931727092642927,
+      "grad_norm": 1.4062586926675278,
+      "learning_rate": 7.495618126462515e-05,
+      "loss": 1.2696,
+      "step": 12876
+    },
+    {
+      "epoch": 0.5932187773529276,
+      "grad_norm": 1.6703726124359368,
+      "learning_rate": 7.49417361469414e-05,
+      "loss": 1.4375,
+      "step": 12877
+    },
+    {
+      "epoch": 0.5932648454415627,
+      "grad_norm": 2.312234114323087,
+      "learning_rate": 7.492729158713686e-05,
+      "loss": 1.1054,
+      "step": 12878
+    },
+    {
+      "epoch": 0.5933109135301976,
+      "grad_norm": 1.697540440031479,
+      "learning_rate": 7.491284758553324e-05,
+      "loss": 1.2615,
+      "step": 12879
+    },
+    {
+      "epoch": 0.5933569816188327,
+      "grad_norm": 2.3640752406964936,
+      "learning_rate": 7.4898404142452e-05,
+      "loss": 1.3912,
+      "step": 12880
+    },
+    {
+      "epoch": 0.5934030497074676,
+      "grad_norm": 1.3764726855723581,
+      "learning_rate": 7.488396125821477e-05,
+      "loss": 1.0237,
+      "step": 12881
+    },
+    {
+      "epoch": 0.5934491177961027,
+      "grad_norm": 1.292921137967115,
+      "learning_rate": 7.486951893314308e-05,
+      "loss": 1.2838,
+      "step": 12882
+    },
+    {
+      "epoch": 0.5934951858847376,
+      "grad_norm": 1.1149467350450644,
+      "learning_rate": 7.485507716755848e-05,
+      "loss": 1.2409,
+      "step": 12883
+    },
+    {
+      "epoch": 0.5935412539733727,
+      "grad_norm": 0.9139819892457134,
+      "learning_rate": 7.484063596178243e-05,
+      "loss": 1.5591,
+      "step": 12884
+    },
+    {
+      "epoch": 0.5935873220620076,
+      "grad_norm": 0.9164809170547302,
+      "learning_rate": 7.482619531613653e-05,
+      "loss": 1.3414,
+      "step": 12885
+    },
+    {
+      "epoch": 0.5936333901506426,
+      "grad_norm": 0.7934362752237134,
+      "learning_rate": 7.48117552309422e-05,
+      "loss": 1.5022,
+      "step": 12886
+    },
+    {
+      "epoch": 0.5936794582392777,
+      "grad_norm": 2.192791829458393,
+      "learning_rate": 7.479731570652095e-05,
+      "loss": 1.3737,
+      "step": 12887
+    },
+    {
+      "epoch": 0.5937255263279126,
+      "grad_norm": 1.100663302601042,
+      "learning_rate": 7.47828767431943e-05,
+      "loss": 1.6168,
+      "step": 12888
+    },
+    {
+      "epoch": 0.5937715944165477,
+      "grad_norm": 1.0162370086307126,
+      "learning_rate": 7.476843834128359e-05,
+      "loss": 1.4479,
+      "step": 12889
+    },
+    {
+      "epoch": 0.5938176625051826,
+      "grad_norm": 0.92797274211966,
+      "learning_rate": 7.47540005011104e-05,
+      "loss": 1.5165,
+      "step": 12890
+    },
+    {
+      "epoch": 0.5938637305938177,
+      "grad_norm": 0.9299238016802013,
+      "learning_rate": 7.473956322299602e-05,
+      "loss": 1.2081,
+      "step": 12891
+    },
+    {
+      "epoch": 0.5939097986824526,
+      "grad_norm": 0.7540140892311099,
+      "learning_rate": 7.472512650726205e-05,
+      "loss": 1.5216,
+      "step": 12892
+    },
+    {
+      "epoch": 0.5939558667710877,
+      "grad_norm": 1.4296958484420634,
+      "learning_rate": 7.471069035422971e-05,
+      "loss": 1.39,
+      "step": 12893
+    },
+    {
+      "epoch": 0.5940019348597226,
+      "grad_norm": 0.9190140162599177,
+      "learning_rate": 7.469625476422055e-05,
+      "loss": 1.2041,
+      "step": 12894
+    },
+    {
+      "epoch": 0.5940480029483577,
+      "grad_norm": 1.3165398356087175,
+      "learning_rate": 7.468181973755586e-05,
+      "loss": 1.1804,
+      "step": 12895
+    },
+    {
+      "epoch": 0.5940940710369926,
+      "grad_norm": 1.163614021181773,
+      "learning_rate": 7.466738527455706e-05,
+      "loss": 1.613,
+      "step": 12896
+    },
+    {
+      "epoch": 0.5941401391256277,
+      "grad_norm": 1.2808914990492264,
+      "learning_rate": 7.465295137554551e-05,
+      "loss": 1.3662,
+      "step": 12897
+    },
+    {
+      "epoch": 0.5941862072142626,
+      "grad_norm": 2.69157409085995,
+      "learning_rate": 7.463851804084248e-05,
+      "loss": 1.1795,
+      "step": 12898
+    },
+    {
+      "epoch": 0.5942322753028977,
+      "grad_norm": 1.442703088607821,
+      "learning_rate": 7.462408527076943e-05,
+      "loss": 1.3052,
+      "step": 12899
+    },
+    {
+      "epoch": 0.5942783433915327,
+      "grad_norm": 1.3029622687090698,
+      "learning_rate": 7.460965306564753e-05,
+      "loss": 1.2909,
+      "step": 12900
+    },
+    {
+      "epoch": 0.5943244114801677,
+      "grad_norm": 2.526674014720588,
+      "learning_rate": 7.459522142579824e-05,
+      "loss": 1.1169,
+      "step": 12901
+    },
+    {
+      "epoch": 0.5943704795688027,
+      "grad_norm": 3.437180365314604,
+      "learning_rate": 7.458079035154272e-05,
+      "loss": 1.1587,
+      "step": 12902
+    },
+    {
+      "epoch": 0.5944165476574377,
+      "grad_norm": 1.6359358687677779,
+      "learning_rate": 7.45663598432024e-05,
+      "loss": 1.3875,
+      "step": 12903
+    },
+    {
+      "epoch": 0.5944626157460727,
+      "grad_norm": 0.9284860909154079,
+      "learning_rate": 7.455192990109838e-05,
+      "loss": 1.0813,
+      "step": 12904
+    },
+    {
+      "epoch": 0.5945086838347077,
+      "grad_norm": 0.8332063105231308,
+      "learning_rate": 7.453750052555208e-05,
+      "loss": 1.6738,
+      "step": 12905
+    },
+    {
+      "epoch": 0.5945547519233427,
+      "grad_norm": 1.4378369302279965,
+      "learning_rate": 7.452307171688466e-05,
+      "loss": 1.3815,
+      "step": 12906
+    },
+    {
+      "epoch": 0.5946008200119777,
+      "grad_norm": 1.2263046588276174,
+      "learning_rate": 7.450864347541732e-05,
+      "loss": 1.3802,
+      "step": 12907
+    },
+    {
+      "epoch": 0.5946468881006127,
+      "grad_norm": 1.8417530993780737,
+      "learning_rate": 7.449421580147137e-05,
+      "loss": 0.9431,
+      "step": 12908
+    },
+    {
+      "epoch": 0.5946929561892477,
+      "grad_norm": 1.5580803560830807,
+      "learning_rate": 7.447978869536794e-05,
+      "loss": 1.4573,
+      "step": 12909
+    },
+    {
+      "epoch": 0.5947390242778827,
+      "grad_norm": 2.8736404462222644,
+      "learning_rate": 7.44653621574283e-05,
+      "loss": 1.4073,
+      "step": 12910
+    },
+    {
+      "epoch": 0.5947850923665177,
+      "grad_norm": 1.6120686819857166,
+      "learning_rate": 7.445093618797352e-05,
+      "loss": 1.3297,
+      "step": 12911
+    },
+    {
+      "epoch": 0.5948311604551527,
+      "grad_norm": 1.2202772689604724,
+      "learning_rate": 7.443651078732492e-05,
+      "loss": 1.2621,
+      "step": 12912
+    },
+    {
+      "epoch": 0.5948772285437878,
+      "grad_norm": 0.8394438364562627,
+      "learning_rate": 7.442208595580348e-05,
+      "loss": 1.2737,
+      "step": 12913
+    },
+    {
+      "epoch": 0.5949232966324227,
+      "grad_norm": 1.7364193187614148,
+      "learning_rate": 7.440766169373054e-05,
+      "loss": 1.2542,
+      "step": 12914
+    },
+    {
+      "epoch": 0.5949693647210578,
+      "grad_norm": 2.214964431027964,
+      "learning_rate": 7.439323800142708e-05,
+      "loss": 1.4111,
+      "step": 12915
+    },
+    {
+      "epoch": 0.5950154328096927,
+      "grad_norm": 1.1729933445656548,
+      "learning_rate": 7.437881487921428e-05,
+      "loss": 1.131,
+      "step": 12916
+    },
+    {
+      "epoch": 0.5950615008983278,
+      "grad_norm": 1.8997321476441882,
+      "learning_rate": 7.436439232741324e-05,
+      "loss": 1.3012,
+      "step": 12917
+    },
+    {
+      "epoch": 0.5951075689869627,
+      "grad_norm": 0.9763921014458722,
+      "learning_rate": 7.434997034634501e-05,
+      "loss": 1.2305,
+      "step": 12918
+    },
+    {
+      "epoch": 0.5951536370755978,
+      "grad_norm": 0.7811172811273605,
+      "learning_rate": 7.433554893633076e-05,
+      "loss": 1.5764,
+      "step": 12919
+    },
+    {
+      "epoch": 0.5951997051642327,
+      "grad_norm": 0.8708324377274425,
+      "learning_rate": 7.432112809769145e-05,
+      "loss": 1.4249,
+      "step": 12920
+    },
+    {
+      "epoch": 0.5952457732528678,
+      "grad_norm": 1.1071628004813014,
+      "learning_rate": 7.430670783074825e-05,
+      "loss": 1.1154,
+      "step": 12921
+    },
+    {
+      "epoch": 0.5952918413415027,
+      "grad_norm": 2.5691526765399253,
+      "learning_rate": 7.42922881358221e-05,
+      "loss": 1.3659,
+      "step": 12922
+    },
+    {
+      "epoch": 0.5953379094301378,
+      "grad_norm": 1.532450343470865,
+      "learning_rate": 7.427786901323414e-05,
+      "loss": 1.3616,
+      "step": 12923
+    },
+    {
+      "epoch": 0.5953839775187727,
+      "grad_norm": 1.2259356235381098,
+      "learning_rate": 7.426345046330528e-05,
+      "loss": 1.205,
+      "step": 12924
+    },
+    {
+      "epoch": 0.5954300456074078,
+      "grad_norm": 2.4290874609065094,
+      "learning_rate": 7.424903248635657e-05,
+      "loss": 1.457,
+      "step": 12925
+    },
+    {
+      "epoch": 0.5954761136960427,
+      "grad_norm": 3.1750113004551186,
+      "learning_rate": 7.4234615082709e-05,
+      "loss": 1.1447,
+      "step": 12926
+    },
+    {
+      "epoch": 0.5955221817846778,
+      "grad_norm": 0.9144070306901151,
+      "learning_rate": 7.422019825268353e-05,
+      "loss": 1.3915,
+      "step": 12927
+    },
+    {
+      "epoch": 0.5955682498733128,
+      "grad_norm": 1.9461729458731543,
+      "learning_rate": 7.420578199660118e-05,
+      "loss": 1.2313,
+      "step": 12928
+    },
+    {
+      "epoch": 0.5956143179619477,
+      "grad_norm": 1.8450522175286486,
+      "learning_rate": 7.419136631478282e-05,
+      "loss": 1.587,
+      "step": 12929
+    },
+    {
+      "epoch": 0.5956603860505828,
+      "grad_norm": 0.9101149937792872,
+      "learning_rate": 7.417695120754948e-05,
+      "loss": 1.2286,
+      "step": 12930
+    },
+    {
+      "epoch": 0.5957064541392177,
+      "grad_norm": 0.9216967396066804,
+      "learning_rate": 7.416253667522202e-05,
+      "loss": 1.5293,
+      "step": 12931
+    },
+    {
+      "epoch": 0.5957525222278528,
+      "grad_norm": 1.3159717590310243,
+      "learning_rate": 7.41481227181214e-05,
+      "loss": 1.4155,
+      "step": 12932
+    },
+    {
+      "epoch": 0.5957985903164877,
+      "grad_norm": 1.1172899130498706,
+      "learning_rate": 7.413370933656848e-05,
+      "loss": 1.2914,
+      "step": 12933
+    },
+    {
+      "epoch": 0.5958446584051228,
+      "grad_norm": 1.1825069903406276,
+      "learning_rate": 7.411929653088416e-05,
+      "loss": 1.0755,
+      "step": 12934
+    },
+    {
+      "epoch": 0.5958907264937577,
+      "grad_norm": 1.5904767080127094,
+      "learning_rate": 7.410488430138935e-05,
+      "loss": 1.2735,
+      "step": 12935
+    },
+    {
+      "epoch": 0.5959367945823928,
+      "grad_norm": 1.959987826676393,
+      "learning_rate": 7.409047264840487e-05,
+      "loss": 1.1649,
+      "step": 12936
+    },
+    {
+      "epoch": 0.5959828626710277,
+      "grad_norm": 2.148045834582729,
+      "learning_rate": 7.407606157225162e-05,
+      "loss": 1.2014,
+      "step": 12937
+    },
+    {
+      "epoch": 0.5960289307596628,
+      "grad_norm": 0.7779943309645317,
+      "learning_rate": 7.406165107325039e-05,
+      "loss": 1.1949,
+      "step": 12938
+    },
+    {
+      "epoch": 0.5960749988482977,
+      "grad_norm": 3.7583151556580012,
+      "learning_rate": 7.404724115172204e-05,
+      "loss": 1.0732,
+      "step": 12939
+    },
+    {
+      "epoch": 0.5961210669369328,
+      "grad_norm": 1.5270715290775216,
+      "learning_rate": 7.403283180798734e-05,
+      "loss": 1.2838,
+      "step": 12940
+    },
+    {
+      "epoch": 0.5961671350255678,
+      "grad_norm": 2.3809868588113368,
+      "learning_rate": 7.401842304236714e-05,
+      "loss": 1.134,
+      "step": 12941
+    },
+    {
+      "epoch": 0.5962132031142028,
+      "grad_norm": 1.2169872877875736,
+      "learning_rate": 7.400401485518221e-05,
+      "loss": 1.3131,
+      "step": 12942
+    },
+    {
+      "epoch": 0.5962592712028378,
+      "grad_norm": 1.0173673212138061,
+      "learning_rate": 7.398960724675327e-05,
+      "loss": 1.1833,
+      "step": 12943
+    },
+    {
+      "epoch": 0.5963053392914728,
+      "grad_norm": 0.8234221636515178,
+      "learning_rate": 7.397520021740117e-05,
+      "loss": 1.3359,
+      "step": 12944
+    },
+    {
+      "epoch": 0.5963514073801078,
+      "grad_norm": 0.9584300454301378,
+      "learning_rate": 7.396079376744658e-05,
+      "loss": 1.3227,
+      "step": 12945
+    },
+    {
+      "epoch": 0.5963974754687428,
+      "grad_norm": 0.9463753604891825,
+      "learning_rate": 7.39463878972103e-05,
+      "loss": 1.0307,
+      "step": 12946
+    },
+    {
+      "epoch": 0.5964435435573778,
+      "grad_norm": 1.2367888687231958,
+      "learning_rate": 7.393198260701299e-05,
+      "loss": 1.1272,
+      "step": 12947
+    },
+    {
+      "epoch": 0.5964896116460128,
+      "grad_norm": 2.7282664879943797,
+      "learning_rate": 7.391757789717543e-05,
+      "loss": 1.3153,
+      "step": 12948
+    },
+    {
+      "epoch": 0.5965356797346478,
+      "grad_norm": 1.0409137980662615,
+      "learning_rate": 7.390317376801824e-05,
+      "loss": 1.1046,
+      "step": 12949
+    },
+    {
+      "epoch": 0.5965817478232828,
+      "grad_norm": 0.8370581317422189,
+      "learning_rate": 7.388877021986215e-05,
+      "loss": 1.3019,
+      "step": 12950
+    },
+    {
+      "epoch": 0.5966278159119178,
+      "grad_norm": 1.157526345411385,
+      "learning_rate": 7.387436725302785e-05,
+      "loss": 1.2457,
+      "step": 12951
+    },
+    {
+      "epoch": 0.5966738840005528,
+      "grad_norm": 1.0467529747513091,
+      "learning_rate": 7.385996486783592e-05,
+      "loss": 1.1446,
+      "step": 12952
+    },
+    {
+      "epoch": 0.5967199520891878,
+      "grad_norm": 1.0919280318493907,
+      "learning_rate": 7.384556306460709e-05,
+      "loss": 1.0794,
+      "step": 12953
+    },
+    {
+      "epoch": 0.5967660201778229,
+      "grad_norm": 0.7500389425127385,
+      "learning_rate": 7.383116184366192e-05,
+      "loss": 1.0202,
+      "step": 12954
+    },
+    {
+      "epoch": 0.5968120882664578,
+      "grad_norm": 1.5620029995315627,
+      "learning_rate": 7.381676120532109e-05,
+      "loss": 1.2411,
+      "step": 12955
+    },
+    {
+      "epoch": 0.5968581563550929,
+      "grad_norm": 1.0104111507114908,
+      "learning_rate": 7.380236114990516e-05,
+      "loss": 1.487,
+      "step": 12956
+    },
+    {
+      "epoch": 0.5969042244437278,
+      "grad_norm": 0.9002601229145066,
+      "learning_rate": 7.378796167773476e-05,
+      "loss": 1.2312,
+      "step": 12957
+    },
+    {
+      "epoch": 0.5969502925323629,
+      "grad_norm": 1.0494619951908002,
+      "learning_rate": 7.377356278913042e-05,
+      "loss": 1.2409,
+      "step": 12958
+    },
+    {
+      "epoch": 0.5969963606209978,
+      "grad_norm": 1.1735307169471068,
+      "learning_rate": 7.375916448441275e-05,
+      "loss": 1.2317,
+      "step": 12959
+    },
+    {
+      "epoch": 0.5970424287096329,
+      "grad_norm": 0.8976837643499902,
+      "learning_rate": 7.374476676390231e-05,
+      "loss": 1.2096,
+      "step": 12960
+    },
+    {
+      "epoch": 0.5970884967982678,
+      "grad_norm": 1.152197914668539,
+      "learning_rate": 7.373036962791958e-05,
+      "loss": 1.2622,
+      "step": 12961
+    },
+    {
+      "epoch": 0.5971345648869029,
+      "grad_norm": 0.7775754258388591,
+      "learning_rate": 7.371597307678516e-05,
+      "loss": 1.6077,
+      "step": 12962
+    },
+    {
+      "epoch": 0.5971806329755378,
+      "grad_norm": 1.0385602512655452,
+      "learning_rate": 7.37015771108195e-05,
+      "loss": 1.2272,
+      "step": 12963
+    },
+    {
+      "epoch": 0.5972267010641729,
+      "grad_norm": 1.1437885667353076,
+      "learning_rate": 7.368718173034316e-05,
+      "loss": 1.2236,
+      "step": 12964
+    },
+    {
+      "epoch": 0.5972727691528078,
+      "grad_norm": 1.590804516274782,
+      "learning_rate": 7.367278693567658e-05,
+      "loss": 1.0625,
+      "step": 12965
+    },
+    {
+      "epoch": 0.5973188372414429,
+      "grad_norm": 1.3705618998928435,
+      "learning_rate": 7.36583927271403e-05,
+      "loss": 1.4407,
+      "step": 12966
+    },
+    {
+      "epoch": 0.5973649053300778,
+      "grad_norm": 2.069184503732537,
+      "learning_rate": 7.36439991050547e-05,
+      "loss": 0.9701,
+      "step": 12967
+    },
+    {
+      "epoch": 0.5974109734187129,
+      "grad_norm": 0.8463974710483819,
+      "learning_rate": 7.362960606974029e-05,
+      "loss": 1.3649,
+      "step": 12968
+    },
+    {
+      "epoch": 0.5974570415073479,
+      "grad_norm": 1.2686200567424748,
+      "learning_rate": 7.361521362151748e-05,
+      "loss": 1.1461,
+      "step": 12969
+    },
+    {
+      "epoch": 0.5975031095959829,
+      "grad_norm": 0.745887890823776,
+      "learning_rate": 7.360082176070668e-05,
+      "loss": 1.5664,
+      "step": 12970
+    },
+    {
+      "epoch": 0.5975491776846179,
+      "grad_norm": 1.0753537472997439,
+      "learning_rate": 7.358643048762834e-05,
+      "loss": 1.3692,
+      "step": 12971
+    },
+    {
+      "epoch": 0.5975952457732528,
+      "grad_norm": 0.9006286121476997,
+      "learning_rate": 7.357203980260282e-05,
+      "loss": 1.1305,
+      "step": 12972
+    },
+    {
+      "epoch": 0.5976413138618879,
+      "grad_norm": 0.8686919384494096,
+      "learning_rate": 7.355764970595054e-05,
+      "loss": 1.3807,
+      "step": 12973
+    },
+    {
+      "epoch": 0.5976873819505228,
+      "grad_norm": 1.9354224739010215,
+      "learning_rate": 7.354326019799183e-05,
+      "loss": 1.472,
+      "step": 12974
+    },
+    {
+      "epoch": 0.5977334500391579,
+      "grad_norm": 0.8727117763336001,
+      "learning_rate": 7.352887127904708e-05,
+      "loss": 1.3113,
+      "step": 12975
+    },
+    {
+      "epoch": 0.5977795181277928,
+      "grad_norm": 1.1822134699355933,
+      "learning_rate": 7.351448294943661e-05,
+      "loss": 1.4564,
+      "step": 12976
+    },
+    {
+      "epoch": 0.5978255862164279,
+      "grad_norm": 1.1288090485145452,
+      "learning_rate": 7.350009520948077e-05,
+      "loss": 1.4945,
+      "step": 12977
+    },
+    {
+      "epoch": 0.5978716543050628,
+      "grad_norm": 1.2664427790019082,
+      "learning_rate": 7.348570805949993e-05,
+      "loss": 1.1553,
+      "step": 12978
+    },
+    {
+      "epoch": 0.5979177223936979,
+      "grad_norm": 0.883181731442718,
+      "learning_rate": 7.347132149981424e-05,
+      "loss": 1.613,
+      "step": 12979
+    },
+    {
+      "epoch": 0.5979637904823328,
+      "grad_norm": 1.1056512688447018,
+      "learning_rate": 7.345693553074415e-05,
+      "loss": 1.1659,
+      "step": 12980
+    },
+    {
+      "epoch": 0.5980098585709679,
+      "grad_norm": 1.0335748984772808,
+      "learning_rate": 7.344255015260985e-05,
+      "loss": 1.4657,
+      "step": 12981
+    },
+    {
+      "epoch": 0.5980559266596029,
+      "grad_norm": 0.9844145212545505,
+      "learning_rate": 7.342816536573165e-05,
+      "loss": 1.2187,
+      "step": 12982
+    },
+    {
+      "epoch": 0.5981019947482379,
+      "grad_norm": 1.3164384738742612,
+      "learning_rate": 7.341378117042977e-05,
+      "loss": 1.0169,
+      "step": 12983
+    },
+    {
+      "epoch": 0.5981480628368729,
+      "grad_norm": 0.9647166563075863,
+      "learning_rate": 7.33993975670245e-05,
+      "loss": 1.4187,
+      "step": 12984
+    },
+    {
+      "epoch": 0.5981941309255079,
+      "grad_norm": 0.7964350176243472,
+      "learning_rate": 7.338501455583599e-05,
+      "loss": 1.267,
+      "step": 12985
+    },
+    {
+      "epoch": 0.5982401990141429,
+      "grad_norm": 0.8488088286125425,
+      "learning_rate": 7.337063213718454e-05,
+      "loss": 1.1683,
+      "step": 12986
+    },
+    {
+      "epoch": 0.5982862671027779,
+      "grad_norm": 1.0192801376487057,
+      "learning_rate": 7.335625031139032e-05,
+      "loss": 1.3706,
+      "step": 12987
+    },
+    {
+      "epoch": 0.5983323351914129,
+      "grad_norm": 0.9692009757884198,
+      "learning_rate": 7.334186907877343e-05,
+      "loss": 1.1732,
+      "step": 12988
+    },
+    {
+      "epoch": 0.5983784032800479,
+      "grad_norm": 1.1023255148125473,
+      "learning_rate": 7.332748843965419e-05,
+      "loss": 1.0031,
+      "step": 12989
+    },
+    {
+      "epoch": 0.5984244713686829,
+      "grad_norm": 1.3512797239291272,
+      "learning_rate": 7.331310839435261e-05,
+      "loss": 1.2507,
+      "step": 12990
+    },
+    {
+      "epoch": 0.5984705394573179,
+      "grad_norm": 0.9537530991316499,
+      "learning_rate": 7.329872894318899e-05,
+      "loss": 1.5539,
+      "step": 12991
+    },
+    {
+      "epoch": 0.5985166075459529,
+      "grad_norm": 0.7743464221729394,
+      "learning_rate": 7.32843500864833e-05,
+      "loss": 1.0877,
+      "step": 12992
+    },
+    {
+      "epoch": 0.5985626756345879,
+      "grad_norm": 0.9800408255977924,
+      "learning_rate": 7.326997182455583e-05,
+      "loss": 1.1403,
+      "step": 12993
+    },
+    {
+      "epoch": 0.5986087437232229,
+      "grad_norm": 1.3714466906356397,
+      "learning_rate": 7.325559415772656e-05,
+      "loss": 1.2575,
+      "step": 12994
+    },
+    {
+      "epoch": 0.598654811811858,
+      "grad_norm": 0.9396636070231349,
+      "learning_rate": 7.324121708631567e-05,
+      "loss": 1.3692,
+      "step": 12995
+    },
+    {
+      "epoch": 0.5987008799004929,
+      "grad_norm": 1.47177416460196,
+      "learning_rate": 7.32268406106432e-05,
+      "loss": 1.3609,
+      "step": 12996
+    },
+    {
+      "epoch": 0.598746947989128,
+      "grad_norm": 1.6063513930703899,
+      "learning_rate": 7.321246473102913e-05,
+      "loss": 1.2551,
+      "step": 12997
+    },
+    {
+      "epoch": 0.5987930160777629,
+      "grad_norm": 11.754981071174923,
+      "learning_rate": 7.319808944779372e-05,
+      "loss": 1.3586,
+      "step": 12998
+    },
+    {
+      "epoch": 0.598839084166398,
+      "grad_norm": 1.3847856824069584,
+      "learning_rate": 7.31837147612568e-05,
+      "loss": 1.1501,
+      "step": 12999
+    },
+    {
+      "epoch": 0.5988851522550329,
+      "grad_norm": 0.7823467955101655,
+      "learning_rate": 7.316934067173856e-05,
+      "loss": 1.5174,
+      "step": 13000
+    },
+    {
+      "epoch": 0.598931220343668,
+      "grad_norm": 0.8952577577635729,
+      "learning_rate": 7.315496717955888e-05,
+      "loss": 1.2198,
+      "step": 13001
+    },
+    {
+      "epoch": 0.5989772884323029,
+      "grad_norm": 1.4856442513043573,
+      "learning_rate": 7.314059428503788e-05,
+      "loss": 1.287,
+      "step": 13002
+    },
+    {
+      "epoch": 0.599023356520938,
+      "grad_norm": 3.571002554745324,
+      "learning_rate": 7.312622198849543e-05,
+      "loss": 1.1025,
+      "step": 13003
+    },
+    {
+      "epoch": 0.5990694246095729,
+      "grad_norm": 1.0951479853005528,
+      "learning_rate": 7.311185029025165e-05,
+      "loss": 1.2746,
+      "step": 13004
+    },
+    {
+      "epoch": 0.599115492698208,
+      "grad_norm": 1.1988374927028582,
+      "learning_rate": 7.30974791906264e-05,
+      "loss": 1.1383,
+      "step": 13005
+    },
+    {
+      "epoch": 0.5991615607868429,
+      "grad_norm": 1.239747494226184,
+      "learning_rate": 7.30831086899396e-05,
+      "loss": 1.2974,
+      "step": 13006
+    },
+    {
+      "epoch": 0.599207628875478,
+      "grad_norm": 1.387917552873531,
+      "learning_rate": 7.306873878851128e-05,
+      "loss": 1.2279,
+      "step": 13007
+    },
+    {
+      "epoch": 0.599253696964113,
+      "grad_norm": 6.39036529361518,
+      "learning_rate": 7.305436948666127e-05,
+      "loss": 1.1474,
+      "step": 13008
+    },
+    {
+      "epoch": 0.599299765052748,
+      "grad_norm": 2.526197966183228,
+      "learning_rate": 7.304000078470958e-05,
+      "loss": 1.2776,
+      "step": 13009
+    },
+    {
+      "epoch": 0.599345833141383,
+      "grad_norm": 1.69637999264793,
+      "learning_rate": 7.302563268297597e-05,
+      "loss": 1.3716,
+      "step": 13010
+    },
+    {
+      "epoch": 0.599391901230018,
+      "grad_norm": 1.1173371119717816,
+      "learning_rate": 7.301126518178047e-05,
+      "loss": 1.2363,
+      "step": 13011
+    },
+    {
+      "epoch": 0.599437969318653,
+      "grad_norm": 1.407695817298745,
+      "learning_rate": 7.299689828144283e-05,
+      "loss": 1.1665,
+      "step": 13012
+    },
+    {
+      "epoch": 0.599484037407288,
+      "grad_norm": 1.4186365075202378,
+      "learning_rate": 7.2982531982283e-05,
+      "loss": 1.3946,
+      "step": 13013
+    },
+    {
+      "epoch": 0.599530105495923,
+      "grad_norm": 1.0451159351100747,
+      "learning_rate": 7.296816628462069e-05,
+      "loss": 1.3591,
+      "step": 13014
+    },
+    {
+      "epoch": 0.599576173584558,
+      "grad_norm": 1.5153161931166195,
+      "learning_rate": 7.295380118877588e-05,
+      "loss": 0.9996,
+      "step": 13015
+    },
+    {
+      "epoch": 0.599622241673193,
+      "grad_norm": 0.9882771489089244,
+      "learning_rate": 7.29394366950683e-05,
+      "loss": 0.9955,
+      "step": 13016
+    },
+    {
+      "epoch": 0.599668309761828,
+      "grad_norm": 0.838961074877163,
+      "learning_rate": 7.292507280381774e-05,
+      "loss": 1.3984,
+      "step": 13017
+    },
+    {
+      "epoch": 0.599714377850463,
+      "grad_norm": 0.8674145753744638,
+      "learning_rate": 7.291070951534402e-05,
+      "loss": 1.4119,
+      "step": 13018
+    },
+    {
+      "epoch": 0.5997604459390979,
+      "grad_norm": 1.3373120892101997,
+      "learning_rate": 7.289634682996687e-05,
+      "loss": 1.3974,
+      "step": 13019
+    },
+    {
+      "epoch": 0.599806514027733,
+      "grad_norm": 0.9536180418067552,
+      "learning_rate": 7.288198474800616e-05,
+      "loss": 1.2982,
+      "step": 13020
+    },
+    {
+      "epoch": 0.5998525821163679,
+      "grad_norm": 1.8183259316015508,
+      "learning_rate": 7.28676232697815e-05,
+      "loss": 1.1581,
+      "step": 13021
+    },
+    {
+      "epoch": 0.599898650205003,
+      "grad_norm": 1.1321100410724394,
+      "learning_rate": 7.285326239561273e-05,
+      "loss": 1.421,
+      "step": 13022
+    },
+    {
+      "epoch": 0.599944718293638,
+      "grad_norm": 0.8297548909599135,
+      "learning_rate": 7.283890212581946e-05,
+      "loss": 1.232,
+      "step": 13023
+    },
+    {
+      "epoch": 0.599990786382273,
+      "grad_norm": 1.0854727225542458,
+      "learning_rate": 7.282454246072154e-05,
+      "loss": 1.0686,
+      "step": 13024
+    },
+    {
+      "epoch": 0.600036854470908,
+      "grad_norm": 0.8227231222032011,
+      "learning_rate": 7.281018340063856e-05,
+      "loss": 1.2806,
+      "step": 13025
+    },
+    {
+      "epoch": 0.600082922559543,
+      "grad_norm": 1.028812643063352,
+      "learning_rate": 7.279582494589021e-05,
+      "loss": 1.126,
+      "step": 13026
+    },
+    {
+      "epoch": 0.600128990648178,
+      "grad_norm": 1.0216411082113408,
+      "learning_rate": 7.278146709679618e-05,
+      "loss": 1.214,
+      "step": 13027
+    },
+    {
+      "epoch": 0.600175058736813,
+      "grad_norm": 3.220099998452181,
+      "learning_rate": 7.276710985367611e-05,
+      "loss": 1.1277,
+      "step": 13028
+    },
+    {
+      "epoch": 0.600221126825448,
+      "grad_norm": 0.9632370896234516,
+      "learning_rate": 7.275275321684966e-05,
+      "loss": 1.4214,
+      "step": 13029
+    },
+    {
+      "epoch": 0.600267194914083,
+      "grad_norm": 1.998592291263483,
+      "learning_rate": 7.273839718663643e-05,
+      "loss": 1.559,
+      "step": 13030
+    },
+    {
+      "epoch": 0.600313263002718,
+      "grad_norm": 1.1443415223203013,
+      "learning_rate": 7.272404176335606e-05,
+      "loss": 1.0659,
+      "step": 13031
+    },
+    {
+      "epoch": 0.600359331091353,
+      "grad_norm": 1.8448517770038653,
+      "learning_rate": 7.27096869473281e-05,
+      "loss": 1.5408,
+      "step": 13032
+    },
+    {
+      "epoch": 0.600405399179988,
+      "grad_norm": 0.7492035860966777,
+      "learning_rate": 7.269533273887219e-05,
+      "loss": 1.5621,
+      "step": 13033
+    },
+    {
+      "epoch": 0.600451467268623,
+      "grad_norm": 1.5588959600080359,
+      "learning_rate": 7.268097913830787e-05,
+      "loss": 1.3669,
+      "step": 13034
+    },
+    {
+      "epoch": 0.600497535357258,
+      "grad_norm": 1.007909709561308,
+      "learning_rate": 7.266662614595469e-05,
+      "loss": 1.3895,
+      "step": 13035
+    },
+    {
+      "epoch": 0.6005436034458931,
+      "grad_norm": 1.182576017716399,
+      "learning_rate": 7.265227376213226e-05,
+      "loss": 1.5978,
+      "step": 13036
+    },
+    {
+      "epoch": 0.600589671534528,
+      "grad_norm": 1.097081790713401,
+      "learning_rate": 7.263792198716002e-05,
+      "loss": 1.4727,
+      "step": 13037
+    },
+    {
+      "epoch": 0.6006357396231631,
+      "grad_norm": 1.8707387811207368,
+      "learning_rate": 7.262357082135756e-05,
+      "loss": 1.1879,
+      "step": 13038
+    },
+    {
+      "epoch": 0.600681807711798,
+      "grad_norm": 1.443554453771784,
+      "learning_rate": 7.260922026504433e-05,
+      "loss": 1.0093,
+      "step": 13039
+    },
+    {
+      "epoch": 0.6007278758004331,
+      "grad_norm": 0.9380127691008773,
+      "learning_rate": 7.259487031853987e-05,
+      "loss": 1.3684,
+      "step": 13040
+    },
+    {
+      "epoch": 0.600773943889068,
+      "grad_norm": 1.9907940481342712,
+      "learning_rate": 7.258052098216359e-05,
+      "loss": 1.0921,
+      "step": 13041
+    },
+    {
+      "epoch": 0.6008200119777031,
+      "grad_norm": 0.9012744842082538,
+      "learning_rate": 7.256617225623504e-05,
+      "loss": 1.2087,
+      "step": 13042
+    },
+    {
+      "epoch": 0.600866080066338,
+      "grad_norm": 1.9475661885778812,
+      "learning_rate": 7.255182414107363e-05,
+      "loss": 1.5264,
+      "step": 13043
+    },
+    {
+      "epoch": 0.6009121481549731,
+      "grad_norm": 0.9497744363171282,
+      "learning_rate": 7.253747663699876e-05,
+      "loss": 1.2157,
+      "step": 13044
+    },
+    {
+      "epoch": 0.600958216243608,
+      "grad_norm": 0.9142443414598771,
+      "learning_rate": 7.252312974432991e-05,
+      "loss": 1.2813,
+      "step": 13045
+    },
+    {
+      "epoch": 0.6010042843322431,
+      "grad_norm": 1.4110668415201897,
+      "learning_rate": 7.250878346338643e-05,
+      "loss": 1.2947,
+      "step": 13046
+    },
+    {
+      "epoch": 0.601050352420878,
+      "grad_norm": 1.3500515269868107,
+      "learning_rate": 7.24944377944878e-05,
+      "loss": 1.3756,
+      "step": 13047
+    },
+    {
+      "epoch": 0.6010964205095131,
+      "grad_norm": 1.3961581593019372,
+      "learning_rate": 7.24800927379533e-05,
+      "loss": 1.3709,
+      "step": 13048
+    },
+    {
+      "epoch": 0.6011424885981481,
+      "grad_norm": 2.1893583208968344,
+      "learning_rate": 7.246574829410237e-05,
+      "loss": 1.4948,
+      "step": 13049
+    },
+    {
+      "epoch": 0.6011885566867831,
+      "grad_norm": 1.0536345106733227,
+      "learning_rate": 7.245140446325434e-05,
+      "loss": 1.3101,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6012346247754181,
+      "grad_norm": 2.0205664284558735,
+      "learning_rate": 7.243706124572855e-05,
+      "loss": 1.2201,
+      "step": 13051
+    },
+    {
+      "epoch": 0.6012806928640531,
+      "grad_norm": 9.473757776396207,
+      "learning_rate": 7.242271864184436e-05,
+      "loss": 1.3042,
+      "step": 13052
+    },
+    {
+      "epoch": 0.6013267609526881,
+      "grad_norm": 1.2270951742697527,
+      "learning_rate": 7.240837665192101e-05,
+      "loss": 1.4191,
+      "step": 13053
+    },
+    {
+      "epoch": 0.601372829041323,
+      "grad_norm": 3.045935741410923,
+      "learning_rate": 7.239403527627788e-05,
+      "loss": 1.2262,
+      "step": 13054
+    },
+    {
+      "epoch": 0.6014188971299581,
+      "grad_norm": 1.1099352628598316,
+      "learning_rate": 7.237969451523422e-05,
+      "loss": 1.1774,
+      "step": 13055
+    },
+    {
+      "epoch": 0.601464965218593,
+      "grad_norm": 2.2060115371938585,
+      "learning_rate": 7.236535436910928e-05,
+      "loss": 1.14,
+      "step": 13056
+    },
+    {
+      "epoch": 0.6015110333072281,
+      "grad_norm": 0.9348529293340571,
+      "learning_rate": 7.235101483822234e-05,
+      "loss": 1.2834,
+      "step": 13057
+    },
+    {
+      "epoch": 0.601557101395863,
+      "grad_norm": 1.2470823043516497,
+      "learning_rate": 7.233667592289267e-05,
+      "loss": 1.0716,
+      "step": 13058
+    },
+    {
+      "epoch": 0.6016031694844981,
+      "grad_norm": 0.9722079365564945,
+      "learning_rate": 7.232233762343945e-05,
+      "loss": 1.3361,
+      "step": 13059
+    },
+    {
+      "epoch": 0.601649237573133,
+      "grad_norm": 2.0420828783405702,
+      "learning_rate": 7.230799994018196e-05,
+      "loss": 1.4242,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6016953056617681,
+      "grad_norm": 1.037975683090143,
+      "learning_rate": 7.229366287343936e-05,
+      "loss": 1.6346,
+      "step": 13061
+    },
+    {
+      "epoch": 0.601741373750403,
+      "grad_norm": 1.5984559748690912,
+      "learning_rate": 7.227932642353084e-05,
+      "loss": 1.3193,
+      "step": 13062
+    },
+    {
+      "epoch": 0.6017874418390381,
+      "grad_norm": 0.7423538588305388,
+      "learning_rate": 7.226499059077558e-05,
+      "loss": 1.1798,
+      "step": 13063
+    },
+    {
+      "epoch": 0.6018335099276731,
+      "grad_norm": 0.8700300332154793,
+      "learning_rate": 7.225065537549276e-05,
+      "loss": 1.0775,
+      "step": 13064
+    },
+    {
+      "epoch": 0.6018795780163081,
+      "grad_norm": 1.5157270826801925,
+      "learning_rate": 7.223632077800152e-05,
+      "loss": 1.2624,
+      "step": 13065
+    },
+    {
+      "epoch": 0.6019256461049431,
+      "grad_norm": 1.3366702816791474,
+      "learning_rate": 7.222198679862098e-05,
+      "loss": 1.5266,
+      "step": 13066
+    },
+    {
+      "epoch": 0.6019717141935781,
+      "grad_norm": 1.50819529317963,
+      "learning_rate": 7.220765343767029e-05,
+      "loss": 1.296,
+      "step": 13067
+    },
+    {
+      "epoch": 0.6020177822822131,
+      "grad_norm": 0.7009796819756731,
+      "learning_rate": 7.219332069546853e-05,
+      "loss": 1.2129,
+      "step": 13068
+    },
+    {
+      "epoch": 0.6020638503708481,
+      "grad_norm": 0.9469779227946197,
+      "learning_rate": 7.217898857233483e-05,
+      "loss": 1.463,
+      "step": 13069
+    },
+    {
+      "epoch": 0.6021099184594831,
+      "grad_norm": 3.287215393684393,
+      "learning_rate": 7.216465706858824e-05,
+      "loss": 1.4202,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6021559865481181,
+      "grad_norm": 1.2497339850662268,
+      "learning_rate": 7.21503261845478e-05,
+      "loss": 1.2271,
+      "step": 13071
+    },
+    {
+      "epoch": 0.6022020546367531,
+      "grad_norm": 0.9316267332760025,
+      "learning_rate": 7.213599592053263e-05,
+      "loss": 1.4705,
+      "step": 13072
+    },
+    {
+      "epoch": 0.6022481227253881,
+      "grad_norm": 0.9845029989887135,
+      "learning_rate": 7.212166627686171e-05,
+      "loss": 1.2054,
+      "step": 13073
+    },
+    {
+      "epoch": 0.6022941908140231,
+      "grad_norm": 1.0401428054371586,
+      "learning_rate": 7.210733725385411e-05,
+      "loss": 1.4613,
+      "step": 13074
+    },
+    {
+      "epoch": 0.6023402589026581,
+      "grad_norm": 0.8336042249477279,
+      "learning_rate": 7.20930088518288e-05,
+      "loss": 1.3902,
+      "step": 13075
+    },
+    {
+      "epoch": 0.6023863269912931,
+      "grad_norm": 1.7313403588586498,
+      "learning_rate": 7.207868107110482e-05,
+      "loss": 1.1616,
+      "step": 13076
+    },
+    {
+      "epoch": 0.6024323950799282,
+      "grad_norm": 0.9925554471536313,
+      "learning_rate": 7.20643539120011e-05,
+      "loss": 1.3972,
+      "step": 13077
+    },
+    {
+      "epoch": 0.6024784631685631,
+      "grad_norm": 0.9709448732004106,
+      "learning_rate": 7.205002737483668e-05,
+      "loss": 0.9939,
+      "step": 13078
+    },
+    {
+      "epoch": 0.6025245312571982,
+      "grad_norm": 0.9326368282140703,
+      "learning_rate": 7.203570145993048e-05,
+      "loss": 1.6227,
+      "step": 13079
+    },
+    {
+      "epoch": 0.6025705993458331,
+      "grad_norm": 1.2179683732193864,
+      "learning_rate": 7.20213761676014e-05,
+      "loss": 1.4454,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6026166674344682,
+      "grad_norm": 1.2024785920164576,
+      "learning_rate": 7.200705149816843e-05,
+      "loss": 1.1762,
+      "step": 13081
+    },
+    {
+      "epoch": 0.6026627355231031,
+      "grad_norm": 3.3703240133360564,
+      "learning_rate": 7.199272745195044e-05,
+      "loss": 1.0796,
+      "step": 13082
+    },
+    {
+      "epoch": 0.6027088036117382,
+      "grad_norm": 0.8248162918471604,
+      "learning_rate": 7.197840402926638e-05,
+      "loss": 1.4898,
+      "step": 13083
+    },
+    {
+      "epoch": 0.6027548717003731,
+      "grad_norm": 0.9311748043896848,
+      "learning_rate": 7.196408123043508e-05,
+      "loss": 1.1802,
+      "step": 13084
+    },
+    {
+      "epoch": 0.6028009397890082,
+      "grad_norm": 1.5191832870965787,
+      "learning_rate": 7.194975905577545e-05,
+      "loss": 1.256,
+      "step": 13085
+    },
+    {
+      "epoch": 0.6028470078776431,
+      "grad_norm": 0.9498326144015017,
+      "learning_rate": 7.193543750560634e-05,
+      "loss": 1.4114,
+      "step": 13086
+    },
+    {
+      "epoch": 0.6028930759662782,
+      "grad_norm": 1.6777558885006445,
+      "learning_rate": 7.19211165802466e-05,
+      "loss": 1.0976,
+      "step": 13087
+    },
+    {
+      "epoch": 0.6029391440549131,
+      "grad_norm": 0.886522321161864,
+      "learning_rate": 7.190679628001508e-05,
+      "loss": 1.4257,
+      "step": 13088
+    },
+    {
+      "epoch": 0.6029852121435482,
+      "grad_norm": 0.8077565950340638,
+      "learning_rate": 7.189247660523052e-05,
+      "loss": 1.2314,
+      "step": 13089
+    },
+    {
+      "epoch": 0.6030312802321832,
+      "grad_norm": 1.292382925951729,
+      "learning_rate": 7.187815755621182e-05,
+      "loss": 1.0251,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6030773483208182,
+      "grad_norm": 0.9713627779798372,
+      "learning_rate": 7.186383913327766e-05,
+      "loss": 1.1119,
+      "step": 13091
+    },
+    {
+      "epoch": 0.6031234164094532,
+      "grad_norm": 1.2508211618290055,
+      "learning_rate": 7.184952133674693e-05,
+      "loss": 1.1091,
+      "step": 13092
+    },
+    {
+      "epoch": 0.6031694844980882,
+      "grad_norm": 0.9410841050547528,
+      "learning_rate": 7.183520416693833e-05,
+      "loss": 1.3687,
+      "step": 13093
+    },
+    {
+      "epoch": 0.6032155525867232,
+      "grad_norm": 0.9235268569242474,
+      "learning_rate": 7.18208876241706e-05,
+      "loss": 1.2654,
+      "step": 13094
+    },
+    {
+      "epoch": 0.6032616206753582,
+      "grad_norm": 1.0130542006351084,
+      "learning_rate": 7.180657170876251e-05,
+      "loss": 1.4565,
+      "step": 13095
+    },
+    {
+      "epoch": 0.6033076887639932,
+      "grad_norm": 0.7396784197204335,
+      "learning_rate": 7.179225642103276e-05,
+      "loss": 1.3069,
+      "step": 13096
+    },
+    {
+      "epoch": 0.6033537568526282,
+      "grad_norm": 1.871837075109175,
+      "learning_rate": 7.177794176130009e-05,
+      "loss": 1.5217,
+      "step": 13097
+    },
+    {
+      "epoch": 0.6033998249412632,
+      "grad_norm": 0.9051614874220961,
+      "learning_rate": 7.176362772988307e-05,
+      "loss": 1.2637,
+      "step": 13098
+    },
+    {
+      "epoch": 0.6034458930298982,
+      "grad_norm": 0.8915090364820804,
+      "learning_rate": 7.174931432710054e-05,
+      "loss": 1.2976,
+      "step": 13099
+    },
+    {
+      "epoch": 0.6034919611185332,
+      "grad_norm": 0.9625418614321436,
+      "learning_rate": 7.173500155327102e-05,
+      "loss": 1.4603,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6035380292071681,
+      "grad_norm": 0.8606828655665715,
+      "learning_rate": 7.17206894087133e-05,
+      "loss": 1.2867,
+      "step": 13101
+    },
+    {
+      "epoch": 0.6035840972958032,
+      "grad_norm": 1.1170935909294044,
+      "learning_rate": 7.170637789374587e-05,
+      "loss": 1.4219,
+      "step": 13102
+    },
+    {
+      "epoch": 0.6036301653844381,
+      "grad_norm": 3.4041749206466734,
+      "learning_rate": 7.169206700868749e-05,
+      "loss": 1.4887,
+      "step": 13103
+    },
+    {
+      "epoch": 0.6036762334730732,
+      "grad_norm": 0.8404144074264622,
+      "learning_rate": 7.167775675385663e-05,
+      "loss": 1.5063,
+      "step": 13104
+    },
+    {
+      "epoch": 0.6037223015617083,
+      "grad_norm": 1.1882111391550196,
+      "learning_rate": 7.166344712957201e-05,
+      "loss": 1.2885,
+      "step": 13105
+    },
+    {
+      "epoch": 0.6037683696503432,
+      "grad_norm": 1.2142836935084649,
+      "learning_rate": 7.164913813615215e-05,
+      "loss": 1.4501,
+      "step": 13106
+    },
+    {
+      "epoch": 0.6038144377389782,
+      "grad_norm": 1.1329600448228065,
+      "learning_rate": 7.163482977391558e-05,
+      "loss": 1.2484,
+      "step": 13107
+    },
+    {
+      "epoch": 0.6038605058276132,
+      "grad_norm": 0.8298767602365517,
+      "learning_rate": 7.162052204318092e-05,
+      "loss": 1.377,
+      "step": 13108
+    },
+    {
+      "epoch": 0.6039065739162482,
+      "grad_norm": 0.966663491493068,
+      "learning_rate": 7.160621494426664e-05,
+      "loss": 0.9551,
+      "step": 13109
+    },
+    {
+      "epoch": 0.6039526420048832,
+      "grad_norm": 6.3191035746726625,
+      "learning_rate": 7.159190847749137e-05,
+      "loss": 1.413,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6039987100935182,
+      "grad_norm": 0.8421436668604885,
+      "learning_rate": 7.157760264317346e-05,
+      "loss": 1.641,
+      "step": 13111
+    },
+    {
+      "epoch": 0.6040447781821532,
+      "grad_norm": 0.7946314958837106,
+      "learning_rate": 7.156329744163157e-05,
+      "loss": 1.3702,
+      "step": 13112
+    },
+    {
+      "epoch": 0.6040908462707882,
+      "grad_norm": 0.9675825796976915,
+      "learning_rate": 7.154899287318405e-05,
+      "loss": 1.4765,
+      "step": 13113
+    },
+    {
+      "epoch": 0.6041369143594232,
+      "grad_norm": 1.185186238340775,
+      "learning_rate": 7.153468893814947e-05,
+      "loss": 1.1762,
+      "step": 13114
+    },
+    {
+      "epoch": 0.6041829824480582,
+      "grad_norm": 0.8154941166912234,
+      "learning_rate": 7.152038563684622e-05,
+      "loss": 1.3188,
+      "step": 13115
+    },
+    {
+      "epoch": 0.6042290505366932,
+      "grad_norm": 0.9762509942194957,
+      "learning_rate": 7.150608296959272e-05,
+      "loss": 1.2164,
+      "step": 13116
+    },
+    {
+      "epoch": 0.6042751186253282,
+      "grad_norm": 0.7771906352480177,
+      "learning_rate": 7.149178093670746e-05,
+      "loss": 1.0101,
+      "step": 13117
+    },
+    {
+      "epoch": 0.6043211867139633,
+      "grad_norm": 0.9518438120876879,
+      "learning_rate": 7.147747953850876e-05,
+      "loss": 1.1946,
+      "step": 13118
+    },
+    {
+      "epoch": 0.6043672548025982,
+      "grad_norm": 2.034265531615791,
+      "learning_rate": 7.146317877531514e-05,
+      "loss": 1.4034,
+      "step": 13119
+    },
+    {
+      "epoch": 0.6044133228912333,
+      "grad_norm": 1.444015263772938,
+      "learning_rate": 7.144887864744487e-05,
+      "loss": 1.361,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6044593909798682,
+      "grad_norm": 0.8564660847918597,
+      "learning_rate": 7.14345791552164e-05,
+      "loss": 1.1973,
+      "step": 13121
+    },
+    {
+      "epoch": 0.6045054590685033,
+      "grad_norm": 2.2039164120345553,
+      "learning_rate": 7.142028029894798e-05,
+      "loss": 1.2248,
+      "step": 13122
+    },
+    {
+      "epoch": 0.6045515271571382,
+      "grad_norm": 1.674570373323103,
+      "learning_rate": 7.14059820789581e-05,
+      "loss": 1.4904,
+      "step": 13123
+    },
+    {
+      "epoch": 0.6045975952457733,
+      "grad_norm": 0.874237444474555,
+      "learning_rate": 7.139168449556497e-05,
+      "loss": 1.3735,
+      "step": 13124
+    },
+    {
+      "epoch": 0.6046436633344082,
+      "grad_norm": 1.0081696493364278,
+      "learning_rate": 7.137738754908692e-05,
+      "loss": 1.2275,
+      "step": 13125
+    },
+    {
+      "epoch": 0.6046897314230433,
+      "grad_norm": 1.1202936214414536,
+      "learning_rate": 7.136309123984226e-05,
+      "loss": 1.3155,
+      "step": 13126
+    },
+    {
+      "epoch": 0.6047357995116782,
+      "grad_norm": 1.3835828568817268,
+      "learning_rate": 7.134879556814926e-05,
+      "loss": 1.46,
+      "step": 13127
+    },
+    {
+      "epoch": 0.6047818676003133,
+      "grad_norm": 1.8507498780529583,
+      "learning_rate": 7.133450053432623e-05,
+      "loss": 1.2137,
+      "step": 13128
+    },
+    {
+      "epoch": 0.6048279356889482,
+      "grad_norm": 0.8618788136566012,
+      "learning_rate": 7.132020613869136e-05,
+      "loss": 1.2416,
+      "step": 13129
+    },
+    {
+      "epoch": 0.6048740037775833,
+      "grad_norm": 4.086704734495591,
+      "learning_rate": 7.130591238156297e-05,
+      "loss": 1.4112,
+      "step": 13130
+    },
+    {
+      "epoch": 0.6049200718662183,
+      "grad_norm": 1.0994894244154354,
+      "learning_rate": 7.129161926325921e-05,
+      "loss": 1.1798,
+      "step": 13131
+    },
+    {
+      "epoch": 0.6049661399548533,
+      "grad_norm": 1.258937936850519,
+      "learning_rate": 7.127732678409833e-05,
+      "loss": 1.2349,
+      "step": 13132
+    },
+    {
+      "epoch": 0.6050122080434883,
+      "grad_norm": 1.5028510428640227,
+      "learning_rate": 7.126303494439847e-05,
+      "loss": 1.352,
+      "step": 13133
+    },
+    {
+      "epoch": 0.6050582761321233,
+      "grad_norm": 1.0719861310845733,
+      "learning_rate": 7.124874374447796e-05,
+      "loss": 1.1014,
+      "step": 13134
+    },
+    {
+      "epoch": 0.6051043442207583,
+      "grad_norm": 1.1636063308583324,
+      "learning_rate": 7.123445318465483e-05,
+      "loss": 1.368,
+      "step": 13135
+    },
+    {
+      "epoch": 0.6051504123093933,
+      "grad_norm": 1.2663924199710515,
+      "learning_rate": 7.122016326524725e-05,
+      "loss": 1.1325,
+      "step": 13136
+    },
+    {
+      "epoch": 0.6051964803980283,
+      "grad_norm": 0.7793802550294265,
+      "learning_rate": 7.120587398657344e-05,
+      "loss": 1.459,
+      "step": 13137
+    },
+    {
+      "epoch": 0.6052425484866633,
+      "grad_norm": 0.9244965279777004,
+      "learning_rate": 7.119158534895143e-05,
+      "loss": 1.4038,
+      "step": 13138
+    },
+    {
+      "epoch": 0.6052886165752983,
+      "grad_norm": 0.9698265871400846,
+      "learning_rate": 7.117729735269939e-05,
+      "loss": 1.1113,
+      "step": 13139
+    },
+    {
+      "epoch": 0.6053346846639333,
+      "grad_norm": 0.8156921545146388,
+      "learning_rate": 7.116300999813539e-05,
+      "loss": 1.167,
+      "step": 13140
+    },
+    {
+      "epoch": 0.6053807527525683,
+      "grad_norm": 1.4450744948008747,
+      "learning_rate": 7.114872328557755e-05,
+      "loss": 1.2462,
+      "step": 13141
+    },
+    {
+      "epoch": 0.6054268208412033,
+      "grad_norm": 0.9589316490861551,
+      "learning_rate": 7.113443721534391e-05,
+      "loss": 1.3073,
+      "step": 13142
+    },
+    {
+      "epoch": 0.6054728889298383,
+      "grad_norm": 0.9435571072763809,
+      "learning_rate": 7.112015178775253e-05,
+      "loss": 1.0689,
+      "step": 13143
+    },
+    {
+      "epoch": 0.6055189570184734,
+      "grad_norm": 0.9146137204943724,
+      "learning_rate": 7.110586700312147e-05,
+      "loss": 1.0618,
+      "step": 13144
+    },
+    {
+      "epoch": 0.6055650251071083,
+      "grad_norm": 0.9464413784584356,
+      "learning_rate": 7.109158286176871e-05,
+      "loss": 0.9721,
+      "step": 13145
+    },
+    {
+      "epoch": 0.6056110931957434,
+      "grad_norm": 1.0695607250435506,
+      "learning_rate": 7.107729936401231e-05,
+      "loss": 1.5085,
+      "step": 13146
+    },
+    {
+      "epoch": 0.6056571612843783,
+      "grad_norm": 1.0066367539742607,
+      "learning_rate": 7.106301651017022e-05,
+      "loss": 1.4987,
+      "step": 13147
+    },
+    {
+      "epoch": 0.6057032293730134,
+      "grad_norm": 1.9231119333311286,
+      "learning_rate": 7.104873430056049e-05,
+      "loss": 1.2951,
+      "step": 13148
+    },
+    {
+      "epoch": 0.6057492974616483,
+      "grad_norm": 2.046254032663886,
+      "learning_rate": 7.103445273550102e-05,
+      "loss": 1.0736,
+      "step": 13149
+    },
+    {
+      "epoch": 0.6057953655502833,
+      "grad_norm": 0.8859332590976188,
+      "learning_rate": 7.102017181530982e-05,
+      "loss": 1.2028,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6058414336389183,
+      "grad_norm": 0.7654164711054533,
+      "learning_rate": 7.100589154030477e-05,
+      "loss": 1.6337,
+      "step": 13151
+    },
+    {
+      "epoch": 0.6058875017275533,
+      "grad_norm": 2.4905931050057135,
+      "learning_rate": 7.099161191080386e-05,
+      "loss": 1.6395,
+      "step": 13152
+    },
+    {
+      "epoch": 0.6059335698161883,
+      "grad_norm": 1.2800652418874678,
+      "learning_rate": 7.097733292712498e-05,
+      "loss": 1.0855,
+      "step": 13153
+    },
+    {
+      "epoch": 0.6059796379048233,
+      "grad_norm": 0.8970432622361142,
+      "learning_rate": 7.096305458958599e-05,
+      "loss": 1.349,
+      "step": 13154
+    },
+    {
+      "epoch": 0.6060257059934583,
+      "grad_norm": 5.3199411542815405,
+      "learning_rate": 7.094877689850481e-05,
+      "loss": 1.1516,
+      "step": 13155
+    },
+    {
+      "epoch": 0.6060717740820933,
+      "grad_norm": 1.1035104039932884,
+      "learning_rate": 7.09344998541993e-05,
+      "loss": 1.0184,
+      "step": 13156
+    },
+    {
+      "epoch": 0.6061178421707283,
+      "grad_norm": 0.8012607450184488,
+      "learning_rate": 7.092022345698732e-05,
+      "loss": 1.2899,
+      "step": 13157
+    },
+    {
+      "epoch": 0.6061639102593633,
+      "grad_norm": 1.128765818825477,
+      "learning_rate": 7.09059477071867e-05,
+      "loss": 1.5728,
+      "step": 13158
+    },
+    {
+      "epoch": 0.6062099783479984,
+      "grad_norm": 1.4252553728393573,
+      "learning_rate": 7.089167260511528e-05,
+      "loss": 1.1122,
+      "step": 13159
+    },
+    {
+      "epoch": 0.6062560464366333,
+      "grad_norm": 2.5052679295497495,
+      "learning_rate": 7.087739815109084e-05,
+      "loss": 1.2303,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6063021145252684,
+      "grad_norm": 1.2501038729616485,
+      "learning_rate": 7.086312434543121e-05,
+      "loss": 1.3073,
+      "step": 13161
+    },
+    {
+      "epoch": 0.6063481826139033,
+      "grad_norm": 0.913297027930066,
+      "learning_rate": 7.084885118845416e-05,
+      "loss": 1.6246,
+      "step": 13162
+    },
+    {
+      "epoch": 0.6063942507025384,
+      "grad_norm": 1.2909141881193213,
+      "learning_rate": 7.083457868047743e-05,
+      "loss": 1.2387,
+      "step": 13163
+    },
+    {
+      "epoch": 0.6064403187911733,
+      "grad_norm": 1.692876413763568,
+      "learning_rate": 7.082030682181884e-05,
+      "loss": 1.5245,
+      "step": 13164
+    },
+    {
+      "epoch": 0.6064863868798084,
+      "grad_norm": 1.4119247418707346,
+      "learning_rate": 7.080603561279604e-05,
+      "loss": 1.0354,
+      "step": 13165
+    },
+    {
+      "epoch": 0.6065324549684433,
+      "grad_norm": 4.932102104079339,
+      "learning_rate": 7.079176505372684e-05,
+      "loss": 1.4814,
+      "step": 13166
+    },
+    {
+      "epoch": 0.6065785230570784,
+      "grad_norm": 0.9396987319948221,
+      "learning_rate": 7.077749514492887e-05,
+      "loss": 1.3745,
+      "step": 13167
+    },
+    {
+      "epoch": 0.6066245911457133,
+      "grad_norm": 0.8323177596934221,
+      "learning_rate": 7.076322588671989e-05,
+      "loss": 1.3675,
+      "step": 13168
+    },
+    {
+      "epoch": 0.6066706592343484,
+      "grad_norm": 0.7965858442025221,
+      "learning_rate": 7.074895727941755e-05,
+      "loss": 1.471,
+      "step": 13169
+    },
+    {
+      "epoch": 0.6067167273229833,
+      "grad_norm": 1.0877722088604924,
+      "learning_rate": 7.073468932333953e-05,
+      "loss": 1.2348,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6067627954116184,
+      "grad_norm": 0.9501262320956071,
+      "learning_rate": 7.072042201880348e-05,
+      "loss": 1.3112,
+      "step": 13171
+    },
+    {
+      "epoch": 0.6068088635002534,
+      "grad_norm": 1.0714301982201826,
+      "learning_rate": 7.0706155366127e-05,
+      "loss": 1.1133,
+      "step": 13172
+    },
+    {
+      "epoch": 0.6068549315888884,
+      "grad_norm": 1.4515226104151282,
+      "learning_rate": 7.069188936562777e-05,
+      "loss": 1.1906,
+      "step": 13173
+    },
+    {
+      "epoch": 0.6069009996775234,
+      "grad_norm": 0.7939590333418719,
+      "learning_rate": 7.067762401762334e-05,
+      "loss": 1.3579,
+      "step": 13174
+    },
+    {
+      "epoch": 0.6069470677661584,
+      "grad_norm": 0.9028187625590045,
+      "learning_rate": 7.066335932243137e-05,
+      "loss": 1.0567,
+      "step": 13175
+    },
+    {
+      "epoch": 0.6069931358547934,
+      "grad_norm": 0.7682335984337257,
+      "learning_rate": 7.064909528036938e-05,
+      "loss": 1.1978,
+      "step": 13176
+    },
+    {
+      "epoch": 0.6070392039434284,
+      "grad_norm": 0.9988750477280093,
+      "learning_rate": 7.063483189175497e-05,
+      "loss": 1.289,
+      "step": 13177
+    },
+    {
+      "epoch": 0.6070852720320634,
+      "grad_norm": 1.0776930643299472,
+      "learning_rate": 7.062056915690566e-05,
+      "loss": 1.3174,
+      "step": 13178
+    },
+    {
+      "epoch": 0.6071313401206984,
+      "grad_norm": 3.082578523327945,
+      "learning_rate": 7.060630707613902e-05,
+      "loss": 1.2265,
+      "step": 13179
+    },
+    {
+      "epoch": 0.6071774082093334,
+      "grad_norm": 1.3485599465239697,
+      "learning_rate": 7.059204564977255e-05,
+      "loss": 1.2943,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6072234762979684,
+      "grad_norm": 1.0951876966465919,
+      "learning_rate": 7.057778487812376e-05,
+      "loss": 1.3263,
+      "step": 13181
+    },
+    {
+      "epoch": 0.6072695443866034,
+      "grad_norm": 1.039801725311745,
+      "learning_rate": 7.056352476151013e-05,
+      "loss": 1.1809,
+      "step": 13182
+    },
+    {
+      "epoch": 0.6073156124752384,
+      "grad_norm": 0.8446129719597382,
+      "learning_rate": 7.054926530024915e-05,
+      "loss": 1.5864,
+      "step": 13183
+    },
+    {
+      "epoch": 0.6073616805638734,
+      "grad_norm": 1.597688432992591,
+      "learning_rate": 7.05350064946583e-05,
+      "loss": 1.3121,
+      "step": 13184
+    },
+    {
+      "epoch": 0.6074077486525085,
+      "grad_norm": 0.8640439741074903,
+      "learning_rate": 7.052074834505499e-05,
+      "loss": 1.4988,
+      "step": 13185
+    },
+    {
+      "epoch": 0.6074538167411434,
+      "grad_norm": 0.7518589266954931,
+      "learning_rate": 7.05064908517567e-05,
+      "loss": 1.4524,
+      "step": 13186
+    },
+    {
+      "epoch": 0.6074998848297785,
+      "grad_norm": 1.7990110424676529,
+      "learning_rate": 7.049223401508078e-05,
+      "loss": 1.5725,
+      "step": 13187
+    },
+    {
+      "epoch": 0.6075459529184134,
+      "grad_norm": 1.4371582694885439,
+      "learning_rate": 7.04779778353447e-05,
+      "loss": 0.8839,
+      "step": 13188
+    },
+    {
+      "epoch": 0.6075920210070485,
+      "grad_norm": 1.2131524476069204,
+      "learning_rate": 7.046372231286584e-05,
+      "loss": 1.44,
+      "step": 13189
+    },
+    {
+      "epoch": 0.6076380890956834,
+      "grad_norm": 1.513308021234583,
+      "learning_rate": 7.044946744796151e-05,
+      "loss": 1.3163,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6076841571843185,
+      "grad_norm": 2.434274320065965,
+      "learning_rate": 7.043521324094916e-05,
+      "loss": 1.3277,
+      "step": 13191
+    },
+    {
+      "epoch": 0.6077302252729534,
+      "grad_norm": 1.0626763019739038,
+      "learning_rate": 7.042095969214607e-05,
+      "loss": 1.1659,
+      "step": 13192
+    },
+    {
+      "epoch": 0.6077762933615884,
+      "grad_norm": 0.9753095832709778,
+      "learning_rate": 7.040670680186961e-05,
+      "loss": 1.1204,
+      "step": 13193
+    },
+    {
+      "epoch": 0.6078223614502234,
+      "grad_norm": 1.2205441012421245,
+      "learning_rate": 7.039245457043708e-05,
+      "loss": 1.4408,
+      "step": 13194
+    },
+    {
+      "epoch": 0.6078684295388584,
+      "grad_norm": 1.2549972851604922,
+      "learning_rate": 7.037820299816577e-05,
+      "loss": 1.1596,
+      "step": 13195
+    },
+    {
+      "epoch": 0.6079144976274934,
+      "grad_norm": 0.9525986909059873,
+      "learning_rate": 7.036395208537299e-05,
+      "loss": 1.0773,
+      "step": 13196
+    },
+    {
+      "epoch": 0.6079605657161284,
+      "grad_norm": 1.5555245324599423,
+      "learning_rate": 7.034970183237602e-05,
+      "loss": 1.1407,
+      "step": 13197
+    },
+    {
+      "epoch": 0.6080066338047634,
+      "grad_norm": 2.5540039435430453,
+      "learning_rate": 7.033545223949211e-05,
+      "loss": 1.477,
+      "step": 13198
+    },
+    {
+      "epoch": 0.6080527018933984,
+      "grad_norm": 0.9597512949125875,
+      "learning_rate": 7.032120330703842e-05,
+      "loss": 1.4846,
+      "step": 13199
+    },
+    {
+      "epoch": 0.6080987699820335,
+      "grad_norm": 0.8350782081263494,
+      "learning_rate": 7.030695503533234e-05,
+      "loss": 1.3752,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6081448380706684,
+      "grad_norm": 5.906194454986765,
+      "learning_rate": 7.02927074246909e-05,
+      "loss": 1.5183,
+      "step": 13201
+    },
+    {
+      "epoch": 0.6081909061593035,
+      "grad_norm": 0.8547701124790644,
+      "learning_rate": 7.027846047543147e-05,
+      "loss": 1.5448,
+      "step": 13202
+    },
+    {
+      "epoch": 0.6082369742479384,
+      "grad_norm": 0.7654135116674962,
+      "learning_rate": 7.02642141878711e-05,
+      "loss": 1.2411,
+      "step": 13203
+    },
+    {
+      "epoch": 0.6082830423365735,
+      "grad_norm": 1.5705030925266141,
+      "learning_rate": 7.024996856232705e-05,
+      "loss": 1.2845,
+      "step": 13204
+    },
+    {
+      "epoch": 0.6083291104252084,
+      "grad_norm": 0.9323732622379325,
+      "learning_rate": 7.023572359911641e-05,
+      "loss": 1.3708,
+      "step": 13205
+    },
+    {
+      "epoch": 0.6083751785138435,
+      "grad_norm": 1.1369032599473623,
+      "learning_rate": 7.022147929855639e-05,
+      "loss": 1.1453,
+      "step": 13206
+    },
+    {
+      "epoch": 0.6084212466024784,
+      "grad_norm": 1.0986983661881935,
+      "learning_rate": 7.020723566096408e-05,
+      "loss": 1.2832,
+      "step": 13207
+    },
+    {
+      "epoch": 0.6084673146911135,
+      "grad_norm": 1.0097244017509237,
+      "learning_rate": 7.019299268665652e-05,
+      "loss": 1.34,
+      "step": 13208
+    },
+    {
+      "epoch": 0.6085133827797484,
+      "grad_norm": 1.5402930159052675,
+      "learning_rate": 7.017875037595094e-05,
+      "loss": 1.1255,
+      "step": 13209
+    },
+    {
+      "epoch": 0.6085594508683835,
+      "grad_norm": 1.5007213318579402,
+      "learning_rate": 7.016450872916429e-05,
+      "loss": 1.203,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6086055189570184,
+      "grad_norm": 0.8264106991949374,
+      "learning_rate": 7.015026774661375e-05,
+      "loss": 1.1139,
+      "step": 13211
+    },
+    {
+      "epoch": 0.6086515870456535,
+      "grad_norm": 1.8704987919622684,
+      "learning_rate": 7.013602742861626e-05,
+      "loss": 1.3778,
+      "step": 13212
+    },
+    {
+      "epoch": 0.6086976551342885,
+      "grad_norm": 3.0198850912387645,
+      "learning_rate": 7.012178777548897e-05,
+      "loss": 0.9185,
+      "step": 13213
+    },
+    {
+      "epoch": 0.6087437232229235,
+      "grad_norm": 3.630276424106299,
+      "learning_rate": 7.01075487875488e-05,
+      "loss": 1.0438,
+      "step": 13214
+    },
+    {
+      "epoch": 0.6087897913115585,
+      "grad_norm": 0.8161129733267588,
+      "learning_rate": 7.009331046511285e-05,
+      "loss": 1.3441,
+      "step": 13215
+    },
+    {
+      "epoch": 0.6088358594001935,
+      "grad_norm": 1.426893232694783,
+      "learning_rate": 7.007907280849803e-05,
+      "loss": 1.1252,
+      "step": 13216
+    },
+    {
+      "epoch": 0.6088819274888285,
+      "grad_norm": 4.875455937087012,
+      "learning_rate": 7.006483581802132e-05,
+      "loss": 1.2798,
+      "step": 13217
+    },
+    {
+      "epoch": 0.6089279955774635,
+      "grad_norm": 1.0849751544606416,
+      "learning_rate": 7.005059949399979e-05,
+      "loss": 1.1957,
+      "step": 13218
+    },
+    {
+      "epoch": 0.6089740636660985,
+      "grad_norm": 1.1846419659567344,
+      "learning_rate": 7.003636383675023e-05,
+      "loss": 1.0271,
+      "step": 13219
+    },
+    {
+      "epoch": 0.6090201317547335,
+      "grad_norm": 1.6822008867542717,
+      "learning_rate": 7.002212884658973e-05,
+      "loss": 1.3039,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6090661998433685,
+      "grad_norm": 0.7323095488240248,
+      "learning_rate": 7.000789452383505e-05,
+      "loss": 1.649,
+      "step": 13221
+    },
+    {
+      "epoch": 0.6091122679320035,
+      "grad_norm": 1.6800782185717955,
+      "learning_rate": 6.999366086880326e-05,
+      "loss": 1.3695,
+      "step": 13222
+    },
+    {
+      "epoch": 0.6091583360206385,
+      "grad_norm": 0.6942941082369142,
+      "learning_rate": 6.997942788181108e-05,
+      "loss": 1.4975,
+      "step": 13223
+    },
+    {
+      "epoch": 0.6092044041092735,
+      "grad_norm": 0.9393553597567932,
+      "learning_rate": 6.996519556317553e-05,
+      "loss": 1.2305,
+      "step": 13224
+    },
+    {
+      "epoch": 0.6092504721979085,
+      "grad_norm": 1.4320522289642756,
+      "learning_rate": 6.995096391321337e-05,
+      "loss": 1.2098,
+      "step": 13225
+    },
+    {
+      "epoch": 0.6092965402865436,
+      "grad_norm": 4.289551003349874,
+      "learning_rate": 6.993673293224147e-05,
+      "loss": 1.3818,
+      "step": 13226
+    },
+    {
+      "epoch": 0.6093426083751785,
+      "grad_norm": 0.7637599168070861,
+      "learning_rate": 6.992250262057668e-05,
+      "loss": 1.3186,
+      "step": 13227
+    },
+    {
+      "epoch": 0.6093886764638136,
+      "grad_norm": 1.1924442219797142,
+      "learning_rate": 6.990827297853577e-05,
+      "loss": 1.1112,
+      "step": 13228
+    },
+    {
+      "epoch": 0.6094347445524485,
+      "grad_norm": 0.7811281736211354,
+      "learning_rate": 6.98940440064356e-05,
+      "loss": 1.2817,
+      "step": 13229
+    },
+    {
+      "epoch": 0.6094808126410836,
+      "grad_norm": 2.6936870376002093,
+      "learning_rate": 6.98798157045929e-05,
+      "loss": 0.9662,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6095268807297185,
+      "grad_norm": 0.9413758014252229,
+      "learning_rate": 6.986558807332446e-05,
+      "loss": 1.2548,
+      "step": 13231
+    },
+    {
+      "epoch": 0.6095729488183536,
+      "grad_norm": 0.6996143216708061,
+      "learning_rate": 6.9851361112947e-05,
+      "loss": 1.4671,
+      "step": 13232
+    },
+    {
+      "epoch": 0.6096190169069885,
+      "grad_norm": 0.8971922169945976,
+      "learning_rate": 6.983713482377736e-05,
+      "loss": 1.2782,
+      "step": 13233
+    },
+    {
+      "epoch": 0.6096650849956236,
+      "grad_norm": 0.9127896541070548,
+      "learning_rate": 6.982290920613217e-05,
+      "loss": 1.2343,
+      "step": 13234
+    },
+    {
+      "epoch": 0.6097111530842585,
+      "grad_norm": 1.0434013345629372,
+      "learning_rate": 6.980868426032813e-05,
+      "loss": 1.3203,
+      "step": 13235
+    },
+    {
+      "epoch": 0.6097572211728935,
+      "grad_norm": 1.0380035184662624,
+      "learning_rate": 6.979445998668202e-05,
+      "loss": 1.14,
+      "step": 13236
+    },
+    {
+      "epoch": 0.6098032892615285,
+      "grad_norm": 0.9714784078174558,
+      "learning_rate": 6.978023638551042e-05,
+      "loss": 0.9772,
+      "step": 13237
+    },
+    {
+      "epoch": 0.6098493573501635,
+      "grad_norm": 0.7734946790399673,
+      "learning_rate": 6.976601345713007e-05,
+      "loss": 1.0563,
+      "step": 13238
+    },
+    {
+      "epoch": 0.6098954254387986,
+      "grad_norm": 0.8217326010568663,
+      "learning_rate": 6.975179120185757e-05,
+      "loss": 1.4175,
+      "step": 13239
+    },
+    {
+      "epoch": 0.6099414935274335,
+      "grad_norm": 1.4146563808624661,
+      "learning_rate": 6.97375696200096e-05,
+      "loss": 1.259,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6099875616160686,
+      "grad_norm": 0.8435785997747631,
+      "learning_rate": 6.972334871190272e-05,
+      "loss": 1.3279,
+      "step": 13241
+    },
+    {
+      "epoch": 0.6100336297047035,
+      "grad_norm": 3.7294713243923323,
+      "learning_rate": 6.970912847785359e-05,
+      "loss": 1.4986,
+      "step": 13242
+    },
+    {
+      "epoch": 0.6100796977933386,
+      "grad_norm": 0.7396715756780871,
+      "learning_rate": 6.969490891817879e-05,
+      "loss": 1.3293,
+      "step": 13243
+    },
+    {
+      "epoch": 0.6101257658819735,
+      "grad_norm": 0.9432463406569311,
+      "learning_rate": 6.968069003319485e-05,
+      "loss": 1.0594,
+      "step": 13244
+    },
+    {
+      "epoch": 0.6101718339706086,
+      "grad_norm": 0.8282746420344341,
+      "learning_rate": 6.966647182321837e-05,
+      "loss": 1.278,
+      "step": 13245
+    },
+    {
+      "epoch": 0.6102179020592435,
+      "grad_norm": 0.8486946280583096,
+      "learning_rate": 6.965225428856587e-05,
+      "loss": 1.3601,
+      "step": 13246
+    },
+    {
+      "epoch": 0.6102639701478786,
+      "grad_norm": 1.4004906597980182,
+      "learning_rate": 6.963803742955392e-05,
+      "loss": 1.4316,
+      "step": 13247
+    },
+    {
+      "epoch": 0.6103100382365135,
+      "grad_norm": 0.7572994461330743,
+      "learning_rate": 6.9623821246499e-05,
+      "loss": 1.3855,
+      "step": 13248
+    },
+    {
+      "epoch": 0.6103561063251486,
+      "grad_norm": 0.7871668084408725,
+      "learning_rate": 6.960960573971763e-05,
+      "loss": 1.6255,
+      "step": 13249
+    },
+    {
+      "epoch": 0.6104021744137835,
+      "grad_norm": 2.3913784311416233,
+      "learning_rate": 6.959539090952625e-05,
+      "loss": 1.292,
+      "step": 13250
+    },
+    {
+      "epoch": 0.6104482425024186,
+      "grad_norm": 0.986524141279261,
+      "learning_rate": 6.958117675624137e-05,
+      "loss": 0.9218,
+      "step": 13251
+    },
+    {
+      "epoch": 0.6104943105910535,
+      "grad_norm": 1.8208397157737328,
+      "learning_rate": 6.956696328017942e-05,
+      "loss": 1.1326,
+      "step": 13252
+    },
+    {
+      "epoch": 0.6105403786796886,
+      "grad_norm": 2.602612024054502,
+      "learning_rate": 6.955275048165688e-05,
+      "loss": 1.4173,
+      "step": 13253
+    },
+    {
+      "epoch": 0.6105864467683236,
+      "grad_norm": 0.8700097966510407,
+      "learning_rate": 6.953853836099013e-05,
+      "loss": 1.6132,
+      "step": 13254
+    },
+    {
+      "epoch": 0.6106325148569586,
+      "grad_norm": 0.7922137529055783,
+      "learning_rate": 6.952432691849559e-05,
+      "loss": 0.9923,
+      "step": 13255
+    },
+    {
+      "epoch": 0.6106785829455936,
+      "grad_norm": 0.9955183592297623,
+      "learning_rate": 6.951011615448966e-05,
+      "loss": 1.3337,
+      "step": 13256
+    },
+    {
+      "epoch": 0.6107246510342286,
+      "grad_norm": 0.7228938152591731,
+      "learning_rate": 6.949590606928871e-05,
+      "loss": 1.3168,
+      "step": 13257
+    },
+    {
+      "epoch": 0.6107707191228636,
+      "grad_norm": 1.1176872659402297,
+      "learning_rate": 6.948169666320912e-05,
+      "loss": 1.1141,
+      "step": 13258
+    },
+    {
+      "epoch": 0.6108167872114986,
+      "grad_norm": 0.8395119493123381,
+      "learning_rate": 6.946748793656721e-05,
+      "loss": 1.1709,
+      "step": 13259
+    },
+    {
+      "epoch": 0.6108628553001336,
+      "grad_norm": 1.087264543258094,
+      "learning_rate": 6.945327988967935e-05,
+      "loss": 1.0728,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6109089233887686,
+      "grad_norm": 1.164541957925953,
+      "learning_rate": 6.943907252286182e-05,
+      "loss": 1.294,
+      "step": 13261
+    },
+    {
+      "epoch": 0.6109549914774036,
+      "grad_norm": 0.7032404101379237,
+      "learning_rate": 6.942486583643095e-05,
+      "loss": 1.2819,
+      "step": 13262
+    },
+    {
+      "epoch": 0.6110010595660386,
+      "grad_norm": 1.548930383309949,
+      "learning_rate": 6.9410659830703e-05,
+      "loss": 0.9803,
+      "step": 13263
+    },
+    {
+      "epoch": 0.6110471276546736,
+      "grad_norm": 1.719163664821074,
+      "learning_rate": 6.939645450599427e-05,
+      "loss": 1.1644,
+      "step": 13264
+    },
+    {
+      "epoch": 0.6110931957433086,
+      "grad_norm": 1.54948491182994,
+      "learning_rate": 6.938224986262101e-05,
+      "loss": 1.1029,
+      "step": 13265
+    },
+    {
+      "epoch": 0.6111392638319436,
+      "grad_norm": 1.0837715337730636,
+      "learning_rate": 6.936804590089945e-05,
+      "loss": 1.4102,
+      "step": 13266
+    },
+    {
+      "epoch": 0.6111853319205787,
+      "grad_norm": 0.9071785742883505,
+      "learning_rate": 6.935384262114584e-05,
+      "loss": 1.0281,
+      "step": 13267
+    },
+    {
+      "epoch": 0.6112314000092136,
+      "grad_norm": 0.7193572320614647,
+      "learning_rate": 6.933964002367636e-05,
+      "loss": 1.1832,
+      "step": 13268
+    },
+    {
+      "epoch": 0.6112774680978487,
+      "grad_norm": 0.6351036028372203,
+      "learning_rate": 6.932543810880724e-05,
+      "loss": 1.1973,
+      "step": 13269
+    },
+    {
+      "epoch": 0.6113235361864836,
+      "grad_norm": 0.9702506568216702,
+      "learning_rate": 6.931123687685463e-05,
+      "loss": 1.2844,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6113696042751187,
+      "grad_norm": 1.3327799554350708,
+      "learning_rate": 6.929703632813473e-05,
+      "loss": 1.2043,
+      "step": 13271
+    },
+    {
+      "epoch": 0.6114156723637536,
+      "grad_norm": 1.0179781181958554,
+      "learning_rate": 6.928283646296366e-05,
+      "loss": 1.121,
+      "step": 13272
+    },
+    {
+      "epoch": 0.6114617404523887,
+      "grad_norm": 0.9384545021080751,
+      "learning_rate": 6.926863728165755e-05,
+      "loss": 1.2,
+      "step": 13273
+    },
+    {
+      "epoch": 0.6115078085410236,
+      "grad_norm": 0.7808513007183828,
+      "learning_rate": 6.925443878453258e-05,
+      "loss": 1.49,
+      "step": 13274
+    },
+    {
+      "epoch": 0.6115538766296587,
+      "grad_norm": 0.9707731391356851,
+      "learning_rate": 6.924024097190476e-05,
+      "loss": 1.3748,
+      "step": 13275
+    },
+    {
+      "epoch": 0.6115999447182936,
+      "grad_norm": 0.719683800929407,
+      "learning_rate": 6.922604384409028e-05,
+      "loss": 1.4075,
+      "step": 13276
+    },
+    {
+      "epoch": 0.6116460128069287,
+      "grad_norm": 1.1239777671411952,
+      "learning_rate": 6.921184740140513e-05,
+      "loss": 1.498,
+      "step": 13277
+    },
+    {
+      "epoch": 0.6116920808955636,
+      "grad_norm": 4.240346836047595,
+      "learning_rate": 6.919765164416544e-05,
+      "loss": 1.2639,
+      "step": 13278
+    },
+    {
+      "epoch": 0.6117381489841986,
+      "grad_norm": 4.858343154305567,
+      "learning_rate": 6.918345657268719e-05,
+      "loss": 1.3728,
+      "step": 13279
+    },
+    {
+      "epoch": 0.6117842170728337,
+      "grad_norm": 1.7130262942230932,
+      "learning_rate": 6.916926218728648e-05,
+      "loss": 1.4743,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6118302851614686,
+      "grad_norm": 1.2387947183925436,
+      "learning_rate": 6.915506848827927e-05,
+      "loss": 1.5752,
+      "step": 13281
+    },
+    {
+      "epoch": 0.6118763532501037,
+      "grad_norm": 0.769736997951734,
+      "learning_rate": 6.914087547598155e-05,
+      "loss": 1.2527,
+      "step": 13282
+    },
+    {
+      "epoch": 0.6119224213387386,
+      "grad_norm": 0.730664516321497,
+      "learning_rate": 6.912668315070934e-05,
+      "loss": 1.5027,
+      "step": 13283
+    },
+    {
+      "epoch": 0.6119684894273737,
+      "grad_norm": 1.3499174466375037,
+      "learning_rate": 6.911249151277858e-05,
+      "loss": 1.2691,
+      "step": 13284
+    },
+    {
+      "epoch": 0.6120145575160086,
+      "grad_norm": 0.94145773001372,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 1.0874,
+      "step": 13285
+    },
+    {
+      "epoch": 0.6120606256046437,
+      "grad_norm": 1.1367460334583692,
+      "learning_rate": 6.908411030020527e-05,
+      "loss": 1.2597,
+      "step": 13286
+    },
+    {
+      "epoch": 0.6121066936932786,
+      "grad_norm": 0.7874866630564753,
+      "learning_rate": 6.906992072619459e-05,
+      "loss": 1.6324,
+      "step": 13287
+    },
+    {
+      "epoch": 0.6121527617819137,
+      "grad_norm": 0.6520964623700792,
+      "learning_rate": 6.905573184078906e-05,
+      "loss": 1.275,
+      "step": 13288
+    },
+    {
+      "epoch": 0.6121988298705486,
+      "grad_norm": 0.6591471343971633,
+      "learning_rate": 6.904154364430461e-05,
+      "loss": 1.3605,
+      "step": 13289
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 1.093230080090717,
+      "learning_rate": 6.902735613705712e-05,
+      "loss": 1.1861,
+      "step": 13290
+    },
+    {
+      "epoch": 0.6122909660478186,
+      "grad_norm": 1.3455233956534898,
+      "learning_rate": 6.901316931936242e-05,
+      "loss": 1.1247,
+      "step": 13291
+    },
+    {
+      "epoch": 0.6123370341364537,
+      "grad_norm": 2.1111774263909253,
+      "learning_rate": 6.89989831915364e-05,
+      "loss": 1.3017,
+      "step": 13292
+    },
+    {
+      "epoch": 0.6123831022250886,
+      "grad_norm": 1.1249920270344487,
+      "learning_rate": 6.898479775389482e-05,
+      "loss": 1.2775,
+      "step": 13293
+    },
+    {
+      "epoch": 0.6124291703137237,
+      "grad_norm": 1.5713424112354157,
+      "learning_rate": 6.897061300675358e-05,
+      "loss": 1.1847,
+      "step": 13294
+    },
+    {
+      "epoch": 0.6124752384023587,
+      "grad_norm": 0.8927089519488861,
+      "learning_rate": 6.89564289504284e-05,
+      "loss": 0.9918,
+      "step": 13295
+    },
+    {
+      "epoch": 0.6125213064909937,
+      "grad_norm": 0.9049507879244181,
+      "learning_rate": 6.894224558523515e-05,
+      "loss": 1.361,
+      "step": 13296
+    },
+    {
+      "epoch": 0.6125673745796287,
+      "grad_norm": 1.044636431566295,
+      "learning_rate": 6.89280629114895e-05,
+      "loss": 0.8907,
+      "step": 13297
+    },
+    {
+      "epoch": 0.6126134426682637,
+      "grad_norm": 0.9209650626991718,
+      "learning_rate": 6.891388092950727e-05,
+      "loss": 1.1404,
+      "step": 13298
+    },
+    {
+      "epoch": 0.6126595107568987,
+      "grad_norm": 1.0614206572414113,
+      "learning_rate": 6.889969963960423e-05,
+      "loss": 1.3114,
+      "step": 13299
+    },
+    {
+      "epoch": 0.6127055788455337,
+      "grad_norm": 0.7871912801693326,
+      "learning_rate": 6.888551904209597e-05,
+      "loss": 1.1727,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6127516469341687,
+      "grad_norm": 0.7907368134837554,
+      "learning_rate": 6.887133913729835e-05,
+      "loss": 1.2962,
+      "step": 13301
+    },
+    {
+      "epoch": 0.6127977150228037,
+      "grad_norm": 2.163121195277429,
+      "learning_rate": 6.885715992552692e-05,
+      "loss": 1.17,
+      "step": 13302
+    },
+    {
+      "epoch": 0.6128437831114387,
+      "grad_norm": 2.724692047671101,
+      "learning_rate": 6.884298140709746e-05,
+      "loss": 1.1037,
+      "step": 13303
+    },
+    {
+      "epoch": 0.6128898512000737,
+      "grad_norm": 0.6808286249074006,
+      "learning_rate": 6.88288035823256e-05,
+      "loss": 1.0144,
+      "step": 13304
+    },
+    {
+      "epoch": 0.6129359192887087,
+      "grad_norm": 1.1395668975089706,
+      "learning_rate": 6.881462645152697e-05,
+      "loss": 1.3568,
+      "step": 13305
+    },
+    {
+      "epoch": 0.6129819873773437,
+      "grad_norm": 1.1879393162354517,
+      "learning_rate": 6.880045001501721e-05,
+      "loss": 1.2501,
+      "step": 13306
+    },
+    {
+      "epoch": 0.6130280554659787,
+      "grad_norm": 0.7984045278687943,
+      "learning_rate": 6.878627427311195e-05,
+      "loss": 1.4732,
+      "step": 13307
+    },
+    {
+      "epoch": 0.6130741235546138,
+      "grad_norm": 1.1293248971446823,
+      "learning_rate": 6.87720992261268e-05,
+      "loss": 1.0134,
+      "step": 13308
+    },
+    {
+      "epoch": 0.6131201916432487,
+      "grad_norm": 1.534366688423286,
+      "learning_rate": 6.875792487437726e-05,
+      "loss": 1.305,
+      "step": 13309
+    },
+    {
+      "epoch": 0.6131662597318838,
+      "grad_norm": 1.0230530148216483,
+      "learning_rate": 6.8743751218179e-05,
+      "loss": 1.3326,
+      "step": 13310
+    },
+    {
+      "epoch": 0.6132123278205187,
+      "grad_norm": 2.8632729507296872,
+      "learning_rate": 6.872957825784747e-05,
+      "loss": 1.4462,
+      "step": 13311
+    },
+    {
+      "epoch": 0.6132583959091538,
+      "grad_norm": 0.6561230258916353,
+      "learning_rate": 6.871540599369832e-05,
+      "loss": 1.4158,
+      "step": 13312
+    },
+    {
+      "epoch": 0.6133044639977887,
+      "grad_norm": 2.634527783793921,
+      "learning_rate": 6.870123442604697e-05,
+      "loss": 1.2719,
+      "step": 13313
+    },
+    {
+      "epoch": 0.6133505320864238,
+      "grad_norm": 1.2428850745080036,
+      "learning_rate": 6.868706355520902e-05,
+      "loss": 1.288,
+      "step": 13314
+    },
+    {
+      "epoch": 0.6133966001750587,
+      "grad_norm": 4.101814654378553,
+      "learning_rate": 6.867289338149985e-05,
+      "loss": 1.2218,
+      "step": 13315
+    },
+    {
+      "epoch": 0.6134426682636938,
+      "grad_norm": 1.2353994595912272,
+      "learning_rate": 6.865872390523505e-05,
+      "loss": 1.507,
+      "step": 13316
+    },
+    {
+      "epoch": 0.6134887363523287,
+      "grad_norm": 0.8664480544940071,
+      "learning_rate": 6.864455512673e-05,
+      "loss": 1.3652,
+      "step": 13317
+    },
+    {
+      "epoch": 0.6135348044409638,
+      "grad_norm": 1.3711701636024234,
+      "learning_rate": 6.863038704630013e-05,
+      "loss": 1.1334,
+      "step": 13318
+    },
+    {
+      "epoch": 0.6135808725295987,
+      "grad_norm": 0.9195036207166053,
+      "learning_rate": 6.861621966426098e-05,
+      "loss": 1.277,
+      "step": 13319
+    },
+    {
+      "epoch": 0.6136269406182338,
+      "grad_norm": 3.3311970156109414,
+      "learning_rate": 6.860205298092781e-05,
+      "loss": 1.1857,
+      "step": 13320
+    },
+    {
+      "epoch": 0.6136730087068688,
+      "grad_norm": 2.1977084322694904,
+      "learning_rate": 6.858788699661616e-05,
+      "loss": 1.3743,
+      "step": 13321
+    },
+    {
+      "epoch": 0.6137190767955037,
+      "grad_norm": 1.6582580692783637,
+      "learning_rate": 6.857372171164128e-05,
+      "loss": 1.1653,
+      "step": 13322
+    },
+    {
+      "epoch": 0.6137651448841388,
+      "grad_norm": 0.7846887991185054,
+      "learning_rate": 6.855955712631868e-05,
+      "loss": 1.3539,
+      "step": 13323
+    },
+    {
+      "epoch": 0.6138112129727737,
+      "grad_norm": 2.1152476408172007,
+      "learning_rate": 6.854539324096354e-05,
+      "loss": 1.2657,
+      "step": 13324
+    },
+    {
+      "epoch": 0.6138572810614088,
+      "grad_norm": 0.8675170230649415,
+      "learning_rate": 6.853123005589137e-05,
+      "loss": 1.0677,
+      "step": 13325
+    },
+    {
+      "epoch": 0.6139033491500437,
+      "grad_norm": 0.7112203690769757,
+      "learning_rate": 6.851706757141738e-05,
+      "loss": 1.5708,
+      "step": 13326
+    },
+    {
+      "epoch": 0.6139494172386788,
+      "grad_norm": 1.3664504840736111,
+      "learning_rate": 6.850290578785687e-05,
+      "loss": 1.386,
+      "step": 13327
+    },
+    {
+      "epoch": 0.6139954853273137,
+      "grad_norm": 2.2235943396890443,
+      "learning_rate": 6.84887447055252e-05,
+      "loss": 1.105,
+      "step": 13328
+    },
+    {
+      "epoch": 0.6140415534159488,
+      "grad_norm": 3.121564908848187,
+      "learning_rate": 6.847458432473757e-05,
+      "loss": 1.3859,
+      "step": 13329
+    },
+    {
+      "epoch": 0.6140876215045837,
+      "grad_norm": 2.176547439585252,
+      "learning_rate": 6.846042464580927e-05,
+      "loss": 1.2048,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6141336895932188,
+      "grad_norm": 0.9954056939327236,
+      "learning_rate": 6.844626566905551e-05,
+      "loss": 1.1461,
+      "step": 13331
+    },
+    {
+      "epoch": 0.6141797576818537,
+      "grad_norm": 4.899543081626873,
+      "learning_rate": 6.84321073947916e-05,
+      "loss": 1.2991,
+      "step": 13332
+    },
+    {
+      "epoch": 0.6142258257704888,
+      "grad_norm": 1.3413721809962003,
+      "learning_rate": 6.841794982333262e-05,
+      "loss": 1.3102,
+      "step": 13333
+    },
+    {
+      "epoch": 0.6142718938591237,
+      "grad_norm": 0.8518283323330844,
+      "learning_rate": 6.840379295499392e-05,
+      "loss": 1.4514,
+      "step": 13334
+    },
+    {
+      "epoch": 0.6143179619477588,
+      "grad_norm": 2.8102620792298483,
+      "learning_rate": 6.838963679009056e-05,
+      "loss": 0.8899,
+      "step": 13335
+    },
+    {
+      "epoch": 0.6143640300363938,
+      "grad_norm": 2.2953887058599856,
+      "learning_rate": 6.837548132893772e-05,
+      "loss": 1.3441,
+      "step": 13336
+    },
+    {
+      "epoch": 0.6144100981250288,
+      "grad_norm": 0.8801636895575271,
+      "learning_rate": 6.836132657185057e-05,
+      "loss": 1.0614,
+      "step": 13337
+    },
+    {
+      "epoch": 0.6144561662136638,
+      "grad_norm": 1.104122214994102,
+      "learning_rate": 6.834717251914424e-05,
+      "loss": 1.4861,
+      "step": 13338
+    },
+    {
+      "epoch": 0.6145022343022988,
+      "grad_norm": 0.975869588531898,
+      "learning_rate": 6.833301917113384e-05,
+      "loss": 1.0371,
+      "step": 13339
+    },
+    {
+      "epoch": 0.6145483023909338,
+      "grad_norm": 1.1200944387478142,
+      "learning_rate": 6.831886652813446e-05,
+      "loss": 1.1803,
+      "step": 13340
+    },
+    {
+      "epoch": 0.6145943704795688,
+      "grad_norm": 0.9669062100384728,
+      "learning_rate": 6.830471459046121e-05,
+      "loss": 1.44,
+      "step": 13341
+    },
+    {
+      "epoch": 0.6146404385682038,
+      "grad_norm": 1.1237936897809884,
+      "learning_rate": 6.829056335842913e-05,
+      "loss": 1.0339,
+      "step": 13342
+    },
+    {
+      "epoch": 0.6146865066568388,
+      "grad_norm": 5.605194467327952,
+      "learning_rate": 6.82764128323533e-05,
+      "loss": 1.3085,
+      "step": 13343
+    },
+    {
+      "epoch": 0.6147325747454738,
+      "grad_norm": 0.6886103255568465,
+      "learning_rate": 6.826226301254874e-05,
+      "loss": 1.393,
+      "step": 13344
+    },
+    {
+      "epoch": 0.6147786428341088,
+      "grad_norm": 1.0843031499390032,
+      "learning_rate": 6.824811389933048e-05,
+      "loss": 1.3447,
+      "step": 13345
+    },
+    {
+      "epoch": 0.6148247109227438,
+      "grad_norm": 0.8968195912867766,
+      "learning_rate": 6.823396549301352e-05,
+      "loss": 1.4315,
+      "step": 13346
+    },
+    {
+      "epoch": 0.6148707790113788,
+      "grad_norm": 0.9409731490332026,
+      "learning_rate": 6.821981779391283e-05,
+      "loss": 1.1972,
+      "step": 13347
+    },
+    {
+      "epoch": 0.6149168471000138,
+      "grad_norm": 2.4057732261070464,
+      "learning_rate": 6.820567080234343e-05,
+      "loss": 1.2536,
+      "step": 13348
+    },
+    {
+      "epoch": 0.6149629151886489,
+      "grad_norm": 0.8446991897542357,
+      "learning_rate": 6.819152451862023e-05,
+      "loss": 1.2579,
+      "step": 13349
+    },
+    {
+      "epoch": 0.6150089832772838,
+      "grad_norm": 0.9737483513611707,
+      "learning_rate": 6.817737894305823e-05,
+      "loss": 1.3938,
+      "step": 13350
+    },
+    {
+      "epoch": 0.6150550513659189,
+      "grad_norm": 0.9286659046227738,
+      "learning_rate": 6.816323407597229e-05,
+      "loss": 1.2803,
+      "step": 13351
+    },
+    {
+      "epoch": 0.6151011194545538,
+      "grad_norm": 0.7543968913428762,
+      "learning_rate": 6.814908991767739e-05,
+      "loss": 1.3409,
+      "step": 13352
+    },
+    {
+      "epoch": 0.6151471875431889,
+      "grad_norm": 0.795531973818243,
+      "learning_rate": 6.813494646848838e-05,
+      "loss": 1.3428,
+      "step": 13353
+    },
+    {
+      "epoch": 0.6151932556318238,
+      "grad_norm": 1.0079326047249126,
+      "learning_rate": 6.812080372872013e-05,
+      "loss": 1.0265,
+      "step": 13354
+    },
+    {
+      "epoch": 0.6152393237204589,
+      "grad_norm": 0.8439586144788888,
+      "learning_rate": 6.810666169868753e-05,
+      "loss": 1.2763,
+      "step": 13355
+    },
+    {
+      "epoch": 0.6152853918090938,
+      "grad_norm": 1.113674551010065,
+      "learning_rate": 6.809252037870543e-05,
+      "loss": 1.3843,
+      "step": 13356
+    },
+    {
+      "epoch": 0.6153314598977289,
+      "grad_norm": 0.8380716982831953,
+      "learning_rate": 6.807837976908866e-05,
+      "loss": 1.3396,
+      "step": 13357
+    },
+    {
+      "epoch": 0.6153775279863638,
+      "grad_norm": 0.8792207006574836,
+      "learning_rate": 6.806423987015201e-05,
+      "loss": 1.1283,
+      "step": 13358
+    },
+    {
+      "epoch": 0.6154235960749989,
+      "grad_norm": 1.5126134910102058,
+      "learning_rate": 6.805010068221032e-05,
+      "loss": 1.5183,
+      "step": 13359
+    },
+    {
+      "epoch": 0.6154696641636338,
+      "grad_norm": 0.836256251944137,
+      "learning_rate": 6.803596220557834e-05,
+      "loss": 1.3162,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6155157322522689,
+      "grad_norm": 2.3241246863432097,
+      "learning_rate": 6.802182444057088e-05,
+      "loss": 1.286,
+      "step": 13361
+    },
+    {
+      "epoch": 0.6155618003409039,
+      "grad_norm": 0.8433718681500431,
+      "learning_rate": 6.800768738750266e-05,
+      "loss": 1.2051,
+      "step": 13362
+    },
+    {
+      "epoch": 0.6156078684295389,
+      "grad_norm": 1.1783760775086227,
+      "learning_rate": 6.799355104668841e-05,
+      "loss": 1.2691,
+      "step": 13363
+    },
+    {
+      "epoch": 0.6156539365181739,
+      "grad_norm": 1.6577915069558011,
+      "learning_rate": 6.797941541844288e-05,
+      "loss": 0.9402,
+      "step": 13364
+    },
+    {
+      "epoch": 0.6157000046068088,
+      "grad_norm": 0.9989998904373719,
+      "learning_rate": 6.796528050308074e-05,
+      "loss": 1.0471,
+      "step": 13365
+    },
+    {
+      "epoch": 0.6157460726954439,
+      "grad_norm": 0.7810818095179395,
+      "learning_rate": 6.795114630091672e-05,
+      "loss": 1.4542,
+      "step": 13366
+    },
+    {
+      "epoch": 0.6157921407840788,
+      "grad_norm": 3.517866281158768,
+      "learning_rate": 6.793701281226547e-05,
+      "loss": 1.4042,
+      "step": 13367
+    },
+    {
+      "epoch": 0.6158382088727139,
+      "grad_norm": 2.1120380568057353,
+      "learning_rate": 6.792288003744167e-05,
+      "loss": 1.2661,
+      "step": 13368
+    },
+    {
+      "epoch": 0.6158842769613488,
+      "grad_norm": 0.6273070952029967,
+      "learning_rate": 6.790874797675992e-05,
+      "loss": 1.3613,
+      "step": 13369
+    },
+    {
+      "epoch": 0.6159303450499839,
+      "grad_norm": 1.2257432604136562,
+      "learning_rate": 6.789461663053489e-05,
+      "loss": 1.4109,
+      "step": 13370
+    },
+    {
+      "epoch": 0.6159764131386188,
+      "grad_norm": 1.0354731886972144,
+      "learning_rate": 6.788048599908117e-05,
+      "loss": 1.1171,
+      "step": 13371
+    },
+    {
+      "epoch": 0.6160224812272539,
+      "grad_norm": 5.201908785726562,
+      "learning_rate": 6.786635608271335e-05,
+      "loss": 1.217,
+      "step": 13372
+    },
+    {
+      "epoch": 0.6160685493158888,
+      "grad_norm": 0.9166949567935571,
+      "learning_rate": 6.785222688174604e-05,
+      "loss": 1.2453,
+      "step": 13373
+    },
+    {
+      "epoch": 0.6161146174045239,
+      "grad_norm": 0.9774186493359162,
+      "learning_rate": 6.783809839649374e-05,
+      "loss": 1.1289,
+      "step": 13374
+    },
+    {
+      "epoch": 0.6161606854931589,
+      "grad_norm": 3.1921497127879674,
+      "learning_rate": 6.782397062727105e-05,
+      "loss": 1.0594,
+      "step": 13375
+    },
+    {
+      "epoch": 0.6162067535817939,
+      "grad_norm": 0.9761569825363047,
+      "learning_rate": 6.780984357439246e-05,
+      "loss": 1.227,
+      "step": 13376
+    },
+    {
+      "epoch": 0.6162528216704289,
+      "grad_norm": 0.7810877807538316,
+      "learning_rate": 6.779571723817255e-05,
+      "loss": 1.1797,
+      "step": 13377
+    },
+    {
+      "epoch": 0.6162988897590639,
+      "grad_norm": 1.2130776331154034,
+      "learning_rate": 6.778159161892575e-05,
+      "loss": 1.0968,
+      "step": 13378
+    },
+    {
+      "epoch": 0.6163449578476989,
+      "grad_norm": 1.1527585051987022,
+      "learning_rate": 6.776746671696659e-05,
+      "loss": 0.9565,
+      "step": 13379
+    },
+    {
+      "epoch": 0.6163910259363339,
+      "grad_norm": 0.7846510862801479,
+      "learning_rate": 6.775334253260948e-05,
+      "loss": 1.5921,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6164370940249689,
+      "grad_norm": 1.0933693499672896,
+      "learning_rate": 6.773921906616893e-05,
+      "loss": 1.4702,
+      "step": 13381
+    },
+    {
+      "epoch": 0.6164831621136039,
+      "grad_norm": 11.422337374964723,
+      "learning_rate": 6.772509631795938e-05,
+      "loss": 1.0214,
+      "step": 13382
+    },
+    {
+      "epoch": 0.6165292302022389,
+      "grad_norm": 2.232518283965765,
+      "learning_rate": 6.771097428829518e-05,
+      "loss": 0.9572,
+      "step": 13383
+    },
+    {
+      "epoch": 0.6165752982908739,
+      "grad_norm": 0.7592707911605449,
+      "learning_rate": 6.76968529774908e-05,
+      "loss": 1.254,
+      "step": 13384
+    },
+    {
+      "epoch": 0.6166213663795089,
+      "grad_norm": 1.6096848550056844,
+      "learning_rate": 6.768273238586057e-05,
+      "loss": 1.2629,
+      "step": 13385
+    },
+    {
+      "epoch": 0.6166674344681439,
+      "grad_norm": 0.8690624718401869,
+      "learning_rate": 6.766861251371892e-05,
+      "loss": 1.2101,
+      "step": 13386
+    },
+    {
+      "epoch": 0.6167135025567789,
+      "grad_norm": 1.7434570786463885,
+      "learning_rate": 6.765449336138015e-05,
+      "loss": 1.1461,
+      "step": 13387
+    },
+    {
+      "epoch": 0.6167595706454139,
+      "grad_norm": 0.6758135080795021,
+      "learning_rate": 6.764037492915864e-05,
+      "loss": 1.6466,
+      "step": 13388
+    },
+    {
+      "epoch": 0.6168056387340489,
+      "grad_norm": 0.6944043791100716,
+      "learning_rate": 6.762625721736871e-05,
+      "loss": 1.5102,
+      "step": 13389
+    },
+    {
+      "epoch": 0.616851706822684,
+      "grad_norm": 1.2214314650194396,
+      "learning_rate": 6.761214022632464e-05,
+      "loss": 0.9644,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6168977749113189,
+      "grad_norm": 1.1085009605051064,
+      "learning_rate": 6.759802395634073e-05,
+      "loss": 1.5033,
+      "step": 13391
+    },
+    {
+      "epoch": 0.616943842999954,
+      "grad_norm": 0.7525251314663917,
+      "learning_rate": 6.758390840773127e-05,
+      "loss": 1.5481,
+      "step": 13392
+    },
+    {
+      "epoch": 0.6169899110885889,
+      "grad_norm": 1.4151710122143737,
+      "learning_rate": 6.75697935808105e-05,
+      "loss": 1.3814,
+      "step": 13393
+    },
+    {
+      "epoch": 0.617035979177224,
+      "grad_norm": 0.8479709117071473,
+      "learning_rate": 6.755567947589268e-05,
+      "loss": 1.2882,
+      "step": 13394
+    },
+    {
+      "epoch": 0.6170820472658589,
+      "grad_norm": 1.315167067494623,
+      "learning_rate": 6.754156609329202e-05,
+      "loss": 1.1836,
+      "step": 13395
+    },
+    {
+      "epoch": 0.617128115354494,
+      "grad_norm": 1.1236836775416617,
+      "learning_rate": 6.752745343332271e-05,
+      "loss": 1.269,
+      "step": 13396
+    },
+    {
+      "epoch": 0.6171741834431289,
+      "grad_norm": 1.11528186285381,
+      "learning_rate": 6.751334149629901e-05,
+      "loss": 1.1772,
+      "step": 13397
+    },
+    {
+      "epoch": 0.617220251531764,
+      "grad_norm": 0.9741006417429234,
+      "learning_rate": 6.749923028253504e-05,
+      "loss": 1.5212,
+      "step": 13398
+    },
+    {
+      "epoch": 0.6172663196203989,
+      "grad_norm": 9.264232848432407,
+      "learning_rate": 6.7485119792345e-05,
+      "loss": 1.3656,
+      "step": 13399
+    },
+    {
+      "epoch": 0.617312387709034,
+      "grad_norm": 1.1098798345321519,
+      "learning_rate": 6.747101002604303e-05,
+      "loss": 1.3889,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6173584557976689,
+      "grad_norm": 0.9684660934729802,
+      "learning_rate": 6.745690098394319e-05,
+      "loss": 1.3819,
+      "step": 13401
+    },
+    {
+      "epoch": 0.617404523886304,
+      "grad_norm": 1.1540200218806143,
+      "learning_rate": 6.744279266635968e-05,
+      "loss": 1.3178,
+      "step": 13402
+    },
+    {
+      "epoch": 0.617450591974939,
+      "grad_norm": 0.9835550251034471,
+      "learning_rate": 6.742868507360656e-05,
+      "loss": 1.4871,
+      "step": 13403
+    },
+    {
+      "epoch": 0.617496660063574,
+      "grad_norm": 1.0030792672730204,
+      "learning_rate": 6.741457820599793e-05,
+      "loss": 1.1347,
+      "step": 13404
+    },
+    {
+      "epoch": 0.617542728152209,
+      "grad_norm": 1.374356767780746,
+      "learning_rate": 6.740047206384784e-05,
+      "loss": 1.363,
+      "step": 13405
+    },
+    {
+      "epoch": 0.617588796240844,
+      "grad_norm": 1.2787822975687106,
+      "learning_rate": 6.738636664747034e-05,
+      "loss": 1.1136,
+      "step": 13406
+    },
+    {
+      "epoch": 0.617634864329479,
+      "grad_norm": 1.4553224128694775,
+      "learning_rate": 6.737226195717948e-05,
+      "loss": 1.3099,
+      "step": 13407
+    },
+    {
+      "epoch": 0.617680932418114,
+      "grad_norm": 0.8318061565853824,
+      "learning_rate": 6.735815799328926e-05,
+      "loss": 1.1454,
+      "step": 13408
+    },
+    {
+      "epoch": 0.617727000506749,
+      "grad_norm": 1.3198416823726857,
+      "learning_rate": 6.734405475611369e-05,
+      "loss": 1.3923,
+      "step": 13409
+    },
+    {
+      "epoch": 0.617773068595384,
+      "grad_norm": 0.7027266586926346,
+      "learning_rate": 6.73299522459667e-05,
+      "loss": 1.2423,
+      "step": 13410
+    },
+    {
+      "epoch": 0.617819136684019,
+      "grad_norm": 1.0651406175009068,
+      "learning_rate": 6.731585046316238e-05,
+      "loss": 1.4962,
+      "step": 13411
+    },
+    {
+      "epoch": 0.6178652047726539,
+      "grad_norm": 1.5910359452518579,
+      "learning_rate": 6.730174940801452e-05,
+      "loss": 1.1852,
+      "step": 13412
+    },
+    {
+      "epoch": 0.617911272861289,
+      "grad_norm": 0.8706620431476781,
+      "learning_rate": 6.728764908083724e-05,
+      "loss": 1.3913,
+      "step": 13413
+    },
+    {
+      "epoch": 0.6179573409499239,
+      "grad_norm": 0.8666436794887068,
+      "learning_rate": 6.727354948194428e-05,
+      "loss": 0.9957,
+      "step": 13414
+    },
+    {
+      "epoch": 0.618003409038559,
+      "grad_norm": 1.337302768048026,
+      "learning_rate": 6.725945061164969e-05,
+      "loss": 1.085,
+      "step": 13415
+    },
+    {
+      "epoch": 0.618049477127194,
+      "grad_norm": 0.962054722718177,
+      "learning_rate": 6.724535247026724e-05,
+      "loss": 1.4447,
+      "step": 13416
+    },
+    {
+      "epoch": 0.618095545215829,
+      "grad_norm": 1.3007734826301198,
+      "learning_rate": 6.723125505811088e-05,
+      "loss": 1.3994,
+      "step": 13417
+    },
+    {
+      "epoch": 0.618141613304464,
+      "grad_norm": 1.6084862558597217,
+      "learning_rate": 6.721715837549449e-05,
+      "loss": 1.0859,
+      "step": 13418
+    },
+    {
+      "epoch": 0.618187681393099,
+      "grad_norm": 1.153426332634578,
+      "learning_rate": 6.720306242273178e-05,
+      "loss": 1.3347,
+      "step": 13419
+    },
+    {
+      "epoch": 0.618233749481734,
+      "grad_norm": 1.024078352061006,
+      "learning_rate": 6.71889672001367e-05,
+      "loss": 1.3684,
+      "step": 13420
+    },
+    {
+      "epoch": 0.618279817570369,
+      "grad_norm": 0.7364733025152164,
+      "learning_rate": 6.717487270802296e-05,
+      "loss": 1.4947,
+      "step": 13421
+    },
+    {
+      "epoch": 0.618325885659004,
+      "grad_norm": 0.9312838257710522,
+      "learning_rate": 6.716077894670445e-05,
+      "loss": 1.1574,
+      "step": 13422
+    },
+    {
+      "epoch": 0.618371953747639,
+      "grad_norm": 1.684795661766313,
+      "learning_rate": 6.714668591649482e-05,
+      "loss": 1.3805,
+      "step": 13423
+    },
+    {
+      "epoch": 0.618418021836274,
+      "grad_norm": 0.7352720595506848,
+      "learning_rate": 6.713259361770799e-05,
+      "loss": 1.3095,
+      "step": 13424
+    },
+    {
+      "epoch": 0.618464089924909,
+      "grad_norm": 0.9522544876195558,
+      "learning_rate": 6.71185020506575e-05,
+      "loss": 1.117,
+      "step": 13425
+    },
+    {
+      "epoch": 0.618510158013544,
+      "grad_norm": 1.4424134911744892,
+      "learning_rate": 6.710441121565728e-05,
+      "loss": 1.1263,
+      "step": 13426
+    },
+    {
+      "epoch": 0.618556226102179,
+      "grad_norm": 1.773724449516095,
+      "learning_rate": 6.709032111302092e-05,
+      "loss": 1.2343,
+      "step": 13427
+    },
+    {
+      "epoch": 0.618602294190814,
+      "grad_norm": 0.8508260729817294,
+      "learning_rate": 6.707623174306208e-05,
+      "loss": 1.0744,
+      "step": 13428
+    },
+    {
+      "epoch": 0.618648362279449,
+      "grad_norm": 0.9720324909595056,
+      "learning_rate": 6.706214310609455e-05,
+      "loss": 1.078,
+      "step": 13429
+    },
+    {
+      "epoch": 0.618694430368084,
+      "grad_norm": 0.737864878222014,
+      "learning_rate": 6.704805520243186e-05,
+      "loss": 1.4095,
+      "step": 13430
+    },
+    {
+      "epoch": 0.6187404984567191,
+      "grad_norm": 0.8155385992321011,
+      "learning_rate": 6.703396803238779e-05,
+      "loss": 1.5418,
+      "step": 13431
+    },
+    {
+      "epoch": 0.618786566545354,
+      "grad_norm": 1.1183426143284856,
+      "learning_rate": 6.701988159627583e-05,
+      "loss": 1.0841,
+      "step": 13432
+    },
+    {
+      "epoch": 0.6188326346339891,
+      "grad_norm": 0.8308878836107862,
+      "learning_rate": 6.700579589440972e-05,
+      "loss": 1.391,
+      "step": 13433
+    },
+    {
+      "epoch": 0.618878702722624,
+      "grad_norm": 1.8670571382137737,
+      "learning_rate": 6.699171092710294e-05,
+      "loss": 0.9724,
+      "step": 13434
+    },
+    {
+      "epoch": 0.6189247708112591,
+      "grad_norm": 0.8799263839880389,
+      "learning_rate": 6.697762669466916e-05,
+      "loss": 1.3511,
+      "step": 13435
+    },
+    {
+      "epoch": 0.618970838899894,
+      "grad_norm": 1.0567525720878321,
+      "learning_rate": 6.696354319742191e-05,
+      "loss": 1.1324,
+      "step": 13436
+    },
+    {
+      "epoch": 0.6190169069885291,
+      "grad_norm": 1.1686636741839374,
+      "learning_rate": 6.694946043567467e-05,
+      "loss": 1.1786,
+      "step": 13437
+    },
+    {
+      "epoch": 0.619062975077164,
+      "grad_norm": 0.9145560012527746,
+      "learning_rate": 6.693537840974106e-05,
+      "loss": 1.4996,
+      "step": 13438
+    },
+    {
+      "epoch": 0.6191090431657991,
+      "grad_norm": 1.1804936089105296,
+      "learning_rate": 6.692129711993455e-05,
+      "loss": 1.3422,
+      "step": 13439
+    },
+    {
+      "epoch": 0.619155111254434,
+      "grad_norm": 1.6517515308937054,
+      "learning_rate": 6.690721656656865e-05,
+      "loss": 1.4403,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6192011793430691,
+      "grad_norm": 1.3644060342828996,
+      "learning_rate": 6.68931367499568e-05,
+      "loss": 1.2801,
+      "step": 13441
+    },
+    {
+      "epoch": 0.619247247431704,
+      "grad_norm": 1.0595866897485697,
+      "learning_rate": 6.687905767041253e-05,
+      "loss": 1.1586,
+      "step": 13442
+    },
+    {
+      "epoch": 0.6192933155203391,
+      "grad_norm": 2.8688660029695043,
+      "learning_rate": 6.68649793282492e-05,
+      "loss": 0.9369,
+      "step": 13443
+    },
+    {
+      "epoch": 0.6193393836089741,
+      "grad_norm": 0.7626530697615387,
+      "learning_rate": 6.685090172378037e-05,
+      "loss": 1.26,
+      "step": 13444
+    },
+    {
+      "epoch": 0.6193854516976091,
+      "grad_norm": 1.6603104045780002,
+      "learning_rate": 6.683682485731934e-05,
+      "loss": 1.2905,
+      "step": 13445
+    },
+    {
+      "epoch": 0.6194315197862441,
+      "grad_norm": 0.7945794243938819,
+      "learning_rate": 6.682274872917952e-05,
+      "loss": 1.3938,
+      "step": 13446
+    },
+    {
+      "epoch": 0.619477587874879,
+      "grad_norm": 0.9442911856931921,
+      "learning_rate": 6.680867333967433e-05,
+      "loss": 1.0506,
+      "step": 13447
+    },
+    {
+      "epoch": 0.6195236559635141,
+      "grad_norm": 1.2660681176211384,
+      "learning_rate": 6.67945986891171e-05,
+      "loss": 1.0755,
+      "step": 13448
+    },
+    {
+      "epoch": 0.619569724052149,
+      "grad_norm": 1.10586461967891,
+      "learning_rate": 6.678052477782124e-05,
+      "loss": 1.2673,
+      "step": 13449
+    },
+    {
+      "epoch": 0.6196157921407841,
+      "grad_norm": 0.8720031756257655,
+      "learning_rate": 6.676645160609997e-05,
+      "loss": 1.3224,
+      "step": 13450
+    },
+    {
+      "epoch": 0.619661860229419,
+      "grad_norm": 1.537320935369197,
+      "learning_rate": 6.675237917426673e-05,
+      "loss": 1.3178,
+      "step": 13451
+    },
+    {
+      "epoch": 0.6197079283180541,
+      "grad_norm": 2.481966020292975,
+      "learning_rate": 6.673830748263471e-05,
+      "loss": 1.4435,
+      "step": 13452
+    },
+    {
+      "epoch": 0.619753996406689,
+      "grad_norm": 0.89205446325058,
+      "learning_rate": 6.67242365315173e-05,
+      "loss": 1.2233,
+      "step": 13453
+    },
+    {
+      "epoch": 0.6198000644953241,
+      "grad_norm": 0.8806401725386387,
+      "learning_rate": 6.671016632122768e-05,
+      "loss": 1.1554,
+      "step": 13454
+    },
+    {
+      "epoch": 0.619846132583959,
+      "grad_norm": 0.6861683540336454,
+      "learning_rate": 6.669609685207912e-05,
+      "loss": 1.2419,
+      "step": 13455
+    },
+    {
+      "epoch": 0.6198922006725941,
+      "grad_norm": 1.9514060464988792,
+      "learning_rate": 6.668202812438487e-05,
+      "loss": 1.2453,
+      "step": 13456
+    },
+    {
+      "epoch": 0.6199382687612291,
+      "grad_norm": 0.9619677254950245,
+      "learning_rate": 6.666796013845812e-05,
+      "loss": 1.027,
+      "step": 13457
+    },
+    {
+      "epoch": 0.6199843368498641,
+      "grad_norm": 0.8293089742840293,
+      "learning_rate": 6.66538928946121e-05,
+      "loss": 1.1693,
+      "step": 13458
+    },
+    {
+      "epoch": 0.6200304049384991,
+      "grad_norm": 0.8208721287874835,
+      "learning_rate": 6.663982639315997e-05,
+      "loss": 1.3828,
+      "step": 13459
+    },
+    {
+      "epoch": 0.6200764730271341,
+      "grad_norm": 0.9290375253138852,
+      "learning_rate": 6.662576063441493e-05,
+      "loss": 0.9721,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6201225411157691,
+      "grad_norm": 0.842857401504628,
+      "learning_rate": 6.661169561869005e-05,
+      "loss": 1.4462,
+      "step": 13461
+    },
+    {
+      "epoch": 0.6201686092044041,
+      "grad_norm": 0.9374883888652258,
+      "learning_rate": 6.659763134629858e-05,
+      "loss": 1.494,
+      "step": 13462
+    },
+    {
+      "epoch": 0.6202146772930391,
+      "grad_norm": 0.9614523013157457,
+      "learning_rate": 6.658356781755355e-05,
+      "loss": 1.1355,
+      "step": 13463
+    },
+    {
+      "epoch": 0.6202607453816741,
+      "grad_norm": 0.8799347938053359,
+      "learning_rate": 6.656950503276807e-05,
+      "loss": 1.2545,
+      "step": 13464
+    },
+    {
+      "epoch": 0.6203068134703091,
+      "grad_norm": 1.0527910520974588,
+      "learning_rate": 6.655544299225527e-05,
+      "loss": 1.4613,
+      "step": 13465
+    },
+    {
+      "epoch": 0.6203528815589441,
+      "grad_norm": 1.2538169199338078,
+      "learning_rate": 6.654138169632815e-05,
+      "loss": 1.2658,
+      "step": 13466
+    },
+    {
+      "epoch": 0.6203989496475791,
+      "grad_norm": 0.9558012303022352,
+      "learning_rate": 6.652732114529983e-05,
+      "loss": 1.2769,
+      "step": 13467
+    },
+    {
+      "epoch": 0.6204450177362141,
+      "grad_norm": 0.7400824180154902,
+      "learning_rate": 6.651326133948328e-05,
+      "loss": 1.5496,
+      "step": 13468
+    },
+    {
+      "epoch": 0.6204910858248491,
+      "grad_norm": 3.797230415559366,
+      "learning_rate": 6.649920227919158e-05,
+      "loss": 1.1529,
+      "step": 13469
+    },
+    {
+      "epoch": 0.6205371539134841,
+      "grad_norm": 0.7988599466628991,
+      "learning_rate": 6.648514396473767e-05,
+      "loss": 1.0893,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6205832220021191,
+      "grad_norm": 8.402038082067545,
+      "learning_rate": 6.64710863964346e-05,
+      "loss": 1.0829,
+      "step": 13471
+    },
+    {
+      "epoch": 0.6206292900907542,
+      "grad_norm": 1.1899278118501955,
+      "learning_rate": 6.645702957459529e-05,
+      "loss": 1.1758,
+      "step": 13472
+    },
+    {
+      "epoch": 0.6206753581793891,
+      "grad_norm": 1.4347599621324763,
+      "learning_rate": 6.644297349953268e-05,
+      "loss": 1.3547,
+      "step": 13473
+    },
+    {
+      "epoch": 0.6207214262680242,
+      "grad_norm": 0.8947401321992032,
+      "learning_rate": 6.642891817155975e-05,
+      "loss": 1.258,
+      "step": 13474
+    },
+    {
+      "epoch": 0.6207674943566591,
+      "grad_norm": 9.243871146442999,
+      "learning_rate": 6.641486359098939e-05,
+      "loss": 1.1494,
+      "step": 13475
+    },
+    {
+      "epoch": 0.6208135624452942,
+      "grad_norm": 1.2482763756041975,
+      "learning_rate": 6.64008097581345e-05,
+      "loss": 1.3302,
+      "step": 13476
+    },
+    {
+      "epoch": 0.6208596305339291,
+      "grad_norm": 1.5399669177072661,
+      "learning_rate": 6.638675667330797e-05,
+      "loss": 1.1089,
+      "step": 13477
+    },
+    {
+      "epoch": 0.6209056986225642,
+      "grad_norm": 1.355155046372048,
+      "learning_rate": 6.637270433682268e-05,
+      "loss": 1.0744,
+      "step": 13478
+    },
+    {
+      "epoch": 0.6209517667111991,
+      "grad_norm": 1.1882249501261173,
+      "learning_rate": 6.635865274899145e-05,
+      "loss": 1.327,
+      "step": 13479
+    },
+    {
+      "epoch": 0.6209978347998342,
+      "grad_norm": 37.82507232825206,
+      "learning_rate": 6.634460191012714e-05,
+      "loss": 1.3594,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6210439028884691,
+      "grad_norm": 0.991291876107477,
+      "learning_rate": 6.633055182054257e-05,
+      "loss": 1.0804,
+      "step": 13481
+    },
+    {
+      "epoch": 0.6210899709771042,
+      "grad_norm": 0.9255128941323468,
+      "learning_rate": 6.631650248055053e-05,
+      "loss": 1.5157,
+      "step": 13482
+    },
+    {
+      "epoch": 0.6211360390657391,
+      "grad_norm": 1.0305412865744352,
+      "learning_rate": 6.63024538904638e-05,
+      "loss": 1.4664,
+      "step": 13483
+    },
+    {
+      "epoch": 0.6211821071543742,
+      "grad_norm": 1.6902241956341086,
+      "learning_rate": 6.628840605059517e-05,
+      "loss": 1.1104,
+      "step": 13484
+    },
+    {
+      "epoch": 0.6212281752430092,
+      "grad_norm": 0.9207526995753814,
+      "learning_rate": 6.627435896125738e-05,
+      "loss": 1.1425,
+      "step": 13485
+    },
+    {
+      "epoch": 0.6212742433316442,
+      "grad_norm": 1.3200301968956947,
+      "learning_rate": 6.626031262276313e-05,
+      "loss": 1.5889,
+      "step": 13486
+    },
+    {
+      "epoch": 0.6213203114202792,
+      "grad_norm": 1.3039932889315595,
+      "learning_rate": 6.62462670354252e-05,
+      "loss": 1.3532,
+      "step": 13487
+    },
+    {
+      "epoch": 0.6213663795089142,
+      "grad_norm": 1.3452778665139884,
+      "learning_rate": 6.623222219955624e-05,
+      "loss": 1.0824,
+      "step": 13488
+    },
+    {
+      "epoch": 0.6214124475975492,
+      "grad_norm": 2.77070755398761,
+      "learning_rate": 6.621817811546897e-05,
+      "loss": 1.4706,
+      "step": 13489
+    },
+    {
+      "epoch": 0.6214585156861842,
+      "grad_norm": 4.107625653119602,
+      "learning_rate": 6.620413478347605e-05,
+      "loss": 1.3255,
+      "step": 13490
+    },
+    {
+      "epoch": 0.6215045837748192,
+      "grad_norm": 1.225784199771488,
+      "learning_rate": 6.619009220389007e-05,
+      "loss": 1.1346,
+      "step": 13491
+    },
+    {
+      "epoch": 0.6215506518634542,
+      "grad_norm": 1.9524392415258864,
+      "learning_rate": 6.617605037702377e-05,
+      "loss": 1.2495,
+      "step": 13492
+    },
+    {
+      "epoch": 0.6215967199520892,
+      "grad_norm": 1.616714111902824,
+      "learning_rate": 6.616200930318968e-05,
+      "loss": 1.4794,
+      "step": 13493
+    },
+    {
+      "epoch": 0.6216427880407241,
+      "grad_norm": 1.2639982230368443,
+      "learning_rate": 6.614796898270046e-05,
+      "loss": 1.1899,
+      "step": 13494
+    },
+    {
+      "epoch": 0.6216888561293592,
+      "grad_norm": 1.3302957123899093,
+      "learning_rate": 6.613392941586862e-05,
+      "loss": 1.3852,
+      "step": 13495
+    },
+    {
+      "epoch": 0.6217349242179941,
+      "grad_norm": 1.6074114089536036,
+      "learning_rate": 6.611989060300682e-05,
+      "loss": 1.5561,
+      "step": 13496
+    },
+    {
+      "epoch": 0.6217809923066292,
+      "grad_norm": 3.1070956908685323,
+      "learning_rate": 6.610585254442754e-05,
+      "loss": 1.2812,
+      "step": 13497
+    },
+    {
+      "epoch": 0.6218270603952643,
+      "grad_norm": 0.8519200198280961,
+      "learning_rate": 6.609181524044335e-05,
+      "loss": 1.3767,
+      "step": 13498
+    },
+    {
+      "epoch": 0.6218731284838992,
+      "grad_norm": 2.234029819293363,
+      "learning_rate": 6.607777869136674e-05,
+      "loss": 1.3143,
+      "step": 13499
+    },
+    {
+      "epoch": 0.6219191965725342,
+      "grad_norm": 2.2678123845418527,
+      "learning_rate": 6.606374289751025e-05,
+      "loss": 1.333,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6219652646611692,
+      "grad_norm": 0.8938766669748149,
+      "learning_rate": 6.604970785918632e-05,
+      "loss": 1.6897,
+      "step": 13501
+    },
+    {
+      "epoch": 0.6220113327498042,
+      "grad_norm": 1.1108685396207179,
+      "learning_rate": 6.603567357670742e-05,
+      "loss": 1.4134,
+      "step": 13502
+    },
+    {
+      "epoch": 0.6220574008384392,
+      "grad_norm": 1.5840748104400024,
+      "learning_rate": 6.602164005038603e-05,
+      "loss": 1.396,
+      "step": 13503
+    },
+    {
+      "epoch": 0.6221034689270742,
+      "grad_norm": 0.9831001902671654,
+      "learning_rate": 6.600760728053454e-05,
+      "loss": 1.2926,
+      "step": 13504
+    },
+    {
+      "epoch": 0.6221495370157092,
+      "grad_norm": 2.7729701399531166,
+      "learning_rate": 6.599357526746543e-05,
+      "loss": 1.3065,
+      "step": 13505
+    },
+    {
+      "epoch": 0.6221956051043442,
+      "grad_norm": 0.8972027099400701,
+      "learning_rate": 6.597954401149102e-05,
+      "loss": 1.4747,
+      "step": 13506
+    },
+    {
+      "epoch": 0.6222416731929792,
+      "grad_norm": 2.2256412196740145,
+      "learning_rate": 6.596551351292375e-05,
+      "loss": 1.2847,
+      "step": 13507
+    },
+    {
+      "epoch": 0.6222877412816142,
+      "grad_norm": 1.27436282311483,
+      "learning_rate": 6.595148377207595e-05,
+      "loss": 1.2118,
+      "step": 13508
+    },
+    {
+      "epoch": 0.6223338093702492,
+      "grad_norm": 0.9139742640646623,
+      "learning_rate": 6.593745478926e-05,
+      "loss": 1.244,
+      "step": 13509
+    },
+    {
+      "epoch": 0.6223798774588842,
+      "grad_norm": 1.2163978957356414,
+      "learning_rate": 6.592342656478824e-05,
+      "loss": 1.0823,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6224259455475193,
+      "grad_norm": 1.2150523437237666,
+      "learning_rate": 6.590939909897289e-05,
+      "loss": 1.5464,
+      "step": 13511
+    },
+    {
+      "epoch": 0.6224720136361542,
+      "grad_norm": 1.0447049365673717,
+      "learning_rate": 6.589537239212638e-05,
+      "loss": 1.7271,
+      "step": 13512
+    },
+    {
+      "epoch": 0.6225180817247893,
+      "grad_norm": 1.5703802116939245,
+      "learning_rate": 6.588134644456088e-05,
+      "loss": 1.3138,
+      "step": 13513
+    },
+    {
+      "epoch": 0.6225641498134242,
+      "grad_norm": 0.8658247101250446,
+      "learning_rate": 6.586732125658874e-05,
+      "loss": 1.511,
+      "step": 13514
+    },
+    {
+      "epoch": 0.6226102179020593,
+      "grad_norm": 1.7904268864537884,
+      "learning_rate": 6.585329682852211e-05,
+      "loss": 1.5929,
+      "step": 13515
+    },
+    {
+      "epoch": 0.6226562859906942,
+      "grad_norm": 1.9012120315624925,
+      "learning_rate": 6.583927316067333e-05,
+      "loss": 1.1034,
+      "step": 13516
+    },
+    {
+      "epoch": 0.6227023540793293,
+      "grad_norm": 0.8907668557854188,
+      "learning_rate": 6.582525025335453e-05,
+      "loss": 1.2301,
+      "step": 13517
+    },
+    {
+      "epoch": 0.6227484221679642,
+      "grad_norm": 0.9116248269687917,
+      "learning_rate": 6.581122810687795e-05,
+      "loss": 1.4004,
+      "step": 13518
+    },
+    {
+      "epoch": 0.6227944902565993,
+      "grad_norm": 1.8912094964610358,
+      "learning_rate": 6.579720672155579e-05,
+      "loss": 1.3677,
+      "step": 13519
+    },
+    {
+      "epoch": 0.6228405583452342,
+      "grad_norm": 0.893873797672546,
+      "learning_rate": 6.57831860977001e-05,
+      "loss": 1.5429,
+      "step": 13520
+    },
+    {
+      "epoch": 0.6228866264338693,
+      "grad_norm": 1.0213311001107452,
+      "learning_rate": 6.576916623562318e-05,
+      "loss": 1.5074,
+      "step": 13521
+    },
+    {
+      "epoch": 0.6229326945225042,
+      "grad_norm": 0.7454507003835683,
+      "learning_rate": 6.575514713563702e-05,
+      "loss": 1.3849,
+      "step": 13522
+    },
+    {
+      "epoch": 0.6229787626111393,
+      "grad_norm": 1.1012147542154773,
+      "learning_rate": 6.574112879805383e-05,
+      "loss": 1.334,
+      "step": 13523
+    },
+    {
+      "epoch": 0.6230248306997742,
+      "grad_norm": 0.8771096954322487,
+      "learning_rate": 6.572711122318563e-05,
+      "loss": 1.3208,
+      "step": 13524
+    },
+    {
+      "epoch": 0.6230708987884093,
+      "grad_norm": 1.222567406642179,
+      "learning_rate": 6.571309441134459e-05,
+      "loss": 1.1692,
+      "step": 13525
+    },
+    {
+      "epoch": 0.6231169668770443,
+      "grad_norm": 1.031141450904415,
+      "learning_rate": 6.569907836284263e-05,
+      "loss": 1.2521,
+      "step": 13526
+    },
+    {
+      "epoch": 0.6231630349656793,
+      "grad_norm": 0.951774622709638,
+      "learning_rate": 6.568506307799196e-05,
+      "loss": 1.0612,
+      "step": 13527
+    },
+    {
+      "epoch": 0.6232091030543143,
+      "grad_norm": 0.9426421783972867,
+      "learning_rate": 6.56710485571045e-05,
+      "loss": 1.3013,
+      "step": 13528
+    },
+    {
+      "epoch": 0.6232551711429493,
+      "grad_norm": 0.7366761387778886,
+      "learning_rate": 6.565703480049224e-05,
+      "loss": 1.3681,
+      "step": 13529
+    },
+    {
+      "epoch": 0.6233012392315843,
+      "grad_norm": 0.856256663745002,
+      "learning_rate": 6.564302180846727e-05,
+      "loss": 1.2161,
+      "step": 13530
+    },
+    {
+      "epoch": 0.6233473073202193,
+      "grad_norm": 1.8751635359860814,
+      "learning_rate": 6.562900958134146e-05,
+      "loss": 1.4406,
+      "step": 13531
+    },
+    {
+      "epoch": 0.6233933754088543,
+      "grad_norm": 1.662030137313673,
+      "learning_rate": 6.561499811942686e-05,
+      "loss": 1.1709,
+      "step": 13532
+    },
+    {
+      "epoch": 0.6234394434974893,
+      "grad_norm": 0.9520227414679967,
+      "learning_rate": 6.560098742303533e-05,
+      "loss": 1.4364,
+      "step": 13533
+    },
+    {
+      "epoch": 0.6234855115861243,
+      "grad_norm": 1.571843973010513,
+      "learning_rate": 6.558697749247888e-05,
+      "loss": 1.1765,
+      "step": 13534
+    },
+    {
+      "epoch": 0.6235315796747593,
+      "grad_norm": 1.4270336197503029,
+      "learning_rate": 6.557296832806931e-05,
+      "loss": 1.386,
+      "step": 13535
+    },
+    {
+      "epoch": 0.6235776477633943,
+      "grad_norm": 0.9332294577200445,
+      "learning_rate": 6.555895993011864e-05,
+      "loss": 1.2205,
+      "step": 13536
+    },
+    {
+      "epoch": 0.6236237158520292,
+      "grad_norm": 1.3401410959040998,
+      "learning_rate": 6.554495229893866e-05,
+      "loss": 1.2029,
+      "step": 13537
+    },
+    {
+      "epoch": 0.6236697839406643,
+      "grad_norm": 0.8896442116741048,
+      "learning_rate": 6.553094543484119e-05,
+      "loss": 1.3453,
+      "step": 13538
+    },
+    {
+      "epoch": 0.6237158520292994,
+      "grad_norm": 0.8120042102099033,
+      "learning_rate": 6.551693933813817e-05,
+      "loss": 1.0278,
+      "step": 13539
+    },
+    {
+      "epoch": 0.6237619201179343,
+      "grad_norm": 1.6754637081945658,
+      "learning_rate": 6.550293400914133e-05,
+      "loss": 1.103,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6238079882065694,
+      "grad_norm": 1.3799069700736772,
+      "learning_rate": 6.548892944816254e-05,
+      "loss": 1.3312,
+      "step": 13541
+    },
+    {
+      "epoch": 0.6238540562952043,
+      "grad_norm": 2.26141932577233,
+      "learning_rate": 6.547492565551353e-05,
+      "loss": 1.5585,
+      "step": 13542
+    },
+    {
+      "epoch": 0.6239001243838393,
+      "grad_norm": 1.3059860002387063,
+      "learning_rate": 6.546092263150616e-05,
+      "loss": 1.5378,
+      "step": 13543
+    },
+    {
+      "epoch": 0.6239461924724743,
+      "grad_norm": 1.274003850858107,
+      "learning_rate": 6.544692037645206e-05,
+      "loss": 1.398,
+      "step": 13544
+    },
+    {
+      "epoch": 0.6239922605611093,
+      "grad_norm": 3.7013958001106624,
+      "learning_rate": 6.54329188906631e-05,
+      "loss": 1.3342,
+      "step": 13545
+    },
+    {
+      "epoch": 0.6240383286497443,
+      "grad_norm": 0.818558604036975,
+      "learning_rate": 6.541891817445089e-05,
+      "loss": 1.5968,
+      "step": 13546
+    },
+    {
+      "epoch": 0.6240843967383793,
+      "grad_norm": 1.479295411151786,
+      "learning_rate": 6.540491822812717e-05,
+      "loss": 1.2153,
+      "step": 13547
+    },
+    {
+      "epoch": 0.6241304648270143,
+      "grad_norm": 1.045422571003721,
+      "learning_rate": 6.539091905200366e-05,
+      "loss": 1.2518,
+      "step": 13548
+    },
+    {
+      "epoch": 0.6241765329156493,
+      "grad_norm": 0.9382668183080013,
+      "learning_rate": 6.537692064639195e-05,
+      "loss": 1.3907,
+      "step": 13549
+    },
+    {
+      "epoch": 0.6242226010042843,
+      "grad_norm": 0.8716153884831983,
+      "learning_rate": 6.536292301160376e-05,
+      "loss": 1.4494,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6242686690929193,
+      "grad_norm": 1.049832873440569,
+      "learning_rate": 6.53489261479507e-05,
+      "loss": 1.3271,
+      "step": 13551
+    },
+    {
+      "epoch": 0.6243147371815544,
+      "grad_norm": 3.4629131422184063,
+      "learning_rate": 6.533493005574438e-05,
+      "loss": 1.6044,
+      "step": 13552
+    },
+    {
+      "epoch": 0.6243608052701893,
+      "grad_norm": 1.1054287456741898,
+      "learning_rate": 6.532093473529639e-05,
+      "loss": 1.2701,
+      "step": 13553
+    },
+    {
+      "epoch": 0.6244068733588244,
+      "grad_norm": 1.0088334437071813,
+      "learning_rate": 6.530694018691836e-05,
+      "loss": 1.23,
+      "step": 13554
+    },
+    {
+      "epoch": 0.6244529414474593,
+      "grad_norm": 2.4604226022438818,
+      "learning_rate": 6.529294641092181e-05,
+      "loss": 1.3169,
+      "step": 13555
+    },
+    {
+      "epoch": 0.6244990095360944,
+      "grad_norm": 1.4509472918771964,
+      "learning_rate": 6.527895340761828e-05,
+      "loss": 1.3339,
+      "step": 13556
+    },
+    {
+      "epoch": 0.6245450776247293,
+      "grad_norm": 2.1679709768163455,
+      "learning_rate": 6.526496117731934e-05,
+      "loss": 1.0548,
+      "step": 13557
+    },
+    {
+      "epoch": 0.6245911457133644,
+      "grad_norm": 1.045745567945279,
+      "learning_rate": 6.525096972033646e-05,
+      "loss": 1.2101,
+      "step": 13558
+    },
+    {
+      "epoch": 0.6246372138019993,
+      "grad_norm": 1.5579641871920298,
+      "learning_rate": 6.523697903698119e-05,
+      "loss": 1.0297,
+      "step": 13559
+    },
+    {
+      "epoch": 0.6246832818906344,
+      "grad_norm": 1.9263276607086206,
+      "learning_rate": 6.522298912756492e-05,
+      "loss": 1.2317,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6247293499792693,
+      "grad_norm": 0.8804462282528723,
+      "learning_rate": 6.520899999239922e-05,
+      "loss": 1.137,
+      "step": 13561
+    },
+    {
+      "epoch": 0.6247754180679044,
+      "grad_norm": 1.1324702366860617,
+      "learning_rate": 6.519501163179546e-05,
+      "loss": 1.1116,
+      "step": 13562
+    },
+    {
+      "epoch": 0.6248214861565393,
+      "grad_norm": 0.6726926981291691,
+      "learning_rate": 6.518102404606509e-05,
+      "loss": 1.3699,
+      "step": 13563
+    },
+    {
+      "epoch": 0.6248675542451744,
+      "grad_norm": 2.53291335585929,
+      "learning_rate": 6.516703723551954e-05,
+      "loss": 1.1978,
+      "step": 13564
+    },
+    {
+      "epoch": 0.6249136223338093,
+      "grad_norm": 1.7552998982990538,
+      "learning_rate": 6.515305120047015e-05,
+      "loss": 1.1131,
+      "step": 13565
+    },
+    {
+      "epoch": 0.6249596904224444,
+      "grad_norm": 0.9731219506119609,
+      "learning_rate": 6.513906594122834e-05,
+      "loss": 1.402,
+      "step": 13566
+    },
+    {
+      "epoch": 0.6250057585110794,
+      "grad_norm": 0.9260596305028266,
+      "learning_rate": 6.512508145810544e-05,
+      "loss": 1.2687,
+      "step": 13567
+    },
+    {
+      "epoch": 0.6250518265997144,
+      "grad_norm": 0.8108152590277721,
+      "learning_rate": 6.511109775141282e-05,
+      "loss": 1.3626,
+      "step": 13568
+    },
+    {
+      "epoch": 0.6250978946883494,
+      "grad_norm": 1.1627431719131032,
+      "learning_rate": 6.509711482146176e-05,
+      "loss": 1.3116,
+      "step": 13569
+    },
+    {
+      "epoch": 0.6251439627769844,
+      "grad_norm": 1.652782852276658,
+      "learning_rate": 6.508313266856364e-05,
+      "loss": 1.0281,
+      "step": 13570
+    },
+    {
+      "epoch": 0.6251900308656194,
+      "grad_norm": 1.2409663891119411,
+      "learning_rate": 6.506915129302967e-05,
+      "loss": 1.2318,
+      "step": 13571
+    },
+    {
+      "epoch": 0.6252360989542544,
+      "grad_norm": 1.1677837161908868,
+      "learning_rate": 6.505517069517115e-05,
+      "loss": 1.1104,
+      "step": 13572
+    },
+    {
+      "epoch": 0.6252821670428894,
+      "grad_norm": 1.9162278805732051,
+      "learning_rate": 6.504119087529936e-05,
+      "loss": 1.3334,
+      "step": 13573
+    },
+    {
+      "epoch": 0.6253282351315244,
+      "grad_norm": 1.1287979614706132,
+      "learning_rate": 6.502721183372549e-05,
+      "loss": 1.2971,
+      "step": 13574
+    },
+    {
+      "epoch": 0.6253743032201594,
+      "grad_norm": 1.1080556383753395,
+      "learning_rate": 6.501323357076079e-05,
+      "loss": 1.111,
+      "step": 13575
+    },
+    {
+      "epoch": 0.6254203713087944,
+      "grad_norm": 2.390598581576574,
+      "learning_rate": 6.499925608671644e-05,
+      "loss": 1.2014,
+      "step": 13576
+    },
+    {
+      "epoch": 0.6254664393974294,
+      "grad_norm": 2.754447667040276,
+      "learning_rate": 6.498527938190367e-05,
+      "loss": 1.3665,
+      "step": 13577
+    },
+    {
+      "epoch": 0.6255125074860644,
+      "grad_norm": 1.1226654830488958,
+      "learning_rate": 6.497130345663359e-05,
+      "loss": 1.3732,
+      "step": 13578
+    },
+    {
+      "epoch": 0.6255585755746994,
+      "grad_norm": 1.0170477981060844,
+      "learning_rate": 6.49573283112174e-05,
+      "loss": 1.2542,
+      "step": 13579
+    },
+    {
+      "epoch": 0.6256046436633345,
+      "grad_norm": 0.8168221676432369,
+      "learning_rate": 6.49433539459662e-05,
+      "loss": 1.3503,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6256507117519694,
+      "grad_norm": 0.9567490222628136,
+      "learning_rate": 6.492938036119111e-05,
+      "loss": 1.3433,
+      "step": 13581
+    },
+    {
+      "epoch": 0.6256967798406045,
+      "grad_norm": 2.0954934940561056,
+      "learning_rate": 6.491540755720325e-05,
+      "loss": 1.3136,
+      "step": 13582
+    },
+    {
+      "epoch": 0.6257428479292394,
+      "grad_norm": 1.3198959291470262,
+      "learning_rate": 6.490143553431366e-05,
+      "loss": 1.2466,
+      "step": 13583
+    },
+    {
+      "epoch": 0.6257889160178745,
+      "grad_norm": 1.1385660148894805,
+      "learning_rate": 6.488746429283346e-05,
+      "loss": 1.3938,
+      "step": 13584
+    },
+    {
+      "epoch": 0.6258349841065094,
+      "grad_norm": 3.736526135350379,
+      "learning_rate": 6.487349383307362e-05,
+      "loss": 1.5654,
+      "step": 13585
+    },
+    {
+      "epoch": 0.6258810521951444,
+      "grad_norm": 1.2661916058576306,
+      "learning_rate": 6.485952415534524e-05,
+      "loss": 1.2726,
+      "step": 13586
+    },
+    {
+      "epoch": 0.6259271202837794,
+      "grad_norm": 1.0425633780386978,
+      "learning_rate": 6.484555525995928e-05,
+      "loss": 1.1704,
+      "step": 13587
+    },
+    {
+      "epoch": 0.6259731883724144,
+      "grad_norm": 0.893417935936355,
+      "learning_rate": 6.483158714722677e-05,
+      "loss": 1.3325,
+      "step": 13588
+    },
+    {
+      "epoch": 0.6260192564610494,
+      "grad_norm": 1.8503030316992335,
+      "learning_rate": 6.481761981745867e-05,
+      "loss": 1.4041,
+      "step": 13589
+    },
+    {
+      "epoch": 0.6260653245496844,
+      "grad_norm": 0.9377546347897041,
+      "learning_rate": 6.480365327096594e-05,
+      "loss": 1.3514,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6261113926383194,
+      "grad_norm": 1.6935428810195126,
+      "learning_rate": 6.478968750805954e-05,
+      "loss": 1.4059,
+      "step": 13591
+    },
+    {
+      "epoch": 0.6261574607269544,
+      "grad_norm": 1.1349177665364794,
+      "learning_rate": 6.477572252905033e-05,
+      "loss": 1.4632,
+      "step": 13592
+    },
+    {
+      "epoch": 0.6262035288155895,
+      "grad_norm": 1.0743705800814565,
+      "learning_rate": 6.47617583342493e-05,
+      "loss": 1.2248,
+      "step": 13593
+    },
+    {
+      "epoch": 0.6262495969042244,
+      "grad_norm": 1.2027619227246023,
+      "learning_rate": 6.474779492396728e-05,
+      "loss": 1.094,
+      "step": 13594
+    },
+    {
+      "epoch": 0.6262956649928595,
+      "grad_norm": 0.7904116963093765,
+      "learning_rate": 6.473383229851517e-05,
+      "loss": 1.4917,
+      "step": 13595
+    },
+    {
+      "epoch": 0.6263417330814944,
+      "grad_norm": 1.6744464470371558,
+      "learning_rate": 6.47198704582038e-05,
+      "loss": 1.617,
+      "step": 13596
+    },
+    {
+      "epoch": 0.6263878011701295,
+      "grad_norm": 0.9252829903229239,
+      "learning_rate": 6.470590940334405e-05,
+      "loss": 1.4731,
+      "step": 13597
+    },
+    {
+      "epoch": 0.6264338692587644,
+      "grad_norm": 1.2252759868150414,
+      "learning_rate": 6.469194913424668e-05,
+      "loss": 1.3164,
+      "step": 13598
+    },
+    {
+      "epoch": 0.6264799373473995,
+      "grad_norm": 1.1340224695599226,
+      "learning_rate": 6.467798965122256e-05,
+      "loss": 1.3581,
+      "step": 13599
+    },
+    {
+      "epoch": 0.6265260054360344,
+      "grad_norm": 4.232636752953746,
+      "learning_rate": 6.466403095458245e-05,
+      "loss": 1.2578,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6265720735246695,
+      "grad_norm": 9.075823551198932,
+      "learning_rate": 6.465007304463702e-05,
+      "loss": 1.1648,
+      "step": 13601
+    },
+    {
+      "epoch": 0.6266181416133044,
+      "grad_norm": 0.7836458523492882,
+      "learning_rate": 6.463611592169716e-05,
+      "loss": 1.3577,
+      "step": 13602
+    },
+    {
+      "epoch": 0.6266642097019395,
+      "grad_norm": 1.516365059038252,
+      "learning_rate": 6.462215958607351e-05,
+      "loss": 1.0886,
+      "step": 13603
+    },
+    {
+      "epoch": 0.6267102777905744,
+      "grad_norm": 0.7457032547807021,
+      "learning_rate": 6.460820403807686e-05,
+      "loss": 1.2607,
+      "step": 13604
+    },
+    {
+      "epoch": 0.6267563458792095,
+      "grad_norm": 1.374809777254989,
+      "learning_rate": 6.459424927801782e-05,
+      "loss": 1.0765,
+      "step": 13605
+    },
+    {
+      "epoch": 0.6268024139678445,
+      "grad_norm": 1.1524689812385611,
+      "learning_rate": 6.458029530620715e-05,
+      "loss": 1.1136,
+      "step": 13606
+    },
+    {
+      "epoch": 0.6268484820564795,
+      "grad_norm": 1.5278954113315846,
+      "learning_rate": 6.456634212295545e-05,
+      "loss": 1.08,
+      "step": 13607
+    },
+    {
+      "epoch": 0.6268945501451145,
+      "grad_norm": 3.9886419597696148,
+      "learning_rate": 6.45523897285734e-05,
+      "loss": 1.4055,
+      "step": 13608
+    },
+    {
+      "epoch": 0.6269406182337495,
+      "grad_norm": 1.048434350023846,
+      "learning_rate": 6.453843812337163e-05,
+      "loss": 1.1945,
+      "step": 13609
+    },
+    {
+      "epoch": 0.6269866863223845,
+      "grad_norm": 3.2719462586131205,
+      "learning_rate": 6.452448730766069e-05,
+      "loss": 1.0599,
+      "step": 13610
+    },
+    {
+      "epoch": 0.6270327544110195,
+      "grad_norm": 0.7993935961366819,
+      "learning_rate": 6.451053728175125e-05,
+      "loss": 1.2184,
+      "step": 13611
+    },
+    {
+      "epoch": 0.6270788224996545,
+      "grad_norm": 1.9950595804566849,
+      "learning_rate": 6.44965880459538e-05,
+      "loss": 1.0852,
+      "step": 13612
+    },
+    {
+      "epoch": 0.6271248905882895,
+      "grad_norm": 1.6447261041941277,
+      "learning_rate": 6.4482639600579e-05,
+      "loss": 1.325,
+      "step": 13613
+    },
+    {
+      "epoch": 0.6271709586769245,
+      "grad_norm": 1.8241875436210169,
+      "learning_rate": 6.446869194593727e-05,
+      "loss": 1.2099,
+      "step": 13614
+    },
+    {
+      "epoch": 0.6272170267655595,
+      "grad_norm": 1.2376067824205002,
+      "learning_rate": 6.445474508233921e-05,
+      "loss": 1.2275,
+      "step": 13615
+    },
+    {
+      "epoch": 0.6272630948541945,
+      "grad_norm": 8.527806759954112,
+      "learning_rate": 6.444079901009532e-05,
+      "loss": 1.431,
+      "step": 13616
+    },
+    {
+      "epoch": 0.6273091629428295,
+      "grad_norm": 1.2704944880413516,
+      "learning_rate": 6.442685372951607e-05,
+      "loss": 1.186,
+      "step": 13617
+    },
+    {
+      "epoch": 0.6273552310314645,
+      "grad_norm": 1.7460310961800476,
+      "learning_rate": 6.44129092409119e-05,
+      "loss": 1.4229,
+      "step": 13618
+    },
+    {
+      "epoch": 0.6274012991200995,
+      "grad_norm": 1.2112042596425114,
+      "learning_rate": 6.439896554459332e-05,
+      "loss": 1.629,
+      "step": 13619
+    },
+    {
+      "epoch": 0.6274473672087345,
+      "grad_norm": 1.3293058522530037,
+      "learning_rate": 6.438502264087075e-05,
+      "loss": 0.9821,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6274934352973696,
+      "grad_norm": 1.6400329279088204,
+      "learning_rate": 6.437108053005452e-05,
+      "loss": 1.1323,
+      "step": 13621
+    },
+    {
+      "epoch": 0.6275395033860045,
+      "grad_norm": 2.0295873784765366,
+      "learning_rate": 6.435713921245514e-05,
+      "loss": 0.9092,
+      "step": 13622
+    },
+    {
+      "epoch": 0.6275855714746396,
+      "grad_norm": 0.8500537725382503,
+      "learning_rate": 6.43431986883829e-05,
+      "loss": 1.2214,
+      "step": 13623
+    },
+    {
+      "epoch": 0.6276316395632745,
+      "grad_norm": 1.7133126268538588,
+      "learning_rate": 6.432925895814825e-05,
+      "loss": 1.1717,
+      "step": 13624
+    },
+    {
+      "epoch": 0.6276777076519096,
+      "grad_norm": 0.9541911071513126,
+      "learning_rate": 6.431532002206143e-05,
+      "loss": 1.2038,
+      "step": 13625
+    },
+    {
+      "epoch": 0.6277237757405445,
+      "grad_norm": 4.929273277614691,
+      "learning_rate": 6.430138188043288e-05,
+      "loss": 1.164,
+      "step": 13626
+    },
+    {
+      "epoch": 0.6277698438291796,
+      "grad_norm": 0.8587265967558081,
+      "learning_rate": 6.428744453357281e-05,
+      "loss": 1.5082,
+      "step": 13627
+    },
+    {
+      "epoch": 0.6278159119178145,
+      "grad_norm": 1.0626621123462936,
+      "learning_rate": 6.427350798179157e-05,
+      "loss": 1.4325,
+      "step": 13628
+    },
+    {
+      "epoch": 0.6278619800064495,
+      "grad_norm": 0.9304856184936691,
+      "learning_rate": 6.425957222539945e-05,
+      "loss": 1.2846,
+      "step": 13629
+    },
+    {
+      "epoch": 0.6279080480950845,
+      "grad_norm": 0.9247784838424356,
+      "learning_rate": 6.424563726470661e-05,
+      "loss": 1.3085,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6279541161837195,
+      "grad_norm": 28.993009989365707,
+      "learning_rate": 6.423170310002342e-05,
+      "loss": 1.3451,
+      "step": 13631
+    },
+    {
+      "epoch": 0.6280001842723545,
+      "grad_norm": 0.8912073605669626,
+      "learning_rate": 6.421776973165996e-05,
+      "loss": 1.3599,
+      "step": 13632
+    },
+    {
+      "epoch": 0.6280462523609895,
+      "grad_norm": 1.903268933207485,
+      "learning_rate": 6.420383715992657e-05,
+      "loss": 1.465,
+      "step": 13633
+    },
+    {
+      "epoch": 0.6280923204496246,
+      "grad_norm": 1.0993333993085261,
+      "learning_rate": 6.41899053851333e-05,
+      "loss": 1.3834,
+      "step": 13634
+    },
+    {
+      "epoch": 0.6281383885382595,
+      "grad_norm": 1.8340110298623278,
+      "learning_rate": 6.417597440759045e-05,
+      "loss": 0.928,
+      "step": 13635
+    },
+    {
+      "epoch": 0.6281844566268946,
+      "grad_norm": 0.8112065758169592,
+      "learning_rate": 6.416204422760804e-05,
+      "loss": 1.6811,
+      "step": 13636
+    },
+    {
+      "epoch": 0.6282305247155295,
+      "grad_norm": 0.970826146802059,
+      "learning_rate": 6.414811484549635e-05,
+      "loss": 1.3941,
+      "step": 13637
+    },
+    {
+      "epoch": 0.6282765928041646,
+      "grad_norm": 1.3196341409730663,
+      "learning_rate": 6.413418626156535e-05,
+      "loss": 1.3267,
+      "step": 13638
+    },
+    {
+      "epoch": 0.6283226608927995,
+      "grad_norm": 1.4156729226656721,
+      "learning_rate": 6.41202584761252e-05,
+      "loss": 1.1896,
+      "step": 13639
+    },
+    {
+      "epoch": 0.6283687289814346,
+      "grad_norm": 1.0386164761538192,
+      "learning_rate": 6.410633148948599e-05,
+      "loss": 1.3447,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6284147970700695,
+      "grad_norm": 0.7064696511910825,
+      "learning_rate": 6.409240530195772e-05,
+      "loss": 1.4905,
+      "step": 13641
+    },
+    {
+      "epoch": 0.6284608651587046,
+      "grad_norm": 1.020160695136311,
+      "learning_rate": 6.407847991385057e-05,
+      "loss": 1.3501,
+      "step": 13642
+    },
+    {
+      "epoch": 0.6285069332473395,
+      "grad_norm": 1.6743448181931015,
+      "learning_rate": 6.406455532547437e-05,
+      "loss": 1.3726,
+      "step": 13643
+    },
+    {
+      "epoch": 0.6285530013359746,
+      "grad_norm": 1.2990770157756637,
+      "learning_rate": 6.40506315371393e-05,
+      "loss": 1.0486,
+      "step": 13644
+    },
+    {
+      "epoch": 0.6285990694246095,
+      "grad_norm": 1.2419517826863489,
+      "learning_rate": 6.403670854915524e-05,
+      "loss": 1.1082,
+      "step": 13645
+    },
+    {
+      "epoch": 0.6286451375132446,
+      "grad_norm": 1.0673919139364392,
+      "learning_rate": 6.402278636183225e-05,
+      "loss": 1.4641,
+      "step": 13646
+    },
+    {
+      "epoch": 0.6286912056018796,
+      "grad_norm": 2.171678180619742,
+      "learning_rate": 6.40088649754802e-05,
+      "loss": 1.251,
+      "step": 13647
+    },
+    {
+      "epoch": 0.6287372736905146,
+      "grad_norm": 0.9195790062455379,
+      "learning_rate": 6.399494439040905e-05,
+      "loss": 1.3007,
+      "step": 13648
+    },
+    {
+      "epoch": 0.6287833417791496,
+      "grad_norm": 1.2815732653166043,
+      "learning_rate": 6.398102460692875e-05,
+      "loss": 1.2385,
+      "step": 13649
+    },
+    {
+      "epoch": 0.6288294098677846,
+      "grad_norm": 0.8678029755235827,
+      "learning_rate": 6.396710562534916e-05,
+      "loss": 1.1577,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6288754779564196,
+      "grad_norm": 0.9561021396974683,
+      "learning_rate": 6.39531874459802e-05,
+      "loss": 1.4278,
+      "step": 13651
+    },
+    {
+      "epoch": 0.6289215460450546,
+      "grad_norm": 1.4666827924364887,
+      "learning_rate": 6.393927006913168e-05,
+      "loss": 1.1769,
+      "step": 13652
+    },
+    {
+      "epoch": 0.6289676141336896,
+      "grad_norm": 0.8321824987604761,
+      "learning_rate": 6.39253534951135e-05,
+      "loss": 1.0726,
+      "step": 13653
+    },
+    {
+      "epoch": 0.6290136822223246,
+      "grad_norm": 6.341865729995115,
+      "learning_rate": 6.391143772423544e-05,
+      "loss": 1.5844,
+      "step": 13654
+    },
+    {
+      "epoch": 0.6290597503109596,
+      "grad_norm": 1.2391201808245067,
+      "learning_rate": 6.389752275680741e-05,
+      "loss": 1.625,
+      "step": 13655
+    },
+    {
+      "epoch": 0.6291058183995946,
+      "grad_norm": 0.7950814379257588,
+      "learning_rate": 6.388360859313911e-05,
+      "loss": 1.5219,
+      "step": 13656
+    },
+    {
+      "epoch": 0.6291518864882296,
+      "grad_norm": 0.8546272553407305,
+      "learning_rate": 6.386969523354031e-05,
+      "loss": 1.5056,
+      "step": 13657
+    },
+    {
+      "epoch": 0.6291979545768646,
+      "grad_norm": 1.0707291352667616,
+      "learning_rate": 6.385578267832081e-05,
+      "loss": 1.1396,
+      "step": 13658
+    },
+    {
+      "epoch": 0.6292440226654996,
+      "grad_norm": 1.157960095269838,
+      "learning_rate": 6.384187092779032e-05,
+      "loss": 1.0702,
+      "step": 13659
+    },
+    {
+      "epoch": 0.6292900907541346,
+      "grad_norm": 0.9332644396113026,
+      "learning_rate": 6.382795998225861e-05,
+      "loss": 1.6236,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6293361588427696,
+      "grad_norm": 0.9144529650890452,
+      "learning_rate": 6.381404984203532e-05,
+      "loss": 1.5522,
+      "step": 13661
+    },
+    {
+      "epoch": 0.6293822269314047,
+      "grad_norm": 2.786098252294943,
+      "learning_rate": 6.380014050743017e-05,
+      "loss": 1.2479,
+      "step": 13662
+    },
+    {
+      "epoch": 0.6294282950200396,
+      "grad_norm": 1.003541548853262,
+      "learning_rate": 6.378623197875282e-05,
+      "loss": 1.4054,
+      "step": 13663
+    },
+    {
+      "epoch": 0.6294743631086747,
+      "grad_norm": 0.9510010636581787,
+      "learning_rate": 6.377232425631295e-05,
+      "loss": 1.2352,
+      "step": 13664
+    },
+    {
+      "epoch": 0.6295204311973096,
+      "grad_norm": 1.7960872970365735,
+      "learning_rate": 6.375841734042014e-05,
+      "loss": 1.193,
+      "step": 13665
+    },
+    {
+      "epoch": 0.6295664992859447,
+      "grad_norm": 1.1427825647364813,
+      "learning_rate": 6.374451123138401e-05,
+      "loss": 1.2459,
+      "step": 13666
+    },
+    {
+      "epoch": 0.6296125673745796,
+      "grad_norm": 0.8911733966099875,
+      "learning_rate": 6.37306059295142e-05,
+      "loss": 1.4073,
+      "step": 13667
+    },
+    {
+      "epoch": 0.6296586354632147,
+      "grad_norm": 1.4086719368652312,
+      "learning_rate": 6.371670143512024e-05,
+      "loss": 1.1982,
+      "step": 13668
+    },
+    {
+      "epoch": 0.6297047035518496,
+      "grad_norm": 1.0054557282274181,
+      "learning_rate": 6.370279774851173e-05,
+      "loss": 1.349,
+      "step": 13669
+    },
+    {
+      "epoch": 0.6297507716404847,
+      "grad_norm": 1.1208608093077477,
+      "learning_rate": 6.368889486999818e-05,
+      "loss": 1.3627,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6297968397291196,
+      "grad_norm": 1.3006471850631374,
+      "learning_rate": 6.367499279988913e-05,
+      "loss": 1.0883,
+      "step": 13671
+    },
+    {
+      "epoch": 0.6298429078177546,
+      "grad_norm": 0.9990802326615882,
+      "learning_rate": 6.366109153849407e-05,
+      "loss": 1.524,
+      "step": 13672
+    },
+    {
+      "epoch": 0.6298889759063896,
+      "grad_norm": 1.1941893961015209,
+      "learning_rate": 6.364719108612252e-05,
+      "loss": 1.2952,
+      "step": 13673
+    },
+    {
+      "epoch": 0.6299350439950246,
+      "grad_norm": 1.1416449906996209,
+      "learning_rate": 6.363329144308392e-05,
+      "loss": 1.2529,
+      "step": 13674
+    },
+    {
+      "epoch": 0.6299811120836597,
+      "grad_norm": 0.8844601235383871,
+      "learning_rate": 6.361939260968772e-05,
+      "loss": 1.3683,
+      "step": 13675
+    },
+    {
+      "epoch": 0.6300271801722946,
+      "grad_norm": 1.0937989963438979,
+      "learning_rate": 6.360549458624338e-05,
+      "loss": 1.1233,
+      "step": 13676
+    },
+    {
+      "epoch": 0.6300732482609297,
+      "grad_norm": 0.8463056385875611,
+      "learning_rate": 6.359159737306027e-05,
+      "loss": 1.5186,
+      "step": 13677
+    },
+    {
+      "epoch": 0.6301193163495646,
+      "grad_norm": 1.3291337934707144,
+      "learning_rate": 6.357770097044786e-05,
+      "loss": 1.555,
+      "step": 13678
+    },
+    {
+      "epoch": 0.6301653844381997,
+      "grad_norm": 1.9912916397745395,
+      "learning_rate": 6.356380537871545e-05,
+      "loss": 1.3778,
+      "step": 13679
+    },
+    {
+      "epoch": 0.6302114525268346,
+      "grad_norm": 1.4908214539332714,
+      "learning_rate": 6.354991059817245e-05,
+      "loss": 1.318,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6302575206154697,
+      "grad_norm": 0.8227471213926085,
+      "learning_rate": 6.35360166291282e-05,
+      "loss": 1.3598,
+      "step": 13681
+    },
+    {
+      "epoch": 0.6303035887041046,
+      "grad_norm": 1.8210918975032804,
+      "learning_rate": 6.352212347189202e-05,
+      "loss": 1.308,
+      "step": 13682
+    },
+    {
+      "epoch": 0.6303496567927397,
+      "grad_norm": 9.859441222388297,
+      "learning_rate": 6.350823112677323e-05,
+      "loss": 1.1679,
+      "step": 13683
+    },
+    {
+      "epoch": 0.6303957248813746,
+      "grad_norm": 0.917856527639862,
+      "learning_rate": 6.349433959408106e-05,
+      "loss": 1.231,
+      "step": 13684
+    },
+    {
+      "epoch": 0.6304417929700097,
+      "grad_norm": 0.8055427937186781,
+      "learning_rate": 6.348044887412485e-05,
+      "loss": 1.3665,
+      "step": 13685
+    },
+    {
+      "epoch": 0.6304878610586446,
+      "grad_norm": 1.0539523452228567,
+      "learning_rate": 6.346655896721382e-05,
+      "loss": 1.4971,
+      "step": 13686
+    },
+    {
+      "epoch": 0.6305339291472797,
+      "grad_norm": 1.1798691094194529,
+      "learning_rate": 6.345266987365724e-05,
+      "loss": 1.3231,
+      "step": 13687
+    },
+    {
+      "epoch": 0.6305799972359147,
+      "grad_norm": 0.7857454415974321,
+      "learning_rate": 6.343878159376426e-05,
+      "loss": 1.3173,
+      "step": 13688
+    },
+    {
+      "epoch": 0.6306260653245497,
+      "grad_norm": 0.7305131451907125,
+      "learning_rate": 6.342489412784413e-05,
+      "loss": 1.2826,
+      "step": 13689
+    },
+    {
+      "epoch": 0.6306721334131847,
+      "grad_norm": 0.89136857736029,
+      "learning_rate": 6.341100747620604e-05,
+      "loss": 1.4897,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6307182015018197,
+      "grad_norm": 1.05343645689023,
+      "learning_rate": 6.339712163915912e-05,
+      "loss": 1.0335,
+      "step": 13691
+    },
+    {
+      "epoch": 0.6307642695904547,
+      "grad_norm": 1.503500267402247,
+      "learning_rate": 6.338323661701254e-05,
+      "loss": 1.2476,
+      "step": 13692
+    },
+    {
+      "epoch": 0.6308103376790897,
+      "grad_norm": 0.9854758248007056,
+      "learning_rate": 6.336935241007537e-05,
+      "loss": 1.2959,
+      "step": 13693
+    },
+    {
+      "epoch": 0.6308564057677247,
+      "grad_norm": 1.4851111477203274,
+      "learning_rate": 6.33554690186568e-05,
+      "loss": 1.157,
+      "step": 13694
+    },
+    {
+      "epoch": 0.6309024738563597,
+      "grad_norm": 1.055913126879732,
+      "learning_rate": 6.334158644306586e-05,
+      "loss": 1.1818,
+      "step": 13695
+    },
+    {
+      "epoch": 0.6309485419449947,
+      "grad_norm": 0.9093288128324898,
+      "learning_rate": 6.332770468361166e-05,
+      "loss": 1.5178,
+      "step": 13696
+    },
+    {
+      "epoch": 0.6309946100336297,
+      "grad_norm": 1.6639735365587665,
+      "learning_rate": 6.331382374060322e-05,
+      "loss": 1.5138,
+      "step": 13697
+    },
+    {
+      "epoch": 0.6310406781222647,
+      "grad_norm": 0.7861524398284557,
+      "learning_rate": 6.329994361434961e-05,
+      "loss": 1.4791,
+      "step": 13698
+    },
+    {
+      "epoch": 0.6310867462108997,
+      "grad_norm": 0.904799883608129,
+      "learning_rate": 6.328606430515979e-05,
+      "loss": 1.4832,
+      "step": 13699
+    },
+    {
+      "epoch": 0.6311328142995347,
+      "grad_norm": 0.871825025096714,
+      "learning_rate": 6.327218581334285e-05,
+      "loss": 1.1932,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6311788823881697,
+      "grad_norm": 0.7945284970589394,
+      "learning_rate": 6.32583081392077e-05,
+      "loss": 1.5208,
+      "step": 13701
+    },
+    {
+      "epoch": 0.6312249504768047,
+      "grad_norm": 0.7952452381173263,
+      "learning_rate": 6.32444312830633e-05,
+      "loss": 0.9926,
+      "step": 13702
+    },
+    {
+      "epoch": 0.6312710185654398,
+      "grad_norm": 0.8258412323114916,
+      "learning_rate": 6.323055524521863e-05,
+      "loss": 1.1874,
+      "step": 13703
+    },
+    {
+      "epoch": 0.6313170866540747,
+      "grad_norm": 0.9063404388027794,
+      "learning_rate": 6.321668002598259e-05,
+      "loss": 1.3289,
+      "step": 13704
+    },
+    {
+      "epoch": 0.6313631547427098,
+      "grad_norm": 5.19486248773381,
+      "learning_rate": 6.320280562566412e-05,
+      "loss": 1.3751,
+      "step": 13705
+    },
+    {
+      "epoch": 0.6314092228313447,
+      "grad_norm": 1.1737949025983665,
+      "learning_rate": 6.318893204457206e-05,
+      "loss": 1.2479,
+      "step": 13706
+    },
+    {
+      "epoch": 0.6314552909199798,
+      "grad_norm": 0.8012372568464518,
+      "learning_rate": 6.317505928301534e-05,
+      "loss": 1.396,
+      "step": 13707
+    },
+    {
+      "epoch": 0.6315013590086147,
+      "grad_norm": 0.8221170542515973,
+      "learning_rate": 6.316118734130276e-05,
+      "loss": 1.1129,
+      "step": 13708
+    },
+    {
+      "epoch": 0.6315474270972498,
+      "grad_norm": 0.9927349731330369,
+      "learning_rate": 6.314731621974319e-05,
+      "loss": 1.3292,
+      "step": 13709
+    },
+    {
+      "epoch": 0.6315934951858847,
+      "grad_norm": 0.8249687008970548,
+      "learning_rate": 6.313344591864547e-05,
+      "loss": 1.3337,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6316395632745198,
+      "grad_norm": 6.362690560990343,
+      "learning_rate": 6.311957643831829e-05,
+      "loss": 1.3988,
+      "step": 13711
+    },
+    {
+      "epoch": 0.6316856313631547,
+      "grad_norm": 1.64367165562748,
+      "learning_rate": 6.31057077790706e-05,
+      "loss": 0.9314,
+      "step": 13712
+    },
+    {
+      "epoch": 0.6317316994517898,
+      "grad_norm": 0.8039011543356321,
+      "learning_rate": 6.309183994121097e-05,
+      "loss": 1.3372,
+      "step": 13713
+    },
+    {
+      "epoch": 0.6317777675404247,
+      "grad_norm": 1.7274872716495946,
+      "learning_rate": 6.307797292504828e-05,
+      "loss": 0.998,
+      "step": 13714
+    },
+    {
+      "epoch": 0.6318238356290597,
+      "grad_norm": 0.787290558581935,
+      "learning_rate": 6.30641067308912e-05,
+      "loss": 1.4371,
+      "step": 13715
+    },
+    {
+      "epoch": 0.6318699037176948,
+      "grad_norm": 1.0103600159730464,
+      "learning_rate": 6.305024135904848e-05,
+      "loss": 1.0894,
+      "step": 13716
+    },
+    {
+      "epoch": 0.6319159718063297,
+      "grad_norm": 0.86812015790994,
+      "learning_rate": 6.303637680982877e-05,
+      "loss": 1.4113,
+      "step": 13717
+    },
+    {
+      "epoch": 0.6319620398949648,
+      "grad_norm": 0.9420320011012684,
+      "learning_rate": 6.302251308354075e-05,
+      "loss": 1.0476,
+      "step": 13718
+    },
+    {
+      "epoch": 0.6320081079835997,
+      "grad_norm": 0.7751161811803885,
+      "learning_rate": 6.300865018049311e-05,
+      "loss": 1.3399,
+      "step": 13719
+    },
+    {
+      "epoch": 0.6320541760722348,
+      "grad_norm": 0.7091924507313948,
+      "learning_rate": 6.299478810099438e-05,
+      "loss": 1.3715,
+      "step": 13720
+    },
+    {
+      "epoch": 0.6321002441608697,
+      "grad_norm": 0.8889411404558756,
+      "learning_rate": 6.29809268453533e-05,
+      "loss": 1.1871,
+      "step": 13721
+    },
+    {
+      "epoch": 0.6321463122495048,
+      "grad_norm": 1.3012969743168976,
+      "learning_rate": 6.296706641387835e-05,
+      "loss": 1.3685,
+      "step": 13722
+    },
+    {
+      "epoch": 0.6321923803381397,
+      "grad_norm": 0.7523684014957236,
+      "learning_rate": 6.295320680687823e-05,
+      "loss": 1.2618,
+      "step": 13723
+    },
+    {
+      "epoch": 0.6322384484267748,
+      "grad_norm": 1.344147356891696,
+      "learning_rate": 6.293934802466136e-05,
+      "loss": 1.3615,
+      "step": 13724
+    },
+    {
+      "epoch": 0.6322845165154097,
+      "grad_norm": 1.6320833914703567,
+      "learning_rate": 6.292549006753645e-05,
+      "loss": 1.2008,
+      "step": 13725
+    },
+    {
+      "epoch": 0.6323305846040448,
+      "grad_norm": 1.674241237720465,
+      "learning_rate": 6.291163293581187e-05,
+      "loss": 1.2715,
+      "step": 13726
+    },
+    {
+      "epoch": 0.6323766526926797,
+      "grad_norm": 0.898436489776203,
+      "learning_rate": 6.289777662979622e-05,
+      "loss": 1.2535,
+      "step": 13727
+    },
+    {
+      "epoch": 0.6324227207813148,
+      "grad_norm": 1.0370365868818354,
+      "learning_rate": 6.288392114979799e-05,
+      "loss": 1.5032,
+      "step": 13728
+    },
+    {
+      "epoch": 0.6324687888699498,
+      "grad_norm": 0.8567625541869948,
+      "learning_rate": 6.287006649612554e-05,
+      "loss": 1.4165,
+      "step": 13729
+    },
+    {
+      "epoch": 0.6325148569585848,
+      "grad_norm": 0.9887652008595067,
+      "learning_rate": 6.285621266908746e-05,
+      "loss": 1.3732,
+      "step": 13730
+    },
+    {
+      "epoch": 0.6325609250472198,
+      "grad_norm": 7.490696375077232,
+      "learning_rate": 6.284235966899205e-05,
+      "loss": 1.0797,
+      "step": 13731
+    },
+    {
+      "epoch": 0.6326069931358548,
+      "grad_norm": 4.442551621571814,
+      "learning_rate": 6.282850749614786e-05,
+      "loss": 1.6377,
+      "step": 13732
+    },
+    {
+      "epoch": 0.6326530612244898,
+      "grad_norm": 1.582594236757945,
+      "learning_rate": 6.281465615086316e-05,
+      "loss": 1.4998,
+      "step": 13733
+    },
+    {
+      "epoch": 0.6326991293131248,
+      "grad_norm": 0.9388176072953104,
+      "learning_rate": 6.280080563344644e-05,
+      "loss": 1.181,
+      "step": 13734
+    },
+    {
+      "epoch": 0.6327451974017598,
+      "grad_norm": 1.6305002612931971,
+      "learning_rate": 6.278695594420593e-05,
+      "loss": 1.6351,
+      "step": 13735
+    },
+    {
+      "epoch": 0.6327912654903948,
+      "grad_norm": 2.9991461287042434,
+      "learning_rate": 6.27731070834501e-05,
+      "loss": 1.1152,
+      "step": 13736
+    },
+    {
+      "epoch": 0.6328373335790298,
+      "grad_norm": 1.016813973936105,
+      "learning_rate": 6.275925905148716e-05,
+      "loss": 1.0953,
+      "step": 13737
+    },
+    {
+      "epoch": 0.6328834016676648,
+      "grad_norm": 0.7639027486021516,
+      "learning_rate": 6.274541184862553e-05,
+      "loss": 1.3676,
+      "step": 13738
+    },
+    {
+      "epoch": 0.6329294697562998,
+      "grad_norm": 1.2913840009596984,
+      "learning_rate": 6.273156547517338e-05,
+      "loss": 1.3366,
+      "step": 13739
+    },
+    {
+      "epoch": 0.6329755378449348,
+      "grad_norm": 0.8932479686575071,
+      "learning_rate": 6.2717719931439e-05,
+      "loss": 1.3912,
+      "step": 13740
+    },
+    {
+      "epoch": 0.6330216059335698,
+      "grad_norm": 1.0011420540631606,
+      "learning_rate": 6.270387521773071e-05,
+      "loss": 1.1551,
+      "step": 13741
+    },
+    {
+      "epoch": 0.6330676740222049,
+      "grad_norm": 0.9439288339812801,
+      "learning_rate": 6.269003133435664e-05,
+      "loss": 0.8816,
+      "step": 13742
+    },
+    {
+      "epoch": 0.6331137421108398,
+      "grad_norm": 1.0965856863222698,
+      "learning_rate": 6.267618828162512e-05,
+      "loss": 1.3124,
+      "step": 13743
+    },
+    {
+      "epoch": 0.6331598101994749,
+      "grad_norm": 0.9050671103177733,
+      "learning_rate": 6.266234605984418e-05,
+      "loss": 1.4726,
+      "step": 13744
+    },
+    {
+      "epoch": 0.6332058782881098,
+      "grad_norm": 1.758816300230811,
+      "learning_rate": 6.264850466932217e-05,
+      "loss": 1.2849,
+      "step": 13745
+    },
+    {
+      "epoch": 0.6332519463767449,
+      "grad_norm": 2.383208362161994,
+      "learning_rate": 6.263466411036709e-05,
+      "loss": 1.547,
+      "step": 13746
+    },
+    {
+      "epoch": 0.6332980144653798,
+      "grad_norm": 1.043362323277461,
+      "learning_rate": 6.26208243832872e-05,
+      "loss": 1.1029,
+      "step": 13747
+    },
+    {
+      "epoch": 0.6333440825540149,
+      "grad_norm": 0.8599982241415159,
+      "learning_rate": 6.260698548839056e-05,
+      "loss": 1.2722,
+      "step": 13748
+    },
+    {
+      "epoch": 0.6333901506426498,
+      "grad_norm": 0.8575089483836662,
+      "learning_rate": 6.259314742598524e-05,
+      "loss": 1.2096,
+      "step": 13749
+    },
+    {
+      "epoch": 0.6334362187312849,
+      "grad_norm": 1.2969387386619993,
+      "learning_rate": 6.257931019637938e-05,
+      "loss": 1.2087,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6334822868199198,
+      "grad_norm": 3.2230636471319167,
+      "learning_rate": 6.2565473799881e-05,
+      "loss": 1.4777,
+      "step": 13751
+    },
+    {
+      "epoch": 0.6335283549085549,
+      "grad_norm": 1.3241093695376436,
+      "learning_rate": 6.255163823679817e-05,
+      "loss": 1.3768,
+      "step": 13752
+    },
+    {
+      "epoch": 0.6335744229971898,
+      "grad_norm": 0.7184038085146092,
+      "learning_rate": 6.253780350743886e-05,
+      "loss": 1.1518,
+      "step": 13753
+    },
+    {
+      "epoch": 0.6336204910858249,
+      "grad_norm": 3.0016035863413153,
+      "learning_rate": 6.25239696121112e-05,
+      "loss": 1.1443,
+      "step": 13754
+    },
+    {
+      "epoch": 0.6336665591744598,
+      "grad_norm": 6.452616701792785,
+      "learning_rate": 6.251013655112303e-05,
+      "loss": 1.3937,
+      "step": 13755
+    },
+    {
+      "epoch": 0.6337126272630949,
+      "grad_norm": 1.15633399357932,
+      "learning_rate": 6.249630432478243e-05,
+      "loss": 1.7276,
+      "step": 13756
+    },
+    {
+      "epoch": 0.6337586953517299,
+      "grad_norm": 0.7913562732826016,
+      "learning_rate": 6.248247293339731e-05,
+      "loss": 1.2417,
+      "step": 13757
+    },
+    {
+      "epoch": 0.6338047634403648,
+      "grad_norm": 2.2693559040055824,
+      "learning_rate": 6.246864237727557e-05,
+      "loss": 1.4469,
+      "step": 13758
+    },
+    {
+      "epoch": 0.6338508315289999,
+      "grad_norm": 0.8548168222207823,
+      "learning_rate": 6.245481265672519e-05,
+      "loss": 1.3364,
+      "step": 13759
+    },
+    {
+      "epoch": 0.6338968996176348,
+      "grad_norm": 3.5736810331422224,
+      "learning_rate": 6.2440983772054e-05,
+      "loss": 1.1863,
+      "step": 13760
+    },
+    {
+      "epoch": 0.6339429677062699,
+      "grad_norm": 0.9370028606115225,
+      "learning_rate": 6.242715572356993e-05,
+      "loss": 1.2178,
+      "step": 13761
+    },
+    {
+      "epoch": 0.6339890357949048,
+      "grad_norm": 0.6943237218073556,
+      "learning_rate": 6.24133285115808e-05,
+      "loss": 1.5251,
+      "step": 13762
+    },
+    {
+      "epoch": 0.6340351038835399,
+      "grad_norm": 0.8036649454504345,
+      "learning_rate": 6.239950213639447e-05,
+      "loss": 1.432,
+      "step": 13763
+    },
+    {
+      "epoch": 0.6340811719721748,
+      "grad_norm": 0.8110418704607355,
+      "learning_rate": 6.238567659831873e-05,
+      "loss": 1.3034,
+      "step": 13764
+    },
+    {
+      "epoch": 0.6341272400608099,
+      "grad_norm": 1.7450294209646589,
+      "learning_rate": 6.237185189766144e-05,
+      "loss": 1.3621,
+      "step": 13765
+    },
+    {
+      "epoch": 0.6341733081494448,
+      "grad_norm": 0.9645147903974912,
+      "learning_rate": 6.235802803473034e-05,
+      "loss": 1.1651,
+      "step": 13766
+    },
+    {
+      "epoch": 0.6342193762380799,
+      "grad_norm": 0.6992778200105886,
+      "learning_rate": 6.234420500983319e-05,
+      "loss": 1.411,
+      "step": 13767
+    },
+    {
+      "epoch": 0.6342654443267148,
+      "grad_norm": 1.0106702819198063,
+      "learning_rate": 6.233038282327777e-05,
+      "loss": 1.3609,
+      "step": 13768
+    },
+    {
+      "epoch": 0.6343115124153499,
+      "grad_norm": 0.8773165800919296,
+      "learning_rate": 6.231656147537176e-05,
+      "loss": 1.5061,
+      "step": 13769
+    },
+    {
+      "epoch": 0.6343575805039849,
+      "grad_norm": 0.7878126003562805,
+      "learning_rate": 6.230274096642292e-05,
+      "loss": 1.2474,
+      "step": 13770
+    },
+    {
+      "epoch": 0.6344036485926199,
+      "grad_norm": 1.1360181840711707,
+      "learning_rate": 6.228892129673889e-05,
+      "loss": 1.1207,
+      "step": 13771
+    },
+    {
+      "epoch": 0.6344497166812549,
+      "grad_norm": 0.7858495494225318,
+      "learning_rate": 6.22751024666274e-05,
+      "loss": 0.9443,
+      "step": 13772
+    },
+    {
+      "epoch": 0.6344957847698899,
+      "grad_norm": 2.0805607707964193,
+      "learning_rate": 6.226128447639603e-05,
+      "loss": 1.3773,
+      "step": 13773
+    },
+    {
+      "epoch": 0.6345418528585249,
+      "grad_norm": 1.3423977142444172,
+      "learning_rate": 6.224746732635249e-05,
+      "loss": 1.4996,
+      "step": 13774
+    },
+    {
+      "epoch": 0.6345879209471599,
+      "grad_norm": 1.7669410030914365,
+      "learning_rate": 6.223365101680436e-05,
+      "loss": 1.1247,
+      "step": 13775
+    },
+    {
+      "epoch": 0.6346339890357949,
+      "grad_norm": 2.5923126131120133,
+      "learning_rate": 6.22198355480592e-05,
+      "loss": 1.3888,
+      "step": 13776
+    },
+    {
+      "epoch": 0.6346800571244299,
+      "grad_norm": 0.6706466640117122,
+      "learning_rate": 6.220602092042465e-05,
+      "loss": 1.342,
+      "step": 13777
+    },
+    {
+      "epoch": 0.6347261252130649,
+      "grad_norm": 1.2836138315262473,
+      "learning_rate": 6.219220713420823e-05,
+      "loss": 1.2989,
+      "step": 13778
+    },
+    {
+      "epoch": 0.6347721933016999,
+      "grad_norm": 1.8675918916917222,
+      "learning_rate": 6.217839418971749e-05,
+      "loss": 1.3589,
+      "step": 13779
+    },
+    {
+      "epoch": 0.6348182613903349,
+      "grad_norm": 0.6833881992774533,
+      "learning_rate": 6.216458208725995e-05,
+      "loss": 1.3443,
+      "step": 13780
+    },
+    {
+      "epoch": 0.6348643294789699,
+      "grad_norm": 0.9450297008191652,
+      "learning_rate": 6.215077082714313e-05,
+      "loss": 1.2263,
+      "step": 13781
+    },
+    {
+      "epoch": 0.6349103975676049,
+      "grad_norm": 0.8599815069669701,
+      "learning_rate": 6.213696040967448e-05,
+      "loss": 1.4932,
+      "step": 13782
+    },
+    {
+      "epoch": 0.63495646565624,
+      "grad_norm": 1.0270962913232613,
+      "learning_rate": 6.212315083516151e-05,
+      "loss": 1.1288,
+      "step": 13783
+    },
+    {
+      "epoch": 0.6350025337448749,
+      "grad_norm": 0.9202519154340395,
+      "learning_rate": 6.210934210391164e-05,
+      "loss": 0.9148,
+      "step": 13784
+    },
+    {
+      "epoch": 0.63504860183351,
+      "grad_norm": 1.0631588286549434,
+      "learning_rate": 6.20955342162323e-05,
+      "loss": 1.1614,
+      "step": 13785
+    },
+    {
+      "epoch": 0.6350946699221449,
+      "grad_norm": 0.9921789251547156,
+      "learning_rate": 6.208172717243088e-05,
+      "loss": 1.226,
+      "step": 13786
+    },
+    {
+      "epoch": 0.63514073801078,
+      "grad_norm": 0.8351841319792113,
+      "learning_rate": 6.206792097281478e-05,
+      "loss": 1.3885,
+      "step": 13787
+    },
+    {
+      "epoch": 0.6351868060994149,
+      "grad_norm": 0.8591716125122683,
+      "learning_rate": 6.20541156176914e-05,
+      "loss": 1.3612,
+      "step": 13788
+    },
+    {
+      "epoch": 0.63523287418805,
+      "grad_norm": 1.2854436079593712,
+      "learning_rate": 6.204031110736807e-05,
+      "loss": 1.1009,
+      "step": 13789
+    },
+    {
+      "epoch": 0.6352789422766849,
+      "grad_norm": 3.260114428015903,
+      "learning_rate": 6.202650744215213e-05,
+      "loss": 1.3257,
+      "step": 13790
+    },
+    {
+      "epoch": 0.63532501036532,
+      "grad_norm": 1.4982502860785816,
+      "learning_rate": 6.201270462235087e-05,
+      "loss": 1.2484,
+      "step": 13791
+    },
+    {
+      "epoch": 0.6353710784539549,
+      "grad_norm": 0.762690353430594,
+      "learning_rate": 6.199890264827164e-05,
+      "loss": 0.8989,
+      "step": 13792
+    },
+    {
+      "epoch": 0.63541714654259,
+      "grad_norm": 2.1007541852164935,
+      "learning_rate": 6.198510152022168e-05,
+      "loss": 1.1894,
+      "step": 13793
+    },
+    {
+      "epoch": 0.6354632146312249,
+      "grad_norm": 1.4006981167979848,
+      "learning_rate": 6.197130123850823e-05,
+      "loss": 1.3345,
+      "step": 13794
+    },
+    {
+      "epoch": 0.63550928271986,
+      "grad_norm": 1.4226596018485274,
+      "learning_rate": 6.195750180343857e-05,
+      "loss": 1.3546,
+      "step": 13795
+    },
+    {
+      "epoch": 0.6355553508084949,
+      "grad_norm": 0.8228394495040178,
+      "learning_rate": 6.194370321531988e-05,
+      "loss": 1.2572,
+      "step": 13796
+    },
+    {
+      "epoch": 0.63560141889713,
+      "grad_norm": 4.752220786459309,
+      "learning_rate": 6.192990547445942e-05,
+      "loss": 1.2228,
+      "step": 13797
+    },
+    {
+      "epoch": 0.635647486985765,
+      "grad_norm": 1.9798948152789178,
+      "learning_rate": 6.19161085811643e-05,
+      "loss": 1.5132,
+      "step": 13798
+    },
+    {
+      "epoch": 0.6356935550744,
+      "grad_norm": 0.8803213293895221,
+      "learning_rate": 6.190231253574177e-05,
+      "loss": 1.2643,
+      "step": 13799
+    },
+    {
+      "epoch": 0.635739623163035,
+      "grad_norm": 0.9582265179492186,
+      "learning_rate": 6.188851733849888e-05,
+      "loss": 1.3985,
+      "step": 13800
+    },
+    {
+      "epoch": 0.63578569125167,
+      "grad_norm": 0.6894959099241658,
+      "learning_rate": 6.187472298974283e-05,
+      "loss": 1.4023,
+      "step": 13801
+    },
+    {
+      "epoch": 0.635831759340305,
+      "grad_norm": 0.7929779164073181,
+      "learning_rate": 6.186092948978071e-05,
+      "loss": 1.3599,
+      "step": 13802
+    },
+    {
+      "epoch": 0.63587782742894,
+      "grad_norm": 1.2798377517119433,
+      "learning_rate": 6.184713683891958e-05,
+      "loss": 1.3523,
+      "step": 13803
+    },
+    {
+      "epoch": 0.635923895517575,
+      "grad_norm": 0.9527369557166238,
+      "learning_rate": 6.183334503746655e-05,
+      "loss": 1.3876,
+      "step": 13804
+    },
+    {
+      "epoch": 0.6359699636062099,
+      "grad_norm": 1.223688826115182,
+      "learning_rate": 6.181955408572863e-05,
+      "loss": 1.1228,
+      "step": 13805
+    },
+    {
+      "epoch": 0.636016031694845,
+      "grad_norm": 2.972274983060988,
+      "learning_rate": 6.180576398401289e-05,
+      "loss": 1.3097,
+      "step": 13806
+    },
+    {
+      "epoch": 0.6360620997834799,
+      "grad_norm": 1.0798219365854325,
+      "learning_rate": 6.17919747326263e-05,
+      "loss": 1.39,
+      "step": 13807
+    },
+    {
+      "epoch": 0.636108167872115,
+      "grad_norm": 0.8922645187629572,
+      "learning_rate": 6.17781863318759e-05,
+      "loss": 1.4284,
+      "step": 13808
+    },
+    {
+      "epoch": 0.6361542359607499,
+      "grad_norm": 0.9434279333529154,
+      "learning_rate": 6.176439878206865e-05,
+      "loss": 1.373,
+      "step": 13809
+    },
+    {
+      "epoch": 0.636200304049385,
+      "grad_norm": 1.101208377562934,
+      "learning_rate": 6.175061208351149e-05,
+      "loss": 1.1894,
+      "step": 13810
+    },
+    {
+      "epoch": 0.63624637213802,
+      "grad_norm": 0.7044233208521267,
+      "learning_rate": 6.17368262365114e-05,
+      "loss": 1.2746,
+      "step": 13811
+    },
+    {
+      "epoch": 0.636292440226655,
+      "grad_norm": 1.0349064636025151,
+      "learning_rate": 6.172304124137523e-05,
+      "loss": 1.2344,
+      "step": 13812
+    },
+    {
+      "epoch": 0.63633850831529,
+      "grad_norm": 6.986589139754427,
+      "learning_rate": 6.170925709840992e-05,
+      "loss": 1.2991,
+      "step": 13813
+    },
+    {
+      "epoch": 0.636384576403925,
+      "grad_norm": 3.764365332815877,
+      "learning_rate": 6.169547380792234e-05,
+      "loss": 1.4623,
+      "step": 13814
+    },
+    {
+      "epoch": 0.63643064449256,
+      "grad_norm": 1.0204147623613675,
+      "learning_rate": 6.168169137021937e-05,
+      "loss": 1.4118,
+      "step": 13815
+    },
+    {
+      "epoch": 0.636476712581195,
+      "grad_norm": 0.7584153964226551,
+      "learning_rate": 6.166790978560783e-05,
+      "loss": 1.5961,
+      "step": 13816
+    },
+    {
+      "epoch": 0.63652278066983,
+      "grad_norm": 1.1724767509257834,
+      "learning_rate": 6.165412905439459e-05,
+      "loss": 1.3111,
+      "step": 13817
+    },
+    {
+      "epoch": 0.636568848758465,
+      "grad_norm": 3.346827158863393,
+      "learning_rate": 6.164034917688637e-05,
+      "loss": 1.3118,
+      "step": 13818
+    },
+    {
+      "epoch": 0.6366149168471,
+      "grad_norm": 1.1896119591251,
+      "learning_rate": 6.162657015339002e-05,
+      "loss": 1.4907,
+      "step": 13819
+    },
+    {
+      "epoch": 0.636660984935735,
+      "grad_norm": 0.80880827001281,
+      "learning_rate": 6.161279198421232e-05,
+      "loss": 1.1941,
+      "step": 13820
+    },
+    {
+      "epoch": 0.63670705302437,
+      "grad_norm": 0.9446914756881445,
+      "learning_rate": 6.159901466965994e-05,
+      "loss": 1.1673,
+      "step": 13821
+    },
+    {
+      "epoch": 0.636753121113005,
+      "grad_norm": 1.0290294283806385,
+      "learning_rate": 6.15852382100397e-05,
+      "loss": 1.4285,
+      "step": 13822
+    },
+    {
+      "epoch": 0.63679918920164,
+      "grad_norm": 1.3681551733756743,
+      "learning_rate": 6.15714626056582e-05,
+      "loss": 1.4494,
+      "step": 13823
+    },
+    {
+      "epoch": 0.6368452572902751,
+      "grad_norm": 1.341018206184873,
+      "learning_rate": 6.155768785682225e-05,
+      "loss": 1.2665,
+      "step": 13824
+    },
+    {
+      "epoch": 0.63689132537891,
+      "grad_norm": 1.213409975443976,
+      "learning_rate": 6.154391396383843e-05,
+      "loss": 1.2036,
+      "step": 13825
+    },
+    {
+      "epoch": 0.6369373934675451,
+      "grad_norm": 2.6047533944149244,
+      "learning_rate": 6.153014092701344e-05,
+      "loss": 1.4366,
+      "step": 13826
+    },
+    {
+      "epoch": 0.63698346155618,
+      "grad_norm": 1.0706305396388587,
+      "learning_rate": 6.15163687466539e-05,
+      "loss": 1.5028,
+      "step": 13827
+    },
+    {
+      "epoch": 0.6370295296448151,
+      "grad_norm": 1.5026771257330216,
+      "learning_rate": 6.150259742306643e-05,
+      "loss": 1.3267,
+      "step": 13828
+    },
+    {
+      "epoch": 0.63707559773345,
+      "grad_norm": 1.6203378149613663,
+      "learning_rate": 6.148882695655765e-05,
+      "loss": 1.3356,
+      "step": 13829
+    },
+    {
+      "epoch": 0.6371216658220851,
+      "grad_norm": 1.1290039212118401,
+      "learning_rate": 6.147505734743404e-05,
+      "loss": 1.1816,
+      "step": 13830
+    },
+    {
+      "epoch": 0.63716773391072,
+      "grad_norm": 6.581997126123518,
+      "learning_rate": 6.146128859600229e-05,
+      "loss": 1.3474,
+      "step": 13831
+    },
+    {
+      "epoch": 0.6372138019993551,
+      "grad_norm": 0.9489079478376553,
+      "learning_rate": 6.14475207025688e-05,
+      "loss": 1.4667,
+      "step": 13832
+    },
+    {
+      "epoch": 0.63725987008799,
+      "grad_norm": 0.8230497072497374,
+      "learning_rate": 6.143375366744023e-05,
+      "loss": 1.0896,
+      "step": 13833
+    },
+    {
+      "epoch": 0.6373059381766251,
+      "grad_norm": 0.8949547532034404,
+      "learning_rate": 6.141998749092295e-05,
+      "loss": 1.3388,
+      "step": 13834
+    },
+    {
+      "epoch": 0.63735200626526,
+      "grad_norm": 0.7494008088042255,
+      "learning_rate": 6.140622217332355e-05,
+      "loss": 1.4396,
+      "step": 13835
+    },
+    {
+      "epoch": 0.6373980743538951,
+      "grad_norm": 0.7174064518712129,
+      "learning_rate": 6.139245771494837e-05,
+      "loss": 1.4881,
+      "step": 13836
+    },
+    {
+      "epoch": 0.63744414244253,
+      "grad_norm": 1.3885305264846686,
+      "learning_rate": 6.137869411610398e-05,
+      "loss": 1.2814,
+      "step": 13837
+    },
+    {
+      "epoch": 0.6374902105311651,
+      "grad_norm": 1.1004376077200655,
+      "learning_rate": 6.136493137709673e-05,
+      "loss": 1.4054,
+      "step": 13838
+    },
+    {
+      "epoch": 0.6375362786198001,
+      "grad_norm": 0.8747579179195165,
+      "learning_rate": 6.135116949823301e-05,
+      "loss": 1.1568,
+      "step": 13839
+    },
+    {
+      "epoch": 0.637582346708435,
+      "grad_norm": 0.8436634133219071,
+      "learning_rate": 6.133740847981928e-05,
+      "loss": 1.4927,
+      "step": 13840
+    },
+    {
+      "epoch": 0.6376284147970701,
+      "grad_norm": 0.7861692936163001,
+      "learning_rate": 6.13236483221618e-05,
+      "loss": 1.3082,
+      "step": 13841
+    },
+    {
+      "epoch": 0.637674482885705,
+      "grad_norm": 0.7266769040153999,
+      "learning_rate": 6.130988902556705e-05,
+      "loss": 1.3229,
+      "step": 13842
+    },
+    {
+      "epoch": 0.6377205509743401,
+      "grad_norm": 1.1698663992099463,
+      "learning_rate": 6.12961305903412e-05,
+      "loss": 1.4792,
+      "step": 13843
+    },
+    {
+      "epoch": 0.637766619062975,
+      "grad_norm": 0.8946540435967779,
+      "learning_rate": 6.12823730167907e-05,
+      "loss": 1.6516,
+      "step": 13844
+    },
+    {
+      "epoch": 0.6378126871516101,
+      "grad_norm": 1.1693808170930633,
+      "learning_rate": 6.126861630522172e-05,
+      "loss": 1.6177,
+      "step": 13845
+    },
+    {
+      "epoch": 0.637858755240245,
+      "grad_norm": 2.523198388053454,
+      "learning_rate": 6.125486045594066e-05,
+      "loss": 1.1554,
+      "step": 13846
+    },
+    {
+      "epoch": 0.6379048233288801,
+      "grad_norm": 0.966279571623004,
+      "learning_rate": 6.124110546925368e-05,
+      "loss": 0.9686,
+      "step": 13847
+    },
+    {
+      "epoch": 0.637950891417515,
+      "grad_norm": 0.9583907032579375,
+      "learning_rate": 6.1227351345467e-05,
+      "loss": 1.3895,
+      "step": 13848
+    },
+    {
+      "epoch": 0.6379969595061501,
+      "grad_norm": 0.729282821213983,
+      "learning_rate": 6.121359808488687e-05,
+      "loss": 1.532,
+      "step": 13849
+    },
+    {
+      "epoch": 0.638043027594785,
+      "grad_norm": 0.7707192091950651,
+      "learning_rate": 6.119984568781946e-05,
+      "loss": 1.3583,
+      "step": 13850
+    },
+    {
+      "epoch": 0.6380890956834201,
+      "grad_norm": 0.9521200140798796,
+      "learning_rate": 6.1186094154571e-05,
+      "loss": 1.0097,
+      "step": 13851
+    },
+    {
+      "epoch": 0.6381351637720551,
+      "grad_norm": 0.8876125788344619,
+      "learning_rate": 6.117234348544754e-05,
+      "loss": 1.4001,
+      "step": 13852
+    },
+    {
+      "epoch": 0.6381812318606901,
+      "grad_norm": 0.8096845502608215,
+      "learning_rate": 6.115859368075535e-05,
+      "loss": 1.1594,
+      "step": 13853
+    },
+    {
+      "epoch": 0.6382272999493251,
+      "grad_norm": 0.9610498475176754,
+      "learning_rate": 6.11448447408004e-05,
+      "loss": 1.5944,
+      "step": 13854
+    },
+    {
+      "epoch": 0.6382733680379601,
+      "grad_norm": 1.2377553209418612,
+      "learning_rate": 6.113109666588894e-05,
+      "loss": 1.2183,
+      "step": 13855
+    },
+    {
+      "epoch": 0.6383194361265951,
+      "grad_norm": 0.7388945931244086,
+      "learning_rate": 6.111734945632692e-05,
+      "loss": 1.3989,
+      "step": 13856
+    },
+    {
+      "epoch": 0.6383655042152301,
+      "grad_norm": 3.5328795582233177,
+      "learning_rate": 6.110360311242045e-05,
+      "loss": 1.3702,
+      "step": 13857
+    },
+    {
+      "epoch": 0.6384115723038651,
+      "grad_norm": 0.7768128021722838,
+      "learning_rate": 6.108985763447556e-05,
+      "loss": 1.3572,
+      "step": 13858
+    },
+    {
+      "epoch": 0.6384576403925001,
+      "grad_norm": 1.442645376214609,
+      "learning_rate": 6.107611302279827e-05,
+      "loss": 1.3839,
+      "step": 13859
+    },
+    {
+      "epoch": 0.6385037084811351,
+      "grad_norm": 0.7764825417428305,
+      "learning_rate": 6.10623692776946e-05,
+      "loss": 1.0969,
+      "step": 13860
+    },
+    {
+      "epoch": 0.6385497765697701,
+      "grad_norm": 0.8176991277322202,
+      "learning_rate": 6.104862639947049e-05,
+      "loss": 1.2675,
+      "step": 13861
+    },
+    {
+      "epoch": 0.6385958446584051,
+      "grad_norm": 0.6648905200232271,
+      "learning_rate": 6.1034884388431946e-05,
+      "loss": 1.0903,
+      "step": 13862
+    },
+    {
+      "epoch": 0.6386419127470401,
+      "grad_norm": 0.6840793484390549,
+      "learning_rate": 6.102114324488487e-05,
+      "loss": 1.2982,
+      "step": 13863
+    },
+    {
+      "epoch": 0.6386879808356751,
+      "grad_norm": 0.7190778696855947,
+      "learning_rate": 6.100740296913523e-05,
+      "loss": 1.4055,
+      "step": 13864
+    },
+    {
+      "epoch": 0.6387340489243102,
+      "grad_norm": 0.9200020827058486,
+      "learning_rate": 6.0993663561488854e-05,
+      "loss": 1.2468,
+      "step": 13865
+    },
+    {
+      "epoch": 0.6387801170129451,
+      "grad_norm": 0.7833607652058346,
+      "learning_rate": 6.097992502225176e-05,
+      "loss": 1.4228,
+      "step": 13866
+    },
+    {
+      "epoch": 0.6388261851015802,
+      "grad_norm": 0.6958133566270482,
+      "learning_rate": 6.0966187351729696e-05,
+      "loss": 1.1286,
+      "step": 13867
+    },
+    {
+      "epoch": 0.6388722531902151,
+      "grad_norm": 1.053984931698078,
+      "learning_rate": 6.0952450550228516e-05,
+      "loss": 1.1867,
+      "step": 13868
+    },
+    {
+      "epoch": 0.6389183212788502,
+      "grad_norm": 2.2450776530447745,
+      "learning_rate": 6.0938714618054095e-05,
+      "loss": 1.5557,
+      "step": 13869
+    },
+    {
+      "epoch": 0.6389643893674851,
+      "grad_norm": 0.856975319659187,
+      "learning_rate": 6.0924979555512206e-05,
+      "loss": 1.2037,
+      "step": 13870
+    },
+    {
+      "epoch": 0.6390104574561202,
+      "grad_norm": 0.9419474095075561,
+      "learning_rate": 6.091124536290866e-05,
+      "loss": 1.3279,
+      "step": 13871
+    },
+    {
+      "epoch": 0.6390565255447551,
+      "grad_norm": 10.231859765189064,
+      "learning_rate": 6.08975120405492e-05,
+      "loss": 1.3831,
+      "step": 13872
+    },
+    {
+      "epoch": 0.6391025936333902,
+      "grad_norm": 0.797630531722436,
+      "learning_rate": 6.088377958873961e-05,
+      "loss": 1.1851,
+      "step": 13873
+    },
+    {
+      "epoch": 0.6391486617220251,
+      "grad_norm": 0.9403864927585904,
+      "learning_rate": 6.0870048007785564e-05,
+      "loss": 0.8578,
+      "step": 13874
+    },
+    {
+      "epoch": 0.6391947298106602,
+      "grad_norm": 0.7788946511930765,
+      "learning_rate": 6.085631729799284e-05,
+      "loss": 1.2314,
+      "step": 13875
+    },
+    {
+      "epoch": 0.6392407978992951,
+      "grad_norm": 0.7327216132463804,
+      "learning_rate": 6.084258745966708e-05,
+      "loss": 1.2795,
+      "step": 13876
+    },
+    {
+      "epoch": 0.6392868659879302,
+      "grad_norm": 0.9788893707753019,
+      "learning_rate": 6.082885849311395e-05,
+      "loss": 1.3647,
+      "step": 13877
+    },
+    {
+      "epoch": 0.6393329340765652,
+      "grad_norm": 0.7533509131448924,
+      "learning_rate": 6.081513039863915e-05,
+      "loss": 1.2956,
+      "step": 13878
+    },
+    {
+      "epoch": 0.6393790021652002,
+      "grad_norm": 12.780667105957653,
+      "learning_rate": 6.080140317654827e-05,
+      "loss": 1.4435,
+      "step": 13879
+    },
+    {
+      "epoch": 0.6394250702538352,
+      "grad_norm": 1.0094527845356955,
+      "learning_rate": 6.0787676827146946e-05,
+      "loss": 0.9048,
+      "step": 13880
+    },
+    {
+      "epoch": 0.6394711383424702,
+      "grad_norm": 1.2979642804052731,
+      "learning_rate": 6.0773951350740755e-05,
+      "loss": 1.3322,
+      "step": 13881
+    },
+    {
+      "epoch": 0.6395172064311052,
+      "grad_norm": 0.7907769249664753,
+      "learning_rate": 6.0760226747635284e-05,
+      "loss": 1.1072,
+      "step": 13882
+    },
+    {
+      "epoch": 0.6395632745197402,
+      "grad_norm": 0.7478977896735222,
+      "learning_rate": 6.0746503018136066e-05,
+      "loss": 1.5646,
+      "step": 13883
+    },
+    {
+      "epoch": 0.6396093426083752,
+      "grad_norm": 0.7172605833811961,
+      "learning_rate": 6.0732780162548686e-05,
+      "loss": 1.4372,
+      "step": 13884
+    },
+    {
+      "epoch": 0.6396554106970102,
+      "grad_norm": 0.8645865530759358,
+      "learning_rate": 6.071905818117861e-05,
+      "loss": 1.1326,
+      "step": 13885
+    },
+    {
+      "epoch": 0.6397014787856452,
+      "grad_norm": 0.8376593829488237,
+      "learning_rate": 6.070533707433134e-05,
+      "loss": 1.5016,
+      "step": 13886
+    },
+    {
+      "epoch": 0.6397475468742801,
+      "grad_norm": 0.7088432247404375,
+      "learning_rate": 6.069161684231238e-05,
+      "loss": 0.9987,
+      "step": 13887
+    },
+    {
+      "epoch": 0.6397936149629152,
+      "grad_norm": 1.4700471351248285,
+      "learning_rate": 6.067789748542716e-05,
+      "loss": 1.5216,
+      "step": 13888
+    },
+    {
+      "epoch": 0.6398396830515501,
+      "grad_norm": 1.0504497774511625,
+      "learning_rate": 6.066417900398114e-05,
+      "loss": 1.324,
+      "step": 13889
+    },
+    {
+      "epoch": 0.6398857511401852,
+      "grad_norm": 1.3044216149372614,
+      "learning_rate": 6.06504613982797e-05,
+      "loss": 1.4731,
+      "step": 13890
+    },
+    {
+      "epoch": 0.6399318192288201,
+      "grad_norm": 1.1400855309365001,
+      "learning_rate": 6.0636744668628295e-05,
+      "loss": 1.3517,
+      "step": 13891
+    },
+    {
+      "epoch": 0.6399778873174552,
+      "grad_norm": 0.7174436249705752,
+      "learning_rate": 6.0623028815332274e-05,
+      "loss": 1.4903,
+      "step": 13892
+    },
+    {
+      "epoch": 0.6400239554060903,
+      "grad_norm": 1.3946061480729914,
+      "learning_rate": 6.0609313838697e-05,
+      "loss": 1.3476,
+      "step": 13893
+    },
+    {
+      "epoch": 0.6400700234947252,
+      "grad_norm": 1.1976332862737697,
+      "learning_rate": 6.059559973902782e-05,
+      "loss": 1.1538,
+      "step": 13894
+    },
+    {
+      "epoch": 0.6401160915833602,
+      "grad_norm": 0.9468385530575917,
+      "learning_rate": 6.0581886516630025e-05,
+      "loss": 1.2575,
+      "step": 13895
+    },
+    {
+      "epoch": 0.6401621596719952,
+      "grad_norm": 0.761849592347987,
+      "learning_rate": 6.0568174171808954e-05,
+      "loss": 1.2431,
+      "step": 13896
+    },
+    {
+      "epoch": 0.6402082277606302,
+      "grad_norm": 1.7783460676775613,
+      "learning_rate": 6.0554462704869865e-05,
+      "loss": 1.4373,
+      "step": 13897
+    },
+    {
+      "epoch": 0.6402542958492652,
+      "grad_norm": 1.0923889951523627,
+      "learning_rate": 6.054075211611805e-05,
+      "loss": 1.2802,
+      "step": 13898
+    },
+    {
+      "epoch": 0.6403003639379002,
+      "grad_norm": 1.0405993154681634,
+      "learning_rate": 6.05270424058587e-05,
+      "loss": 1.0463,
+      "step": 13899
+    },
+    {
+      "epoch": 0.6403464320265352,
+      "grad_norm": 1.0459607697368827,
+      "learning_rate": 6.05133335743971e-05,
+      "loss": 1.5273,
+      "step": 13900
+    },
+    {
+      "epoch": 0.6403925001151702,
+      "grad_norm": 1.1684379573221653,
+      "learning_rate": 6.049962562203839e-05,
+      "loss": 1.0422,
+      "step": 13901
+    },
+    {
+      "epoch": 0.6404385682038052,
+      "grad_norm": 0.7603441227449346,
+      "learning_rate": 6.048591854908782e-05,
+      "loss": 1.4694,
+      "step": 13902
+    },
+    {
+      "epoch": 0.6404846362924402,
+      "grad_norm": 1.0450262037623144,
+      "learning_rate": 6.047221235585051e-05,
+      "loss": 1.2694,
+      "step": 13903
+    },
+    {
+      "epoch": 0.6405307043810752,
+      "grad_norm": 0.748134754333469,
+      "learning_rate": 6.0458507042631605e-05,
+      "loss": 1.2611,
+      "step": 13904
+    },
+    {
+      "epoch": 0.6405767724697102,
+      "grad_norm": 1.0113843859578604,
+      "learning_rate": 6.044480260973626e-05,
+      "loss": 1.5112,
+      "step": 13905
+    },
+    {
+      "epoch": 0.6406228405583453,
+      "grad_norm": 0.919759620554638,
+      "learning_rate": 6.0431099057469534e-05,
+      "loss": 1.1615,
+      "step": 13906
+    },
+    {
+      "epoch": 0.6406689086469802,
+      "grad_norm": 0.9191469587307933,
+      "learning_rate": 6.041739638613657e-05,
+      "loss": 1.459,
+      "step": 13907
+    },
+    {
+      "epoch": 0.6407149767356153,
+      "grad_norm": 0.7512537633742292,
+      "learning_rate": 6.040369459604238e-05,
+      "loss": 1.3214,
+      "step": 13908
+    },
+    {
+      "epoch": 0.6407610448242502,
+      "grad_norm": 1.1688844841481998,
+      "learning_rate": 6.038999368749207e-05,
+      "loss": 1.1537,
+      "step": 13909
+    },
+    {
+      "epoch": 0.6408071129128853,
+      "grad_norm": 0.8604707996341308,
+      "learning_rate": 6.03762936607906e-05,
+      "loss": 1.1365,
+      "step": 13910
+    },
+    {
+      "epoch": 0.6408531810015202,
+      "grad_norm": 0.8941002621868334,
+      "learning_rate": 6.036259451624304e-05,
+      "loss": 1.2572,
+      "step": 13911
+    },
+    {
+      "epoch": 0.6408992490901553,
+      "grad_norm": 1.0566080054231295,
+      "learning_rate": 6.034889625415434e-05,
+      "loss": 1.2722,
+      "step": 13912
+    },
+    {
+      "epoch": 0.6409453171787902,
+      "grad_norm": 0.9795352550699825,
+      "learning_rate": 6.033519887482947e-05,
+      "loss": 1.6616,
+      "step": 13913
+    },
+    {
+      "epoch": 0.6409913852674253,
+      "grad_norm": 1.0169678282189023,
+      "learning_rate": 6.0321502378573394e-05,
+      "loss": 1.447,
+      "step": 13914
+    },
+    {
+      "epoch": 0.6410374533560602,
+      "grad_norm": 0.7206225715927163,
+      "learning_rate": 6.030780676569102e-05,
+      "loss": 1.5276,
+      "step": 13915
+    },
+    {
+      "epoch": 0.6410835214446953,
+      "grad_norm": 0.651409878698151,
+      "learning_rate": 6.0294112036487295e-05,
+      "loss": 1.5908,
+      "step": 13916
+    },
+    {
+      "epoch": 0.6411295895333302,
+      "grad_norm": 6.574786221586101,
+      "learning_rate": 6.028041819126706e-05,
+      "loss": 1.2897,
+      "step": 13917
+    },
+    {
+      "epoch": 0.6411756576219653,
+      "grad_norm": 0.770860752634593,
+      "learning_rate": 6.026672523033523e-05,
+      "loss": 0.8505,
+      "step": 13918
+    },
+    {
+      "epoch": 0.6412217257106003,
+      "grad_norm": 0.7657981064378337,
+      "learning_rate": 6.0253033153996605e-05,
+      "loss": 1.2456,
+      "step": 13919
+    },
+    {
+      "epoch": 0.6412677937992353,
+      "grad_norm": 1.1104686113664448,
+      "learning_rate": 6.0239341962556074e-05,
+      "loss": 1.1734,
+      "step": 13920
+    },
+    {
+      "epoch": 0.6413138618878703,
+      "grad_norm": 0.8859932817530448,
+      "learning_rate": 6.022565165631844e-05,
+      "loss": 1.3879,
+      "step": 13921
+    },
+    {
+      "epoch": 0.6413599299765053,
+      "grad_norm": 0.7043825447426686,
+      "learning_rate": 6.0211962235588426e-05,
+      "loss": 1.4068,
+      "step": 13922
+    },
+    {
+      "epoch": 0.6414059980651403,
+      "grad_norm": 1.051109291666219,
+      "learning_rate": 6.0198273700670894e-05,
+      "loss": 1.2272,
+      "step": 13923
+    },
+    {
+      "epoch": 0.6414520661537753,
+      "grad_norm": 0.6501522729748931,
+      "learning_rate": 6.01845860518705e-05,
+      "loss": 1.4248,
+      "step": 13924
+    },
+    {
+      "epoch": 0.6414981342424103,
+      "grad_norm": 1.4316327081545652,
+      "learning_rate": 6.0170899289492064e-05,
+      "loss": 1.3818,
+      "step": 13925
+    },
+    {
+      "epoch": 0.6415442023310453,
+      "grad_norm": 0.96161893853215,
+      "learning_rate": 6.015721341384024e-05,
+      "loss": 1.2159,
+      "step": 13926
+    },
+    {
+      "epoch": 0.6415902704196803,
+      "grad_norm": 0.6929385877610295,
+      "learning_rate": 6.0143528425219777e-05,
+      "loss": 1.4543,
+      "step": 13927
+    },
+    {
+      "epoch": 0.6416363385083153,
+      "grad_norm": 0.8524805984213788,
+      "learning_rate": 6.012984432393527e-05,
+      "loss": 1.3041,
+      "step": 13928
+    },
+    {
+      "epoch": 0.6416824065969503,
+      "grad_norm": 0.9576641341204977,
+      "learning_rate": 6.011616111029146e-05,
+      "loss": 0.9724,
+      "step": 13929
+    },
+    {
+      "epoch": 0.6417284746855852,
+      "grad_norm": 0.7277639592666734,
+      "learning_rate": 6.010247878459294e-05,
+      "loss": 1.4622,
+      "step": 13930
+    },
+    {
+      "epoch": 0.6417745427742203,
+      "grad_norm": 0.9664018342769243,
+      "learning_rate": 6.0088797347144275e-05,
+      "loss": 1.3511,
+      "step": 13931
+    },
+    {
+      "epoch": 0.6418206108628552,
+      "grad_norm": 0.9967878664072691,
+      "learning_rate": 6.0075116798250166e-05,
+      "loss": 1.2932,
+      "step": 13932
+    },
+    {
+      "epoch": 0.6418666789514903,
+      "grad_norm": 1.442589309368219,
+      "learning_rate": 6.006143713821506e-05,
+      "loss": 1.2481,
+      "step": 13933
+    },
+    {
+      "epoch": 0.6419127470401254,
+      "grad_norm": 0.8157556145110918,
+      "learning_rate": 6.004775836734362e-05,
+      "loss": 1.1989,
+      "step": 13934
+    },
+    {
+      "epoch": 0.6419588151287603,
+      "grad_norm": 1.2784570367200616,
+      "learning_rate": 6.00340804859403e-05,
+      "loss": 1.3573,
+      "step": 13935
+    },
+    {
+      "epoch": 0.6420048832173954,
+      "grad_norm": 0.7499797550268832,
+      "learning_rate": 6.00204034943097e-05,
+      "loss": 1.3299,
+      "step": 13936
+    },
+    {
+      "epoch": 0.6420509513060303,
+      "grad_norm": 0.8102077391284603,
+      "learning_rate": 6.000672739275621e-05,
+      "loss": 1.2702,
+      "step": 13937
+    },
+    {
+      "epoch": 0.6420970193946653,
+      "grad_norm": 0.8360053028622643,
+      "learning_rate": 5.99930521815844e-05,
+      "loss": 0.9434,
+      "step": 13938
+    },
+    {
+      "epoch": 0.6421430874833003,
+      "grad_norm": 0.9935744944640988,
+      "learning_rate": 5.9979377861098696e-05,
+      "loss": 1.1247,
+      "step": 13939
+    },
+    {
+      "epoch": 0.6421891555719353,
+      "grad_norm": 0.8227965027586115,
+      "learning_rate": 5.9965704431603475e-05,
+      "loss": 1.4227,
+      "step": 13940
+    },
+    {
+      "epoch": 0.6422352236605703,
+      "grad_norm": 1.44055167359274,
+      "learning_rate": 5.995203189340326e-05,
+      "loss": 0.9235,
+      "step": 13941
+    },
+    {
+      "epoch": 0.6422812917492053,
+      "grad_norm": 0.796435060988716,
+      "learning_rate": 5.993836024680231e-05,
+      "loss": 1.0723,
+      "step": 13942
+    },
+    {
+      "epoch": 0.6423273598378403,
+      "grad_norm": 1.0660285907988059,
+      "learning_rate": 5.9924689492105165e-05,
+      "loss": 1.1613,
+      "step": 13943
+    },
+    {
+      "epoch": 0.6423734279264753,
+      "grad_norm": 0.7724225170396337,
+      "learning_rate": 5.991101962961602e-05,
+      "loss": 1.2254,
+      "step": 13944
+    },
+    {
+      "epoch": 0.6424194960151103,
+      "grad_norm": 2.4230825349522407,
+      "learning_rate": 5.989735065963935e-05,
+      "loss": 1.2448,
+      "step": 13945
+    },
+    {
+      "epoch": 0.6424655641037453,
+      "grad_norm": 1.001234426164676,
+      "learning_rate": 5.988368258247934e-05,
+      "loss": 1.4187,
+      "step": 13946
+    },
+    {
+      "epoch": 0.6425116321923804,
+      "grad_norm": 0.7999437032412251,
+      "learning_rate": 5.987001539844042e-05,
+      "loss": 1.3458,
+      "step": 13947
+    },
+    {
+      "epoch": 0.6425577002810153,
+      "grad_norm": 1.7262491015740389,
+      "learning_rate": 5.985634910782676e-05,
+      "loss": 1.3571,
+      "step": 13948
+    },
+    {
+      "epoch": 0.6426037683696504,
+      "grad_norm": 0.8699251564978325,
+      "learning_rate": 5.984268371094266e-05,
+      "loss": 1.4452,
+      "step": 13949
+    },
+    {
+      "epoch": 0.6426498364582853,
+      "grad_norm": 2.2205181764388944,
+      "learning_rate": 5.9829019208092364e-05,
+      "loss": 1.3425,
+      "step": 13950
+    },
+    {
+      "epoch": 0.6426959045469204,
+      "grad_norm": 0.9898912616182879,
+      "learning_rate": 5.9815355599580024e-05,
+      "loss": 1.3893,
+      "step": 13951
+    },
+    {
+      "epoch": 0.6427419726355553,
+      "grad_norm": 0.811279768412141,
+      "learning_rate": 5.980169288570996e-05,
+      "loss": 1.3728,
+      "step": 13952
+    },
+    {
+      "epoch": 0.6427880407241904,
+      "grad_norm": 2.5854533456464264,
+      "learning_rate": 5.978803106678622e-05,
+      "loss": 1.4541,
+      "step": 13953
+    },
+    {
+      "epoch": 0.6428341088128253,
+      "grad_norm": 1.0267021050463225,
+      "learning_rate": 5.977437014311309e-05,
+      "loss": 1.1787,
+      "step": 13954
+    },
+    {
+      "epoch": 0.6428801769014604,
+      "grad_norm": 0.7725258277140501,
+      "learning_rate": 5.9760710114994554e-05,
+      "loss": 1.2428,
+      "step": 13955
+    },
+    {
+      "epoch": 0.6429262449900953,
+      "grad_norm": 1.2145222809469765,
+      "learning_rate": 5.974705098273489e-05,
+      "loss": 1.0231,
+      "step": 13956
+    },
+    {
+      "epoch": 0.6429723130787304,
+      "grad_norm": 0.732641428624338,
+      "learning_rate": 5.9733392746638094e-05,
+      "loss": 1.0753,
+      "step": 13957
+    },
+    {
+      "epoch": 0.6430183811673653,
+      "grad_norm": 0.9035399909037811,
+      "learning_rate": 5.9719735407008234e-05,
+      "loss": 1.0891,
+      "step": 13958
+    },
+    {
+      "epoch": 0.6430644492560004,
+      "grad_norm": 1.1345799363722566,
+      "learning_rate": 5.970607896414943e-05,
+      "loss": 1.2226,
+      "step": 13959
+    },
+    {
+      "epoch": 0.6431105173446354,
+      "grad_norm": 1.0388329067696163,
+      "learning_rate": 5.969242341836566e-05,
+      "loss": 1.3539,
+      "step": 13960
+    },
+    {
+      "epoch": 0.6431565854332704,
+      "grad_norm": 1.0333752691233302,
+      "learning_rate": 5.9678768769960994e-05,
+      "loss": 1.1864,
+      "step": 13961
+    },
+    {
+      "epoch": 0.6432026535219054,
+      "grad_norm": 0.9280326794168492,
+      "learning_rate": 5.966511501923939e-05,
+      "loss": 1.1437,
+      "step": 13962
+    },
+    {
+      "epoch": 0.6432487216105404,
+      "grad_norm": 1.821154972720638,
+      "learning_rate": 5.965146216650486e-05,
+      "loss": 1.1173,
+      "step": 13963
+    },
+    {
+      "epoch": 0.6432947896991754,
+      "grad_norm": 1.0071417099445479,
+      "learning_rate": 5.963781021206129e-05,
+      "loss": 1.4397,
+      "step": 13964
+    },
+    {
+      "epoch": 0.6433408577878104,
+      "grad_norm": 0.7054891959700024,
+      "learning_rate": 5.9624159156212756e-05,
+      "loss": 1.3695,
+      "step": 13965
+    },
+    {
+      "epoch": 0.6433869258764454,
+      "grad_norm": 1.4805099115453613,
+      "learning_rate": 5.961050899926304e-05,
+      "loss": 1.29,
+      "step": 13966
+    },
+    {
+      "epoch": 0.6434329939650804,
+      "grad_norm": 0.7762832447285888,
+      "learning_rate": 5.9596859741516074e-05,
+      "loss": 1.3145,
+      "step": 13967
+    },
+    {
+      "epoch": 0.6434790620537154,
+      "grad_norm": 1.0275255808698474,
+      "learning_rate": 5.958321138327578e-05,
+      "loss": 1.175,
+      "step": 13968
+    },
+    {
+      "epoch": 0.6435251301423504,
+      "grad_norm": 0.749185777046689,
+      "learning_rate": 5.956956392484596e-05,
+      "loss": 1.1993,
+      "step": 13969
+    },
+    {
+      "epoch": 0.6435711982309854,
+      "grad_norm": 0.8949726986405077,
+      "learning_rate": 5.95559173665305e-05,
+      "loss": 1.1499,
+      "step": 13970
+    },
+    {
+      "epoch": 0.6436172663196204,
+      "grad_norm": 0.7022970648047032,
+      "learning_rate": 5.9542271708633166e-05,
+      "loss": 1.3765,
+      "step": 13971
+    },
+    {
+      "epoch": 0.6436633344082554,
+      "grad_norm": 0.8960313011755442,
+      "learning_rate": 5.95286269514578e-05,
+      "loss": 1.1888,
+      "step": 13972
+    },
+    {
+      "epoch": 0.6437094024968905,
+      "grad_norm": 0.882645060883647,
+      "learning_rate": 5.951498309530815e-05,
+      "loss": 1.238,
+      "step": 13973
+    },
+    {
+      "epoch": 0.6437554705855254,
+      "grad_norm": 1.1142009288312062,
+      "learning_rate": 5.9501340140488006e-05,
+      "loss": 1.3343,
+      "step": 13974
+    },
+    {
+      "epoch": 0.6438015386741605,
+      "grad_norm": 0.6694217456452806,
+      "learning_rate": 5.948769808730108e-05,
+      "loss": 1.1309,
+      "step": 13975
+    },
+    {
+      "epoch": 0.6438476067627954,
+      "grad_norm": 0.7068357660489011,
+      "learning_rate": 5.947405693605106e-05,
+      "loss": 1.3807,
+      "step": 13976
+    },
+    {
+      "epoch": 0.6438936748514305,
+      "grad_norm": 1.4942960439260076,
+      "learning_rate": 5.9460416687041706e-05,
+      "loss": 1.1998,
+      "step": 13977
+    },
+    {
+      "epoch": 0.6439397429400654,
+      "grad_norm": 0.6921661163093155,
+      "learning_rate": 5.9446777340576646e-05,
+      "loss": 1.2809,
+      "step": 13978
+    },
+    {
+      "epoch": 0.6439858110287005,
+      "grad_norm": 0.788795757332369,
+      "learning_rate": 5.943313889695957e-05,
+      "loss": 1.3078,
+      "step": 13979
+    },
+    {
+      "epoch": 0.6440318791173354,
+      "grad_norm": 1.0329370632278998,
+      "learning_rate": 5.9419501356494075e-05,
+      "loss": 1.1285,
+      "step": 13980
+    },
+    {
+      "epoch": 0.6440779472059704,
+      "grad_norm": 0.7869257406386018,
+      "learning_rate": 5.940586471948384e-05,
+      "loss": 1.3035,
+      "step": 13981
+    },
+    {
+      "epoch": 0.6441240152946054,
+      "grad_norm": 0.7835392628417958,
+      "learning_rate": 5.939222898623239e-05,
+      "loss": 1.1147,
+      "step": 13982
+    },
+    {
+      "epoch": 0.6441700833832404,
+      "grad_norm": 1.555400265809659,
+      "learning_rate": 5.9378594157043355e-05,
+      "loss": 1.1713,
+      "step": 13983
+    },
+    {
+      "epoch": 0.6442161514718754,
+      "grad_norm": 0.8884884578572739,
+      "learning_rate": 5.9364960232220244e-05,
+      "loss": 1.232,
+      "step": 13984
+    },
+    {
+      "epoch": 0.6442622195605104,
+      "grad_norm": 0.7724212510077253,
+      "learning_rate": 5.9351327212066654e-05,
+      "loss": 1.4188,
+      "step": 13985
+    },
+    {
+      "epoch": 0.6443082876491454,
+      "grad_norm": 1.5369461921705312,
+      "learning_rate": 5.933769509688606e-05,
+      "loss": 1.4736,
+      "step": 13986
+    },
+    {
+      "epoch": 0.6443543557377804,
+      "grad_norm": 0.6685821506396328,
+      "learning_rate": 5.932406388698194e-05,
+      "loss": 1.3516,
+      "step": 13987
+    },
+    {
+      "epoch": 0.6444004238264155,
+      "grad_norm": 0.7918224679388631,
+      "learning_rate": 5.931043358265782e-05,
+      "loss": 1.3632,
+      "step": 13988
+    },
+    {
+      "epoch": 0.6444464919150504,
+      "grad_norm": 1.5736348468091108,
+      "learning_rate": 5.9296804184217105e-05,
+      "loss": 1.246,
+      "step": 13989
+    },
+    {
+      "epoch": 0.6444925600036855,
+      "grad_norm": 1.5434550283470236,
+      "learning_rate": 5.9283175691963265e-05,
+      "loss": 1.4348,
+      "step": 13990
+    },
+    {
+      "epoch": 0.6445386280923204,
+      "grad_norm": 0.7974062729387204,
+      "learning_rate": 5.9269548106199694e-05,
+      "loss": 1.1968,
+      "step": 13991
+    },
+    {
+      "epoch": 0.6445846961809555,
+      "grad_norm": 1.2347161936518474,
+      "learning_rate": 5.925592142722981e-05,
+      "loss": 1.0372,
+      "step": 13992
+    },
+    {
+      "epoch": 0.6446307642695904,
+      "grad_norm": 0.837025505566972,
+      "learning_rate": 5.924229565535696e-05,
+      "loss": 1.0638,
+      "step": 13993
+    },
+    {
+      "epoch": 0.6446768323582255,
+      "grad_norm": 1.6980177040803568,
+      "learning_rate": 5.922867079088453e-05,
+      "loss": 1.3615,
+      "step": 13994
+    },
+    {
+      "epoch": 0.6447229004468604,
+      "grad_norm": 0.7257376207422566,
+      "learning_rate": 5.921504683411584e-05,
+      "loss": 1.3515,
+      "step": 13995
+    },
+    {
+      "epoch": 0.6447689685354955,
+      "grad_norm": 0.9202552089436754,
+      "learning_rate": 5.9201423785354185e-05,
+      "loss": 1.4393,
+      "step": 13996
+    },
+    {
+      "epoch": 0.6448150366241304,
+      "grad_norm": 0.8257741238478213,
+      "learning_rate": 5.9187801644902885e-05,
+      "loss": 1.2203,
+      "step": 13997
+    },
+    {
+      "epoch": 0.6448611047127655,
+      "grad_norm": 1.7111793800251072,
+      "learning_rate": 5.9174180413065183e-05,
+      "loss": 1.1024,
+      "step": 13998
+    },
+    {
+      "epoch": 0.6449071728014004,
+      "grad_norm": 1.1529636693741734,
+      "learning_rate": 5.916056009014438e-05,
+      "loss": 1.3134,
+      "step": 13999
+    },
+    {
+      "epoch": 0.6449532408900355,
+      "grad_norm": 1.9229546991454214,
+      "learning_rate": 5.914694067644366e-05,
+      "loss": 1.0349,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6449993089786705,
+      "grad_norm": 2.8768116206126626,
+      "learning_rate": 5.913332217226628e-05,
+      "loss": 0.8906,
+      "step": 14001
+    },
+    {
+      "epoch": 0.6450453770673055,
+      "grad_norm": 1.9226511766633294,
+      "learning_rate": 5.911970457791538e-05,
+      "loss": 1.3585,
+      "step": 14002
+    },
+    {
+      "epoch": 0.6450914451559405,
+      "grad_norm": 0.7909699308522606,
+      "learning_rate": 5.9106087893694194e-05,
+      "loss": 1.2189,
+      "step": 14003
+    },
+    {
+      "epoch": 0.6451375132445755,
+      "grad_norm": 4.1071518350379534,
+      "learning_rate": 5.909247211990584e-05,
+      "loss": 1.5407,
+      "step": 14004
+    },
+    {
+      "epoch": 0.6451835813332105,
+      "grad_norm": 0.7332912335086336,
+      "learning_rate": 5.907885725685343e-05,
+      "loss": 0.9207,
+      "step": 14005
+    },
+    {
+      "epoch": 0.6452296494218455,
+      "grad_norm": 0.7308801438110953,
+      "learning_rate": 5.9065243304840127e-05,
+      "loss": 1.1513,
+      "step": 14006
+    },
+    {
+      "epoch": 0.6452757175104805,
+      "grad_norm": 0.765041913915635,
+      "learning_rate": 5.9051630264168954e-05,
+      "loss": 1.5341,
+      "step": 14007
+    },
+    {
+      "epoch": 0.6453217855991155,
+      "grad_norm": 0.9220529680572143,
+      "learning_rate": 5.903801813514307e-05,
+      "loss": 1.234,
+      "step": 14008
+    },
+    {
+      "epoch": 0.6453678536877505,
+      "grad_norm": 0.7424443668175992,
+      "learning_rate": 5.9024406918065435e-05,
+      "loss": 1.4573,
+      "step": 14009
+    },
+    {
+      "epoch": 0.6454139217763855,
+      "grad_norm": 4.019068968566757,
+      "learning_rate": 5.901079661323916e-05,
+      "loss": 1.3965,
+      "step": 14010
+    },
+    {
+      "epoch": 0.6454599898650205,
+      "grad_norm": 0.8467302838403727,
+      "learning_rate": 5.899718722096719e-05,
+      "loss": 1.2708,
+      "step": 14011
+    },
+    {
+      "epoch": 0.6455060579536555,
+      "grad_norm": 0.8529230496234057,
+      "learning_rate": 5.8983578741552546e-05,
+      "loss": 1.2705,
+      "step": 14012
+    },
+    {
+      "epoch": 0.6455521260422905,
+      "grad_norm": 0.645890839837048,
+      "learning_rate": 5.896997117529821e-05,
+      "loss": 1.4918,
+      "step": 14013
+    },
+    {
+      "epoch": 0.6455981941309256,
+      "grad_norm": 0.8077374181170591,
+      "learning_rate": 5.895636452250709e-05,
+      "loss": 1.3807,
+      "step": 14014
+    },
+    {
+      "epoch": 0.6456442622195605,
+      "grad_norm": 1.0272838278446212,
+      "learning_rate": 5.894275878348215e-05,
+      "loss": 1.4177,
+      "step": 14015
+    },
+    {
+      "epoch": 0.6456903303081956,
+      "grad_norm": 0.6712514949864995,
+      "learning_rate": 5.8929153958526284e-05,
+      "loss": 1.3748,
+      "step": 14016
+    },
+    {
+      "epoch": 0.6457363983968305,
+      "grad_norm": 0.7360913056530708,
+      "learning_rate": 5.89155500479424e-05,
+      "loss": 1.3701,
+      "step": 14017
+    },
+    {
+      "epoch": 0.6457824664854656,
+      "grad_norm": 1.114849050622612,
+      "learning_rate": 5.890194705203332e-05,
+      "loss": 1.5723,
+      "step": 14018
+    },
+    {
+      "epoch": 0.6458285345741005,
+      "grad_norm": 1.1100719152749556,
+      "learning_rate": 5.888834497110195e-05,
+      "loss": 0.9778,
+      "step": 14019
+    },
+    {
+      "epoch": 0.6458746026627356,
+      "grad_norm": 2.661942965328453,
+      "learning_rate": 5.887474380545106e-05,
+      "loss": 1.262,
+      "step": 14020
+    },
+    {
+      "epoch": 0.6459206707513705,
+      "grad_norm": 0.8436330299714814,
+      "learning_rate": 5.8861143555383514e-05,
+      "loss": 1.1671,
+      "step": 14021
+    },
+    {
+      "epoch": 0.6459667388400055,
+      "grad_norm": 0.6688282218024006,
+      "learning_rate": 5.884754422120209e-05,
+      "loss": 1.3061,
+      "step": 14022
+    },
+    {
+      "epoch": 0.6460128069286405,
+      "grad_norm": 0.8111860768225032,
+      "learning_rate": 5.8833945803209464e-05,
+      "loss": 1.2622,
+      "step": 14023
+    },
+    {
+      "epoch": 0.6460588750172755,
+      "grad_norm": 1.1493968374777825,
+      "learning_rate": 5.882034830170848e-05,
+      "loss": 1.121,
+      "step": 14024
+    },
+    {
+      "epoch": 0.6461049431059105,
+      "grad_norm": 3.0839770030307347,
+      "learning_rate": 5.880675171700183e-05,
+      "loss": 1.2255,
+      "step": 14025
+    },
+    {
+      "epoch": 0.6461510111945455,
+      "grad_norm": 0.776525123782563,
+      "learning_rate": 5.879315604939224e-05,
+      "loss": 1.2087,
+      "step": 14026
+    },
+    {
+      "epoch": 0.6461970792831805,
+      "grad_norm": 0.9946929742605601,
+      "learning_rate": 5.877956129918236e-05,
+      "loss": 1.5271,
+      "step": 14027
+    },
+    {
+      "epoch": 0.6462431473718155,
+      "grad_norm": 0.8182877197936056,
+      "learning_rate": 5.8765967466674885e-05,
+      "loss": 1.4904,
+      "step": 14028
+    },
+    {
+      "epoch": 0.6462892154604506,
+      "grad_norm": 1.1440536735778655,
+      "learning_rate": 5.8752374552172415e-05,
+      "loss": 1.1577,
+      "step": 14029
+    },
+    {
+      "epoch": 0.6463352835490855,
+      "grad_norm": 0.6883403213019682,
+      "learning_rate": 5.8738782555977625e-05,
+      "loss": 1.2598,
+      "step": 14030
+    },
+    {
+      "epoch": 0.6463813516377206,
+      "grad_norm": 1.0405025827434242,
+      "learning_rate": 5.872519147839313e-05,
+      "loss": 1.2074,
+      "step": 14031
+    },
+    {
+      "epoch": 0.6464274197263555,
+      "grad_norm": 0.9315522705424355,
+      "learning_rate": 5.87116013197214e-05,
+      "loss": 1.3685,
+      "step": 14032
+    },
+    {
+      "epoch": 0.6464734878149906,
+      "grad_norm": 0.9910058490464285,
+      "learning_rate": 5.869801208026513e-05,
+      "loss": 1.2889,
+      "step": 14033
+    },
+    {
+      "epoch": 0.6465195559036255,
+      "grad_norm": 1.0910182076052535,
+      "learning_rate": 5.8684423760326745e-05,
+      "loss": 1.4362,
+      "step": 14034
+    },
+    {
+      "epoch": 0.6465656239922606,
+      "grad_norm": 1.021318523178161,
+      "learning_rate": 5.867083636020889e-05,
+      "loss": 1.2919,
+      "step": 14035
+    },
+    {
+      "epoch": 0.6466116920808955,
+      "grad_norm": 0.8319497378816799,
+      "learning_rate": 5.865724988021393e-05,
+      "loss": 0.8597,
+      "step": 14036
+    },
+    {
+      "epoch": 0.6466577601695306,
+      "grad_norm": 0.8736644514627426,
+      "learning_rate": 5.8643664320644455e-05,
+      "loss": 1.3197,
+      "step": 14037
+    },
+    {
+      "epoch": 0.6467038282581655,
+      "grad_norm": 0.9148805735836523,
+      "learning_rate": 5.8630079681802876e-05,
+      "loss": 1.2746,
+      "step": 14038
+    },
+    {
+      "epoch": 0.6467498963468006,
+      "grad_norm": 1.2541789692216838,
+      "learning_rate": 5.861649596399163e-05,
+      "loss": 1.3998,
+      "step": 14039
+    },
+    {
+      "epoch": 0.6467959644354355,
+      "grad_norm": 1.9455647408477958,
+      "learning_rate": 5.8602913167513196e-05,
+      "loss": 1.2811,
+      "step": 14040
+    },
+    {
+      "epoch": 0.6468420325240706,
+      "grad_norm": 0.9699576124466287,
+      "learning_rate": 5.858933129266985e-05,
+      "loss": 1.2395,
+      "step": 14041
+    },
+    {
+      "epoch": 0.6468881006127056,
+      "grad_norm": 3.401250298024971,
+      "learning_rate": 5.8575750339764105e-05,
+      "loss": 1.0514,
+      "step": 14042
+    },
+    {
+      "epoch": 0.6469341687013406,
+      "grad_norm": 0.7408966297591238,
+      "learning_rate": 5.856217030909816e-05,
+      "loss": 1.0482,
+      "step": 14043
+    },
+    {
+      "epoch": 0.6469802367899756,
+      "grad_norm": 0.8160951261630703,
+      "learning_rate": 5.854859120097453e-05,
+      "loss": 1.296,
+      "step": 14044
+    },
+    {
+      "epoch": 0.6470263048786106,
+      "grad_norm": 2.0276361308421884,
+      "learning_rate": 5.853501301569538e-05,
+      "loss": 1.2853,
+      "step": 14045
+    },
+    {
+      "epoch": 0.6470723729672456,
+      "grad_norm": 0.7526390762374097,
+      "learning_rate": 5.852143575356313e-05,
+      "loss": 1.4102,
+      "step": 14046
+    },
+    {
+      "epoch": 0.6471184410558806,
+      "grad_norm": 1.333228207013245,
+      "learning_rate": 5.8507859414879926e-05,
+      "loss": 1.33,
+      "step": 14047
+    },
+    {
+      "epoch": 0.6471645091445156,
+      "grad_norm": 1.0482134144607398,
+      "learning_rate": 5.849428399994815e-05,
+      "loss": 1.3592,
+      "step": 14048
+    },
+    {
+      "epoch": 0.6472105772331506,
+      "grad_norm": 1.099145962935947,
+      "learning_rate": 5.8480709509069954e-05,
+      "loss": 1.0978,
+      "step": 14049
+    },
+    {
+      "epoch": 0.6472566453217856,
+      "grad_norm": 1.147412020323743,
+      "learning_rate": 5.846713594254754e-05,
+      "loss": 1.4916,
+      "step": 14050
+    },
+    {
+      "epoch": 0.6473027134104206,
+      "grad_norm": 0.8013545955311668,
+      "learning_rate": 5.845356330068317e-05,
+      "loss": 1.1139,
+      "step": 14051
+    },
+    {
+      "epoch": 0.6473487814990556,
+      "grad_norm": 2.276514690699294,
+      "learning_rate": 5.843999158377893e-05,
+      "loss": 1.2086,
+      "step": 14052
+    },
+    {
+      "epoch": 0.6473948495876906,
+      "grad_norm": 0.9675503920207656,
+      "learning_rate": 5.842642079213707e-05,
+      "loss": 1.1928,
+      "step": 14053
+    },
+    {
+      "epoch": 0.6474409176763256,
+      "grad_norm": 0.7075218025039234,
+      "learning_rate": 5.841285092605962e-05,
+      "loss": 1.3377,
+      "step": 14054
+    },
+    {
+      "epoch": 0.6474869857649607,
+      "grad_norm": 0.7929043918665594,
+      "learning_rate": 5.839928198584879e-05,
+      "loss": 1.4521,
+      "step": 14055
+    },
+    {
+      "epoch": 0.6475330538535956,
+      "grad_norm": 1.3595609312479031,
+      "learning_rate": 5.838571397180655e-05,
+      "loss": 1.082,
+      "step": 14056
+    },
+    {
+      "epoch": 0.6475791219422307,
+      "grad_norm": 0.9768840151317285,
+      "learning_rate": 5.837214688423511e-05,
+      "loss": 1.3086,
+      "step": 14057
+    },
+    {
+      "epoch": 0.6476251900308656,
+      "grad_norm": 0.9141896931387326,
+      "learning_rate": 5.8358580723436386e-05,
+      "loss": 1.0161,
+      "step": 14058
+    },
+    {
+      "epoch": 0.6476712581195007,
+      "grad_norm": 0.9976150851758463,
+      "learning_rate": 5.834501548971247e-05,
+      "loss": 1.4555,
+      "step": 14059
+    },
+    {
+      "epoch": 0.6477173262081356,
+      "grad_norm": 1.008471666852524,
+      "learning_rate": 5.83314511833654e-05,
+      "loss": 1.2695,
+      "step": 14060
+    },
+    {
+      "epoch": 0.6477633942967707,
+      "grad_norm": 1.8501710827094946,
+      "learning_rate": 5.831788780469709e-05,
+      "loss": 1.2127,
+      "step": 14061
+    },
+    {
+      "epoch": 0.6478094623854056,
+      "grad_norm": 0.9542198665596505,
+      "learning_rate": 5.830432535400954e-05,
+      "loss": 1.2986,
+      "step": 14062
+    },
+    {
+      "epoch": 0.6478555304740407,
+      "grad_norm": 1.2684796805545304,
+      "learning_rate": 5.8290763831604676e-05,
+      "loss": 1.2457,
+      "step": 14063
+    },
+    {
+      "epoch": 0.6479015985626756,
+      "grad_norm": 1.7932554575198107,
+      "learning_rate": 5.8277203237784496e-05,
+      "loss": 1.3621,
+      "step": 14064
+    },
+    {
+      "epoch": 0.6479476666513106,
+      "grad_norm": 1.351079318416742,
+      "learning_rate": 5.8263643572850815e-05,
+      "loss": 0.9875,
+      "step": 14065
+    },
+    {
+      "epoch": 0.6479937347399456,
+      "grad_norm": 1.0545735912214593,
+      "learning_rate": 5.825008483710552e-05,
+      "loss": 1.1162,
+      "step": 14066
+    },
+    {
+      "epoch": 0.6480398028285806,
+      "grad_norm": 0.7541897957409628,
+      "learning_rate": 5.8236527030850564e-05,
+      "loss": 1.2309,
+      "step": 14067
+    },
+    {
+      "epoch": 0.6480858709172156,
+      "grad_norm": 0.8234779442040893,
+      "learning_rate": 5.822297015438767e-05,
+      "loss": 1.7292,
+      "step": 14068
+    },
+    {
+      "epoch": 0.6481319390058506,
+      "grad_norm": 1.1220658246833017,
+      "learning_rate": 5.820941420801871e-05,
+      "loss": 1.4244,
+      "step": 14069
+    },
+    {
+      "epoch": 0.6481780070944857,
+      "grad_norm": 0.9645854316348607,
+      "learning_rate": 5.81958591920455e-05,
+      "loss": 1.2517,
+      "step": 14070
+    },
+    {
+      "epoch": 0.6482240751831206,
+      "grad_norm": 1.3806608566939254,
+      "learning_rate": 5.818230510676984e-05,
+      "loss": 1.4184,
+      "step": 14071
+    },
+    {
+      "epoch": 0.6482701432717557,
+      "grad_norm": 1.489189256775977,
+      "learning_rate": 5.816875195249342e-05,
+      "loss": 1.285,
+      "step": 14072
+    },
+    {
+      "epoch": 0.6483162113603906,
+      "grad_norm": 2.398049364653923,
+      "learning_rate": 5.815519972951801e-05,
+      "loss": 1.4024,
+      "step": 14073
+    },
+    {
+      "epoch": 0.6483622794490257,
+      "grad_norm": 1.545941449814051,
+      "learning_rate": 5.8141648438145326e-05,
+      "loss": 1.5489,
+      "step": 14074
+    },
+    {
+      "epoch": 0.6484083475376606,
+      "grad_norm": 2.334974169658527,
+      "learning_rate": 5.812809807867712e-05,
+      "loss": 1.275,
+      "step": 14075
+    },
+    {
+      "epoch": 0.6484544156262957,
+      "grad_norm": 1.0351390719568299,
+      "learning_rate": 5.811454865141499e-05,
+      "loss": 1.4603,
+      "step": 14076
+    },
+    {
+      "epoch": 0.6485004837149306,
+      "grad_norm": 0.9533842233428415,
+      "learning_rate": 5.810100015666056e-05,
+      "loss": 1.566,
+      "step": 14077
+    },
+    {
+      "epoch": 0.6485465518035657,
+      "grad_norm": 2.4871041261233207,
+      "learning_rate": 5.80874525947156e-05,
+      "loss": 1.3041,
+      "step": 14078
+    },
+    {
+      "epoch": 0.6485926198922006,
+      "grad_norm": 1.3712063141063728,
+      "learning_rate": 5.8073905965881584e-05,
+      "loss": 0.8496,
+      "step": 14079
+    },
+    {
+      "epoch": 0.6486386879808357,
+      "grad_norm": 0.8974605001924681,
+      "learning_rate": 5.8060360270460176e-05,
+      "loss": 1.0055,
+      "step": 14080
+    },
+    {
+      "epoch": 0.6486847560694706,
+      "grad_norm": 0.7745146246572913,
+      "learning_rate": 5.804681550875293e-05,
+      "loss": 1.4865,
+      "step": 14081
+    },
+    {
+      "epoch": 0.6487308241581057,
+      "grad_norm": 1.007882042669442,
+      "learning_rate": 5.803327168106145e-05,
+      "loss": 1.0444,
+      "step": 14082
+    },
+    {
+      "epoch": 0.6487768922467407,
+      "grad_norm": 1.08432613803041,
+      "learning_rate": 5.8019728787687176e-05,
+      "loss": 1.5259,
+      "step": 14083
+    },
+    {
+      "epoch": 0.6488229603353757,
+      "grad_norm": 0.891399099335354,
+      "learning_rate": 5.8006186828931666e-05,
+      "loss": 1.1607,
+      "step": 14084
+    },
+    {
+      "epoch": 0.6488690284240107,
+      "grad_norm": 1.522350777533674,
+      "learning_rate": 5.799264580509642e-05,
+      "loss": 1.1619,
+      "step": 14085
+    },
+    {
+      "epoch": 0.6489150965126457,
+      "grad_norm": 1.2683829105761752,
+      "learning_rate": 5.797910571648287e-05,
+      "loss": 1.4579,
+      "step": 14086
+    },
+    {
+      "epoch": 0.6489611646012807,
+      "grad_norm": 2.109174516932318,
+      "learning_rate": 5.796556656339251e-05,
+      "loss": 1.5373,
+      "step": 14087
+    },
+    {
+      "epoch": 0.6490072326899157,
+      "grad_norm": 1.1686905337357911,
+      "learning_rate": 5.795202834612667e-05,
+      "loss": 1.32,
+      "step": 14088
+    },
+    {
+      "epoch": 0.6490533007785507,
+      "grad_norm": 1.133740461286054,
+      "learning_rate": 5.79384910649869e-05,
+      "loss": 1.5908,
+      "step": 14089
+    },
+    {
+      "epoch": 0.6490993688671857,
+      "grad_norm": 0.902040170986817,
+      "learning_rate": 5.792495472027446e-05,
+      "loss": 1.4982,
+      "step": 14090
+    },
+    {
+      "epoch": 0.6491454369558207,
+      "grad_norm": 1.0802263573342732,
+      "learning_rate": 5.7911419312290796e-05,
+      "loss": 1.2303,
+      "step": 14091
+    },
+    {
+      "epoch": 0.6491915050444557,
+      "grad_norm": 1.4889048179330169,
+      "learning_rate": 5.7897884841337135e-05,
+      "loss": 1.0285,
+      "step": 14092
+    },
+    {
+      "epoch": 0.6492375731330907,
+      "grad_norm": 0.8826953374263,
+      "learning_rate": 5.788435130771497e-05,
+      "loss": 1.0516,
+      "step": 14093
+    },
+    {
+      "epoch": 0.6492836412217257,
+      "grad_norm": 1.055992824739771,
+      "learning_rate": 5.7870818711725496e-05,
+      "loss": 1.4436,
+      "step": 14094
+    },
+    {
+      "epoch": 0.6493297093103607,
+      "grad_norm": 0.8265610159958425,
+      "learning_rate": 5.785728705366993e-05,
+      "loss": 1.4746,
+      "step": 14095
+    },
+    {
+      "epoch": 0.6493757773989958,
+      "grad_norm": 0.7786527315285634,
+      "learning_rate": 5.7843756333849684e-05,
+      "loss": 1.3003,
+      "step": 14096
+    },
+    {
+      "epoch": 0.6494218454876307,
+      "grad_norm": 0.7998073080737453,
+      "learning_rate": 5.783022655256589e-05,
+      "loss": 1.2663,
+      "step": 14097
+    },
+    {
+      "epoch": 0.6494679135762658,
+      "grad_norm": 0.9887382822428489,
+      "learning_rate": 5.781669771011984e-05,
+      "loss": 1.0687,
+      "step": 14098
+    },
+    {
+      "epoch": 0.6495139816649007,
+      "grad_norm": 2.5125725739947504,
+      "learning_rate": 5.7803169806812595e-05,
+      "loss": 1.2853,
+      "step": 14099
+    },
+    {
+      "epoch": 0.6495600497535358,
+      "grad_norm": 1.0527855467561362,
+      "learning_rate": 5.778964284294553e-05,
+      "loss": 0.9103,
+      "step": 14100
+    },
+    {
+      "epoch": 0.6496061178421707,
+      "grad_norm": 0.9190442112185383,
+      "learning_rate": 5.7776116818819616e-05,
+      "loss": 1.0395,
+      "step": 14101
+    },
+    {
+      "epoch": 0.6496521859308058,
+      "grad_norm": 2.5010869526812396,
+      "learning_rate": 5.7762591734736115e-05,
+      "loss": 1.3633,
+      "step": 14102
+    },
+    {
+      "epoch": 0.6496982540194407,
+      "grad_norm": 2.149123896138001,
+      "learning_rate": 5.774906759099601e-05,
+      "loss": 1.2907,
+      "step": 14103
+    },
+    {
+      "epoch": 0.6497443221080758,
+      "grad_norm": 1.3052065569260174,
+      "learning_rate": 5.773554438790055e-05,
+      "loss": 1.1866,
+      "step": 14104
+    },
+    {
+      "epoch": 0.6497903901967107,
+      "grad_norm": 0.890807428183744,
+      "learning_rate": 5.772202212575073e-05,
+      "loss": 1.156,
+      "step": 14105
+    },
+    {
+      "epoch": 0.6498364582853458,
+      "grad_norm": 0.691891414847943,
+      "learning_rate": 5.7708500804847506e-05,
+      "loss": 1.2467,
+      "step": 14106
+    },
+    {
+      "epoch": 0.6498825263739807,
+      "grad_norm": 1.0641992268499831,
+      "learning_rate": 5.769498042549209e-05,
+      "loss": 0.9469,
+      "step": 14107
+    },
+    {
+      "epoch": 0.6499285944626157,
+      "grad_norm": 1.0893992670539945,
+      "learning_rate": 5.7681460987985346e-05,
+      "loss": 1.247,
+      "step": 14108
+    },
+    {
+      "epoch": 0.6499746625512508,
+      "grad_norm": 1.3030860297588385,
+      "learning_rate": 5.766794249262836e-05,
+      "loss": 1.3927,
+      "step": 14109
+    },
+    {
+      "epoch": 0.6500207306398857,
+      "grad_norm": 1.0797555622038113,
+      "learning_rate": 5.7654424939721974e-05,
+      "loss": 1.0424,
+      "step": 14110
+    },
+    {
+      "epoch": 0.6500667987285208,
+      "grad_norm": 1.169181763983255,
+      "learning_rate": 5.764090832956728e-05,
+      "loss": 1.0462,
+      "step": 14111
+    },
+    {
+      "epoch": 0.6501128668171557,
+      "grad_norm": 0.9412143256273827,
+      "learning_rate": 5.7627392662465086e-05,
+      "loss": 1.2606,
+      "step": 14112
+    },
+    {
+      "epoch": 0.6501589349057908,
+      "grad_norm": 2.862021437741857,
+      "learning_rate": 5.76138779387164e-05,
+      "loss": 1.2676,
+      "step": 14113
+    },
+    {
+      "epoch": 0.6502050029944257,
+      "grad_norm": 0.7315756033402655,
+      "learning_rate": 5.760036415862199e-05,
+      "loss": 1.5013,
+      "step": 14114
+    },
+    {
+      "epoch": 0.6502510710830608,
+      "grad_norm": 1.3800732339446737,
+      "learning_rate": 5.7586851322482785e-05,
+      "loss": 1.1873,
+      "step": 14115
+    },
+    {
+      "epoch": 0.6502971391716957,
+      "grad_norm": 0.9809576310575365,
+      "learning_rate": 5.757333943059964e-05,
+      "loss": 1.1093,
+      "step": 14116
+    },
+    {
+      "epoch": 0.6503432072603308,
+      "grad_norm": 0.8535736101331448,
+      "learning_rate": 5.755982848327331e-05,
+      "loss": 1.1438,
+      "step": 14117
+    },
+    {
+      "epoch": 0.6503892753489657,
+      "grad_norm": 0.7712169446084659,
+      "learning_rate": 5.754631848080463e-05,
+      "loss": 1.6294,
+      "step": 14118
+    },
+    {
+      "epoch": 0.6504353434376008,
+      "grad_norm": 0.847915562153767,
+      "learning_rate": 5.753280942349437e-05,
+      "loss": 1.2527,
+      "step": 14119
+    },
+    {
+      "epoch": 0.6504814115262357,
+      "grad_norm": 0.7271427341620057,
+      "learning_rate": 5.7519301311643336e-05,
+      "loss": 1.3158,
+      "step": 14120
+    },
+    {
+      "epoch": 0.6505274796148708,
+      "grad_norm": 0.9113246055803043,
+      "learning_rate": 5.750579414555215e-05,
+      "loss": 1.3885,
+      "step": 14121
+    },
+    {
+      "epoch": 0.6505735477035057,
+      "grad_norm": 2.4040237558153628,
+      "learning_rate": 5.749228792552166e-05,
+      "loss": 1.0592,
+      "step": 14122
+    },
+    {
+      "epoch": 0.6506196157921408,
+      "grad_norm": 0.9470591918141777,
+      "learning_rate": 5.7478782651852514e-05,
+      "loss": 1.2364,
+      "step": 14123
+    },
+    {
+      "epoch": 0.6506656838807758,
+      "grad_norm": 0.8204411390610624,
+      "learning_rate": 5.746527832484531e-05,
+      "loss": 1.3662,
+      "step": 14124
+    },
+    {
+      "epoch": 0.6507117519694108,
+      "grad_norm": 0.706671351565998,
+      "learning_rate": 5.745177494480075e-05,
+      "loss": 1.3358,
+      "step": 14125
+    },
+    {
+      "epoch": 0.6507578200580458,
+      "grad_norm": 0.8144937466492954,
+      "learning_rate": 5.7438272512019475e-05,
+      "loss": 0.9814,
+      "step": 14126
+    },
+    {
+      "epoch": 0.6508038881466808,
+      "grad_norm": 0.854969098024431,
+      "learning_rate": 5.742477102680213e-05,
+      "loss": 1.424,
+      "step": 14127
+    },
+    {
+      "epoch": 0.6508499562353158,
+      "grad_norm": 3.6266695765662478,
+      "learning_rate": 5.74112704894492e-05,
+      "loss": 1.3517,
+      "step": 14128
+    },
+    {
+      "epoch": 0.6508960243239508,
+      "grad_norm": 0.8265915817400513,
+      "learning_rate": 5.739777090026132e-05,
+      "loss": 1.3056,
+      "step": 14129
+    },
+    {
+      "epoch": 0.6509420924125858,
+      "grad_norm": 1.2930908754376258,
+      "learning_rate": 5.7384272259539015e-05,
+      "loss": 1.0479,
+      "step": 14130
+    },
+    {
+      "epoch": 0.6509881605012208,
+      "grad_norm": 0.8047764574059056,
+      "learning_rate": 5.7370774567582865e-05,
+      "loss": 1.1297,
+      "step": 14131
+    },
+    {
+      "epoch": 0.6510342285898558,
+      "grad_norm": 0.7723263860587246,
+      "learning_rate": 5.735727782469329e-05,
+      "loss": 1.0712,
+      "step": 14132
+    },
+    {
+      "epoch": 0.6510802966784908,
+      "grad_norm": 0.8040976540005403,
+      "learning_rate": 5.73437820311708e-05,
+      "loss": 1.4455,
+      "step": 14133
+    },
+    {
+      "epoch": 0.6511263647671258,
+      "grad_norm": 1.441581960801826,
+      "learning_rate": 5.73302871873159e-05,
+      "loss": 1.2045,
+      "step": 14134
+    },
+    {
+      "epoch": 0.6511724328557608,
+      "grad_norm": 0.6424792171460241,
+      "learning_rate": 5.731679329342896e-05,
+      "loss": 0.978,
+      "step": 14135
+    },
+    {
+      "epoch": 0.6512185009443958,
+      "grad_norm": 0.7805811544813231,
+      "learning_rate": 5.730330034981044e-05,
+      "loss": 1.468,
+      "step": 14136
+    },
+    {
+      "epoch": 0.6512645690330309,
+      "grad_norm": 1.0693455113276782,
+      "learning_rate": 5.728980835676071e-05,
+      "loss": 1.0048,
+      "step": 14137
+    },
+    {
+      "epoch": 0.6513106371216658,
+      "grad_norm": 0.7632146728338812,
+      "learning_rate": 5.727631731458021e-05,
+      "loss": 1.3095,
+      "step": 14138
+    },
+    {
+      "epoch": 0.6513567052103009,
+      "grad_norm": 1.8849969266854016,
+      "learning_rate": 5.7262827223569214e-05,
+      "loss": 0.8923,
+      "step": 14139
+    },
+    {
+      "epoch": 0.6514027732989358,
+      "grad_norm": 0.7056706880701583,
+      "learning_rate": 5.724933808402809e-05,
+      "loss": 1.3699,
+      "step": 14140
+    },
+    {
+      "epoch": 0.6514488413875709,
+      "grad_norm": 1.047451869437187,
+      "learning_rate": 5.723584989625719e-05,
+      "loss": 1.1794,
+      "step": 14141
+    },
+    {
+      "epoch": 0.6514949094762058,
+      "grad_norm": 0.9481589596685231,
+      "learning_rate": 5.722236266055674e-05,
+      "loss": 1.3247,
+      "step": 14142
+    },
+    {
+      "epoch": 0.6515409775648409,
+      "grad_norm": 0.7264010868810099,
+      "learning_rate": 5.7208876377227026e-05,
+      "loss": 1.1798,
+      "step": 14143
+    },
+    {
+      "epoch": 0.6515870456534758,
+      "grad_norm": 1.0944544934927372,
+      "learning_rate": 5.7195391046568306e-05,
+      "loss": 1.4796,
+      "step": 14144
+    },
+    {
+      "epoch": 0.6516331137421109,
+      "grad_norm": 0.616089028424064,
+      "learning_rate": 5.718190666888085e-05,
+      "loss": 1.4712,
+      "step": 14145
+    },
+    {
+      "epoch": 0.6516791818307458,
+      "grad_norm": 0.7629601104430774,
+      "learning_rate": 5.7168423244464784e-05,
+      "loss": 1.7097,
+      "step": 14146
+    },
+    {
+      "epoch": 0.6517252499193809,
+      "grad_norm": 1.0799006938608162,
+      "learning_rate": 5.715494077362034e-05,
+      "loss": 1.1699,
+      "step": 14147
+    },
+    {
+      "epoch": 0.6517713180080158,
+      "grad_norm": 0.800201743747409,
+      "learning_rate": 5.7141459256647666e-05,
+      "loss": 1.4173,
+      "step": 14148
+    },
+    {
+      "epoch": 0.6518173860966509,
+      "grad_norm": 1.1143833032158164,
+      "learning_rate": 5.712797869384697e-05,
+      "loss": 1.5328,
+      "step": 14149
+    },
+    {
+      "epoch": 0.6518634541852859,
+      "grad_norm": 1.306660186450059,
+      "learning_rate": 5.711449908551827e-05,
+      "loss": 1.4228,
+      "step": 14150
+    },
+    {
+      "epoch": 0.6519095222739208,
+      "grad_norm": 0.9517892070155782,
+      "learning_rate": 5.7101020431961727e-05,
+      "loss": 1.5524,
+      "step": 14151
+    },
+    {
+      "epoch": 0.6519555903625559,
+      "grad_norm": 1.9910898634810772,
+      "learning_rate": 5.708754273347743e-05,
+      "loss": 1.437,
+      "step": 14152
+    },
+    {
+      "epoch": 0.6520016584511908,
+      "grad_norm": 0.7435530862306466,
+      "learning_rate": 5.7074065990365376e-05,
+      "loss": 1.3888,
+      "step": 14153
+    },
+    {
+      "epoch": 0.6520477265398259,
+      "grad_norm": 1.0771619227060365,
+      "learning_rate": 5.7060590202925655e-05,
+      "loss": 1.1493,
+      "step": 14154
+    },
+    {
+      "epoch": 0.6520937946284608,
+      "grad_norm": 0.8405310257344318,
+      "learning_rate": 5.704711537145826e-05,
+      "loss": 1.1797,
+      "step": 14155
+    },
+    {
+      "epoch": 0.6521398627170959,
+      "grad_norm": 1.3757916697869883,
+      "learning_rate": 5.7033641496263235e-05,
+      "loss": 1.3664,
+      "step": 14156
+    },
+    {
+      "epoch": 0.6521859308057308,
+      "grad_norm": 1.5140237129661234,
+      "learning_rate": 5.702016857764047e-05,
+      "loss": 1.3993,
+      "step": 14157
+    },
+    {
+      "epoch": 0.6522319988943659,
+      "grad_norm": 0.7804708716259683,
+      "learning_rate": 5.7006696615889955e-05,
+      "loss": 1.3644,
+      "step": 14158
+    },
+    {
+      "epoch": 0.6522780669830008,
+      "grad_norm": 5.072324735657279,
+      "learning_rate": 5.6993225611311664e-05,
+      "loss": 1.2306,
+      "step": 14159
+    },
+    {
+      "epoch": 0.6523241350716359,
+      "grad_norm": 1.2962540237897295,
+      "learning_rate": 5.697975556420542e-05,
+      "loss": 1.2164,
+      "step": 14160
+    },
+    {
+      "epoch": 0.6523702031602708,
+      "grad_norm": 0.7356763283844373,
+      "learning_rate": 5.696628647487116e-05,
+      "loss": 1.1662,
+      "step": 14161
+    },
+    {
+      "epoch": 0.6524162712489059,
+      "grad_norm": 2.332370211561007,
+      "learning_rate": 5.695281834360874e-05,
+      "loss": 1.154,
+      "step": 14162
+    },
+    {
+      "epoch": 0.6524623393375408,
+      "grad_norm": 0.9196917671960722,
+      "learning_rate": 5.6939351170718045e-05,
+      "loss": 1.4283,
+      "step": 14163
+    },
+    {
+      "epoch": 0.6525084074261759,
+      "grad_norm": 0.9996133824172708,
+      "learning_rate": 5.692588495649883e-05,
+      "loss": 1.332,
+      "step": 14164
+    },
+    {
+      "epoch": 0.6525544755148109,
+      "grad_norm": 1.1349823392013867,
+      "learning_rate": 5.691241970125094e-05,
+      "loss": 1.2882,
+      "step": 14165
+    },
+    {
+      "epoch": 0.6526005436034459,
+      "grad_norm": 1.0451834329925043,
+      "learning_rate": 5.6898955405274124e-05,
+      "loss": 0.9043,
+      "step": 14166
+    },
+    {
+      "epoch": 0.6526466116920809,
+      "grad_norm": 0.9362658163058217,
+      "learning_rate": 5.688549206886822e-05,
+      "loss": 1.3763,
+      "step": 14167
+    },
+    {
+      "epoch": 0.6526926797807159,
+      "grad_norm": 0.8863395537898163,
+      "learning_rate": 5.687202969233286e-05,
+      "loss": 1.1359,
+      "step": 14168
+    },
+    {
+      "epoch": 0.6527387478693509,
+      "grad_norm": 0.7850824153653071,
+      "learning_rate": 5.685856827596782e-05,
+      "loss": 1.2336,
+      "step": 14169
+    },
+    {
+      "epoch": 0.6527848159579859,
+      "grad_norm": 0.9169869723794262,
+      "learning_rate": 5.684510782007282e-05,
+      "loss": 1.4859,
+      "step": 14170
+    },
+    {
+      "epoch": 0.6528308840466209,
+      "grad_norm": 1.1835493235275318,
+      "learning_rate": 5.683164832494745e-05,
+      "loss": 1.2219,
+      "step": 14171
+    },
+    {
+      "epoch": 0.6528769521352559,
+      "grad_norm": 1.3527552816588277,
+      "learning_rate": 5.681818979089143e-05,
+      "loss": 1.3713,
+      "step": 14172
+    },
+    {
+      "epoch": 0.6529230202238909,
+      "grad_norm": 0.7756062668018351,
+      "learning_rate": 5.680473221820437e-05,
+      "loss": 1.0584,
+      "step": 14173
+    },
+    {
+      "epoch": 0.6529690883125259,
+      "grad_norm": 0.8715932537216511,
+      "learning_rate": 5.679127560718591e-05,
+      "loss": 1.1878,
+      "step": 14174
+    },
+    {
+      "epoch": 0.6530151564011609,
+      "grad_norm": 0.7490180138322747,
+      "learning_rate": 5.677781995813558e-05,
+      "loss": 1.2847,
+      "step": 14175
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 0.8487480705758274,
+      "learning_rate": 5.676436527135298e-05,
+      "loss": 1.5116,
+      "step": 14176
+    },
+    {
+      "epoch": 0.6531072925784309,
+      "grad_norm": 0.8231849745946085,
+      "learning_rate": 5.6750911547137696e-05,
+      "loss": 1.5224,
+      "step": 14177
+    },
+    {
+      "epoch": 0.653153360667066,
+      "grad_norm": 1.2975937335969405,
+      "learning_rate": 5.673745878578917e-05,
+      "loss": 1.1794,
+      "step": 14178
+    },
+    {
+      "epoch": 0.6531994287557009,
+      "grad_norm": 1.00583192929205,
+      "learning_rate": 5.672400698760694e-05,
+      "loss": 1.3405,
+      "step": 14179
+    },
+    {
+      "epoch": 0.653245496844336,
+      "grad_norm": 0.7894993263424579,
+      "learning_rate": 5.6710556152890496e-05,
+      "loss": 1.1969,
+      "step": 14180
+    },
+    {
+      "epoch": 0.6532915649329709,
+      "grad_norm": 0.7683041723703811,
+      "learning_rate": 5.669710628193933e-05,
+      "loss": 1.3801,
+      "step": 14181
+    },
+    {
+      "epoch": 0.653337633021606,
+      "grad_norm": 0.6952393604724785,
+      "learning_rate": 5.66836573750528e-05,
+      "loss": 0.9957,
+      "step": 14182
+    },
+    {
+      "epoch": 0.6533837011102409,
+      "grad_norm": 1.0889026522045706,
+      "learning_rate": 5.667020943253038e-05,
+      "loss": 1.0569,
+      "step": 14183
+    },
+    {
+      "epoch": 0.653429769198876,
+      "grad_norm": 0.8099895492576018,
+      "learning_rate": 5.665676245467144e-05,
+      "loss": 1.1763,
+      "step": 14184
+    },
+    {
+      "epoch": 0.6534758372875109,
+      "grad_norm": 0.9192705001566512,
+      "learning_rate": 5.6643316441775405e-05,
+      "loss": 1.374,
+      "step": 14185
+    },
+    {
+      "epoch": 0.653521905376146,
+      "grad_norm": 0.9052641248815689,
+      "learning_rate": 5.66298713941416e-05,
+      "loss": 1.4278,
+      "step": 14186
+    },
+    {
+      "epoch": 0.6535679734647809,
+      "grad_norm": 0.898547220119679,
+      "learning_rate": 5.661642731206926e-05,
+      "loss": 1.2355,
+      "step": 14187
+    },
+    {
+      "epoch": 0.653614041553416,
+      "grad_norm": 0.7381705514764306,
+      "learning_rate": 5.660298419585785e-05,
+      "loss": 1.5302,
+      "step": 14188
+    },
+    {
+      "epoch": 0.6536601096420509,
+      "grad_norm": 1.3021194529513866,
+      "learning_rate": 5.6589542045806565e-05,
+      "loss": 1.5713,
+      "step": 14189
+    },
+    {
+      "epoch": 0.653706177730686,
+      "grad_norm": 0.8454817635487571,
+      "learning_rate": 5.657610086221472e-05,
+      "loss": 1.0314,
+      "step": 14190
+    },
+    {
+      "epoch": 0.653752245819321,
+      "grad_norm": 0.6706165773437048,
+      "learning_rate": 5.656266064538146e-05,
+      "loss": 0.8833,
+      "step": 14191
+    },
+    {
+      "epoch": 0.653798313907956,
+      "grad_norm": 1.0049978469858742,
+      "learning_rate": 5.654922139560616e-05,
+      "loss": 1.3302,
+      "step": 14192
+    },
+    {
+      "epoch": 0.653844381996591,
+      "grad_norm": 0.7417595540300739,
+      "learning_rate": 5.65357831131879e-05,
+      "loss": 1.2656,
+      "step": 14193
+    },
+    {
+      "epoch": 0.653890450085226,
+      "grad_norm": 0.701719801610173,
+      "learning_rate": 5.65223457984259e-05,
+      "loss": 1.4287,
+      "step": 14194
+    },
+    {
+      "epoch": 0.653936518173861,
+      "grad_norm": 1.1810650544316055,
+      "learning_rate": 5.6508909451619375e-05,
+      "loss": 1.5053,
+      "step": 14195
+    },
+    {
+      "epoch": 0.653982586262496,
+      "grad_norm": 0.7166879081979616,
+      "learning_rate": 5.649547407306737e-05,
+      "loss": 1.0989,
+      "step": 14196
+    },
+    {
+      "epoch": 0.654028654351131,
+      "grad_norm": 0.9753994674027175,
+      "learning_rate": 5.6482039663069066e-05,
+      "loss": 1.3315,
+      "step": 14197
+    },
+    {
+      "epoch": 0.6540747224397659,
+      "grad_norm": 0.6878671816852716,
+      "learning_rate": 5.646860622192346e-05,
+      "loss": 1.4299,
+      "step": 14198
+    },
+    {
+      "epoch": 0.654120790528401,
+      "grad_norm": 1.163948484505679,
+      "learning_rate": 5.645517374992977e-05,
+      "loss": 1.1804,
+      "step": 14199
+    },
+    {
+      "epoch": 0.6541668586170359,
+      "grad_norm": 1.0680370583300096,
+      "learning_rate": 5.644174224738694e-05,
+      "loss": 1.0852,
+      "step": 14200
+    },
+    {
+      "epoch": 0.654212926705671,
+      "grad_norm": 1.3914062240062455,
+      "learning_rate": 5.6428311714594074e-05,
+      "loss": 1.2953,
+      "step": 14201
+    },
+    {
+      "epoch": 0.6542589947943059,
+      "grad_norm": 0.8218807840326591,
+      "learning_rate": 5.6414882151850055e-05,
+      "loss": 1.0787,
+      "step": 14202
+    },
+    {
+      "epoch": 0.654305062882941,
+      "grad_norm": 1.000656821310893,
+      "learning_rate": 5.640145355945403e-05,
+      "loss": 0.9745,
+      "step": 14203
+    },
+    {
+      "epoch": 0.6543511309715759,
+      "grad_norm": 0.941227801034748,
+      "learning_rate": 5.638802593770489e-05,
+      "loss": 1.3268,
+      "step": 14204
+    },
+    {
+      "epoch": 0.654397199060211,
+      "grad_norm": 0.9019850235477307,
+      "learning_rate": 5.6374599286901495e-05,
+      "loss": 1.3757,
+      "step": 14205
+    },
+    {
+      "epoch": 0.654443267148846,
+      "grad_norm": 0.8787011439035262,
+      "learning_rate": 5.636117360734292e-05,
+      "loss": 0.8304,
+      "step": 14206
+    },
+    {
+      "epoch": 0.654489335237481,
+      "grad_norm": 0.9245939906256062,
+      "learning_rate": 5.6347748899327956e-05,
+      "loss": 1.2813,
+      "step": 14207
+    },
+    {
+      "epoch": 0.654535403326116,
+      "grad_norm": 0.9225588684890637,
+      "learning_rate": 5.6334325163155556e-05,
+      "loss": 1.3491,
+      "step": 14208
+    },
+    {
+      "epoch": 0.654581471414751,
+      "grad_norm": 1.3778907181342657,
+      "learning_rate": 5.632090239912444e-05,
+      "loss": 1.1349,
+      "step": 14209
+    },
+    {
+      "epoch": 0.654627539503386,
+      "grad_norm": 3.1433408344984106,
+      "learning_rate": 5.630748060753364e-05,
+      "loss": 1.2472,
+      "step": 14210
+    },
+    {
+      "epoch": 0.654673607592021,
+      "grad_norm": 0.8089843156694441,
+      "learning_rate": 5.629405978868182e-05,
+      "loss": 1.2671,
+      "step": 14211
+    },
+    {
+      "epoch": 0.654719675680656,
+      "grad_norm": 0.8069422835974186,
+      "learning_rate": 5.628063994286785e-05,
+      "loss": 1.2247,
+      "step": 14212
+    },
+    {
+      "epoch": 0.654765743769291,
+      "grad_norm": 1.4998635925097241,
+      "learning_rate": 5.626722107039044e-05,
+      "loss": 1.5849,
+      "step": 14213
+    },
+    {
+      "epoch": 0.654811811857926,
+      "grad_norm": 0.8714463300337437,
+      "learning_rate": 5.625380317154836e-05,
+      "loss": 1.2111,
+      "step": 14214
+    },
+    {
+      "epoch": 0.654857879946561,
+      "grad_norm": 0.6578274167182611,
+      "learning_rate": 5.624038624664039e-05,
+      "loss": 1.5575,
+      "step": 14215
+    },
+    {
+      "epoch": 0.654903948035196,
+      "grad_norm": 1.037829921358594,
+      "learning_rate": 5.622697029596515e-05,
+      "loss": 1.3183,
+      "step": 14216
+    },
+    {
+      "epoch": 0.654950016123831,
+      "grad_norm": 0.811745867974791,
+      "learning_rate": 5.6213555319821355e-05,
+      "loss": 1.1908,
+      "step": 14217
+    },
+    {
+      "epoch": 0.654996084212466,
+      "grad_norm": 1.2778575414102,
+      "learning_rate": 5.620014131850768e-05,
+      "loss": 1.3451,
+      "step": 14218
+    },
+    {
+      "epoch": 0.6550421523011011,
+      "grad_norm": 2.255516952823987,
+      "learning_rate": 5.6186728292322785e-05,
+      "loss": 1.329,
+      "step": 14219
+    },
+    {
+      "epoch": 0.655088220389736,
+      "grad_norm": 1.4645956479875808,
+      "learning_rate": 5.617331624156519e-05,
+      "loss": 1.3558,
+      "step": 14220
+    },
+    {
+      "epoch": 0.6551342884783711,
+      "grad_norm": 0.7802221498768651,
+      "learning_rate": 5.615990516653364e-05,
+      "loss": 1.5274,
+      "step": 14221
+    },
+    {
+      "epoch": 0.655180356567006,
+      "grad_norm": 1.1705479020559553,
+      "learning_rate": 5.614649506752664e-05,
+      "loss": 1.5436,
+      "step": 14222
+    },
+    {
+      "epoch": 0.6552264246556411,
+      "grad_norm": 0.8327311881421846,
+      "learning_rate": 5.6133085944842686e-05,
+      "loss": 1.0859,
+      "step": 14223
+    },
+    {
+      "epoch": 0.655272492744276,
+      "grad_norm": 0.7380059171285288,
+      "learning_rate": 5.6119677798780355e-05,
+      "loss": 1.3054,
+      "step": 14224
+    },
+    {
+      "epoch": 0.6553185608329111,
+      "grad_norm": 1.3930187186677743,
+      "learning_rate": 5.610627062963818e-05,
+      "loss": 1.3862,
+      "step": 14225
+    },
+    {
+      "epoch": 0.655364628921546,
+      "grad_norm": 0.9221838153093393,
+      "learning_rate": 5.6092864437714644e-05,
+      "loss": 0.9866,
+      "step": 14226
+    },
+    {
+      "epoch": 0.6554106970101811,
+      "grad_norm": 2.448180453365222,
+      "learning_rate": 5.607945922330817e-05,
+      "loss": 1.1607,
+      "step": 14227
+    },
+    {
+      "epoch": 0.655456765098816,
+      "grad_norm": 0.7122993854371095,
+      "learning_rate": 5.606605498671722e-05,
+      "loss": 1.1793,
+      "step": 14228
+    },
+    {
+      "epoch": 0.6555028331874511,
+      "grad_norm": 0.7057696678848918,
+      "learning_rate": 5.6052651728240234e-05,
+      "loss": 1.4472,
+      "step": 14229
+    },
+    {
+      "epoch": 0.655548901276086,
+      "grad_norm": 1.2221511081120717,
+      "learning_rate": 5.603924944817563e-05,
+      "loss": 1.3845,
+      "step": 14230
+    },
+    {
+      "epoch": 0.6555949693647211,
+      "grad_norm": 0.7092922678008062,
+      "learning_rate": 5.602584814682169e-05,
+      "loss": 1.4105,
+      "step": 14231
+    },
+    {
+      "epoch": 0.6556410374533561,
+      "grad_norm": 1.0375446402516575,
+      "learning_rate": 5.601244782447692e-05,
+      "loss": 1.442,
+      "step": 14232
+    },
+    {
+      "epoch": 0.655687105541991,
+      "grad_norm": 0.9145808267612752,
+      "learning_rate": 5.5999048481439577e-05,
+      "loss": 1.0268,
+      "step": 14233
+    },
+    {
+      "epoch": 0.6557331736306261,
+      "grad_norm": 0.916553575149989,
+      "learning_rate": 5.598565011800794e-05,
+      "loss": 1.0048,
+      "step": 14234
+    },
+    {
+      "epoch": 0.655779241719261,
+      "grad_norm": 0.6389495710586394,
+      "learning_rate": 5.597225273448032e-05,
+      "loss": 1.415,
+      "step": 14235
+    },
+    {
+      "epoch": 0.6558253098078961,
+      "grad_norm": 1.1324003247765169,
+      "learning_rate": 5.5958856331155015e-05,
+      "loss": 1.0785,
+      "step": 14236
+    },
+    {
+      "epoch": 0.655871377896531,
+      "grad_norm": 3.7607730820315304,
+      "learning_rate": 5.594546090833031e-05,
+      "loss": 1.2153,
+      "step": 14237
+    },
+    {
+      "epoch": 0.6559174459851661,
+      "grad_norm": 1.1380294428963424,
+      "learning_rate": 5.593206646630431e-05,
+      "loss": 1.1378,
+      "step": 14238
+    },
+    {
+      "epoch": 0.655963514073801,
+      "grad_norm": 1.4538232129118709,
+      "learning_rate": 5.591867300537531e-05,
+      "loss": 1.218,
+      "step": 14239
+    },
+    {
+      "epoch": 0.6560095821624361,
+      "grad_norm": 1.0413017373787976,
+      "learning_rate": 5.5905280525841476e-05,
+      "loss": 0.9142,
+      "step": 14240
+    },
+    {
+      "epoch": 0.656055650251071,
+      "grad_norm": 1.1833611759524574,
+      "learning_rate": 5.589188902800099e-05,
+      "loss": 1.7214,
+      "step": 14241
+    },
+    {
+      "epoch": 0.6561017183397061,
+      "grad_norm": 0.9764155724884523,
+      "learning_rate": 5.5878498512151924e-05,
+      "loss": 1.1746,
+      "step": 14242
+    },
+    {
+      "epoch": 0.656147786428341,
+      "grad_norm": 1.0471433902730292,
+      "learning_rate": 5.586510897859245e-05,
+      "loss": 1.2492,
+      "step": 14243
+    },
+    {
+      "epoch": 0.6561938545169761,
+      "grad_norm": 0.986916621856756,
+      "learning_rate": 5.585172042762068e-05,
+      "loss": 1.3369,
+      "step": 14244
+    },
+    {
+      "epoch": 0.6562399226056111,
+      "grad_norm": 0.7366954488365783,
+      "learning_rate": 5.583833285953462e-05,
+      "loss": 1.2486,
+      "step": 14245
+    },
+    {
+      "epoch": 0.6562859906942461,
+      "grad_norm": 0.7678257558699019,
+      "learning_rate": 5.582494627463236e-05,
+      "loss": 1.3729,
+      "step": 14246
+    },
+    {
+      "epoch": 0.6563320587828811,
+      "grad_norm": 0.7371102686867349,
+      "learning_rate": 5.581156067321194e-05,
+      "loss": 1.2102,
+      "step": 14247
+    },
+    {
+      "epoch": 0.6563781268715161,
+      "grad_norm": 1.0531852443552183,
+      "learning_rate": 5.579817605557137e-05,
+      "loss": 1.4138,
+      "step": 14248
+    },
+    {
+      "epoch": 0.6564241949601511,
+      "grad_norm": 0.99925959054787,
+      "learning_rate": 5.578479242200859e-05,
+      "loss": 1.3145,
+      "step": 14249
+    },
+    {
+      "epoch": 0.6564702630487861,
+      "grad_norm": 0.8940631021088392,
+      "learning_rate": 5.57714097728216e-05,
+      "loss": 0.8387,
+      "step": 14250
+    },
+    {
+      "epoch": 0.6565163311374211,
+      "grad_norm": 1.1420736926620905,
+      "learning_rate": 5.575802810830837e-05,
+      "loss": 1.1686,
+      "step": 14251
+    },
+    {
+      "epoch": 0.6565623992260561,
+      "grad_norm": 0.864861280339885,
+      "learning_rate": 5.574464742876674e-05,
+      "loss": 0.9714,
+      "step": 14252
+    },
+    {
+      "epoch": 0.6566084673146911,
+      "grad_norm": 0.8108994332431593,
+      "learning_rate": 5.573126773449466e-05,
+      "loss": 1.3201,
+      "step": 14253
+    },
+    {
+      "epoch": 0.6566545354033261,
+      "grad_norm": 0.9215787115046814,
+      "learning_rate": 5.5717889025790006e-05,
+      "loss": 1.2065,
+      "step": 14254
+    },
+    {
+      "epoch": 0.6567006034919611,
+      "grad_norm": 0.8475533539291479,
+      "learning_rate": 5.570451130295065e-05,
+      "loss": 0.9379,
+      "step": 14255
+    },
+    {
+      "epoch": 0.6567466715805961,
+      "grad_norm": 0.7486246935250799,
+      "learning_rate": 5.569113456627437e-05,
+      "loss": 1.1179,
+      "step": 14256
+    },
+    {
+      "epoch": 0.6567927396692311,
+      "grad_norm": 0.6786892776393301,
+      "learning_rate": 5.5677758816059e-05,
+      "loss": 1.5836,
+      "step": 14257
+    },
+    {
+      "epoch": 0.6568388077578661,
+      "grad_norm": 0.7515644389318062,
+      "learning_rate": 5.5664384052602337e-05,
+      "loss": 1.3684,
+      "step": 14258
+    },
+    {
+      "epoch": 0.6568848758465011,
+      "grad_norm": 0.8872921266048616,
+      "learning_rate": 5.5651010276202184e-05,
+      "loss": 1.2191,
+      "step": 14259
+    },
+    {
+      "epoch": 0.6569309439351362,
+      "grad_norm": 1.102910599662564,
+      "learning_rate": 5.5637637487156206e-05,
+      "loss": 1.3483,
+      "step": 14260
+    },
+    {
+      "epoch": 0.6569770120237711,
+      "grad_norm": 0.8078645595592364,
+      "learning_rate": 5.5624265685762156e-05,
+      "loss": 1.0427,
+      "step": 14261
+    },
+    {
+      "epoch": 0.6570230801124062,
+      "grad_norm": 0.9750560384224529,
+      "learning_rate": 5.5610894872317785e-05,
+      "loss": 1.1381,
+      "step": 14262
+    },
+    {
+      "epoch": 0.6570691482010411,
+      "grad_norm": 0.7582441423093352,
+      "learning_rate": 5.55975250471207e-05,
+      "loss": 0.9075,
+      "step": 14263
+    },
+    {
+      "epoch": 0.6571152162896762,
+      "grad_norm": 1.2539685295713028,
+      "learning_rate": 5.5584156210468573e-05,
+      "loss": 1.2698,
+      "step": 14264
+    },
+    {
+      "epoch": 0.6571612843783111,
+      "grad_norm": 1.3422272111873774,
+      "learning_rate": 5.5570788362659054e-05,
+      "loss": 1.1451,
+      "step": 14265
+    },
+    {
+      "epoch": 0.6572073524669462,
+      "grad_norm": 1.9166175108644619,
+      "learning_rate": 5.5557421503989796e-05,
+      "loss": 1.4487,
+      "step": 14266
+    },
+    {
+      "epoch": 0.6572534205555811,
+      "grad_norm": 1.2151786776765237,
+      "learning_rate": 5.5544055634758304e-05,
+      "loss": 0.966,
+      "step": 14267
+    },
+    {
+      "epoch": 0.6572994886442162,
+      "grad_norm": 0.8290031948952832,
+      "learning_rate": 5.553069075526219e-05,
+      "loss": 1.3514,
+      "step": 14268
+    },
+    {
+      "epoch": 0.6573455567328511,
+      "grad_norm": 1.148198032785902,
+      "learning_rate": 5.551732686579903e-05,
+      "loss": 1.3037,
+      "step": 14269
+    },
+    {
+      "epoch": 0.6573916248214862,
+      "grad_norm": 0.8796987624435045,
+      "learning_rate": 5.550396396666628e-05,
+      "loss": 1.0954,
+      "step": 14270
+    },
+    {
+      "epoch": 0.6574376929101211,
+      "grad_norm": 1.2356668026813422,
+      "learning_rate": 5.549060205816149e-05,
+      "loss": 0.9694,
+      "step": 14271
+    },
+    {
+      "epoch": 0.6574837609987562,
+      "grad_norm": 1.0623337298019369,
+      "learning_rate": 5.5477241140582116e-05,
+      "loss": 1.3098,
+      "step": 14272
+    },
+    {
+      "epoch": 0.6575298290873912,
+      "grad_norm": 0.9438780117034069,
+      "learning_rate": 5.546388121422567e-05,
+      "loss": 1.3515,
+      "step": 14273
+    },
+    {
+      "epoch": 0.6575758971760262,
+      "grad_norm": 0.7156637075284286,
+      "learning_rate": 5.54505222793895e-05,
+      "loss": 1.6625,
+      "step": 14274
+    },
+    {
+      "epoch": 0.6576219652646612,
+      "grad_norm": 0.8689771129684043,
+      "learning_rate": 5.5437164336371074e-05,
+      "loss": 1.194,
+      "step": 14275
+    },
+    {
+      "epoch": 0.6576680333532962,
+      "grad_norm": 0.9626682531174533,
+      "learning_rate": 5.5423807385467776e-05,
+      "loss": 0.9388,
+      "step": 14276
+    },
+    {
+      "epoch": 0.6577141014419312,
+      "grad_norm": 0.9568917037571456,
+      "learning_rate": 5.5410451426977e-05,
+      "loss": 1.3246,
+      "step": 14277
+    },
+    {
+      "epoch": 0.6577601695305662,
+      "grad_norm": 0.8832102588520244,
+      "learning_rate": 5.5397096461196044e-05,
+      "loss": 1.2078,
+      "step": 14278
+    },
+    {
+      "epoch": 0.6578062376192012,
+      "grad_norm": 0.9341556261653805,
+      "learning_rate": 5.5383742488422255e-05,
+      "loss": 1.0845,
+      "step": 14279
+    },
+    {
+      "epoch": 0.6578523057078361,
+      "grad_norm": 1.0913333233648337,
+      "learning_rate": 5.537038950895297e-05,
+      "loss": 1.2842,
+      "step": 14280
+    },
+    {
+      "epoch": 0.6578983737964712,
+      "grad_norm": 1.109855619516993,
+      "learning_rate": 5.5357037523085407e-05,
+      "loss": 1.2811,
+      "step": 14281
+    },
+    {
+      "epoch": 0.6579444418851061,
+      "grad_norm": 0.8612665619940686,
+      "learning_rate": 5.5343686531116856e-05,
+      "loss": 1.1888,
+      "step": 14282
+    },
+    {
+      "epoch": 0.6579905099737412,
+      "grad_norm": 0.7968378037378353,
+      "learning_rate": 5.533033653334456e-05,
+      "loss": 1.6474,
+      "step": 14283
+    },
+    {
+      "epoch": 0.6580365780623761,
+      "grad_norm": 1.1033557630536606,
+      "learning_rate": 5.5316987530065755e-05,
+      "loss": 1.455,
+      "step": 14284
+    },
+    {
+      "epoch": 0.6580826461510112,
+      "grad_norm": 0.8249733673252155,
+      "learning_rate": 5.5303639521577576e-05,
+      "loss": 0.806,
+      "step": 14285
+    },
+    {
+      "epoch": 0.6581287142396463,
+      "grad_norm": 1.126404964906607,
+      "learning_rate": 5.529029250817723e-05,
+      "loss": 1.2167,
+      "step": 14286
+    },
+    {
+      "epoch": 0.6581747823282812,
+      "grad_norm": 0.8767060355655152,
+      "learning_rate": 5.5276946490161905e-05,
+      "loss": 1.5929,
+      "step": 14287
+    },
+    {
+      "epoch": 0.6582208504169162,
+      "grad_norm": 0.7320074171730001,
+      "learning_rate": 5.526360146782863e-05,
+      "loss": 1.4546,
+      "step": 14288
+    },
+    {
+      "epoch": 0.6582669185055512,
+      "grad_norm": 0.8557168692337267,
+      "learning_rate": 5.5250257441474606e-05,
+      "loss": 1.2161,
+      "step": 14289
+    },
+    {
+      "epoch": 0.6583129865941862,
+      "grad_norm": 0.9553416539466315,
+      "learning_rate": 5.523691441139679e-05,
+      "loss": 1.3093,
+      "step": 14290
+    },
+    {
+      "epoch": 0.6583590546828212,
+      "grad_norm": 0.8259480902207021,
+      "learning_rate": 5.5223572377892393e-05,
+      "loss": 1.1745,
+      "step": 14291
+    },
+    {
+      "epoch": 0.6584051227714562,
+      "grad_norm": 1.0344610119631445,
+      "learning_rate": 5.521023134125835e-05,
+      "loss": 1.0032,
+      "step": 14292
+    },
+    {
+      "epoch": 0.6584511908600912,
+      "grad_norm": 0.9211772402766624,
+      "learning_rate": 5.519689130179173e-05,
+      "loss": 1.3128,
+      "step": 14293
+    },
+    {
+      "epoch": 0.6584972589487262,
+      "grad_norm": 0.7732511978510529,
+      "learning_rate": 5.5183552259789484e-05,
+      "loss": 1.4293,
+      "step": 14294
+    },
+    {
+      "epoch": 0.6585433270373612,
+      "grad_norm": 1.2172964092948222,
+      "learning_rate": 5.517021421554864e-05,
+      "loss": 1.2037,
+      "step": 14295
+    },
+    {
+      "epoch": 0.6585893951259962,
+      "grad_norm": 0.8469756692876789,
+      "learning_rate": 5.5156877169366105e-05,
+      "loss": 1.2125,
+      "step": 14296
+    },
+    {
+      "epoch": 0.6586354632146312,
+      "grad_norm": 0.7322487511350628,
+      "learning_rate": 5.514354112153874e-05,
+      "loss": 1.4149,
+      "step": 14297
+    },
+    {
+      "epoch": 0.6586815313032662,
+      "grad_norm": 0.7155883477137753,
+      "learning_rate": 5.513020607236361e-05,
+      "loss": 1.1441,
+      "step": 14298
+    },
+    {
+      "epoch": 0.6587275993919012,
+      "grad_norm": 0.956555807655296,
+      "learning_rate": 5.511687202213745e-05,
+      "loss": 1.045,
+      "step": 14299
+    },
+    {
+      "epoch": 0.6587736674805362,
+      "grad_norm": 0.8691393100458359,
+      "learning_rate": 5.510353897115722e-05,
+      "loss": 1.5505,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6588197355691713,
+      "grad_norm": 1.4273186054873284,
+      "learning_rate": 5.509020691971964e-05,
+      "loss": 1.1898,
+      "step": 14301
+    },
+    {
+      "epoch": 0.6588658036578062,
+      "grad_norm": 0.8599236614297266,
+      "learning_rate": 5.507687586812167e-05,
+      "loss": 1.2024,
+      "step": 14302
+    },
+    {
+      "epoch": 0.6589118717464413,
+      "grad_norm": 0.6845919796320906,
+      "learning_rate": 5.506354581666e-05,
+      "loss": 1.5136,
+      "step": 14303
+    },
+    {
+      "epoch": 0.6589579398350762,
+      "grad_norm": 1.1114363660846125,
+      "learning_rate": 5.505021676563142e-05,
+      "loss": 1.019,
+      "step": 14304
+    },
+    {
+      "epoch": 0.6590040079237113,
+      "grad_norm": 0.6468222901571736,
+      "learning_rate": 5.503688871533274e-05,
+      "loss": 1.4951,
+      "step": 14305
+    },
+    {
+      "epoch": 0.6590500760123462,
+      "grad_norm": 1.116074577991044,
+      "learning_rate": 5.502356166606062e-05,
+      "loss": 1.1813,
+      "step": 14306
+    },
+    {
+      "epoch": 0.6590961441009813,
+      "grad_norm": 0.6889009951594997,
+      "learning_rate": 5.501023561811178e-05,
+      "loss": 1.1434,
+      "step": 14307
+    },
+    {
+      "epoch": 0.6591422121896162,
+      "grad_norm": 0.6794162625282145,
+      "learning_rate": 5.499691057178286e-05,
+      "loss": 1.232,
+      "step": 14308
+    },
+    {
+      "epoch": 0.6591882802782513,
+      "grad_norm": 0.7698853031362702,
+      "learning_rate": 5.498358652737061e-05,
+      "loss": 1.2099,
+      "step": 14309
+    },
+    {
+      "epoch": 0.6592343483668862,
+      "grad_norm": 0.9494018904072199,
+      "learning_rate": 5.4970263485171605e-05,
+      "loss": 1.0264,
+      "step": 14310
+    },
+    {
+      "epoch": 0.6592804164555213,
+      "grad_norm": 1.3190020992893805,
+      "learning_rate": 5.49569414454825e-05,
+      "loss": 0.8532,
+      "step": 14311
+    },
+    {
+      "epoch": 0.6593264845441562,
+      "grad_norm": 1.022685400130411,
+      "learning_rate": 5.494362040859979e-05,
+      "loss": 1.0046,
+      "step": 14312
+    },
+    {
+      "epoch": 0.6593725526327913,
+      "grad_norm": 1.2836013392861554,
+      "learning_rate": 5.493030037482019e-05,
+      "loss": 1.1385,
+      "step": 14313
+    },
+    {
+      "epoch": 0.6594186207214263,
+      "grad_norm": 0.9546512274322921,
+      "learning_rate": 5.491698134444018e-05,
+      "loss": 1.3625,
+      "step": 14314
+    },
+    {
+      "epoch": 0.6594646888100613,
+      "grad_norm": 0.8143435485017609,
+      "learning_rate": 5.4903663317756224e-05,
+      "loss": 1.3253,
+      "step": 14315
+    },
+    {
+      "epoch": 0.6595107568986963,
+      "grad_norm": 0.6822511066694773,
+      "learning_rate": 5.489034629506491e-05,
+      "loss": 1.4093,
+      "step": 14316
+    },
+    {
+      "epoch": 0.6595568249873313,
+      "grad_norm": 1.335671223812395,
+      "learning_rate": 5.4877030276662664e-05,
+      "loss": 1.1828,
+      "step": 14317
+    },
+    {
+      "epoch": 0.6596028930759663,
+      "grad_norm": 0.7025944431458857,
+      "learning_rate": 5.4863715262846015e-05,
+      "loss": 1.4394,
+      "step": 14318
+    },
+    {
+      "epoch": 0.6596489611646013,
+      "grad_norm": 0.6154822853223835,
+      "learning_rate": 5.4850401253911274e-05,
+      "loss": 1.2222,
+      "step": 14319
+    },
+    {
+      "epoch": 0.6596950292532363,
+      "grad_norm": 0.6527512956762364,
+      "learning_rate": 5.4837088250155024e-05,
+      "loss": 1.1518,
+      "step": 14320
+    },
+    {
+      "epoch": 0.6597410973418713,
+      "grad_norm": 0.9589181242345369,
+      "learning_rate": 5.482377625187352e-05,
+      "loss": 0.8905,
+      "step": 14321
+    },
+    {
+      "epoch": 0.6597871654305063,
+      "grad_norm": 0.8184972339428982,
+      "learning_rate": 5.481046525936321e-05,
+      "loss": 1.041,
+      "step": 14322
+    },
+    {
+      "epoch": 0.6598332335191412,
+      "grad_norm": 0.9937014918463907,
+      "learning_rate": 5.479715527292037e-05,
+      "loss": 1.158,
+      "step": 14323
+    },
+    {
+      "epoch": 0.6598793016077763,
+      "grad_norm": 0.6876517624023842,
+      "learning_rate": 5.478384629284137e-05,
+      "loss": 1.3778,
+      "step": 14324
+    },
+    {
+      "epoch": 0.6599253696964112,
+      "grad_norm": 1.0040808833392412,
+      "learning_rate": 5.477053831942254e-05,
+      "loss": 1.1521,
+      "step": 14325
+    },
+    {
+      "epoch": 0.6599714377850463,
+      "grad_norm": 0.8140879210680434,
+      "learning_rate": 5.475723135296007e-05,
+      "loss": 1.1153,
+      "step": 14326
+    },
+    {
+      "epoch": 0.6600175058736814,
+      "grad_norm": 0.7590972015898876,
+      "learning_rate": 5.4743925393750284e-05,
+      "loss": 1.4614,
+      "step": 14327
+    },
+    {
+      "epoch": 0.6600635739623163,
+      "grad_norm": 1.0200471694272508,
+      "learning_rate": 5.4730620442089394e-05,
+      "loss": 1.5901,
+      "step": 14328
+    },
+    {
+      "epoch": 0.6601096420509514,
+      "grad_norm": 5.3303210488376775,
+      "learning_rate": 5.4717316498273666e-05,
+      "loss": 1.2042,
+      "step": 14329
+    },
+    {
+      "epoch": 0.6601557101395863,
+      "grad_norm": 0.7477098676894072,
+      "learning_rate": 5.4704013562599154e-05,
+      "loss": 1.3142,
+      "step": 14330
+    },
+    {
+      "epoch": 0.6602017782282213,
+      "grad_norm": 4.400884033068185,
+      "learning_rate": 5.4690711635362204e-05,
+      "loss": 1.3118,
+      "step": 14331
+    },
+    {
+      "epoch": 0.6602478463168563,
+      "grad_norm": 0.7371420420025471,
+      "learning_rate": 5.467741071685887e-05,
+      "loss": 1.3882,
+      "step": 14332
+    },
+    {
+      "epoch": 0.6602939144054913,
+      "grad_norm": 5.129715656631132,
+      "learning_rate": 5.466411080738523e-05,
+      "loss": 1.3421,
+      "step": 14333
+    },
+    {
+      "epoch": 0.6603399824941263,
+      "grad_norm": 0.638446376423647,
+      "learning_rate": 5.465081190723742e-05,
+      "loss": 1.2002,
+      "step": 14334
+    },
+    {
+      "epoch": 0.6603860505827613,
+      "grad_norm": 1.3386487859798957,
+      "learning_rate": 5.4637514016711545e-05,
+      "loss": 1.2854,
+      "step": 14335
+    },
+    {
+      "epoch": 0.6604321186713963,
+      "grad_norm": 0.8795929842799673,
+      "learning_rate": 5.4624217136103664e-05,
+      "loss": 1.4739,
+      "step": 14336
+    },
+    {
+      "epoch": 0.6604781867600313,
+      "grad_norm": 0.8654728918709805,
+      "learning_rate": 5.4610921265709756e-05,
+      "loss": 1.2682,
+      "step": 14337
+    },
+    {
+      "epoch": 0.6605242548486663,
+      "grad_norm": 0.8673086255669261,
+      "learning_rate": 5.4597626405825864e-05,
+      "loss": 1.2334,
+      "step": 14338
+    },
+    {
+      "epoch": 0.6605703229373013,
+      "grad_norm": 1.050654455385326,
+      "learning_rate": 5.458433255674795e-05,
+      "loss": 1.2009,
+      "step": 14339
+    },
+    {
+      "epoch": 0.6606163910259364,
+      "grad_norm": 0.7706952569490473,
+      "learning_rate": 5.4571039718772064e-05,
+      "loss": 0.9206,
+      "step": 14340
+    },
+    {
+      "epoch": 0.6606624591145713,
+      "grad_norm": 2.378666157265115,
+      "learning_rate": 5.4557747892194037e-05,
+      "loss": 1.2087,
+      "step": 14341
+    },
+    {
+      "epoch": 0.6607085272032064,
+      "grad_norm": 0.8343849985146115,
+      "learning_rate": 5.454445707730983e-05,
+      "loss": 1.1532,
+      "step": 14342
+    },
+    {
+      "epoch": 0.6607545952918413,
+      "grad_norm": 0.9785715389986911,
+      "learning_rate": 5.45311672744154e-05,
+      "loss": 1.5235,
+      "step": 14343
+    },
+    {
+      "epoch": 0.6608006633804764,
+      "grad_norm": 0.8271506712920622,
+      "learning_rate": 5.451787848380652e-05,
+      "loss": 1.1434,
+      "step": 14344
+    },
+    {
+      "epoch": 0.6608467314691113,
+      "grad_norm": 1.1210316433689167,
+      "learning_rate": 5.450459070577909e-05,
+      "loss": 1.4479,
+      "step": 14345
+    },
+    {
+      "epoch": 0.6608927995577464,
+      "grad_norm": 0.8087924703697256,
+      "learning_rate": 5.4491303940628935e-05,
+      "loss": 1.2539,
+      "step": 14346
+    },
+    {
+      "epoch": 0.6609388676463813,
+      "grad_norm": 1.6880175310285521,
+      "learning_rate": 5.4478018188651915e-05,
+      "loss": 1.6456,
+      "step": 14347
+    },
+    {
+      "epoch": 0.6609849357350164,
+      "grad_norm": 0.9410002442334503,
+      "learning_rate": 5.446473345014371e-05,
+      "loss": 1.0979,
+      "step": 14348
+    },
+    {
+      "epoch": 0.6610310038236513,
+      "grad_norm": 1.0173134869607046,
+      "learning_rate": 5.445144972540016e-05,
+      "loss": 1.338,
+      "step": 14349
+    },
+    {
+      "epoch": 0.6610770719122864,
+      "grad_norm": 1.7070053460142307,
+      "learning_rate": 5.443816701471698e-05,
+      "loss": 1.1279,
+      "step": 14350
+    },
+    {
+      "epoch": 0.6611231400009213,
+      "grad_norm": 0.7523701562720502,
+      "learning_rate": 5.442488531838992e-05,
+      "loss": 1.4599,
+      "step": 14351
+    },
+    {
+      "epoch": 0.6611692080895564,
+      "grad_norm": 0.705756479707106,
+      "learning_rate": 5.441160463671462e-05,
+      "loss": 1.2484,
+      "step": 14352
+    },
+    {
+      "epoch": 0.6612152761781913,
+      "grad_norm": 1.0811223248423463,
+      "learning_rate": 5.4398324969986755e-05,
+      "loss": 1.2142,
+      "step": 14353
+    },
+    {
+      "epoch": 0.6612613442668264,
+      "grad_norm": 1.3587715435087244,
+      "learning_rate": 5.4385046318502055e-05,
+      "loss": 1.396,
+      "step": 14354
+    },
+    {
+      "epoch": 0.6613074123554614,
+      "grad_norm": 0.8803949592127849,
+      "learning_rate": 5.437176868255602e-05,
+      "loss": 1.3601,
+      "step": 14355
+    },
+    {
+      "epoch": 0.6613534804440964,
+      "grad_norm": 1.035646470308575,
+      "learning_rate": 5.435849206244434e-05,
+      "loss": 0.9917,
+      "step": 14356
+    },
+    {
+      "epoch": 0.6613995485327314,
+      "grad_norm": 4.463685069929477,
+      "learning_rate": 5.4345216458462566e-05,
+      "loss": 1.2061,
+      "step": 14357
+    },
+    {
+      "epoch": 0.6614456166213664,
+      "grad_norm": 0.9273664154750388,
+      "learning_rate": 5.433194187090631e-05,
+      "loss": 1.2785,
+      "step": 14358
+    },
+    {
+      "epoch": 0.6614916847100014,
+      "grad_norm": 0.869325770047276,
+      "learning_rate": 5.431866830007103e-05,
+      "loss": 1.0325,
+      "step": 14359
+    },
+    {
+      "epoch": 0.6615377527986364,
+      "grad_norm": 0.9755509904124222,
+      "learning_rate": 5.4305395746252266e-05,
+      "loss": 1.3286,
+      "step": 14360
+    },
+    {
+      "epoch": 0.6615838208872714,
+      "grad_norm": 1.8911644881894387,
+      "learning_rate": 5.4292124209745565e-05,
+      "loss": 1.3615,
+      "step": 14361
+    },
+    {
+      "epoch": 0.6616298889759064,
+      "grad_norm": 1.0594732006436756,
+      "learning_rate": 5.427885369084629e-05,
+      "loss": 1.2586,
+      "step": 14362
+    },
+    {
+      "epoch": 0.6616759570645414,
+      "grad_norm": 1.369625114704906,
+      "learning_rate": 5.426558418984995e-05,
+      "loss": 1.2452,
+      "step": 14363
+    },
+    {
+      "epoch": 0.6617220251531764,
+      "grad_norm": 0.6853677677400088,
+      "learning_rate": 5.425231570705197e-05,
+      "loss": 1.4634,
+      "step": 14364
+    },
+    {
+      "epoch": 0.6617680932418114,
+      "grad_norm": 1.020975058741279,
+      "learning_rate": 5.423904824274777e-05,
+      "loss": 1.3588,
+      "step": 14365
+    },
+    {
+      "epoch": 0.6618141613304463,
+      "grad_norm": 1.0219292246087799,
+      "learning_rate": 5.4225781797232655e-05,
+      "loss": 1.127,
+      "step": 14366
+    },
+    {
+      "epoch": 0.6618602294190814,
+      "grad_norm": 0.8188879945854216,
+      "learning_rate": 5.4212516370802036e-05,
+      "loss": 1.2423,
+      "step": 14367
+    },
+    {
+      "epoch": 0.6619062975077165,
+      "grad_norm": 0.8251128104539746,
+      "learning_rate": 5.419925196375122e-05,
+      "loss": 1.3289,
+      "step": 14368
+    },
+    {
+      "epoch": 0.6619523655963514,
+      "grad_norm": 1.1531837106747793,
+      "learning_rate": 5.4185988576375576e-05,
+      "loss": 1.085,
+      "step": 14369
+    },
+    {
+      "epoch": 0.6619984336849865,
+      "grad_norm": 1.16312955194683,
+      "learning_rate": 5.4172726208970294e-05,
+      "loss": 1.4049,
+      "step": 14370
+    },
+    {
+      "epoch": 0.6620445017736214,
+      "grad_norm": 0.7498193747277461,
+      "learning_rate": 5.415946486183069e-05,
+      "loss": 1.4005,
+      "step": 14371
+    },
+    {
+      "epoch": 0.6620905698622565,
+      "grad_norm": 1.4980257759130495,
+      "learning_rate": 5.414620453525204e-05,
+      "loss": 1.1246,
+      "step": 14372
+    },
+    {
+      "epoch": 0.6621366379508914,
+      "grad_norm": 1.0801611653650345,
+      "learning_rate": 5.4132945229529484e-05,
+      "loss": 0.9824,
+      "step": 14373
+    },
+    {
+      "epoch": 0.6621827060395264,
+      "grad_norm": 0.9374222859167242,
+      "learning_rate": 5.411968694495826e-05,
+      "loss": 1.1927,
+      "step": 14374
+    },
+    {
+      "epoch": 0.6622287741281614,
+      "grad_norm": 0.7225548916734373,
+      "learning_rate": 5.410642968183354e-05,
+      "loss": 1.6664,
+      "step": 14375
+    },
+    {
+      "epoch": 0.6622748422167964,
+      "grad_norm": 0.5919429558119484,
+      "learning_rate": 5.4093173440450504e-05,
+      "loss": 1.2794,
+      "step": 14376
+    },
+    {
+      "epoch": 0.6623209103054314,
+      "grad_norm": 0.780753341060828,
+      "learning_rate": 5.407991822110421e-05,
+      "loss": 1.6778,
+      "step": 14377
+    },
+    {
+      "epoch": 0.6623669783940664,
+      "grad_norm": 3.1137232057348605,
+      "learning_rate": 5.40666640240898e-05,
+      "loss": 1.0873,
+      "step": 14378
+    },
+    {
+      "epoch": 0.6624130464827014,
+      "grad_norm": 1.7992419385501124,
+      "learning_rate": 5.40534108497024e-05,
+      "loss": 1.2437,
+      "step": 14379
+    },
+    {
+      "epoch": 0.6624591145713364,
+      "grad_norm": 0.7846034647359252,
+      "learning_rate": 5.404015869823699e-05,
+      "loss": 1.2092,
+      "step": 14380
+    },
+    {
+      "epoch": 0.6625051826599715,
+      "grad_norm": 0.7994538307627936,
+      "learning_rate": 5.402690756998864e-05,
+      "loss": 1.544,
+      "step": 14381
+    },
+    {
+      "epoch": 0.6625512507486064,
+      "grad_norm": 1.9984866887040789,
+      "learning_rate": 5.401365746525235e-05,
+      "loss": 1.3344,
+      "step": 14382
+    },
+    {
+      "epoch": 0.6625973188372415,
+      "grad_norm": 0.707542135599105,
+      "learning_rate": 5.400040838432319e-05,
+      "loss": 1.2672,
+      "step": 14383
+    },
+    {
+      "epoch": 0.6626433869258764,
+      "grad_norm": 2.52442461022207,
+      "learning_rate": 5.398716032749602e-05,
+      "loss": 1.1506,
+      "step": 14384
+    },
+    {
+      "epoch": 0.6626894550145115,
+      "grad_norm": 0.7814448633053994,
+      "learning_rate": 5.397391329506583e-05,
+      "loss": 1.5608,
+      "step": 14385
+    },
+    {
+      "epoch": 0.6627355231031464,
+      "grad_norm": 1.5027505262145455,
+      "learning_rate": 5.396066728732754e-05,
+      "loss": 1.126,
+      "step": 14386
+    },
+    {
+      "epoch": 0.6627815911917815,
+      "grad_norm": 0.8075769525488479,
+      "learning_rate": 5.394742230457608e-05,
+      "loss": 1.3036,
+      "step": 14387
+    },
+    {
+      "epoch": 0.6628276592804164,
+      "grad_norm": 0.8894010483894693,
+      "learning_rate": 5.393417834710631e-05,
+      "loss": 1.2612,
+      "step": 14388
+    },
+    {
+      "epoch": 0.6628737273690515,
+      "grad_norm": 0.8391323018878251,
+      "learning_rate": 5.392093541521299e-05,
+      "loss": 1.3282,
+      "step": 14389
+    },
+    {
+      "epoch": 0.6629197954576864,
+      "grad_norm": 1.2207318435093544,
+      "learning_rate": 5.3907693509191125e-05,
+      "loss": 1.3687,
+      "step": 14390
+    },
+    {
+      "epoch": 0.6629658635463215,
+      "grad_norm": 1.1045096372852667,
+      "learning_rate": 5.389445262933539e-05,
+      "loss": 1.092,
+      "step": 14391
+    },
+    {
+      "epoch": 0.6630119316349564,
+      "grad_norm": 0.6459441844757946,
+      "learning_rate": 5.3881212775940606e-05,
+      "loss": 1.3641,
+      "step": 14392
+    },
+    {
+      "epoch": 0.6630579997235915,
+      "grad_norm": 1.1229000251574233,
+      "learning_rate": 5.386797394930154e-05,
+      "loss": 1.1899,
+      "step": 14393
+    },
+    {
+      "epoch": 0.6631040678122264,
+      "grad_norm": 5.348554696892905,
+      "learning_rate": 5.385473614971298e-05,
+      "loss": 1.3156,
+      "step": 14394
+    },
+    {
+      "epoch": 0.6631501359008615,
+      "grad_norm": 0.7571358214675621,
+      "learning_rate": 5.384149937746956e-05,
+      "loss": 1.1532,
+      "step": 14395
+    },
+    {
+      "epoch": 0.6631962039894965,
+      "grad_norm": 1.1045503968495547,
+      "learning_rate": 5.3828263632866006e-05,
+      "loss": 1.0484,
+      "step": 14396
+    },
+    {
+      "epoch": 0.6632422720781315,
+      "grad_norm": 0.9332597221073956,
+      "learning_rate": 5.381502891619704e-05,
+      "loss": 1.274,
+      "step": 14397
+    },
+    {
+      "epoch": 0.6632883401667665,
+      "grad_norm": 1.1307297792458315,
+      "learning_rate": 5.380179522775722e-05,
+      "loss": 1.3641,
+      "step": 14398
+    },
+    {
+      "epoch": 0.6633344082554015,
+      "grad_norm": 0.9456270604055681,
+      "learning_rate": 5.3788562567841264e-05,
+      "loss": 1.2725,
+      "step": 14399
+    },
+    {
+      "epoch": 0.6633804763440365,
+      "grad_norm": 0.7715359520214532,
+      "learning_rate": 5.377533093674364e-05,
+      "loss": 1.5178,
+      "step": 14400
+    },
+    {
+      "epoch": 0.6634265444326715,
+      "grad_norm": 1.0425126615298212,
+      "learning_rate": 5.3762100334759094e-05,
+      "loss": 1.1588,
+      "step": 14401
+    },
+    {
+      "epoch": 0.6634726125213065,
+      "grad_norm": 0.839886694068636,
+      "learning_rate": 5.3748870762182066e-05,
+      "loss": 1.3012,
+      "step": 14402
+    },
+    {
+      "epoch": 0.6635186806099415,
+      "grad_norm": 0.8632194905626215,
+      "learning_rate": 5.3735642219307126e-05,
+      "loss": 1.581,
+      "step": 14403
+    },
+    {
+      "epoch": 0.6635647486985765,
+      "grad_norm": 0.8374532927537872,
+      "learning_rate": 5.372241470642878e-05,
+      "loss": 1.2779,
+      "step": 14404
+    },
+    {
+      "epoch": 0.6636108167872115,
+      "grad_norm": 1.3446599389001066,
+      "learning_rate": 5.3709188223841566e-05,
+      "loss": 1.4403,
+      "step": 14405
+    },
+    {
+      "epoch": 0.6636568848758465,
+      "grad_norm": 1.114592171174378,
+      "learning_rate": 5.36959627718399e-05,
+      "loss": 1.4463,
+      "step": 14406
+    },
+    {
+      "epoch": 0.6637029529644815,
+      "grad_norm": 1.0412342042749598,
+      "learning_rate": 5.3682738350718163e-05,
+      "loss": 1.2245,
+      "step": 14407
+    },
+    {
+      "epoch": 0.6637490210531165,
+      "grad_norm": 1.168747345767826,
+      "learning_rate": 5.366951496077091e-05,
+      "loss": 1.4775,
+      "step": 14408
+    },
+    {
+      "epoch": 0.6637950891417516,
+      "grad_norm": 1.0257860744624128,
+      "learning_rate": 5.365629260229242e-05,
+      "loss": 1.3546,
+      "step": 14409
+    },
+    {
+      "epoch": 0.6638411572303865,
+      "grad_norm": 1.0258623322312215,
+      "learning_rate": 5.364307127557717e-05,
+      "loss": 1.0769,
+      "step": 14410
+    },
+    {
+      "epoch": 0.6638872253190216,
+      "grad_norm": 1.732796790913698,
+      "learning_rate": 5.362985098091936e-05,
+      "loss": 1.1257,
+      "step": 14411
+    },
+    {
+      "epoch": 0.6639332934076565,
+      "grad_norm": 0.750078729651733,
+      "learning_rate": 5.3616631718613485e-05,
+      "loss": 1.0308,
+      "step": 14412
+    },
+    {
+      "epoch": 0.6639793614962916,
+      "grad_norm": 0.9470094389218706,
+      "learning_rate": 5.3603413488953744e-05,
+      "loss": 1.4403,
+      "step": 14413
+    },
+    {
+      "epoch": 0.6640254295849265,
+      "grad_norm": 1.0718104276130263,
+      "learning_rate": 5.3590196292234495e-05,
+      "loss": 1.2114,
+      "step": 14414
+    },
+    {
+      "epoch": 0.6640714976735616,
+      "grad_norm": 1.0922709788089373,
+      "learning_rate": 5.35769801287499e-05,
+      "loss": 1.2148,
+      "step": 14415
+    },
+    {
+      "epoch": 0.6641175657621965,
+      "grad_norm": 0.8232090125037996,
+      "learning_rate": 5.356376499879424e-05,
+      "loss": 1.0965,
+      "step": 14416
+    },
+    {
+      "epoch": 0.6641636338508315,
+      "grad_norm": 0.9940728670807322,
+      "learning_rate": 5.355055090266178e-05,
+      "loss": 1.4356,
+      "step": 14417
+    },
+    {
+      "epoch": 0.6642097019394665,
+      "grad_norm": 1.6646290163882136,
+      "learning_rate": 5.3537337840646584e-05,
+      "loss": 0.908,
+      "step": 14418
+    },
+    {
+      "epoch": 0.6642557700281015,
+      "grad_norm": 0.8954251448547281,
+      "learning_rate": 5.352412581304296e-05,
+      "loss": 1.2446,
+      "step": 14419
+    },
+    {
+      "epoch": 0.6643018381167365,
+      "grad_norm": 1.7275839077482493,
+      "learning_rate": 5.351091482014495e-05,
+      "loss": 1.2224,
+      "step": 14420
+    },
+    {
+      "epoch": 0.6643479062053715,
+      "grad_norm": 1.0924188083849578,
+      "learning_rate": 5.3497704862246755e-05,
+      "loss": 1.5143,
+      "step": 14421
+    },
+    {
+      "epoch": 0.6643939742940066,
+      "grad_norm": 0.8030246009857365,
+      "learning_rate": 5.348449593964235e-05,
+      "loss": 1.0472,
+      "step": 14422
+    },
+    {
+      "epoch": 0.6644400423826415,
+      "grad_norm": 4.236619359556944,
+      "learning_rate": 5.347128805262596e-05,
+      "loss": 1.099,
+      "step": 14423
+    },
+    {
+      "epoch": 0.6644861104712766,
+      "grad_norm": 0.9213246479877406,
+      "learning_rate": 5.3458081201491564e-05,
+      "loss": 1.5067,
+      "step": 14424
+    },
+    {
+      "epoch": 0.6645321785599115,
+      "grad_norm": 1.0287966128542811,
+      "learning_rate": 5.3444875386533144e-05,
+      "loss": 1.0114,
+      "step": 14425
+    },
+    {
+      "epoch": 0.6645782466485466,
+      "grad_norm": 1.2512179341685525,
+      "learning_rate": 5.3431670608044746e-05,
+      "loss": 1.5677,
+      "step": 14426
+    },
+    {
+      "epoch": 0.6646243147371815,
+      "grad_norm": 1.8170250864760726,
+      "learning_rate": 5.341846686632037e-05,
+      "loss": 1.5359,
+      "step": 14427
+    },
+    {
+      "epoch": 0.6646703828258166,
+      "grad_norm": 1.4172682825877916,
+      "learning_rate": 5.340526416165399e-05,
+      "loss": 1.0373,
+      "step": 14428
+    },
+    {
+      "epoch": 0.6647164509144515,
+      "grad_norm": 0.9195440428374082,
+      "learning_rate": 5.3392062494339435e-05,
+      "loss": 1.2122,
+      "step": 14429
+    },
+    {
+      "epoch": 0.6647625190030866,
+      "grad_norm": 1.9169646916998484,
+      "learning_rate": 5.337886186467076e-05,
+      "loss": 1.1621,
+      "step": 14430
+    },
+    {
+      "epoch": 0.6648085870917215,
+      "grad_norm": 1.6406764538857712,
+      "learning_rate": 5.336566227294175e-05,
+      "loss": 1.2693,
+      "step": 14431
+    },
+    {
+      "epoch": 0.6648546551803566,
+      "grad_norm": 0.9086332568188413,
+      "learning_rate": 5.335246371944635e-05,
+      "loss": 1.3667,
+      "step": 14432
+    },
+    {
+      "epoch": 0.6649007232689915,
+      "grad_norm": 0.8943316849824468,
+      "learning_rate": 5.3339266204478334e-05,
+      "loss": 1.506,
+      "step": 14433
+    },
+    {
+      "epoch": 0.6649467913576266,
+      "grad_norm": 0.7183449303670774,
+      "learning_rate": 5.332606972833154e-05,
+      "loss": 1.2753,
+      "step": 14434
+    },
+    {
+      "epoch": 0.6649928594462615,
+      "grad_norm": 12.018071630671804,
+      "learning_rate": 5.331287429129982e-05,
+      "loss": 1.1547,
+      "step": 14435
+    },
+    {
+      "epoch": 0.6650389275348966,
+      "grad_norm": 1.0639729719436093,
+      "learning_rate": 5.329967989367686e-05,
+      "loss": 0.9618,
+      "step": 14436
+    },
+    {
+      "epoch": 0.6650849956235316,
+      "grad_norm": 1.5618363016743064,
+      "learning_rate": 5.328648653575645e-05,
+      "loss": 1.5359,
+      "step": 14437
+    },
+    {
+      "epoch": 0.6651310637121666,
+      "grad_norm": 1.3994436659821845,
+      "learning_rate": 5.327329421783232e-05,
+      "loss": 1.3621,
+      "step": 14438
+    },
+    {
+      "epoch": 0.6651771318008016,
+      "grad_norm": 0.8615304229485548,
+      "learning_rate": 5.326010294019822e-05,
+      "loss": 0.9322,
+      "step": 14439
+    },
+    {
+      "epoch": 0.6652231998894366,
+      "grad_norm": 1.053029324694184,
+      "learning_rate": 5.3246912703147745e-05,
+      "loss": 0.9889,
+      "step": 14440
+    },
+    {
+      "epoch": 0.6652692679780716,
+      "grad_norm": 1.2783489421681875,
+      "learning_rate": 5.32337235069746e-05,
+      "loss": 1.3032,
+      "step": 14441
+    },
+    {
+      "epoch": 0.6653153360667066,
+      "grad_norm": 1.4524134097691574,
+      "learning_rate": 5.3220535351972444e-05,
+      "loss": 1.1189,
+      "step": 14442
+    },
+    {
+      "epoch": 0.6653614041553416,
+      "grad_norm": 0.7943335865003579,
+      "learning_rate": 5.320734823843482e-05,
+      "loss": 1.2788,
+      "step": 14443
+    },
+    {
+      "epoch": 0.6654074722439766,
+      "grad_norm": 0.9911616035479328,
+      "learning_rate": 5.319416216665536e-05,
+      "loss": 1.2435,
+      "step": 14444
+    },
+    {
+      "epoch": 0.6654535403326116,
+      "grad_norm": 4.052608610484512,
+      "learning_rate": 5.318097713692762e-05,
+      "loss": 1.0746,
+      "step": 14445
+    },
+    {
+      "epoch": 0.6654996084212466,
+      "grad_norm": 0.9402177071483172,
+      "learning_rate": 5.316779314954519e-05,
+      "loss": 1.5838,
+      "step": 14446
+    },
+    {
+      "epoch": 0.6655456765098816,
+      "grad_norm": 1.3530777491729447,
+      "learning_rate": 5.31546102048015e-05,
+      "loss": 1.2415,
+      "step": 14447
+    },
+    {
+      "epoch": 0.6655917445985166,
+      "grad_norm": 1.4886959848224328,
+      "learning_rate": 5.3141428302990105e-05,
+      "loss": 1.1949,
+      "step": 14448
+    },
+    {
+      "epoch": 0.6656378126871516,
+      "grad_norm": 2.4143412310243955,
+      "learning_rate": 5.3128247444404454e-05,
+      "loss": 1.3435,
+      "step": 14449
+    },
+    {
+      "epoch": 0.6656838807757867,
+      "grad_norm": 1.212867416207893,
+      "learning_rate": 5.3115067629338045e-05,
+      "loss": 1.3952,
+      "step": 14450
+    },
+    {
+      "epoch": 0.6657299488644216,
+      "grad_norm": 0.7918796895912911,
+      "learning_rate": 5.310188885808423e-05,
+      "loss": 1.305,
+      "step": 14451
+    },
+    {
+      "epoch": 0.6657760169530567,
+      "grad_norm": 1.046972085548711,
+      "learning_rate": 5.308871113093643e-05,
+      "loss": 1.0627,
+      "step": 14452
+    },
+    {
+      "epoch": 0.6658220850416916,
+      "grad_norm": 0.9199279251984717,
+      "learning_rate": 5.30755344481881e-05,
+      "loss": 1.1558,
+      "step": 14453
+    },
+    {
+      "epoch": 0.6658681531303267,
+      "grad_norm": 1.3164313626971127,
+      "learning_rate": 5.306235881013249e-05,
+      "loss": 1.0202,
+      "step": 14454
+    },
+    {
+      "epoch": 0.6659142212189616,
+      "grad_norm": 0.7836095347224861,
+      "learning_rate": 5.3049184217062984e-05,
+      "loss": 1.0722,
+      "step": 14455
+    },
+    {
+      "epoch": 0.6659602893075967,
+      "grad_norm": 1.0102572942627557,
+      "learning_rate": 5.3036010669272884e-05,
+      "loss": 1.3847,
+      "step": 14456
+    },
+    {
+      "epoch": 0.6660063573962316,
+      "grad_norm": 1.0529055131481426,
+      "learning_rate": 5.3022838167055525e-05,
+      "loss": 1.2755,
+      "step": 14457
+    },
+    {
+      "epoch": 0.6660524254848667,
+      "grad_norm": 0.9667589605387343,
+      "learning_rate": 5.300966671070409e-05,
+      "loss": 1.2153,
+      "step": 14458
+    },
+    {
+      "epoch": 0.6660984935735016,
+      "grad_norm": 0.9286619386566125,
+      "learning_rate": 5.299649630051184e-05,
+      "loss": 1.1594,
+      "step": 14459
+    },
+    {
+      "epoch": 0.6661445616621366,
+      "grad_norm": 1.3902338474923275,
+      "learning_rate": 5.298332693677206e-05,
+      "loss": 1.5082,
+      "step": 14460
+    },
+    {
+      "epoch": 0.6661906297507716,
+      "grad_norm": 0.7405935356673496,
+      "learning_rate": 5.297015861977783e-05,
+      "loss": 1.1711,
+      "step": 14461
+    },
+    {
+      "epoch": 0.6662366978394066,
+      "grad_norm": 2.181363060070056,
+      "learning_rate": 5.2956991349822395e-05,
+      "loss": 1.4292,
+      "step": 14462
+    },
+    {
+      "epoch": 0.6662827659280417,
+      "grad_norm": 0.851038321724406,
+      "learning_rate": 5.2943825127198885e-05,
+      "loss": 1.1458,
+      "step": 14463
+    },
+    {
+      "epoch": 0.6663288340166766,
+      "grad_norm": 0.8451494303564595,
+      "learning_rate": 5.2930659952200455e-05,
+      "loss": 1.3109,
+      "step": 14464
+    },
+    {
+      "epoch": 0.6663749021053117,
+      "grad_norm": 0.7526229377185952,
+      "learning_rate": 5.291749582512015e-05,
+      "loss": 1.2245,
+      "step": 14465
+    },
+    {
+      "epoch": 0.6664209701939466,
+      "grad_norm": 0.8388174336296972,
+      "learning_rate": 5.2904332746251065e-05,
+      "loss": 1.3721,
+      "step": 14466
+    },
+    {
+      "epoch": 0.6664670382825817,
+      "grad_norm": 0.8175339001023012,
+      "learning_rate": 5.289117071588625e-05,
+      "loss": 1.4072,
+      "step": 14467
+    },
+    {
+      "epoch": 0.6665131063712166,
+      "grad_norm": 1.0604442882958667,
+      "learning_rate": 5.2878009734318804e-05,
+      "loss": 1.1187,
+      "step": 14468
+    },
+    {
+      "epoch": 0.6665591744598517,
+      "grad_norm": 0.7381951605241126,
+      "learning_rate": 5.286484980184162e-05,
+      "loss": 1.5791,
+      "step": 14469
+    },
+    {
+      "epoch": 0.6666052425484866,
+      "grad_norm": 0.8269416146463138,
+      "learning_rate": 5.285169091874773e-05,
+      "loss": 1.0119,
+      "step": 14470
+    },
+    {
+      "epoch": 0.6666513106371217,
+      "grad_norm": 0.8228487153920166,
+      "learning_rate": 5.2838533085330156e-05,
+      "loss": 1.3604,
+      "step": 14471
+    },
+    {
+      "epoch": 0.6666973787257566,
+      "grad_norm": 0.7599612193141987,
+      "learning_rate": 5.2825376301881734e-05,
+      "loss": 1.1578,
+      "step": 14472
+    },
+    {
+      "epoch": 0.6667434468143917,
+      "grad_norm": 0.7990947618176418,
+      "learning_rate": 5.281222056869542e-05,
+      "loss": 1.3397,
+      "step": 14473
+    },
+    {
+      "epoch": 0.6667895149030266,
+      "grad_norm": 0.8271423876489409,
+      "learning_rate": 5.27990658860641e-05,
+      "loss": 1.2336,
+      "step": 14474
+    },
+    {
+      "epoch": 0.6668355829916617,
+      "grad_norm": 1.0149716302441247,
+      "learning_rate": 5.278591225428069e-05,
+      "loss": 1.4223,
+      "step": 14475
+    },
+    {
+      "epoch": 0.6668816510802967,
+      "grad_norm": 1.003605494416176,
+      "learning_rate": 5.277275967363795e-05,
+      "loss": 1.3101,
+      "step": 14476
+    },
+    {
+      "epoch": 0.6669277191689317,
+      "grad_norm": 0.8082220026501017,
+      "learning_rate": 5.275960814442873e-05,
+      "loss": 1.4201,
+      "step": 14477
+    },
+    {
+      "epoch": 0.6669737872575667,
+      "grad_norm": 1.1745428408518745,
+      "learning_rate": 5.2746457666945845e-05,
+      "loss": 1.2251,
+      "step": 14478
+    },
+    {
+      "epoch": 0.6670198553462017,
+      "grad_norm": 0.8689704815587439,
+      "learning_rate": 5.273330824148211e-05,
+      "loss": 1.4314,
+      "step": 14479
+    },
+    {
+      "epoch": 0.6670659234348367,
+      "grad_norm": 0.7000581819449598,
+      "learning_rate": 5.272015986833015e-05,
+      "loss": 1.2866,
+      "step": 14480
+    },
+    {
+      "epoch": 0.6671119915234717,
+      "grad_norm": 0.9133657437641617,
+      "learning_rate": 5.270701254778279e-05,
+      "loss": 1.352,
+      "step": 14481
+    },
+    {
+      "epoch": 0.6671580596121067,
+      "grad_norm": 0.8822128262221131,
+      "learning_rate": 5.269386628013273e-05,
+      "loss": 1.1928,
+      "step": 14482
+    },
+    {
+      "epoch": 0.6672041277007417,
+      "grad_norm": 1.4870616880863579,
+      "learning_rate": 5.268072106567259e-05,
+      "loss": 1.2309,
+      "step": 14483
+    },
+    {
+      "epoch": 0.6672501957893767,
+      "grad_norm": 2.012414434654418,
+      "learning_rate": 5.266757690469506e-05,
+      "loss": 1.0247,
+      "step": 14484
+    },
+    {
+      "epoch": 0.6672962638780117,
+      "grad_norm": 0.7009862963324982,
+      "learning_rate": 5.2654433797492774e-05,
+      "loss": 1.6489,
+      "step": 14485
+    },
+    {
+      "epoch": 0.6673423319666467,
+      "grad_norm": 1.2350892558488142,
+      "learning_rate": 5.264129174435837e-05,
+      "loss": 1.0625,
+      "step": 14486
+    },
+    {
+      "epoch": 0.6673884000552817,
+      "grad_norm": 0.93693385652825,
+      "learning_rate": 5.262815074558438e-05,
+      "loss": 1.2962,
+      "step": 14487
+    },
+    {
+      "epoch": 0.6674344681439167,
+      "grad_norm": 0.9100371427024263,
+      "learning_rate": 5.2615010801463385e-05,
+      "loss": 1.2754,
+      "step": 14488
+    },
+    {
+      "epoch": 0.6674805362325517,
+      "grad_norm": 0.7365715252686935,
+      "learning_rate": 5.260187191228795e-05,
+      "loss": 1.0906,
+      "step": 14489
+    },
+    {
+      "epoch": 0.6675266043211867,
+      "grad_norm": 1.264130338219427,
+      "learning_rate": 5.258873407835054e-05,
+      "loss": 0.9454,
+      "step": 14490
+    },
+    {
+      "epoch": 0.6675726724098218,
+      "grad_norm": 1.587383448521216,
+      "learning_rate": 5.257559729994367e-05,
+      "loss": 1.212,
+      "step": 14491
+    },
+    {
+      "epoch": 0.6676187404984567,
+      "grad_norm": 0.7597726664428712,
+      "learning_rate": 5.25624615773598e-05,
+      "loss": 1.1747,
+      "step": 14492
+    },
+    {
+      "epoch": 0.6676648085870918,
+      "grad_norm": 1.2680343837650554,
+      "learning_rate": 5.254932691089144e-05,
+      "loss": 1.3164,
+      "step": 14493
+    },
+    {
+      "epoch": 0.6677108766757267,
+      "grad_norm": 0.6519102872727627,
+      "learning_rate": 5.253619330083091e-05,
+      "loss": 1.4013,
+      "step": 14494
+    },
+    {
+      "epoch": 0.6677569447643618,
+      "grad_norm": 1.3612517164685725,
+      "learning_rate": 5.2523060747470664e-05,
+      "loss": 1.2827,
+      "step": 14495
+    },
+    {
+      "epoch": 0.6678030128529967,
+      "grad_norm": 0.801096544486594,
+      "learning_rate": 5.250992925110305e-05,
+      "loss": 1.3481,
+      "step": 14496
+    },
+    {
+      "epoch": 0.6678490809416318,
+      "grad_norm": 0.9877619555381594,
+      "learning_rate": 5.249679881202048e-05,
+      "loss": 1.2346,
+      "step": 14497
+    },
+    {
+      "epoch": 0.6678951490302667,
+      "grad_norm": 0.8409989428756427,
+      "learning_rate": 5.248366943051524e-05,
+      "loss": 1.2753,
+      "step": 14498
+    },
+    {
+      "epoch": 0.6679412171189018,
+      "grad_norm": 0.9887109511094616,
+      "learning_rate": 5.247054110687952e-05,
+      "loss": 1.2717,
+      "step": 14499
+    },
+    {
+      "epoch": 0.6679872852075367,
+      "grad_norm": 0.7797407688099892,
+      "learning_rate": 5.245741384140581e-05,
+      "loss": 1.3494,
+      "step": 14500
+    },
+    {
+      "epoch": 0.6680333532961718,
+      "grad_norm": 0.9401037475596101,
+      "learning_rate": 5.244428763438623e-05,
+      "loss": 1.0452,
+      "step": 14501
+    },
+    {
+      "epoch": 0.6680794213848067,
+      "grad_norm": 0.8846725288838383,
+      "learning_rate": 5.2431162486113025e-05,
+      "loss": 1.2401,
+      "step": 14502
+    },
+    {
+      "epoch": 0.6681254894734417,
+      "grad_norm": 0.7786615224156954,
+      "learning_rate": 5.241803839687844e-05,
+      "loss": 1.1266,
+      "step": 14503
+    },
+    {
+      "epoch": 0.6681715575620768,
+      "grad_norm": 1.321936537773121,
+      "learning_rate": 5.240491536697467e-05,
+      "loss": 1.2104,
+      "step": 14504
+    },
+    {
+      "epoch": 0.6682176256507117,
+      "grad_norm": 0.8641276623294437,
+      "learning_rate": 5.239179339669382e-05,
+      "loss": 1.2578,
+      "step": 14505
+    },
+    {
+      "epoch": 0.6682636937393468,
+      "grad_norm": 1.4465164638297432,
+      "learning_rate": 5.2378672486328063e-05,
+      "loss": 1.235,
+      "step": 14506
+    },
+    {
+      "epoch": 0.6683097618279817,
+      "grad_norm": 2.9438892003247847,
+      "learning_rate": 5.2365552636169544e-05,
+      "loss": 0.8512,
+      "step": 14507
+    },
+    {
+      "epoch": 0.6683558299166168,
+      "grad_norm": 0.9412486820459084,
+      "learning_rate": 5.235243384651029e-05,
+      "loss": 1.3241,
+      "step": 14508
+    },
+    {
+      "epoch": 0.6684018980052517,
+      "grad_norm": 28.25995555368446,
+      "learning_rate": 5.233931611764242e-05,
+      "loss": 1.19,
+      "step": 14509
+    },
+    {
+      "epoch": 0.6684479660938868,
+      "grad_norm": 0.8370551037476637,
+      "learning_rate": 5.2326199449857895e-05,
+      "loss": 0.9857,
+      "step": 14510
+    },
+    {
+      "epoch": 0.6684940341825217,
+      "grad_norm": 0.9003903250129565,
+      "learning_rate": 5.2313083843448865e-05,
+      "loss": 1.2035,
+      "step": 14511
+    },
+    {
+      "epoch": 0.6685401022711568,
+      "grad_norm": 0.933183503773496,
+      "learning_rate": 5.229996929870722e-05,
+      "loss": 1.41,
+      "step": 14512
+    },
+    {
+      "epoch": 0.6685861703597917,
+      "grad_norm": 0.7983688082139131,
+      "learning_rate": 5.2286855815925004e-05,
+      "loss": 1.147,
+      "step": 14513
+    },
+    {
+      "epoch": 0.6686322384484268,
+      "grad_norm": 0.6944459105881143,
+      "learning_rate": 5.2273743395394056e-05,
+      "loss": 1.6165,
+      "step": 14514
+    },
+    {
+      "epoch": 0.6686783065370617,
+      "grad_norm": 0.743999932137126,
+      "learning_rate": 5.226063203740648e-05,
+      "loss": 1.5397,
+      "step": 14515
+    },
+    {
+      "epoch": 0.6687243746256968,
+      "grad_norm": 1.031969549196954,
+      "learning_rate": 5.2247521742254045e-05,
+      "loss": 1.1226,
+      "step": 14516
+    },
+    {
+      "epoch": 0.6687704427143318,
+      "grad_norm": 1.0310447479035338,
+      "learning_rate": 5.22344125102286e-05,
+      "loss": 1.1224,
+      "step": 14517
+    },
+    {
+      "epoch": 0.6688165108029668,
+      "grad_norm": 1.0788284058199524,
+      "learning_rate": 5.2221304341622135e-05,
+      "loss": 1.3386,
+      "step": 14518
+    },
+    {
+      "epoch": 0.6688625788916018,
+      "grad_norm": 0.8152568539513299,
+      "learning_rate": 5.220819723672637e-05,
+      "loss": 1.0669,
+      "step": 14519
+    },
+    {
+      "epoch": 0.6689086469802368,
+      "grad_norm": 1.0201224843925032,
+      "learning_rate": 5.219509119583317e-05,
+      "loss": 1.2588,
+      "step": 14520
+    },
+    {
+      "epoch": 0.6689547150688718,
+      "grad_norm": 0.9860014485287428,
+      "learning_rate": 5.218198621923422e-05,
+      "loss": 1.4903,
+      "step": 14521
+    },
+    {
+      "epoch": 0.6690007831575068,
+      "grad_norm": 1.9298049020917074,
+      "learning_rate": 5.216888230722146e-05,
+      "loss": 1.5037,
+      "step": 14522
+    },
+    {
+      "epoch": 0.6690468512461418,
+      "grad_norm": 1.3956009818354118,
+      "learning_rate": 5.215577946008647e-05,
+      "loss": 1.2356,
+      "step": 14523
+    },
+    {
+      "epoch": 0.6690929193347768,
+      "grad_norm": 1.2330013466729732,
+      "learning_rate": 5.214267767812104e-05,
+      "loss": 1.3064,
+      "step": 14524
+    },
+    {
+      "epoch": 0.6691389874234118,
+      "grad_norm": 1.9364226618817555,
+      "learning_rate": 5.2129576961616824e-05,
+      "loss": 1.258,
+      "step": 14525
+    },
+    {
+      "epoch": 0.6691850555120468,
+      "grad_norm": 4.032032388073026,
+      "learning_rate": 5.2116477310865474e-05,
+      "loss": 1.0746,
+      "step": 14526
+    },
+    {
+      "epoch": 0.6692311236006818,
+      "grad_norm": 1.3479873213857074,
+      "learning_rate": 5.210337872615869e-05,
+      "loss": 1.0615,
+      "step": 14527
+    },
+    {
+      "epoch": 0.6692771916893168,
+      "grad_norm": 1.7207366555590105,
+      "learning_rate": 5.2090281207787997e-05,
+      "loss": 1.2251,
+      "step": 14528
+    },
+    {
+      "epoch": 0.6693232597779518,
+      "grad_norm": 1.3247913505511673,
+      "learning_rate": 5.207718475604511e-05,
+      "loss": 1.2462,
+      "step": 14529
+    },
+    {
+      "epoch": 0.6693693278665868,
+      "grad_norm": 1.1114494859872708,
+      "learning_rate": 5.206408937122149e-05,
+      "loss": 1.5473,
+      "step": 14530
+    },
+    {
+      "epoch": 0.6694153959552218,
+      "grad_norm": 1.1155417554860707,
+      "learning_rate": 5.2050995053608776e-05,
+      "loss": 1.4365,
+      "step": 14531
+    },
+    {
+      "epoch": 0.6694614640438569,
+      "grad_norm": 1.5019261385538563,
+      "learning_rate": 5.203790180349837e-05,
+      "loss": 1.3356,
+      "step": 14532
+    },
+    {
+      "epoch": 0.6695075321324918,
+      "grad_norm": 1.1071518349350715,
+      "learning_rate": 5.20248096211819e-05,
+      "loss": 1.2328,
+      "step": 14533
+    },
+    {
+      "epoch": 0.6695536002211269,
+      "grad_norm": 1.1834296540329716,
+      "learning_rate": 5.20117185069508e-05,
+      "loss": 1.2388,
+      "step": 14534
+    },
+    {
+      "epoch": 0.6695996683097618,
+      "grad_norm": 0.7355044786200979,
+      "learning_rate": 5.199862846109646e-05,
+      "loss": 1.7022,
+      "step": 14535
+    },
+    {
+      "epoch": 0.6696457363983969,
+      "grad_norm": 1.7180763513997492,
+      "learning_rate": 5.198553948391036e-05,
+      "loss": 1.3756,
+      "step": 14536
+    },
+    {
+      "epoch": 0.6696918044870318,
+      "grad_norm": 1.2671885411096027,
+      "learning_rate": 5.1972451575683886e-05,
+      "loss": 1.3404,
+      "step": 14537
+    },
+    {
+      "epoch": 0.6697378725756669,
+      "grad_norm": 1.1123386294156485,
+      "learning_rate": 5.195936473670848e-05,
+      "loss": 0.9597,
+      "step": 14538
+    },
+    {
+      "epoch": 0.6697839406643018,
+      "grad_norm": 0.9259173850208627,
+      "learning_rate": 5.1946278967275394e-05,
+      "loss": 1.2232,
+      "step": 14539
+    },
+    {
+      "epoch": 0.6698300087529369,
+      "grad_norm": 1.3782397431760152,
+      "learning_rate": 5.193319426767602e-05,
+      "loss": 1.5268,
+      "step": 14540
+    },
+    {
+      "epoch": 0.6698760768415718,
+      "grad_norm": 0.9551682284530577,
+      "learning_rate": 5.1920110638201655e-05,
+      "loss": 1.5943,
+      "step": 14541
+    },
+    {
+      "epoch": 0.6699221449302069,
+      "grad_norm": 1.557635612873799,
+      "learning_rate": 5.190702807914363e-05,
+      "loss": 0.9638,
+      "step": 14542
+    },
+    {
+      "epoch": 0.6699682130188418,
+      "grad_norm": 0.7662159037037483,
+      "learning_rate": 5.189394659079312e-05,
+      "loss": 1.5169,
+      "step": 14543
+    },
+    {
+      "epoch": 0.6700142811074769,
+      "grad_norm": 2.031509173032336,
+      "learning_rate": 5.1880866173441414e-05,
+      "loss": 1.3335,
+      "step": 14544
+    },
+    {
+      "epoch": 0.6700603491961119,
+      "grad_norm": 0.8817664384508462,
+      "learning_rate": 5.1867786827379736e-05,
+      "loss": 1.0765,
+      "step": 14545
+    },
+    {
+      "epoch": 0.6701064172847468,
+      "grad_norm": 1.049677389412653,
+      "learning_rate": 5.1854708552899244e-05,
+      "loss": 1.0624,
+      "step": 14546
+    },
+    {
+      "epoch": 0.6701524853733819,
+      "grad_norm": 1.4073898544069128,
+      "learning_rate": 5.18416313502911e-05,
+      "loss": 1.1744,
+      "step": 14547
+    },
+    {
+      "epoch": 0.6701985534620168,
+      "grad_norm": 1.2024467726084707,
+      "learning_rate": 5.182855521984645e-05,
+      "loss": 1.5326,
+      "step": 14548
+    },
+    {
+      "epoch": 0.6702446215506519,
+      "grad_norm": 2.1412298571151136,
+      "learning_rate": 5.181548016185647e-05,
+      "loss": 1.0288,
+      "step": 14549
+    },
+    {
+      "epoch": 0.6702906896392868,
+      "grad_norm": 1.1097388036308478,
+      "learning_rate": 5.1802406176612164e-05,
+      "loss": 1.4993,
+      "step": 14550
+    },
+    {
+      "epoch": 0.6703367577279219,
+      "grad_norm": 1.102623281707363,
+      "learning_rate": 5.178933326440465e-05,
+      "loss": 1.277,
+      "step": 14551
+    },
+    {
+      "epoch": 0.6703828258165568,
+      "grad_norm": 0.8810732980047197,
+      "learning_rate": 5.1776261425525007e-05,
+      "loss": 1.2011,
+      "step": 14552
+    },
+    {
+      "epoch": 0.6704288939051919,
+      "grad_norm": 1.16659426389969,
+      "learning_rate": 5.176319066026417e-05,
+      "loss": 1.1552,
+      "step": 14553
+    },
+    {
+      "epoch": 0.6704749619938268,
+      "grad_norm": 1.1477551153045737,
+      "learning_rate": 5.175012096891319e-05,
+      "loss": 1.1468,
+      "step": 14554
+    },
+    {
+      "epoch": 0.6705210300824619,
+      "grad_norm": 0.8811579142949687,
+      "learning_rate": 5.1737052351763024e-05,
+      "loss": 1.2746,
+      "step": 14555
+    },
+    {
+      "epoch": 0.6705670981710968,
+      "grad_norm": 1.0962659464304165,
+      "learning_rate": 5.1723984809104687e-05,
+      "loss": 1.1287,
+      "step": 14556
+    },
+    {
+      "epoch": 0.6706131662597319,
+      "grad_norm": 0.9461471206283428,
+      "learning_rate": 5.171091834122901e-05,
+      "loss": 1.2284,
+      "step": 14557
+    },
+    {
+      "epoch": 0.6706592343483669,
+      "grad_norm": 2.7513287334889607,
+      "learning_rate": 5.169785294842694e-05,
+      "loss": 1.1737,
+      "step": 14558
+    },
+    {
+      "epoch": 0.6707053024370019,
+      "grad_norm": 1.4807721102869325,
+      "learning_rate": 5.168478863098936e-05,
+      "loss": 1.4564,
+      "step": 14559
+    },
+    {
+      "epoch": 0.6707513705256369,
+      "grad_norm": 1.2494850509707411,
+      "learning_rate": 5.167172538920714e-05,
+      "loss": 1.2994,
+      "step": 14560
+    },
+    {
+      "epoch": 0.6707974386142719,
+      "grad_norm": 0.8980822137560616,
+      "learning_rate": 5.165866322337105e-05,
+      "loss": 1.2134,
+      "step": 14561
+    },
+    {
+      "epoch": 0.6708435067029069,
+      "grad_norm": 1.309864643255989,
+      "learning_rate": 5.164560213377195e-05,
+      "loss": 1.1737,
+      "step": 14562
+    },
+    {
+      "epoch": 0.6708895747915419,
+      "grad_norm": 0.821951974642687,
+      "learning_rate": 5.163254212070064e-05,
+      "loss": 1.2406,
+      "step": 14563
+    },
+    {
+      "epoch": 0.6709356428801769,
+      "grad_norm": 0.8918075063790977,
+      "learning_rate": 5.161948318444782e-05,
+      "loss": 1.2435,
+      "step": 14564
+    },
+    {
+      "epoch": 0.6709817109688119,
+      "grad_norm": 0.9168655619772499,
+      "learning_rate": 5.160642532530424e-05,
+      "loss": 1.3508,
+      "step": 14565
+    },
+    {
+      "epoch": 0.6710277790574469,
+      "grad_norm": 1.2804237242257126,
+      "learning_rate": 5.1593368543560624e-05,
+      "loss": 1.2752,
+      "step": 14566
+    },
+    {
+      "epoch": 0.6710738471460819,
+      "grad_norm": 0.88399146537528,
+      "learning_rate": 5.1580312839507694e-05,
+      "loss": 1.187,
+      "step": 14567
+    },
+    {
+      "epoch": 0.6711199152347169,
+      "grad_norm": 2.410027072232538,
+      "learning_rate": 5.156725821343604e-05,
+      "loss": 1.4513,
+      "step": 14568
+    },
+    {
+      "epoch": 0.6711659833233519,
+      "grad_norm": 0.9738803829196016,
+      "learning_rate": 5.155420466563634e-05,
+      "loss": 1.2278,
+      "step": 14569
+    },
+    {
+      "epoch": 0.6712120514119869,
+      "grad_norm": 0.9424523569734672,
+      "learning_rate": 5.154115219639924e-05,
+      "loss": 1.1788,
+      "step": 14570
+    },
+    {
+      "epoch": 0.6712581195006219,
+      "grad_norm": 0.7910199417836364,
+      "learning_rate": 5.152810080601527e-05,
+      "loss": 1.4649,
+      "step": 14571
+    },
+    {
+      "epoch": 0.6713041875892569,
+      "grad_norm": 0.7435371339988375,
+      "learning_rate": 5.1515050494775005e-05,
+      "loss": 1.1441,
+      "step": 14572
+    },
+    {
+      "epoch": 0.671350255677892,
+      "grad_norm": 0.8152502138954659,
+      "learning_rate": 5.150200126296902e-05,
+      "loss": 1.4237,
+      "step": 14573
+    },
+    {
+      "epoch": 0.6713963237665269,
+      "grad_norm": 0.8735502270245369,
+      "learning_rate": 5.148895311088786e-05,
+      "loss": 1.3599,
+      "step": 14574
+    },
+    {
+      "epoch": 0.671442391855162,
+      "grad_norm": 0.8221467489814124,
+      "learning_rate": 5.1475906038821955e-05,
+      "loss": 1.4619,
+      "step": 14575
+    },
+    {
+      "epoch": 0.6714884599437969,
+      "grad_norm": 1.4823407686958554,
+      "learning_rate": 5.1462860047061776e-05,
+      "loss": 1.4338,
+      "step": 14576
+    },
+    {
+      "epoch": 0.671534528032432,
+      "grad_norm": 1.7978422839440884,
+      "learning_rate": 5.144981513589781e-05,
+      "loss": 1.2911,
+      "step": 14577
+    },
+    {
+      "epoch": 0.6715805961210669,
+      "grad_norm": 0.7174409516880105,
+      "learning_rate": 5.14367713056205e-05,
+      "loss": 1.3432,
+      "step": 14578
+    },
+    {
+      "epoch": 0.671626664209702,
+      "grad_norm": 1.3238419272047088,
+      "learning_rate": 5.142372855652016e-05,
+      "loss": 1.2629,
+      "step": 14579
+    },
+    {
+      "epoch": 0.6716727322983369,
+      "grad_norm": 1.0942956843881133,
+      "learning_rate": 5.141068688888722e-05,
+      "loss": 1.1351,
+      "step": 14580
+    },
+    {
+      "epoch": 0.671718800386972,
+      "grad_norm": 0.9723717235658497,
+      "learning_rate": 5.139764630301207e-05,
+      "loss": 1.3862,
+      "step": 14581
+    },
+    {
+      "epoch": 0.6717648684756069,
+      "grad_norm": 0.970954338480689,
+      "learning_rate": 5.138460679918494e-05,
+      "loss": 1.3476,
+      "step": 14582
+    },
+    {
+      "epoch": 0.671810936564242,
+      "grad_norm": 0.7191223397480122,
+      "learning_rate": 5.1371568377696186e-05,
+      "loss": 1.4572,
+      "step": 14583
+    },
+    {
+      "epoch": 0.6718570046528769,
+      "grad_norm": 1.2625519175477722,
+      "learning_rate": 5.135853103883608e-05,
+      "loss": 1.3098,
+      "step": 14584
+    },
+    {
+      "epoch": 0.671903072741512,
+      "grad_norm": 0.93594896046385,
+      "learning_rate": 5.134549478289492e-05,
+      "loss": 1.5345,
+      "step": 14585
+    },
+    {
+      "epoch": 0.671949140830147,
+      "grad_norm": 0.9214420099675007,
+      "learning_rate": 5.1332459610162865e-05,
+      "loss": 1.233,
+      "step": 14586
+    },
+    {
+      "epoch": 0.671995208918782,
+      "grad_norm": 1.1571895303693445,
+      "learning_rate": 5.131942552093013e-05,
+      "loss": 1.1882,
+      "step": 14587
+    },
+    {
+      "epoch": 0.672041277007417,
+      "grad_norm": 1.4229675050813546,
+      "learning_rate": 5.1306392515486923e-05,
+      "loss": 0.9668,
+      "step": 14588
+    },
+    {
+      "epoch": 0.672087345096052,
+      "grad_norm": 0.8240953945468007,
+      "learning_rate": 5.129336059412344e-05,
+      "loss": 1.2205,
+      "step": 14589
+    },
+    {
+      "epoch": 0.672133413184687,
+      "grad_norm": 0.922172426781291,
+      "learning_rate": 5.128032975712973e-05,
+      "loss": 1.3509,
+      "step": 14590
+    },
+    {
+      "epoch": 0.6721794812733219,
+      "grad_norm": 2.144506009286274,
+      "learning_rate": 5.126730000479594e-05,
+      "loss": 1.3364,
+      "step": 14591
+    },
+    {
+      "epoch": 0.672225549361957,
+      "grad_norm": 0.9833118591557344,
+      "learning_rate": 5.1254271337412195e-05,
+      "loss": 1.2735,
+      "step": 14592
+    },
+    {
+      "epoch": 0.6722716174505919,
+      "grad_norm": 1.799618646185124,
+      "learning_rate": 5.124124375526848e-05,
+      "loss": 1.2,
+      "step": 14593
+    },
+    {
+      "epoch": 0.672317685539227,
+      "grad_norm": 0.7605883647474673,
+      "learning_rate": 5.1228217258654865e-05,
+      "loss": 1.4319,
+      "step": 14594
+    },
+    {
+      "epoch": 0.6723637536278619,
+      "grad_norm": 1.0097959502994815,
+      "learning_rate": 5.1215191847861365e-05,
+      "loss": 1.2003,
+      "step": 14595
+    },
+    {
+      "epoch": 0.672409821716497,
+      "grad_norm": 0.7491881957663729,
+      "learning_rate": 5.120216752317801e-05,
+      "loss": 1.2799,
+      "step": 14596
+    },
+    {
+      "epoch": 0.6724558898051319,
+      "grad_norm": 8.703835874297353,
+      "learning_rate": 5.1189144284894695e-05,
+      "loss": 1.3096,
+      "step": 14597
+    },
+    {
+      "epoch": 0.672501957893767,
+      "grad_norm": 0.8880942646667533,
+      "learning_rate": 5.117612213330138e-05,
+      "loss": 1.349,
+      "step": 14598
+    },
+    {
+      "epoch": 0.672548025982402,
+      "grad_norm": 0.7705414615133218,
+      "learning_rate": 5.1163101068688044e-05,
+      "loss": 1.3055,
+      "step": 14599
+    },
+    {
+      "epoch": 0.672594094071037,
+      "grad_norm": 1.0551196364925486,
+      "learning_rate": 5.115008109134447e-05,
+      "loss": 1.3329,
+      "step": 14600
+    },
+    {
+      "epoch": 0.672640162159672,
+      "grad_norm": 0.6653046201492457,
+      "learning_rate": 5.11370622015606e-05,
+      "loss": 1.4033,
+      "step": 14601
+    },
+    {
+      "epoch": 0.672686230248307,
+      "grad_norm": 1.7076291288822714,
+      "learning_rate": 5.1124044399626235e-05,
+      "loss": 1.2824,
+      "step": 14602
+    },
+    {
+      "epoch": 0.672732298336942,
+      "grad_norm": 1.7759752915754128,
+      "learning_rate": 5.111102768583127e-05,
+      "loss": 0.8853,
+      "step": 14603
+    },
+    {
+      "epoch": 0.672778366425577,
+      "grad_norm": 0.7145563281527224,
+      "learning_rate": 5.1098012060465403e-05,
+      "loss": 1.4291,
+      "step": 14604
+    },
+    {
+      "epoch": 0.672824434514212,
+      "grad_norm": 1.0043402645587802,
+      "learning_rate": 5.108499752381846e-05,
+      "loss": 1.3431,
+      "step": 14605
+    },
+    {
+      "epoch": 0.672870502602847,
+      "grad_norm": 0.7071548429701353,
+      "learning_rate": 5.107198407618017e-05,
+      "loss": 1.3123,
+      "step": 14606
+    },
+    {
+      "epoch": 0.672916570691482,
+      "grad_norm": 1.075211605589372,
+      "learning_rate": 5.1058971717840285e-05,
+      "loss": 1.1917,
+      "step": 14607
+    },
+    {
+      "epoch": 0.672962638780117,
+      "grad_norm": 1.1084062303491373,
+      "learning_rate": 5.104596044908849e-05,
+      "loss": 0.9688,
+      "step": 14608
+    },
+    {
+      "epoch": 0.673008706868752,
+      "grad_norm": 0.9352697749205287,
+      "learning_rate": 5.103295027021437e-05,
+      "loss": 1.3777,
+      "step": 14609
+    },
+    {
+      "epoch": 0.673054774957387,
+      "grad_norm": 2.3019395483606724,
+      "learning_rate": 5.101994118150774e-05,
+      "loss": 1.2396,
+      "step": 14610
+    },
+    {
+      "epoch": 0.673100843046022,
+      "grad_norm": 0.8200105631393557,
+      "learning_rate": 5.100693318325809e-05,
+      "loss": 1.4363,
+      "step": 14611
+    },
+    {
+      "epoch": 0.6731469111346571,
+      "grad_norm": 0.8301028606561324,
+      "learning_rate": 5.09939262757551e-05,
+      "loss": 1.213,
+      "step": 14612
+    },
+    {
+      "epoch": 0.673192979223292,
+      "grad_norm": 1.1319720437392204,
+      "learning_rate": 5.098092045928824e-05,
+      "loss": 1.3739,
+      "step": 14613
+    },
+    {
+      "epoch": 0.6732390473119271,
+      "grad_norm": 0.8843775275361965,
+      "learning_rate": 5.096791573414721e-05,
+      "loss": 1.2178,
+      "step": 14614
+    },
+    {
+      "epoch": 0.673285115400562,
+      "grad_norm": 0.9078851821168039,
+      "learning_rate": 5.095491210062143e-05,
+      "loss": 1.3134,
+      "step": 14615
+    },
+    {
+      "epoch": 0.6733311834891971,
+      "grad_norm": 0.9615880982285595,
+      "learning_rate": 5.094190955900044e-05,
+      "loss": 1.5837,
+      "step": 14616
+    },
+    {
+      "epoch": 0.673377251577832,
+      "grad_norm": 0.8794528945941044,
+      "learning_rate": 5.0928908109573756e-05,
+      "loss": 1.4075,
+      "step": 14617
+    },
+    {
+      "epoch": 0.6734233196664671,
+      "grad_norm": 0.6931766765195356,
+      "learning_rate": 5.091590775263076e-05,
+      "loss": 1.2126,
+      "step": 14618
+    },
+    {
+      "epoch": 0.673469387755102,
+      "grad_norm": 0.9586959375130474,
+      "learning_rate": 5.0902908488460935e-05,
+      "loss": 1.2635,
+      "step": 14619
+    },
+    {
+      "epoch": 0.6735154558437371,
+      "grad_norm": 0.8711199905096675,
+      "learning_rate": 5.088991031735361e-05,
+      "loss": 1.2036,
+      "step": 14620
+    },
+    {
+      "epoch": 0.673561523932372,
+      "grad_norm": 4.257805240284082,
+      "learning_rate": 5.087691323959829e-05,
+      "loss": 1.0011,
+      "step": 14621
+    },
+    {
+      "epoch": 0.6736075920210071,
+      "grad_norm": 1.5071360203326418,
+      "learning_rate": 5.086391725548425e-05,
+      "loss": 1.6115,
+      "step": 14622
+    },
+    {
+      "epoch": 0.673653660109642,
+      "grad_norm": 0.603146248130678,
+      "learning_rate": 5.0850922365300866e-05,
+      "loss": 1.4812,
+      "step": 14623
+    },
+    {
+      "epoch": 0.6736997281982771,
+      "grad_norm": 1.0090691520604003,
+      "learning_rate": 5.083792856933734e-05,
+      "loss": 1.0995,
+      "step": 14624
+    },
+    {
+      "epoch": 0.673745796286912,
+      "grad_norm": 0.7647663327988596,
+      "learning_rate": 5.082493586788313e-05,
+      "loss": 1.3361,
+      "step": 14625
+    },
+    {
+      "epoch": 0.6737918643755471,
+      "grad_norm": 0.9637415501917592,
+      "learning_rate": 5.0811944261227394e-05,
+      "loss": 1.2861,
+      "step": 14626
+    },
+    {
+      "epoch": 0.6738379324641821,
+      "grad_norm": 1.3115465379726174,
+      "learning_rate": 5.079895374965935e-05,
+      "loss": 1.1558,
+      "step": 14627
+    },
+    {
+      "epoch": 0.673884000552817,
+      "grad_norm": 0.9294099992134268,
+      "learning_rate": 5.078596433346825e-05,
+      "loss": 1.2564,
+      "step": 14628
+    },
+    {
+      "epoch": 0.6739300686414521,
+      "grad_norm": 0.9494763296512706,
+      "learning_rate": 5.077297601294325e-05,
+      "loss": 1.2752,
+      "step": 14629
+    },
+    {
+      "epoch": 0.673976136730087,
+      "grad_norm": 0.9526963511355787,
+      "learning_rate": 5.075998878837358e-05,
+      "loss": 1.1704,
+      "step": 14630
+    },
+    {
+      "epoch": 0.6740222048187221,
+      "grad_norm": 1.2247067961913536,
+      "learning_rate": 5.074700266004826e-05,
+      "loss": 1.1482,
+      "step": 14631
+    },
+    {
+      "epoch": 0.674068272907357,
+      "grad_norm": 0.8228976609143064,
+      "learning_rate": 5.073401762825653e-05,
+      "loss": 1.4044,
+      "step": 14632
+    },
+    {
+      "epoch": 0.6741143409959921,
+      "grad_norm": 1.3618537367629393,
+      "learning_rate": 5.07210336932874e-05,
+      "loss": 0.986,
+      "step": 14633
+    },
+    {
+      "epoch": 0.674160409084627,
+      "grad_norm": 2.833485647297411,
+      "learning_rate": 5.070805085543e-05,
+      "loss": 1.2696,
+      "step": 14634
+    },
+    {
+      "epoch": 0.6742064771732621,
+      "grad_norm": 0.866806320628466,
+      "learning_rate": 5.06950691149733e-05,
+      "loss": 1.1532,
+      "step": 14635
+    },
+    {
+      "epoch": 0.674252545261897,
+      "grad_norm": 0.8017958190846609,
+      "learning_rate": 5.068208847220632e-05,
+      "loss": 1.3061,
+      "step": 14636
+    },
+    {
+      "epoch": 0.6742986133505321,
+      "grad_norm": 0.6880232407175445,
+      "learning_rate": 5.0669108927418144e-05,
+      "loss": 1.3591,
+      "step": 14637
+    },
+    {
+      "epoch": 0.674344681439167,
+      "grad_norm": 0.8981549581712941,
+      "learning_rate": 5.065613048089762e-05,
+      "loss": 1.2236,
+      "step": 14638
+    },
+    {
+      "epoch": 0.6743907495278021,
+      "grad_norm": 1.1316087093188756,
+      "learning_rate": 5.064315313293375e-05,
+      "loss": 0.9259,
+      "step": 14639
+    },
+    {
+      "epoch": 0.6744368176164371,
+      "grad_norm": 0.7518512705856861,
+      "learning_rate": 5.063017688381545e-05,
+      "loss": 1.3114,
+      "step": 14640
+    },
+    {
+      "epoch": 0.6744828857050721,
+      "grad_norm": 0.9059330500754942,
+      "learning_rate": 5.0617201733831646e-05,
+      "loss": 1.2195,
+      "step": 14641
+    },
+    {
+      "epoch": 0.6745289537937071,
+      "grad_norm": 0.8392328370273288,
+      "learning_rate": 5.060422768327109e-05,
+      "loss": 1.2764,
+      "step": 14642
+    },
+    {
+      "epoch": 0.6745750218823421,
+      "grad_norm": 2.119333995059813,
+      "learning_rate": 5.059125473242279e-05,
+      "loss": 1.3445,
+      "step": 14643
+    },
+    {
+      "epoch": 0.6746210899709771,
+      "grad_norm": 1.0097167685778783,
+      "learning_rate": 5.057828288157549e-05,
+      "loss": 1.2861,
+      "step": 14644
+    },
+    {
+      "epoch": 0.6746671580596121,
+      "grad_norm": 2.254481464765927,
+      "learning_rate": 5.056531213101794e-05,
+      "loss": 1.2824,
+      "step": 14645
+    },
+    {
+      "epoch": 0.6747132261482471,
+      "grad_norm": 0.9619185437245185,
+      "learning_rate": 5.055234248103894e-05,
+      "loss": 1.1682,
+      "step": 14646
+    },
+    {
+      "epoch": 0.6747592942368821,
+      "grad_norm": 0.7159745357274666,
+      "learning_rate": 5.053937393192727e-05,
+      "loss": 1.2527,
+      "step": 14647
+    },
+    {
+      "epoch": 0.6748053623255171,
+      "grad_norm": 0.7371173829236931,
+      "learning_rate": 5.0526406483971656e-05,
+      "loss": 1.4062,
+      "step": 14648
+    },
+    {
+      "epoch": 0.6748514304141521,
+      "grad_norm": 0.8818564700918133,
+      "learning_rate": 5.0513440137460734e-05,
+      "loss": 1.5217,
+      "step": 14649
+    },
+    {
+      "epoch": 0.6748974985027871,
+      "grad_norm": 1.268346458133708,
+      "learning_rate": 5.050047489268322e-05,
+      "loss": 1.4052,
+      "step": 14650
+    },
+    {
+      "epoch": 0.6749435665914221,
+      "grad_norm": 0.9185901108449829,
+      "learning_rate": 5.048751074992777e-05,
+      "loss": 1.1896,
+      "step": 14651
+    },
+    {
+      "epoch": 0.6749896346800571,
+      "grad_norm": 1.6447055338493053,
+      "learning_rate": 5.047454770948303e-05,
+      "loss": 1.0306,
+      "step": 14652
+    },
+    {
+      "epoch": 0.6750357027686922,
+      "grad_norm": 0.8375269686573675,
+      "learning_rate": 5.046158577163752e-05,
+      "loss": 1.3471,
+      "step": 14653
+    },
+    {
+      "epoch": 0.6750817708573271,
+      "grad_norm": 1.1554010607629674,
+      "learning_rate": 5.044862493667988e-05,
+      "loss": 1.2123,
+      "step": 14654
+    },
+    {
+      "epoch": 0.6751278389459622,
+      "grad_norm": 0.8893996864117191,
+      "learning_rate": 5.0435665204898666e-05,
+      "loss": 1.4455,
+      "step": 14655
+    },
+    {
+      "epoch": 0.6751739070345971,
+      "grad_norm": 1.0474051361275833,
+      "learning_rate": 5.0422706576582345e-05,
+      "loss": 1.1985,
+      "step": 14656
+    },
+    {
+      "epoch": 0.6752199751232322,
+      "grad_norm": 0.7098274092812551,
+      "learning_rate": 5.040974905201945e-05,
+      "loss": 1.6351,
+      "step": 14657
+    },
+    {
+      "epoch": 0.6752660432118671,
+      "grad_norm": 0.69263058010809,
+      "learning_rate": 5.039679263149848e-05,
+      "loss": 1.2147,
+      "step": 14658
+    },
+    {
+      "epoch": 0.6753121113005022,
+      "grad_norm": 0.6593399111820082,
+      "learning_rate": 5.038383731530789e-05,
+      "loss": 1.3299,
+      "step": 14659
+    },
+    {
+      "epoch": 0.6753581793891371,
+      "grad_norm": 0.7228550700083873,
+      "learning_rate": 5.037088310373607e-05,
+      "loss": 1.0637,
+      "step": 14660
+    },
+    {
+      "epoch": 0.6754042474777722,
+      "grad_norm": 0.9523398765087734,
+      "learning_rate": 5.035792999707143e-05,
+      "loss": 1.2387,
+      "step": 14661
+    },
+    {
+      "epoch": 0.6754503155664071,
+      "grad_norm": 0.7195145526738909,
+      "learning_rate": 5.034497799560241e-05,
+      "loss": 1.0482,
+      "step": 14662
+    },
+    {
+      "epoch": 0.6754963836550422,
+      "grad_norm": 0.9891082715048399,
+      "learning_rate": 5.033202709961726e-05,
+      "loss": 1.1254,
+      "step": 14663
+    },
+    {
+      "epoch": 0.6755424517436771,
+      "grad_norm": 0.7816811069840258,
+      "learning_rate": 5.031907730940437e-05,
+      "loss": 1.3135,
+      "step": 14664
+    },
+    {
+      "epoch": 0.6755885198323122,
+      "grad_norm": 0.8530059051011161,
+      "learning_rate": 5.030612862525205e-05,
+      "loss": 1.3768,
+      "step": 14665
+    },
+    {
+      "epoch": 0.6756345879209471,
+      "grad_norm": 0.8187921225915074,
+      "learning_rate": 5.02931810474486e-05,
+      "loss": 1.1133,
+      "step": 14666
+    },
+    {
+      "epoch": 0.6756806560095822,
+      "grad_norm": 0.9125881224040145,
+      "learning_rate": 5.028023457628219e-05,
+      "loss": 1.1841,
+      "step": 14667
+    },
+    {
+      "epoch": 0.6757267240982172,
+      "grad_norm": 0.62647627042598,
+      "learning_rate": 5.0267289212041135e-05,
+      "loss": 1.4571,
+      "step": 14668
+    },
+    {
+      "epoch": 0.6757727921868522,
+      "grad_norm": 1.2028360183894289,
+      "learning_rate": 5.02543449550136e-05,
+      "loss": 1.3019,
+      "step": 14669
+    },
+    {
+      "epoch": 0.6758188602754872,
+      "grad_norm": 1.1405791138653358,
+      "learning_rate": 5.024140180548782e-05,
+      "loss": 1.3388,
+      "step": 14670
+    },
+    {
+      "epoch": 0.6758649283641222,
+      "grad_norm": 0.7719164333527554,
+      "learning_rate": 5.022845976375186e-05,
+      "loss": 0.9773,
+      "step": 14671
+    },
+    {
+      "epoch": 0.6759109964527572,
+      "grad_norm": 1.0146493494896536,
+      "learning_rate": 5.0215518830093925e-05,
+      "loss": 1.3146,
+      "step": 14672
+    },
+    {
+      "epoch": 0.6759570645413921,
+      "grad_norm": 0.7198088465860797,
+      "learning_rate": 5.0202579004802116e-05,
+      "loss": 1.2449,
+      "step": 14673
+    },
+    {
+      "epoch": 0.6760031326300272,
+      "grad_norm": 0.6776046434021858,
+      "learning_rate": 5.0189640288164485e-05,
+      "loss": 1.4273,
+      "step": 14674
+    },
+    {
+      "epoch": 0.6760492007186621,
+      "grad_norm": 1.0819827139767688,
+      "learning_rate": 5.017670268046909e-05,
+      "loss": 1.3017,
+      "step": 14675
+    },
+    {
+      "epoch": 0.6760952688072972,
+      "grad_norm": 0.784180678940416,
+      "learning_rate": 5.016376618200398e-05,
+      "loss": 1.3694,
+      "step": 14676
+    },
+    {
+      "epoch": 0.6761413368959321,
+      "grad_norm": 0.6471886674073822,
+      "learning_rate": 5.0150830793057226e-05,
+      "loss": 1.2778,
+      "step": 14677
+    },
+    {
+      "epoch": 0.6761874049845672,
+      "grad_norm": 0.7143021941087798,
+      "learning_rate": 5.013789651391669e-05,
+      "loss": 1.3863,
+      "step": 14678
+    },
+    {
+      "epoch": 0.6762334730732021,
+      "grad_norm": 0.9691628068124839,
+      "learning_rate": 5.0124963344870404e-05,
+      "loss": 1.4308,
+      "step": 14679
+    },
+    {
+      "epoch": 0.6762795411618372,
+      "grad_norm": 0.6626994612953706,
+      "learning_rate": 5.0112031286206316e-05,
+      "loss": 1.2733,
+      "step": 14680
+    },
+    {
+      "epoch": 0.6763256092504722,
+      "grad_norm": 0.7467958401563528,
+      "learning_rate": 5.009910033821229e-05,
+      "loss": 1.1262,
+      "step": 14681
+    },
+    {
+      "epoch": 0.6763716773391072,
+      "grad_norm": 0.6679514606401071,
+      "learning_rate": 5.0086170501176235e-05,
+      "loss": 1.52,
+      "step": 14682
+    },
+    {
+      "epoch": 0.6764177454277422,
+      "grad_norm": 0.8327600530953635,
+      "learning_rate": 5.0073241775385994e-05,
+      "loss": 1.4274,
+      "step": 14683
+    },
+    {
+      "epoch": 0.6764638135163772,
+      "grad_norm": 0.823426978431261,
+      "learning_rate": 5.006031416112946e-05,
+      "loss": 1.0228,
+      "step": 14684
+    },
+    {
+      "epoch": 0.6765098816050122,
+      "grad_norm": 0.8315305697470027,
+      "learning_rate": 5.004738765869439e-05,
+      "loss": 1.043,
+      "step": 14685
+    },
+    {
+      "epoch": 0.6765559496936472,
+      "grad_norm": 0.7327728253864018,
+      "learning_rate": 5.003446226836855e-05,
+      "loss": 1.4525,
+      "step": 14686
+    },
+    {
+      "epoch": 0.6766020177822822,
+      "grad_norm": 0.7719924365150075,
+      "learning_rate": 5.002153799043976e-05,
+      "loss": 1.4105,
+      "step": 14687
+    },
+    {
+      "epoch": 0.6766480858709172,
+      "grad_norm": 0.8691409559219051,
+      "learning_rate": 5.000861482519575e-05,
+      "loss": 1.4501,
+      "step": 14688
+    },
+    {
+      "epoch": 0.6766941539595522,
+      "grad_norm": 0.5974052872168153,
+      "learning_rate": 4.999569277292418e-05,
+      "loss": 1.4606,
+      "step": 14689
+    },
+    {
+      "epoch": 0.6767402220481872,
+      "grad_norm": 0.7414303602178929,
+      "learning_rate": 4.998277183391279e-05,
+      "loss": 1.3882,
+      "step": 14690
+    },
+    {
+      "epoch": 0.6767862901368222,
+      "grad_norm": 0.7744907276731139,
+      "learning_rate": 4.996985200844925e-05,
+      "loss": 1.1967,
+      "step": 14691
+    },
+    {
+      "epoch": 0.6768323582254572,
+      "grad_norm": 0.7480496551819584,
+      "learning_rate": 4.995693329682113e-05,
+      "loss": 1.5952,
+      "step": 14692
+    },
+    {
+      "epoch": 0.6768784263140922,
+      "grad_norm": 0.8272738229631587,
+      "learning_rate": 4.994401569931608e-05,
+      "loss": 1.1292,
+      "step": 14693
+    },
+    {
+      "epoch": 0.6769244944027273,
+      "grad_norm": 0.7359580758426871,
+      "learning_rate": 4.99310992162217e-05,
+      "loss": 1.1351,
+      "step": 14694
+    },
+    {
+      "epoch": 0.6769705624913622,
+      "grad_norm": 0.721158354465387,
+      "learning_rate": 4.9918183847825587e-05,
+      "loss": 1.2343,
+      "step": 14695
+    },
+    {
+      "epoch": 0.6770166305799973,
+      "grad_norm": 1.2464408430928222,
+      "learning_rate": 4.9905269594415196e-05,
+      "loss": 0.9773,
+      "step": 14696
+    },
+    {
+      "epoch": 0.6770626986686322,
+      "grad_norm": 0.8046719612991228,
+      "learning_rate": 4.989235645627809e-05,
+      "loss": 1.4229,
+      "step": 14697
+    },
+    {
+      "epoch": 0.6771087667572673,
+      "grad_norm": 0.9674992580782763,
+      "learning_rate": 4.987944443370178e-05,
+      "loss": 1.278,
+      "step": 14698
+    },
+    {
+      "epoch": 0.6771548348459022,
+      "grad_norm": 0.760165800594797,
+      "learning_rate": 4.986653352697367e-05,
+      "loss": 1.3554,
+      "step": 14699
+    },
+    {
+      "epoch": 0.6772009029345373,
+      "grad_norm": 0.6891915350070974,
+      "learning_rate": 4.985362373638123e-05,
+      "loss": 1.5028,
+      "step": 14700
+    },
+    {
+      "epoch": 0.6772469710231722,
+      "grad_norm": 1.1448147350004658,
+      "learning_rate": 4.984071506221187e-05,
+      "loss": 1.2257,
+      "step": 14701
+    },
+    {
+      "epoch": 0.6772930391118073,
+      "grad_norm": 0.7237327421019593,
+      "learning_rate": 4.982780750475302e-05,
+      "loss": 1.3176,
+      "step": 14702
+    },
+    {
+      "epoch": 0.6773391072004422,
+      "grad_norm": 0.9186246803120718,
+      "learning_rate": 4.981490106429199e-05,
+      "loss": 1.2814,
+      "step": 14703
+    },
+    {
+      "epoch": 0.6773851752890773,
+      "grad_norm": 0.8803351100146443,
+      "learning_rate": 4.9801995741116114e-05,
+      "loss": 1.0248,
+      "step": 14704
+    },
+    {
+      "epoch": 0.6774312433777122,
+      "grad_norm": 0.8285148118449283,
+      "learning_rate": 4.9789091535512745e-05,
+      "loss": 1.2977,
+      "step": 14705
+    },
+    {
+      "epoch": 0.6774773114663473,
+      "grad_norm": 0.6605718587230033,
+      "learning_rate": 4.97761884477692e-05,
+      "loss": 1.563,
+      "step": 14706
+    },
+    {
+      "epoch": 0.6775233795549823,
+      "grad_norm": 0.7841375119749474,
+      "learning_rate": 4.97632864781727e-05,
+      "loss": 1.4351,
+      "step": 14707
+    },
+    {
+      "epoch": 0.6775694476436173,
+      "grad_norm": 1.0607520772166306,
+      "learning_rate": 4.975038562701041e-05,
+      "loss": 1.338,
+      "step": 14708
+    },
+    {
+      "epoch": 0.6776155157322523,
+      "grad_norm": 1.4984678824599187,
+      "learning_rate": 4.9737485894569705e-05,
+      "loss": 1.3842,
+      "step": 14709
+    },
+    {
+      "epoch": 0.6776615838208873,
+      "grad_norm": 0.9543164229879008,
+      "learning_rate": 4.9724587281137655e-05,
+      "loss": 1.0323,
+      "step": 14710
+    },
+    {
+      "epoch": 0.6777076519095223,
+      "grad_norm": 0.7152957699098275,
+      "learning_rate": 4.9711689787001503e-05,
+      "loss": 1.1777,
+      "step": 14711
+    },
+    {
+      "epoch": 0.6777537199981573,
+      "grad_norm": 1.3843869138982208,
+      "learning_rate": 4.9698793412448274e-05,
+      "loss": 1.1184,
+      "step": 14712
+    },
+    {
+      "epoch": 0.6777997880867923,
+      "grad_norm": 1.3851050753602556,
+      "learning_rate": 4.9685898157765246e-05,
+      "loss": 1.099,
+      "step": 14713
+    },
+    {
+      "epoch": 0.6778458561754273,
+      "grad_norm": 0.8919184031619107,
+      "learning_rate": 4.967300402323938e-05,
+      "loss": 1.1564,
+      "step": 14714
+    },
+    {
+      "epoch": 0.6778919242640623,
+      "grad_norm": 0.8998552698958944,
+      "learning_rate": 4.966011100915778e-05,
+      "loss": 1.0968,
+      "step": 14715
+    },
+    {
+      "epoch": 0.6779379923526972,
+      "grad_norm": 0.769812573654472,
+      "learning_rate": 4.964721911580751e-05,
+      "loss": 1.1283,
+      "step": 14716
+    },
+    {
+      "epoch": 0.6779840604413323,
+      "grad_norm": 0.8761147415127059,
+      "learning_rate": 4.96343283434756e-05,
+      "loss": 1.1061,
+      "step": 14717
+    },
+    {
+      "epoch": 0.6780301285299672,
+      "grad_norm": 0.9419051873140659,
+      "learning_rate": 4.962143869244901e-05,
+      "loss": 0.949,
+      "step": 14718
+    },
+    {
+      "epoch": 0.6780761966186023,
+      "grad_norm": 5.289416263531327,
+      "learning_rate": 4.9608550163014636e-05,
+      "loss": 1.2816,
+      "step": 14719
+    },
+    {
+      "epoch": 0.6781222647072372,
+      "grad_norm": 0.998345349515545,
+      "learning_rate": 4.959566275545956e-05,
+      "loss": 1.2822,
+      "step": 14720
+    },
+    {
+      "epoch": 0.6781683327958723,
+      "grad_norm": 0.9912054237274808,
+      "learning_rate": 4.9582776470070594e-05,
+      "loss": 1.2687,
+      "step": 14721
+    },
+    {
+      "epoch": 0.6782144008845074,
+      "grad_norm": 0.638255881505565,
+      "learning_rate": 4.95698913071347e-05,
+      "loss": 1.249,
+      "step": 14722
+    },
+    {
+      "epoch": 0.6782604689731423,
+      "grad_norm": 1.2012928940307461,
+      "learning_rate": 4.9557007266938624e-05,
+      "loss": 1.1055,
+      "step": 14723
+    },
+    {
+      "epoch": 0.6783065370617773,
+      "grad_norm": 0.8353245861132302,
+      "learning_rate": 4.954412434976937e-05,
+      "loss": 1.1731,
+      "step": 14724
+    },
+    {
+      "epoch": 0.6783526051504123,
+      "grad_norm": 0.8138626909236024,
+      "learning_rate": 4.953124255591364e-05,
+      "loss": 1.3264,
+      "step": 14725
+    },
+    {
+      "epoch": 0.6783986732390473,
+      "grad_norm": 0.7753219513413502,
+      "learning_rate": 4.951836188565829e-05,
+      "loss": 1.203,
+      "step": 14726
+    },
+    {
+      "epoch": 0.6784447413276823,
+      "grad_norm": 0.5571745475702435,
+      "learning_rate": 4.9505482339290024e-05,
+      "loss": 1.2981,
+      "step": 14727
+    },
+    {
+      "epoch": 0.6784908094163173,
+      "grad_norm": 0.8628954450799433,
+      "learning_rate": 4.949260391709561e-05,
+      "loss": 1.349,
+      "step": 14728
+    },
+    {
+      "epoch": 0.6785368775049523,
+      "grad_norm": 0.8894142096324028,
+      "learning_rate": 4.947972661936181e-05,
+      "loss": 0.9707,
+      "step": 14729
+    },
+    {
+      "epoch": 0.6785829455935873,
+      "grad_norm": 1.276454875844669,
+      "learning_rate": 4.94668504463752e-05,
+      "loss": 1.3246,
+      "step": 14730
+    },
+    {
+      "epoch": 0.6786290136822223,
+      "grad_norm": 2.5361819891233384,
+      "learning_rate": 4.945397539842259e-05,
+      "loss": 1.4331,
+      "step": 14731
+    },
+    {
+      "epoch": 0.6786750817708573,
+      "grad_norm": 1.1643262219200836,
+      "learning_rate": 4.944110147579051e-05,
+      "loss": 1.0489,
+      "step": 14732
+    },
+    {
+      "epoch": 0.6787211498594923,
+      "grad_norm": 2.821841529326987,
+      "learning_rate": 4.9428228678765657e-05,
+      "loss": 1.2984,
+      "step": 14733
+    },
+    {
+      "epoch": 0.6787672179481273,
+      "grad_norm": 1.3502294841441047,
+      "learning_rate": 4.94153570076345e-05,
+      "loss": 0.9809,
+      "step": 14734
+    },
+    {
+      "epoch": 0.6788132860367624,
+      "grad_norm": 0.9318483757609038,
+      "learning_rate": 4.940248646268376e-05,
+      "loss": 1.2272,
+      "step": 14735
+    },
+    {
+      "epoch": 0.6788593541253973,
+      "grad_norm": 0.7867064519385663,
+      "learning_rate": 4.938961704419991e-05,
+      "loss": 1.3605,
+      "step": 14736
+    },
+    {
+      "epoch": 0.6789054222140324,
+      "grad_norm": 0.8780730190136948,
+      "learning_rate": 4.9376748752469426e-05,
+      "loss": 1.3272,
+      "step": 14737
+    },
+    {
+      "epoch": 0.6789514903026673,
+      "grad_norm": 1.8887856881831626,
+      "learning_rate": 4.936388158777881e-05,
+      "loss": 1.3923,
+      "step": 14738
+    },
+    {
+      "epoch": 0.6789975583913024,
+      "grad_norm": 0.7426314247080567,
+      "learning_rate": 4.935101555041457e-05,
+      "loss": 0.8814,
+      "step": 14739
+    },
+    {
+      "epoch": 0.6790436264799373,
+      "grad_norm": 0.7828351539903593,
+      "learning_rate": 4.933815064066315e-05,
+      "loss": 1.1487,
+      "step": 14740
+    },
+    {
+      "epoch": 0.6790896945685724,
+      "grad_norm": 0.9854480164558659,
+      "learning_rate": 4.932528685881087e-05,
+      "loss": 1.2962,
+      "step": 14741
+    },
+    {
+      "epoch": 0.6791357626572073,
+      "grad_norm": 1.0783407303990638,
+      "learning_rate": 4.9312424205144256e-05,
+      "loss": 1.2291,
+      "step": 14742
+    },
+    {
+      "epoch": 0.6791818307458424,
+      "grad_norm": 0.897791209201859,
+      "learning_rate": 4.9299562679949573e-05,
+      "loss": 1.2593,
+      "step": 14743
+    },
+    {
+      "epoch": 0.6792278988344773,
+      "grad_norm": 0.7109995526506993,
+      "learning_rate": 4.928670228351324e-05,
+      "loss": 1.1186,
+      "step": 14744
+    },
+    {
+      "epoch": 0.6792739669231124,
+      "grad_norm": 6.028298477618974,
+      "learning_rate": 4.9273843016121476e-05,
+      "loss": 1.5061,
+      "step": 14745
+    },
+    {
+      "epoch": 0.6793200350117473,
+      "grad_norm": 0.789184470627623,
+      "learning_rate": 4.9260984878060636e-05,
+      "loss": 1.5879,
+      "step": 14746
+    },
+    {
+      "epoch": 0.6793661031003824,
+      "grad_norm": 1.7829257949552588,
+      "learning_rate": 4.924812786961699e-05,
+      "loss": 1.3825,
+      "step": 14747
+    },
+    {
+      "epoch": 0.6794121711890174,
+      "grad_norm": 2.2034822964568854,
+      "learning_rate": 4.923527199107673e-05,
+      "loss": 1.1471,
+      "step": 14748
+    },
+    {
+      "epoch": 0.6794582392776524,
+      "grad_norm": 1.2151957711540713,
+      "learning_rate": 4.922241724272609e-05,
+      "loss": 1.059,
+      "step": 14749
+    },
+    {
+      "epoch": 0.6795043073662874,
+      "grad_norm": 0.7756297225564676,
+      "learning_rate": 4.920956362485126e-05,
+      "loss": 1.1893,
+      "step": 14750
+    },
+    {
+      "epoch": 0.6795503754549224,
+      "grad_norm": 0.8212853834908806,
+      "learning_rate": 4.919671113773846e-05,
+      "loss": 1.3637,
+      "step": 14751
+    },
+    {
+      "epoch": 0.6795964435435574,
+      "grad_norm": 0.8929899105833188,
+      "learning_rate": 4.918385978167373e-05,
+      "loss": 1.3387,
+      "step": 14752
+    },
+    {
+      "epoch": 0.6796425116321924,
+      "grad_norm": 0.7188098188287638,
+      "learning_rate": 4.9171009556943237e-05,
+      "loss": 1.5088,
+      "step": 14753
+    },
+    {
+      "epoch": 0.6796885797208274,
+      "grad_norm": 0.7512651428589586,
+      "learning_rate": 4.9158160463833104e-05,
+      "loss": 1.0804,
+      "step": 14754
+    },
+    {
+      "epoch": 0.6797346478094624,
+      "grad_norm": 1.3915362227804486,
+      "learning_rate": 4.91453125026293e-05,
+      "loss": 1.1801,
+      "step": 14755
+    },
+    {
+      "epoch": 0.6797807158980974,
+      "grad_norm": 0.9285892780661337,
+      "learning_rate": 4.9132465673617936e-05,
+      "loss": 1.223,
+      "step": 14756
+    },
+    {
+      "epoch": 0.6798267839867324,
+      "grad_norm": 1.5167117262284213,
+      "learning_rate": 4.9119619977085e-05,
+      "loss": 1.2798,
+      "step": 14757
+    },
+    {
+      "epoch": 0.6798728520753674,
+      "grad_norm": 0.8944437805597705,
+      "learning_rate": 4.910677541331651e-05,
+      "loss": 1.2061,
+      "step": 14758
+    },
+    {
+      "epoch": 0.6799189201640023,
+      "grad_norm": 1.0322626263765193,
+      "learning_rate": 4.9093931982598374e-05,
+      "loss": 1.6674,
+      "step": 14759
+    },
+    {
+      "epoch": 0.6799649882526374,
+      "grad_norm": 0.7679088136055726,
+      "learning_rate": 4.9081089685216556e-05,
+      "loss": 1.3221,
+      "step": 14760
+    },
+    {
+      "epoch": 0.6800110563412723,
+      "grad_norm": 0.7938748063620255,
+      "learning_rate": 4.906824852145696e-05,
+      "loss": 1.26,
+      "step": 14761
+    },
+    {
+      "epoch": 0.6800571244299074,
+      "grad_norm": 0.9670468257246304,
+      "learning_rate": 4.905540849160553e-05,
+      "loss": 1.3593,
+      "step": 14762
+    },
+    {
+      "epoch": 0.6801031925185425,
+      "grad_norm": 1.5825918182628531,
+      "learning_rate": 4.904256959594803e-05,
+      "loss": 1.4498,
+      "step": 14763
+    },
+    {
+      "epoch": 0.6801492606071774,
+      "grad_norm": 1.2336984076452076,
+      "learning_rate": 4.902973183477036e-05,
+      "loss": 1.2107,
+      "step": 14764
+    },
+    {
+      "epoch": 0.6801953286958125,
+      "grad_norm": 1.2469384264456813,
+      "learning_rate": 4.9016895208358336e-05,
+      "loss": 1.0587,
+      "step": 14765
+    },
+    {
+      "epoch": 0.6802413967844474,
+      "grad_norm": 0.7437965086519022,
+      "learning_rate": 4.90040597169977e-05,
+      "loss": 1.3407,
+      "step": 14766
+    },
+    {
+      "epoch": 0.6802874648730824,
+      "grad_norm": 0.9305487760214723,
+      "learning_rate": 4.8991225360974225e-05,
+      "loss": 1.243,
+      "step": 14767
+    },
+    {
+      "epoch": 0.6803335329617174,
+      "grad_norm": 1.2744953713951115,
+      "learning_rate": 4.897839214057367e-05,
+      "loss": 1.2368,
+      "step": 14768
+    },
+    {
+      "epoch": 0.6803796010503524,
+      "grad_norm": 0.7249471550487785,
+      "learning_rate": 4.896556005608175e-05,
+      "loss": 1.3315,
+      "step": 14769
+    },
+    {
+      "epoch": 0.6804256691389874,
+      "grad_norm": 0.7432913673561334,
+      "learning_rate": 4.89527291077841e-05,
+      "loss": 1.4298,
+      "step": 14770
+    },
+    {
+      "epoch": 0.6804717372276224,
+      "grad_norm": 0.842384918423537,
+      "learning_rate": 4.8939899295966416e-05,
+      "loss": 1.1893,
+      "step": 14771
+    },
+    {
+      "epoch": 0.6805178053162574,
+      "grad_norm": 0.9541058518918027,
+      "learning_rate": 4.8927070620914364e-05,
+      "loss": 1.3565,
+      "step": 14772
+    },
+    {
+      "epoch": 0.6805638734048924,
+      "grad_norm": 1.1723458737762944,
+      "learning_rate": 4.891424308291346e-05,
+      "loss": 1.0045,
+      "step": 14773
+    },
+    {
+      "epoch": 0.6806099414935274,
+      "grad_norm": 1.3699124620152792,
+      "learning_rate": 4.8901416682249345e-05,
+      "loss": 0.9408,
+      "step": 14774
+    },
+    {
+      "epoch": 0.6806560095821624,
+      "grad_norm": 1.173051787697727,
+      "learning_rate": 4.888859141920758e-05,
+      "loss": 1.3299,
+      "step": 14775
+    },
+    {
+      "epoch": 0.6807020776707975,
+      "grad_norm": 1.526284428145409,
+      "learning_rate": 4.8875767294073726e-05,
+      "loss": 1.4008,
+      "step": 14776
+    },
+    {
+      "epoch": 0.6807481457594324,
+      "grad_norm": 4.399191943933764,
+      "learning_rate": 4.886294430713322e-05,
+      "loss": 1.377,
+      "step": 14777
+    },
+    {
+      "epoch": 0.6807942138480675,
+      "grad_norm": 1.2113309442309683,
+      "learning_rate": 4.885012245867157e-05,
+      "loss": 0.9816,
+      "step": 14778
+    },
+    {
+      "epoch": 0.6808402819367024,
+      "grad_norm": 1.2267224574326066,
+      "learning_rate": 4.883730174897424e-05,
+      "loss": 1.1013,
+      "step": 14779
+    },
+    {
+      "epoch": 0.6808863500253375,
+      "grad_norm": 2.186047500320811,
+      "learning_rate": 4.8824482178326704e-05,
+      "loss": 0.8901,
+      "step": 14780
+    },
+    {
+      "epoch": 0.6809324181139724,
+      "grad_norm": 2.321533920804338,
+      "learning_rate": 4.881166374701428e-05,
+      "loss": 1.2096,
+      "step": 14781
+    },
+    {
+      "epoch": 0.6809784862026075,
+      "grad_norm": 1.0130960045904989,
+      "learning_rate": 4.87988464553224e-05,
+      "loss": 1.1532,
+      "step": 14782
+    },
+    {
+      "epoch": 0.6810245542912424,
+      "grad_norm": 0.740934147809423,
+      "learning_rate": 4.8786030303536444e-05,
+      "loss": 1.4816,
+      "step": 14783
+    },
+    {
+      "epoch": 0.6810706223798775,
+      "grad_norm": 1.2691244122738115,
+      "learning_rate": 4.877321529194168e-05,
+      "loss": 1.1388,
+      "step": 14784
+    },
+    {
+      "epoch": 0.6811166904685124,
+      "grad_norm": 0.7303718273415659,
+      "learning_rate": 4.8760401420823435e-05,
+      "loss": 1.1592,
+      "step": 14785
+    },
+    {
+      "epoch": 0.6811627585571475,
+      "grad_norm": 0.7761451081205957,
+      "learning_rate": 4.8747588690467006e-05,
+      "loss": 1.1948,
+      "step": 14786
+    },
+    {
+      "epoch": 0.6812088266457824,
+      "grad_norm": 2.7749924793022522,
+      "learning_rate": 4.873477710115766e-05,
+      "loss": 1.0227,
+      "step": 14787
+    },
+    {
+      "epoch": 0.6812548947344175,
+      "grad_norm": 27.210835293733513,
+      "learning_rate": 4.872196665318057e-05,
+      "loss": 1.2604,
+      "step": 14788
+    },
+    {
+      "epoch": 0.6813009628230525,
+      "grad_norm": 0.9246488165386386,
+      "learning_rate": 4.870915734682097e-05,
+      "loss": 1.3837,
+      "step": 14789
+    },
+    {
+      "epoch": 0.6813470309116875,
+      "grad_norm": 0.8415689251477357,
+      "learning_rate": 4.869634918236408e-05,
+      "loss": 1.2906,
+      "step": 14790
+    },
+    {
+      "epoch": 0.6813930990003225,
+      "grad_norm": 0.9885571188158805,
+      "learning_rate": 4.8683542160094964e-05,
+      "loss": 0.7343,
+      "step": 14791
+    },
+    {
+      "epoch": 0.6814391670889575,
+      "grad_norm": 0.9427280158770692,
+      "learning_rate": 4.867073628029879e-05,
+      "loss": 1.319,
+      "step": 14792
+    },
+    {
+      "epoch": 0.6814852351775925,
+      "grad_norm": 0.7780809848822869,
+      "learning_rate": 4.865793154326066e-05,
+      "loss": 1.2196,
+      "step": 14793
+    },
+    {
+      "epoch": 0.6815313032662275,
+      "grad_norm": 0.9789639972860805,
+      "learning_rate": 4.8645127949265704e-05,
+      "loss": 1.1302,
+      "step": 14794
+    },
+    {
+      "epoch": 0.6815773713548625,
+      "grad_norm": 1.397416268781441,
+      "learning_rate": 4.8632325498598863e-05,
+      "loss": 1.4777,
+      "step": 14795
+    },
+    {
+      "epoch": 0.6816234394434975,
+      "grad_norm": 1.0804013739947624,
+      "learning_rate": 4.861952419154523e-05,
+      "loss": 1.0343,
+      "step": 14796
+    },
+    {
+      "epoch": 0.6816695075321325,
+      "grad_norm": 1.12402397004112,
+      "learning_rate": 4.8606724028389796e-05,
+      "loss": 1.666,
+      "step": 14797
+    },
+    {
+      "epoch": 0.6817155756207675,
+      "grad_norm": 1.0658255498424896,
+      "learning_rate": 4.859392500941756e-05,
+      "loss": 1.6176,
+      "step": 14798
+    },
+    {
+      "epoch": 0.6817616437094025,
+      "grad_norm": 1.3393644462111687,
+      "learning_rate": 4.8581127134913404e-05,
+      "loss": 1.4694,
+      "step": 14799
+    },
+    {
+      "epoch": 0.6818077117980375,
+      "grad_norm": 0.8121202795760705,
+      "learning_rate": 4.856833040516229e-05,
+      "loss": 1.3134,
+      "step": 14800
+    },
+    {
+      "epoch": 0.6818537798866725,
+      "grad_norm": 1.0809754212673757,
+      "learning_rate": 4.8555534820449156e-05,
+      "loss": 1.3758,
+      "step": 14801
+    },
+    {
+      "epoch": 0.6818998479753074,
+      "grad_norm": 1.6657816149383877,
+      "learning_rate": 4.854274038105878e-05,
+      "loss": 1.3373,
+      "step": 14802
+    },
+    {
+      "epoch": 0.6819459160639425,
+      "grad_norm": 1.069761629787109,
+      "learning_rate": 4.852994708727607e-05,
+      "loss": 1.4492,
+      "step": 14803
+    },
+    {
+      "epoch": 0.6819919841525776,
+      "grad_norm": 1.0452206742411505,
+      "learning_rate": 4.8517154939385835e-05,
+      "loss": 1.2269,
+      "step": 14804
+    },
+    {
+      "epoch": 0.6820380522412125,
+      "grad_norm": 0.7577168525278347,
+      "learning_rate": 4.85043639376729e-05,
+      "loss": 1.1141,
+      "step": 14805
+    },
+    {
+      "epoch": 0.6820841203298476,
+      "grad_norm": 1.1524612131205025,
+      "learning_rate": 4.8491574082421975e-05,
+      "loss": 1.1649,
+      "step": 14806
+    },
+    {
+      "epoch": 0.6821301884184825,
+      "grad_norm": 0.7763681573861931,
+      "learning_rate": 4.8478785373917836e-05,
+      "loss": 1.4954,
+      "step": 14807
+    },
+    {
+      "epoch": 0.6821762565071176,
+      "grad_norm": 1.351163781754468,
+      "learning_rate": 4.846599781244524e-05,
+      "loss": 1.4254,
+      "step": 14808
+    },
+    {
+      "epoch": 0.6822223245957525,
+      "grad_norm": 0.9937915489795822,
+      "learning_rate": 4.84532113982888e-05,
+      "loss": 1.2835,
+      "step": 14809
+    },
+    {
+      "epoch": 0.6822683926843875,
+      "grad_norm": 2.305836953522628,
+      "learning_rate": 4.844042613173326e-05,
+      "loss": 1.4624,
+      "step": 14810
+    },
+    {
+      "epoch": 0.6823144607730225,
+      "grad_norm": 1.6321888043523423,
+      "learning_rate": 4.842764201306316e-05,
+      "loss": 1.2131,
+      "step": 14811
+    },
+    {
+      "epoch": 0.6823605288616575,
+      "grad_norm": 1.4022225144690488,
+      "learning_rate": 4.841485904256324e-05,
+      "loss": 1.2583,
+      "step": 14812
+    },
+    {
+      "epoch": 0.6824065969502925,
+      "grad_norm": 0.9328059831998673,
+      "learning_rate": 4.8402077220518013e-05,
+      "loss": 1.287,
+      "step": 14813
+    },
+    {
+      "epoch": 0.6824526650389275,
+      "grad_norm": 0.7815014504269056,
+      "learning_rate": 4.838929654721206e-05,
+      "loss": 1.2499,
+      "step": 14814
+    },
+    {
+      "epoch": 0.6824987331275625,
+      "grad_norm": 2.313708664554234,
+      "learning_rate": 4.8376517022929934e-05,
+      "loss": 1.1497,
+      "step": 14815
+    },
+    {
+      "epoch": 0.6825448012161975,
+      "grad_norm": 1.1487304233558275,
+      "learning_rate": 4.836373864795617e-05,
+      "loss": 1.4405,
+      "step": 14816
+    },
+    {
+      "epoch": 0.6825908693048326,
+      "grad_norm": 1.1801614401155565,
+      "learning_rate": 4.835096142257522e-05,
+      "loss": 1.2376,
+      "step": 14817
+    },
+    {
+      "epoch": 0.6826369373934675,
+      "grad_norm": 2.120446707979522,
+      "learning_rate": 4.8338185347071496e-05,
+      "loss": 1.1398,
+      "step": 14818
+    },
+    {
+      "epoch": 0.6826830054821026,
+      "grad_norm": 0.8868633691092835,
+      "learning_rate": 4.8325410421729557e-05,
+      "loss": 0.909,
+      "step": 14819
+    },
+    {
+      "epoch": 0.6827290735707375,
+      "grad_norm": 1.2528060770403904,
+      "learning_rate": 4.831263664683372e-05,
+      "loss": 1.245,
+      "step": 14820
+    },
+    {
+      "epoch": 0.6827751416593726,
+      "grad_norm": 0.8853318415775955,
+      "learning_rate": 4.829986402266844e-05,
+      "loss": 1.3036,
+      "step": 14821
+    },
+    {
+      "epoch": 0.6828212097480075,
+      "grad_norm": 0.9495361136092273,
+      "learning_rate": 4.828709254951796e-05,
+      "loss": 1.682,
+      "step": 14822
+    },
+    {
+      "epoch": 0.6828672778366426,
+      "grad_norm": 1.8727717810957938,
+      "learning_rate": 4.827432222766679e-05,
+      "loss": 1.4872,
+      "step": 14823
+    },
+    {
+      "epoch": 0.6829133459252775,
+      "grad_norm": 0.863778005975373,
+      "learning_rate": 4.826155305739909e-05,
+      "loss": 1.2455,
+      "step": 14824
+    },
+    {
+      "epoch": 0.6829594140139126,
+      "grad_norm": 0.8367049962267841,
+      "learning_rate": 4.8248785038999244e-05,
+      "loss": 1.2257,
+      "step": 14825
+    },
+    {
+      "epoch": 0.6830054821025475,
+      "grad_norm": 0.7662805900658831,
+      "learning_rate": 4.823601817275142e-05,
+      "loss": 1.4408,
+      "step": 14826
+    },
+    {
+      "epoch": 0.6830515501911826,
+      "grad_norm": 1.155656116707937,
+      "learning_rate": 4.8223252458939905e-05,
+      "loss": 1.4243,
+      "step": 14827
+    },
+    {
+      "epoch": 0.6830976182798175,
+      "grad_norm": 0.8476715955397771,
+      "learning_rate": 4.821048789784892e-05,
+      "loss": 1.4525,
+      "step": 14828
+    },
+    {
+      "epoch": 0.6831436863684526,
+      "grad_norm": 2.81281559272129,
+      "learning_rate": 4.8197724489762555e-05,
+      "loss": 1.3016,
+      "step": 14829
+    },
+    {
+      "epoch": 0.6831897544570876,
+      "grad_norm": 0.9854068556100315,
+      "learning_rate": 4.818496223496511e-05,
+      "loss": 1.1553,
+      "step": 14830
+    },
+    {
+      "epoch": 0.6832358225457226,
+      "grad_norm": 1.0982276351214202,
+      "learning_rate": 4.817220113374059e-05,
+      "loss": 0.8492,
+      "step": 14831
+    },
+    {
+      "epoch": 0.6832818906343576,
+      "grad_norm": 2.322888934077733,
+      "learning_rate": 4.81594411863732e-05,
+      "loss": 1.0203,
+      "step": 14832
+    },
+    {
+      "epoch": 0.6833279587229926,
+      "grad_norm": 1.0532530721766573,
+      "learning_rate": 4.814668239314687e-05,
+      "loss": 1.4477,
+      "step": 14833
+    },
+    {
+      "epoch": 0.6833740268116276,
+      "grad_norm": 1.8277118227796985,
+      "learning_rate": 4.813392475434584e-05,
+      "loss": 1.2537,
+      "step": 14834
+    },
+    {
+      "epoch": 0.6834200949002626,
+      "grad_norm": 1.1458398877868903,
+      "learning_rate": 4.812116827025399e-05,
+      "loss": 1.2515,
+      "step": 14835
+    },
+    {
+      "epoch": 0.6834661629888976,
+      "grad_norm": 0.9728803695849638,
+      "learning_rate": 4.8108412941155426e-05,
+      "loss": 1.146,
+      "step": 14836
+    },
+    {
+      "epoch": 0.6835122310775326,
+      "grad_norm": 1.1442161050246944,
+      "learning_rate": 4.809565876733403e-05,
+      "loss": 1.4407,
+      "step": 14837
+    },
+    {
+      "epoch": 0.6835582991661676,
+      "grad_norm": 1.0319612674391136,
+      "learning_rate": 4.808290574907379e-05,
+      "loss": 1.4603,
+      "step": 14838
+    },
+    {
+      "epoch": 0.6836043672548026,
+      "grad_norm": 0.8778206496486151,
+      "learning_rate": 4.8070153886658684e-05,
+      "loss": 1.3372,
+      "step": 14839
+    },
+    {
+      "epoch": 0.6836504353434376,
+      "grad_norm": 0.8294104311063393,
+      "learning_rate": 4.8057403180372474e-05,
+      "loss": 1.1156,
+      "step": 14840
+    },
+    {
+      "epoch": 0.6836965034320726,
+      "grad_norm": 0.8530769934540458,
+      "learning_rate": 4.80446536304992e-05,
+      "loss": 1.207,
+      "step": 14841
+    },
+    {
+      "epoch": 0.6837425715207076,
+      "grad_norm": 0.8519945824754969,
+      "learning_rate": 4.8031905237322596e-05,
+      "loss": 1.3976,
+      "step": 14842
+    },
+    {
+      "epoch": 0.6837886396093427,
+      "grad_norm": 1.0656160859690536,
+      "learning_rate": 4.8019158001126555e-05,
+      "loss": 1.4301,
+      "step": 14843
+    },
+    {
+      "epoch": 0.6838347076979776,
+      "grad_norm": 0.7711476980693893,
+      "learning_rate": 4.8006411922194746e-05,
+      "loss": 1.0895,
+      "step": 14844
+    },
+    {
+      "epoch": 0.6838807757866127,
+      "grad_norm": 0.7694732390628035,
+      "learning_rate": 4.79936670008111e-05,
+      "loss": 1.5498,
+      "step": 14845
+    },
+    {
+      "epoch": 0.6839268438752476,
+      "grad_norm": 1.1976844020920314,
+      "learning_rate": 4.79809232372593e-05,
+      "loss": 1.2356,
+      "step": 14846
+    },
+    {
+      "epoch": 0.6839729119638827,
+      "grad_norm": 8.92699691153447,
+      "learning_rate": 4.7968180631823e-05,
+      "loss": 1.3175,
+      "step": 14847
+    },
+    {
+      "epoch": 0.6840189800525176,
+      "grad_norm": 0.7375399341400273,
+      "learning_rate": 4.7955439184785955e-05,
+      "loss": 1.3938,
+      "step": 14848
+    },
+    {
+      "epoch": 0.6840650481411527,
+      "grad_norm": 5.393517316514261,
+      "learning_rate": 4.794269889643181e-05,
+      "loss": 1.4592,
+      "step": 14849
+    },
+    {
+      "epoch": 0.6841111162297876,
+      "grad_norm": 0.7683415944106056,
+      "learning_rate": 4.792995976704425e-05,
+      "loss": 1.2248,
+      "step": 14850
+    },
+    {
+      "epoch": 0.6841571843184227,
+      "grad_norm": 0.8463880817658309,
+      "learning_rate": 4.791722179690683e-05,
+      "loss": 1.1925,
+      "step": 14851
+    },
+    {
+      "epoch": 0.6842032524070576,
+      "grad_norm": 0.8129456655532676,
+      "learning_rate": 4.790448498630315e-05,
+      "loss": 1.4711,
+      "step": 14852
+    },
+    {
+      "epoch": 0.6842493204956926,
+      "grad_norm": 1.1574649446943956,
+      "learning_rate": 4.78917493355168e-05,
+      "loss": 1.0511,
+      "step": 14853
+    },
+    {
+      "epoch": 0.6842953885843276,
+      "grad_norm": 1.2690160094750687,
+      "learning_rate": 4.787901484483134e-05,
+      "loss": 1.1239,
+      "step": 14854
+    },
+    {
+      "epoch": 0.6843414566729626,
+      "grad_norm": 20.372333295269783,
+      "learning_rate": 4.78662815145302e-05,
+      "loss": 1.0971,
+      "step": 14855
+    },
+    {
+      "epoch": 0.6843875247615976,
+      "grad_norm": 1.9920408972472652,
+      "learning_rate": 4.785354934489692e-05,
+      "loss": 1.4627,
+      "step": 14856
+    },
+    {
+      "epoch": 0.6844335928502326,
+      "grad_norm": 0.8973976999224081,
+      "learning_rate": 4.784081833621499e-05,
+      "loss": 1.3803,
+      "step": 14857
+    },
+    {
+      "epoch": 0.6844796609388677,
+      "grad_norm": 0.8698289080331489,
+      "learning_rate": 4.7828088488767756e-05,
+      "loss": 1.2548,
+      "step": 14858
+    },
+    {
+      "epoch": 0.6845257290275026,
+      "grad_norm": 0.8417835205915745,
+      "learning_rate": 4.781535980283869e-05,
+      "loss": 0.8784,
+      "step": 14859
+    },
+    {
+      "epoch": 0.6845717971161377,
+      "grad_norm": 1.159966791952023,
+      "learning_rate": 4.7802632278711146e-05,
+      "loss": 1.2754,
+      "step": 14860
+    },
+    {
+      "epoch": 0.6846178652047726,
+      "grad_norm": 4.300880675728729,
+      "learning_rate": 4.7789905916668544e-05,
+      "loss": 1.1473,
+      "step": 14861
+    },
+    {
+      "epoch": 0.6846639332934077,
+      "grad_norm": 0.7866019953908934,
+      "learning_rate": 4.777718071699413e-05,
+      "loss": 1.4923,
+      "step": 14862
+    },
+    {
+      "epoch": 0.6847100013820426,
+      "grad_norm": 0.9543335451767029,
+      "learning_rate": 4.7764456679971236e-05,
+      "loss": 1.1692,
+      "step": 14863
+    },
+    {
+      "epoch": 0.6847560694706777,
+      "grad_norm": 1.0518843111767164,
+      "learning_rate": 4.7751733805883195e-05,
+      "loss": 1.0863,
+      "step": 14864
+    },
+    {
+      "epoch": 0.6848021375593126,
+      "grad_norm": 5.198469604611438,
+      "learning_rate": 4.773901209501317e-05,
+      "loss": 1.3889,
+      "step": 14865
+    },
+    {
+      "epoch": 0.6848482056479477,
+      "grad_norm": 1.1046047742268257,
+      "learning_rate": 4.772629154764443e-05,
+      "loss": 1.4838,
+      "step": 14866
+    },
+    {
+      "epoch": 0.6848942737365826,
+      "grad_norm": 0.838938746828939,
+      "learning_rate": 4.7713572164060185e-05,
+      "loss": 1.6535,
+      "step": 14867
+    },
+    {
+      "epoch": 0.6849403418252177,
+      "grad_norm": 1.1440280007890116,
+      "learning_rate": 4.7700853944543645e-05,
+      "loss": 1.2645,
+      "step": 14868
+    },
+    {
+      "epoch": 0.6849864099138526,
+      "grad_norm": 1.6032300628517002,
+      "learning_rate": 4.7688136889377866e-05,
+      "loss": 1.1304,
+      "step": 14869
+    },
+    {
+      "epoch": 0.6850324780024877,
+      "grad_norm": 1.2999589077304814,
+      "learning_rate": 4.7675420998846035e-05,
+      "loss": 1.0878,
+      "step": 14870
+    },
+    {
+      "epoch": 0.6850785460911227,
+      "grad_norm": 1.1595914423321445,
+      "learning_rate": 4.7662706273231236e-05,
+      "loss": 0.9986,
+      "step": 14871
+    },
+    {
+      "epoch": 0.6851246141797577,
+      "grad_norm": 0.9339413655712477,
+      "learning_rate": 4.764999271281658e-05,
+      "loss": 1.1717,
+      "step": 14872
+    },
+    {
+      "epoch": 0.6851706822683927,
+      "grad_norm": 1.7367052584168388,
+      "learning_rate": 4.763728031788504e-05,
+      "loss": 1.2242,
+      "step": 14873
+    },
+    {
+      "epoch": 0.6852167503570277,
+      "grad_norm": 0.9893603312853051,
+      "learning_rate": 4.7624569088719674e-05,
+      "loss": 1.4428,
+      "step": 14874
+    },
+    {
+      "epoch": 0.6852628184456627,
+      "grad_norm": 1.0407717060913735,
+      "learning_rate": 4.76118590256035e-05,
+      "loss": 1.4072,
+      "step": 14875
+    },
+    {
+      "epoch": 0.6853088865342977,
+      "grad_norm": 1.2843555899319068,
+      "learning_rate": 4.759915012881943e-05,
+      "loss": 1.4187,
+      "step": 14876
+    },
+    {
+      "epoch": 0.6853549546229327,
+      "grad_norm": 0.7962174623912616,
+      "learning_rate": 4.758644239865042e-05,
+      "loss": 1.4861,
+      "step": 14877
+    },
+    {
+      "epoch": 0.6854010227115677,
+      "grad_norm": 1.8673632374522589,
+      "learning_rate": 4.7573735835379404e-05,
+      "loss": 1.3645,
+      "step": 14878
+    },
+    {
+      "epoch": 0.6854470908002027,
+      "grad_norm": 0.9900036549604918,
+      "learning_rate": 4.756103043928932e-05,
+      "loss": 1.1866,
+      "step": 14879
+    },
+    {
+      "epoch": 0.6854931588888377,
+      "grad_norm": 1.4933357818278556,
+      "learning_rate": 4.754832621066293e-05,
+      "loss": 1.2278,
+      "step": 14880
+    },
+    {
+      "epoch": 0.6855392269774727,
+      "grad_norm": 1.3714399244564066,
+      "learning_rate": 4.753562314978313e-05,
+      "loss": 1.3085,
+      "step": 14881
+    },
+    {
+      "epoch": 0.6855852950661077,
+      "grad_norm": 0.9716202882416329,
+      "learning_rate": 4.752292125693275e-05,
+      "loss": 1.4271,
+      "step": 14882
+    },
+    {
+      "epoch": 0.6856313631547427,
+      "grad_norm": 2.183517630235747,
+      "learning_rate": 4.751022053239452e-05,
+      "loss": 1.3781,
+      "step": 14883
+    },
+    {
+      "epoch": 0.6856774312433778,
+      "grad_norm": 1.0379471188853449,
+      "learning_rate": 4.7497520976451225e-05,
+      "loss": 1.055,
+      "step": 14884
+    },
+    {
+      "epoch": 0.6857234993320127,
+      "grad_norm": 1.2666421470548428,
+      "learning_rate": 4.74848225893856e-05,
+      "loss": 1.2108,
+      "step": 14885
+    },
+    {
+      "epoch": 0.6857695674206478,
+      "grad_norm": 0.8151924470076938,
+      "learning_rate": 4.7472125371480405e-05,
+      "loss": 1.5586,
+      "step": 14886
+    },
+    {
+      "epoch": 0.6858156355092827,
+      "grad_norm": 0.8421511644799624,
+      "learning_rate": 4.745942932301823e-05,
+      "loss": 1.113,
+      "step": 14887
+    },
+    {
+      "epoch": 0.6858617035979178,
+      "grad_norm": 0.8697503124966118,
+      "learning_rate": 4.744673444428177e-05,
+      "loss": 1.4741,
+      "step": 14888
+    },
+    {
+      "epoch": 0.6859077716865527,
+      "grad_norm": 0.8563417951979407,
+      "learning_rate": 4.743404073555366e-05,
+      "loss": 1.2986,
+      "step": 14889
+    },
+    {
+      "epoch": 0.6859538397751878,
+      "grad_norm": 1.0012912924287498,
+      "learning_rate": 4.7421348197116545e-05,
+      "loss": 1.5272,
+      "step": 14890
+    },
+    {
+      "epoch": 0.6859999078638227,
+      "grad_norm": 1.9334257757159867,
+      "learning_rate": 4.7408656829252904e-05,
+      "loss": 1.2368,
+      "step": 14891
+    },
+    {
+      "epoch": 0.6860459759524578,
+      "grad_norm": 1.459996020221619,
+      "learning_rate": 4.739596663224536e-05,
+      "loss": 1.2807,
+      "step": 14892
+    },
+    {
+      "epoch": 0.6860920440410927,
+      "grad_norm": 0.9011383154722177,
+      "learning_rate": 4.7383277606376455e-05,
+      "loss": 1.5179,
+      "step": 14893
+    },
+    {
+      "epoch": 0.6861381121297278,
+      "grad_norm": 1.270907693877038,
+      "learning_rate": 4.7370589751928616e-05,
+      "loss": 1.635,
+      "step": 14894
+    },
+    {
+      "epoch": 0.6861841802183627,
+      "grad_norm": 1.4756203091280329,
+      "learning_rate": 4.735790306918436e-05,
+      "loss": 1.3332,
+      "step": 14895
+    },
+    {
+      "epoch": 0.6862302483069977,
+      "grad_norm": 0.925889021695867,
+      "learning_rate": 4.734521755842614e-05,
+      "loss": 1.2092,
+      "step": 14896
+    },
+    {
+      "epoch": 0.6862763163956327,
+      "grad_norm": 1.0719160764974303,
+      "learning_rate": 4.733253321993638e-05,
+      "loss": 1.3076,
+      "step": 14897
+    },
+    {
+      "epoch": 0.6863223844842677,
+      "grad_norm": 0.7498573504174182,
+      "learning_rate": 4.731985005399745e-05,
+      "loss": 1.0826,
+      "step": 14898
+    },
+    {
+      "epoch": 0.6863684525729028,
+      "grad_norm": 0.630715220596266,
+      "learning_rate": 4.7307168060891714e-05,
+      "loss": 1.5106,
+      "step": 14899
+    },
+    {
+      "epoch": 0.6864145206615377,
+      "grad_norm": 0.7953456337882342,
+      "learning_rate": 4.7294487240901575e-05,
+      "loss": 1.4763,
+      "step": 14900
+    },
+    {
+      "epoch": 0.6864605887501728,
+      "grad_norm": 1.0537647567536865,
+      "learning_rate": 4.728180759430927e-05,
+      "loss": 1.0726,
+      "step": 14901
+    },
+    {
+      "epoch": 0.6865066568388077,
+      "grad_norm": 1.0427686953942947,
+      "learning_rate": 4.7269129121397125e-05,
+      "loss": 1.0575,
+      "step": 14902
+    },
+    {
+      "epoch": 0.6865527249274428,
+      "grad_norm": 0.7412930359703325,
+      "learning_rate": 4.725645182244741e-05,
+      "loss": 1.4493,
+      "step": 14903
+    },
+    {
+      "epoch": 0.6865987930160777,
+      "grad_norm": 1.1062596517944452,
+      "learning_rate": 4.7243775697742396e-05,
+      "loss": 1.2318,
+      "step": 14904
+    },
+    {
+      "epoch": 0.6866448611047128,
+      "grad_norm": 1.0262785624598971,
+      "learning_rate": 4.7231100747564224e-05,
+      "loss": 1.1322,
+      "step": 14905
+    },
+    {
+      "epoch": 0.6866909291933477,
+      "grad_norm": 1.308671591670559,
+      "learning_rate": 4.721842697219511e-05,
+      "loss": 1.0946,
+      "step": 14906
+    },
+    {
+      "epoch": 0.6867369972819828,
+      "grad_norm": 0.7941856748749635,
+      "learning_rate": 4.720575437191722e-05,
+      "loss": 1.1035,
+      "step": 14907
+    },
+    {
+      "epoch": 0.6867830653706177,
+      "grad_norm": 0.839231777760418,
+      "learning_rate": 4.719308294701272e-05,
+      "loss": 0.9506,
+      "step": 14908
+    },
+    {
+      "epoch": 0.6868291334592528,
+      "grad_norm": 1.3342284235724087,
+      "learning_rate": 4.7180412697763684e-05,
+      "loss": 1.4565,
+      "step": 14909
+    },
+    {
+      "epoch": 0.6868752015478877,
+      "grad_norm": 0.9947463097659257,
+      "learning_rate": 4.716774362445213e-05,
+      "loss": 1.1979,
+      "step": 14910
+    },
+    {
+      "epoch": 0.6869212696365228,
+      "grad_norm": 1.559889537045598,
+      "learning_rate": 4.715507572736024e-05,
+      "loss": 1.3307,
+      "step": 14911
+    },
+    {
+      "epoch": 0.6869673377251578,
+      "grad_norm": 0.9540800155286735,
+      "learning_rate": 4.714240900676995e-05,
+      "loss": 1.4431,
+      "step": 14912
+    },
+    {
+      "epoch": 0.6870134058137928,
+      "grad_norm": 0.8112101103111865,
+      "learning_rate": 4.712974346296328e-05,
+      "loss": 1.1677,
+      "step": 14913
+    },
+    {
+      "epoch": 0.6870594739024278,
+      "grad_norm": 0.8798568098131702,
+      "learning_rate": 4.711707909622223e-05,
+      "loss": 1.5875,
+      "step": 14914
+    },
+    {
+      "epoch": 0.6871055419910628,
+      "grad_norm": 0.7941478117339472,
+      "learning_rate": 4.710441590682877e-05,
+      "loss": 1.4169,
+      "step": 14915
+    },
+    {
+      "epoch": 0.6871516100796978,
+      "grad_norm": 0.7467445322690974,
+      "learning_rate": 4.7091753895064764e-05,
+      "loss": 1.4285,
+      "step": 14916
+    },
+    {
+      "epoch": 0.6871976781683328,
+      "grad_norm": 5.806898611109548,
+      "learning_rate": 4.707909306121212e-05,
+      "loss": 1.0691,
+      "step": 14917
+    },
+    {
+      "epoch": 0.6872437462569678,
+      "grad_norm": 0.9818677266726188,
+      "learning_rate": 4.706643340555278e-05,
+      "loss": 1.5577,
+      "step": 14918
+    },
+    {
+      "epoch": 0.6872898143456028,
+      "grad_norm": 0.989644189190374,
+      "learning_rate": 4.70537749283685e-05,
+      "loss": 1.1703,
+      "step": 14919
+    },
+    {
+      "epoch": 0.6873358824342378,
+      "grad_norm": 1.946082171962431,
+      "learning_rate": 4.704111762994118e-05,
+      "loss": 1.3242,
+      "step": 14920
+    },
+    {
+      "epoch": 0.6873819505228728,
+      "grad_norm": 0.9972314268018247,
+      "learning_rate": 4.702846151055249e-05,
+      "loss": 1.2768,
+      "step": 14921
+    },
+    {
+      "epoch": 0.6874280186115078,
+      "grad_norm": 0.7670518168823525,
+      "learning_rate": 4.7015806570484355e-05,
+      "loss": 1.3236,
+      "step": 14922
+    },
+    {
+      "epoch": 0.6874740867001428,
+      "grad_norm": 0.7930897657659906,
+      "learning_rate": 4.70031528100184e-05,
+      "loss": 1.3883,
+      "step": 14923
+    },
+    {
+      "epoch": 0.6875201547887778,
+      "grad_norm": 0.8343569363174727,
+      "learning_rate": 4.699050022943642e-05,
+      "loss": 1.4601,
+      "step": 14924
+    },
+    {
+      "epoch": 0.6875662228774129,
+      "grad_norm": 0.8884178019718902,
+      "learning_rate": 4.697784882901998e-05,
+      "loss": 1.0398,
+      "step": 14925
+    },
+    {
+      "epoch": 0.6876122909660478,
+      "grad_norm": 0.9950414734171352,
+      "learning_rate": 4.69651986090509e-05,
+      "loss": 1.4745,
+      "step": 14926
+    },
+    {
+      "epoch": 0.6876583590546829,
+      "grad_norm": 0.9274369362856084,
+      "learning_rate": 4.6952549569810735e-05,
+      "loss": 1.4972,
+      "step": 14927
+    },
+    {
+      "epoch": 0.6877044271433178,
+      "grad_norm": 0.6130885899348104,
+      "learning_rate": 4.6939901711581026e-05,
+      "loss": 1.5506,
+      "step": 14928
+    },
+    {
+      "epoch": 0.6877504952319529,
+      "grad_norm": 0.768871639418429,
+      "learning_rate": 4.692725503464349e-05,
+      "loss": 1.276,
+      "step": 14929
+    },
+    {
+      "epoch": 0.6877965633205878,
+      "grad_norm": 0.8334684903888088,
+      "learning_rate": 4.69146095392796e-05,
+      "loss": 1.6371,
+      "step": 14930
+    },
+    {
+      "epoch": 0.6878426314092229,
+      "grad_norm": 0.9716061708158209,
+      "learning_rate": 4.690196522577094e-05,
+      "loss": 1.0666,
+      "step": 14931
+    },
+    {
+      "epoch": 0.6878886994978578,
+      "grad_norm": 2.0452272817546424,
+      "learning_rate": 4.688932209439889e-05,
+      "loss": 1.2356,
+      "step": 14932
+    },
+    {
+      "epoch": 0.6879347675864929,
+      "grad_norm": 1.0270523941502225,
+      "learning_rate": 4.687668014544511e-05,
+      "loss": 1.0774,
+      "step": 14933
+    },
+    {
+      "epoch": 0.6879808356751278,
+      "grad_norm": 0.8942149813106729,
+      "learning_rate": 4.686403937919092e-05,
+      "loss": 0.9682,
+      "step": 14934
+    },
+    {
+      "epoch": 0.6880269037637629,
+      "grad_norm": 1.0046306225526276,
+      "learning_rate": 4.6851399795917816e-05,
+      "loss": 1.1145,
+      "step": 14935
+    },
+    {
+      "epoch": 0.6880729718523978,
+      "grad_norm": 1.1769612569213075,
+      "learning_rate": 4.6838761395907126e-05,
+      "loss": 1.1303,
+      "step": 14936
+    },
+    {
+      "epoch": 0.6881190399410329,
+      "grad_norm": 1.9725799321672748,
+      "learning_rate": 4.682612417944028e-05,
+      "loss": 1.2019,
+      "step": 14937
+    },
+    {
+      "epoch": 0.6881651080296678,
+      "grad_norm": 0.7846437137937851,
+      "learning_rate": 4.6813488146798625e-05,
+      "loss": 1.24,
+      "step": 14938
+    },
+    {
+      "epoch": 0.6882111761183028,
+      "grad_norm": 0.9743326012553994,
+      "learning_rate": 4.6800853298263393e-05,
+      "loss": 1.3328,
+      "step": 14939
+    },
+    {
+      "epoch": 0.6882572442069379,
+      "grad_norm": 0.7439187150898919,
+      "learning_rate": 4.6788219634116016e-05,
+      "loss": 1.3772,
+      "step": 14940
+    },
+    {
+      "epoch": 0.6883033122955728,
+      "grad_norm": 1.478103821755281,
+      "learning_rate": 4.677558715463766e-05,
+      "loss": 1.3222,
+      "step": 14941
+    },
+    {
+      "epoch": 0.6883493803842079,
+      "grad_norm": 1.0059782130080228,
+      "learning_rate": 4.676295586010963e-05,
+      "loss": 1.3358,
+      "step": 14942
+    },
+    {
+      "epoch": 0.6883954484728428,
+      "grad_norm": 1.132117919889504,
+      "learning_rate": 4.675032575081302e-05,
+      "loss": 1.3009,
+      "step": 14943
+    },
+    {
+      "epoch": 0.6884415165614779,
+      "grad_norm": 0.9273415673917366,
+      "learning_rate": 4.673769682702919e-05,
+      "loss": 1.2102,
+      "step": 14944
+    },
+    {
+      "epoch": 0.6884875846501128,
+      "grad_norm": 1.223935658671459,
+      "learning_rate": 4.672506908903921e-05,
+      "loss": 1.3537,
+      "step": 14945
+    },
+    {
+      "epoch": 0.6885336527387479,
+      "grad_norm": 1.1764784013467555,
+      "learning_rate": 4.6712442537124177e-05,
+      "loss": 1.1256,
+      "step": 14946
+    },
+    {
+      "epoch": 0.6885797208273828,
+      "grad_norm": 1.5322505639825386,
+      "learning_rate": 4.669981717156525e-05,
+      "loss": 1.3848,
+      "step": 14947
+    },
+    {
+      "epoch": 0.6886257889160179,
+      "grad_norm": 0.9584686956627472,
+      "learning_rate": 4.6687192992643515e-05,
+      "loss": 1.5271,
+      "step": 14948
+    },
+    {
+      "epoch": 0.6886718570046528,
+      "grad_norm": 0.9188703018866489,
+      "learning_rate": 4.667457000064004e-05,
+      "loss": 1.2481,
+      "step": 14949
+    },
+    {
+      "epoch": 0.6887179250932879,
+      "grad_norm": 0.7190781781027508,
+      "learning_rate": 4.666194819583579e-05,
+      "loss": 1.2047,
+      "step": 14950
+    },
+    {
+      "epoch": 0.6887639931819228,
+      "grad_norm": 0.9327529564248664,
+      "learning_rate": 4.6649327578511815e-05,
+      "loss": 1.4096,
+      "step": 14951
+    },
+    {
+      "epoch": 0.6888100612705579,
+      "grad_norm": 0.792222287159354,
+      "learning_rate": 4.663670814894908e-05,
+      "loss": 1.5495,
+      "step": 14952
+    },
+    {
+      "epoch": 0.6888561293591929,
+      "grad_norm": 3.9072294435787946,
+      "learning_rate": 4.662408990742858e-05,
+      "loss": 1.1057,
+      "step": 14953
+    },
+    {
+      "epoch": 0.6889021974478279,
+      "grad_norm": 1.0277052408707374,
+      "learning_rate": 4.661147285423112e-05,
+      "loss": 1.2096,
+      "step": 14954
+    },
+    {
+      "epoch": 0.6889482655364629,
+      "grad_norm": 0.9105231050713941,
+      "learning_rate": 4.659885698963776e-05,
+      "loss": 0.9482,
+      "step": 14955
+    },
+    {
+      "epoch": 0.6889943336250979,
+      "grad_norm": 0.8788276474909976,
+      "learning_rate": 4.6586242313929275e-05,
+      "loss": 1.007,
+      "step": 14956
+    },
+    {
+      "epoch": 0.6890404017137329,
+      "grad_norm": 1.2222740248924404,
+      "learning_rate": 4.6573628827386486e-05,
+      "loss": 1.2392,
+      "step": 14957
+    },
+    {
+      "epoch": 0.6890864698023679,
+      "grad_norm": 1.189958033214866,
+      "learning_rate": 4.656101653029025e-05,
+      "loss": 1.3376,
+      "step": 14958
+    },
+    {
+      "epoch": 0.6891325378910029,
+      "grad_norm": 0.7693294762714554,
+      "learning_rate": 4.654840542292135e-05,
+      "loss": 1.3393,
+      "step": 14959
+    },
+    {
+      "epoch": 0.6891786059796379,
+      "grad_norm": 4.0100899885812735,
+      "learning_rate": 4.653579550556059e-05,
+      "loss": 1.3638,
+      "step": 14960
+    },
+    {
+      "epoch": 0.6892246740682729,
+      "grad_norm": 0.7749310336597054,
+      "learning_rate": 4.6523186778488626e-05,
+      "loss": 1.1434,
+      "step": 14961
+    },
+    {
+      "epoch": 0.6892707421569079,
+      "grad_norm": 0.6688427689534726,
+      "learning_rate": 4.651057924198622e-05,
+      "loss": 1.2742,
+      "step": 14962
+    },
+    {
+      "epoch": 0.6893168102455429,
+      "grad_norm": 1.4358202840133119,
+      "learning_rate": 4.649797289633405e-05,
+      "loss": 1.3901,
+      "step": 14963
+    },
+    {
+      "epoch": 0.6893628783341779,
+      "grad_norm": 1.0416192300288878,
+      "learning_rate": 4.6485367741812824e-05,
+      "loss": 1.2568,
+      "step": 14964
+    },
+    {
+      "epoch": 0.6894089464228129,
+      "grad_norm": 1.2896864254327836,
+      "learning_rate": 4.647276377870309e-05,
+      "loss": 1.2534,
+      "step": 14965
+    },
+    {
+      "epoch": 0.689455014511448,
+      "grad_norm": 2.4994551214391616,
+      "learning_rate": 4.6460161007285486e-05,
+      "loss": 1.1013,
+      "step": 14966
+    },
+    {
+      "epoch": 0.6895010826000829,
+      "grad_norm": 2.0832297420560844,
+      "learning_rate": 4.644755942784064e-05,
+      "loss": 1.3941,
+      "step": 14967
+    },
+    {
+      "epoch": 0.689547150688718,
+      "grad_norm": 0.779868407325636,
+      "learning_rate": 4.643495904064901e-05,
+      "loss": 1.4014,
+      "step": 14968
+    },
+    {
+      "epoch": 0.6895932187773529,
+      "grad_norm": 0.9410115389197533,
+      "learning_rate": 4.6422359845991184e-05,
+      "loss": 1.2743,
+      "step": 14969
+    },
+    {
+      "epoch": 0.689639286865988,
+      "grad_norm": 1.4119912072528504,
+      "learning_rate": 4.640976184414764e-05,
+      "loss": 1.2129,
+      "step": 14970
+    },
+    {
+      "epoch": 0.6896853549546229,
+      "grad_norm": 1.27733111913193,
+      "learning_rate": 4.6397165035398904e-05,
+      "loss": 1.0966,
+      "step": 14971
+    },
+    {
+      "epoch": 0.689731423043258,
+      "grad_norm": 1.7346104552355543,
+      "learning_rate": 4.6384569420025334e-05,
+      "loss": 1.2312,
+      "step": 14972
+    },
+    {
+      "epoch": 0.6897774911318929,
+      "grad_norm": 0.9109909244700218,
+      "learning_rate": 4.637197499830741e-05,
+      "loss": 1.3731,
+      "step": 14973
+    },
+    {
+      "epoch": 0.689823559220528,
+      "grad_norm": 0.9454198870631142,
+      "learning_rate": 4.635938177052555e-05,
+      "loss": 1.3624,
+      "step": 14974
+    },
+    {
+      "epoch": 0.6898696273091629,
+      "grad_norm": 1.1929074668845217,
+      "learning_rate": 4.634678973696003e-05,
+      "loss": 1.4924,
+      "step": 14975
+    },
+    {
+      "epoch": 0.689915695397798,
+      "grad_norm": 2.03914600305857,
+      "learning_rate": 4.6334198897891234e-05,
+      "loss": 1.3151,
+      "step": 14976
+    },
+    {
+      "epoch": 0.6899617634864329,
+      "grad_norm": 5.013818737220303,
+      "learning_rate": 4.6321609253599494e-05,
+      "loss": 1.0719,
+      "step": 14977
+    },
+    {
+      "epoch": 0.690007831575068,
+      "grad_norm": 0.8194320889234155,
+      "learning_rate": 4.630902080436512e-05,
+      "loss": 1.2805,
+      "step": 14978
+    },
+    {
+      "epoch": 0.690053899663703,
+      "grad_norm": 1.2794767427581701,
+      "learning_rate": 4.629643355046831e-05,
+      "loss": 1.3435,
+      "step": 14979
+    },
+    {
+      "epoch": 0.690099967752338,
+      "grad_norm": 0.9698318785787725,
+      "learning_rate": 4.628384749218931e-05,
+      "loss": 1.2523,
+      "step": 14980
+    },
+    {
+      "epoch": 0.690146035840973,
+      "grad_norm": 1.9733049502358628,
+      "learning_rate": 4.6271262629808354e-05,
+      "loss": 1.3541,
+      "step": 14981
+    },
+    {
+      "epoch": 0.690192103929608,
+      "grad_norm": 0.9417243751784635,
+      "learning_rate": 4.625867896360565e-05,
+      "loss": 1.3181,
+      "step": 14982
+    },
+    {
+      "epoch": 0.690238172018243,
+      "grad_norm": 1.0739485786878178,
+      "learning_rate": 4.6246096493861266e-05,
+      "loss": 1.2096,
+      "step": 14983
+    },
+    {
+      "epoch": 0.6902842401068779,
+      "grad_norm": 0.8936935059887372,
+      "learning_rate": 4.6233515220855376e-05,
+      "loss": 1.2282,
+      "step": 14984
+    },
+    {
+      "epoch": 0.690330308195513,
+      "grad_norm": 0.9942290120799262,
+      "learning_rate": 4.622093514486812e-05,
+      "loss": 1.4241,
+      "step": 14985
+    },
+    {
+      "epoch": 0.6903763762841479,
+      "grad_norm": 0.9548374698618673,
+      "learning_rate": 4.620835626617949e-05,
+      "loss": 1.0851,
+      "step": 14986
+    },
+    {
+      "epoch": 0.690422444372783,
+      "grad_norm": 1.7931106297742063,
+      "learning_rate": 4.619577858506958e-05,
+      "loss": 1.2297,
+      "step": 14987
+    },
+    {
+      "epoch": 0.6904685124614179,
+      "grad_norm": 1.3276598229227503,
+      "learning_rate": 4.61832021018184e-05,
+      "loss": 1.1034,
+      "step": 14988
+    },
+    {
+      "epoch": 0.690514580550053,
+      "grad_norm": 0.9035367274897286,
+      "learning_rate": 4.6170626816705996e-05,
+      "loss": 1.3905,
+      "step": 14989
+    },
+    {
+      "epoch": 0.6905606486386879,
+      "grad_norm": 1.490208421160709,
+      "learning_rate": 4.6158052730012245e-05,
+      "loss": 1.1339,
+      "step": 14990
+    },
+    {
+      "epoch": 0.690606716727323,
+      "grad_norm": 2.336941350361134,
+      "learning_rate": 4.6145479842017126e-05,
+      "loss": 1.0035,
+      "step": 14991
+    },
+    {
+      "epoch": 0.6906527848159579,
+      "grad_norm": 0.9978940690302496,
+      "learning_rate": 4.61329081530006e-05,
+      "loss": 1.1786,
+      "step": 14992
+    },
+    {
+      "epoch": 0.690698852904593,
+      "grad_norm": 0.9471902665907035,
+      "learning_rate": 4.612033766324247e-05,
+      "loss": 1.3894,
+      "step": 14993
+    },
+    {
+      "epoch": 0.690744920993228,
+      "grad_norm": 0.708995786113125,
+      "learning_rate": 4.6107768373022656e-05,
+      "loss": 1.33,
+      "step": 14994
+    },
+    {
+      "epoch": 0.690790989081863,
+      "grad_norm": 1.0650593751594613,
+      "learning_rate": 4.609520028262095e-05,
+      "loss": 1.4012,
+      "step": 14995
+    },
+    {
+      "epoch": 0.690837057170498,
+      "grad_norm": 0.9744886803355353,
+      "learning_rate": 4.608263339231723e-05,
+      "loss": 1.3578,
+      "step": 14996
+    },
+    {
+      "epoch": 0.690883125259133,
+      "grad_norm": 0.8877680764894826,
+      "learning_rate": 4.6070067702391197e-05,
+      "loss": 1.3521,
+      "step": 14997
+    },
+    {
+      "epoch": 0.690929193347768,
+      "grad_norm": 1.424275470222331,
+      "learning_rate": 4.605750321312263e-05,
+      "loss": 1.4462,
+      "step": 14998
+    },
+    {
+      "epoch": 0.690975261436403,
+      "grad_norm": 1.3439665152807863,
+      "learning_rate": 4.604493992479127e-05,
+      "loss": 1.2805,
+      "step": 14999
+    },
+    {
+      "epoch": 0.691021329525038,
+      "grad_norm": 0.8764427638649156,
+      "learning_rate": 4.603237783767683e-05,
+      "loss": 1.4611,
+      "step": 15000
+    },
+    {
+      "epoch": 0.691067397613673,
+      "grad_norm": 0.726562509135377,
+      "learning_rate": 4.6019816952058934e-05,
+      "loss": 1.6445,
+      "step": 15001
+    },
+    {
+      "epoch": 0.691113465702308,
+      "grad_norm": 1.2073227799453226,
+      "learning_rate": 4.6007257268217263e-05,
+      "loss": 1.3752,
+      "step": 15002
+    },
+    {
+      "epoch": 0.691159533790943,
+      "grad_norm": 1.6909129663183073,
+      "learning_rate": 4.599469878643147e-05,
+      "loss": 1.433,
+      "step": 15003
+    },
+    {
+      "epoch": 0.691205601879578,
+      "grad_norm": 3.9075132298618818,
+      "learning_rate": 4.598214150698107e-05,
+      "loss": 1.2243,
+      "step": 15004
+    },
+    {
+      "epoch": 0.691251669968213,
+      "grad_norm": 1.6497243589655801,
+      "learning_rate": 4.596958543014566e-05,
+      "loss": 1.1906,
+      "step": 15005
+    },
+    {
+      "epoch": 0.691297738056848,
+      "grad_norm": 1.9652918186139616,
+      "learning_rate": 4.59570305562048e-05,
+      "loss": 0.8353,
+      "step": 15006
+    },
+    {
+      "epoch": 0.6913438061454831,
+      "grad_norm": 1.0070977323420576,
+      "learning_rate": 4.594447688543801e-05,
+      "loss": 1.3219,
+      "step": 15007
+    },
+    {
+      "epoch": 0.691389874234118,
+      "grad_norm": 0.8369335551136645,
+      "learning_rate": 4.593192441812474e-05,
+      "loss": 1.426,
+      "step": 15008
+    },
+    {
+      "epoch": 0.6914359423227531,
+      "grad_norm": 0.8072739562224308,
+      "learning_rate": 4.5919373154544453e-05,
+      "loss": 1.3544,
+      "step": 15009
+    },
+    {
+      "epoch": 0.691482010411388,
+      "grad_norm": 0.8583875121539697,
+      "learning_rate": 4.5906823094976634e-05,
+      "loss": 1.0843,
+      "step": 15010
+    },
+    {
+      "epoch": 0.6915280785000231,
+      "grad_norm": 1.9069413997371787,
+      "learning_rate": 4.589427423970062e-05,
+      "loss": 1.346,
+      "step": 15011
+    },
+    {
+      "epoch": 0.691574146588658,
+      "grad_norm": 2.0770293528840136,
+      "learning_rate": 4.588172658899581e-05,
+      "loss": 1.0548,
+      "step": 15012
+    },
+    {
+      "epoch": 0.6916202146772931,
+      "grad_norm": 2.6310182767578842,
+      "learning_rate": 4.5869180143141565e-05,
+      "loss": 1.2788,
+      "step": 15013
+    },
+    {
+      "epoch": 0.691666282765928,
+      "grad_norm": 0.9146213425845101,
+      "learning_rate": 4.585663490241725e-05,
+      "loss": 1.2538,
+      "step": 15014
+    },
+    {
+      "epoch": 0.6917123508545631,
+      "grad_norm": 0.8155006440838484,
+      "learning_rate": 4.584409086710209e-05,
+      "loss": 1.2846,
+      "step": 15015
+    },
+    {
+      "epoch": 0.691758418943198,
+      "grad_norm": 1.3731070513961638,
+      "learning_rate": 4.583154803747538e-05,
+      "loss": 1.2294,
+      "step": 15016
+    },
+    {
+      "epoch": 0.6918044870318331,
+      "grad_norm": 0.9302664539330991,
+      "learning_rate": 4.581900641381638e-05,
+      "loss": 1.5669,
+      "step": 15017
+    },
+    {
+      "epoch": 0.691850555120468,
+      "grad_norm": 1.0037482235823556,
+      "learning_rate": 4.5806465996404335e-05,
+      "loss": 1.2033,
+      "step": 15018
+    },
+    {
+      "epoch": 0.6918966232091031,
+      "grad_norm": 1.7225757725123665,
+      "learning_rate": 4.5793926785518406e-05,
+      "loss": 1.019,
+      "step": 15019
+    },
+    {
+      "epoch": 0.6919426912977381,
+      "grad_norm": 1.2583005986958375,
+      "learning_rate": 4.5781388781437675e-05,
+      "loss": 1.4317,
+      "step": 15020
+    },
+    {
+      "epoch": 0.691988759386373,
+      "grad_norm": 0.8844139337353114,
+      "learning_rate": 4.576885198444143e-05,
+      "loss": 1.339,
+      "step": 15021
+    },
+    {
+      "epoch": 0.6920348274750081,
+      "grad_norm": 0.8647984556135649,
+      "learning_rate": 4.5756316394808675e-05,
+      "loss": 1.3552,
+      "step": 15022
+    },
+    {
+      "epoch": 0.692080895563643,
+      "grad_norm": 0.9117773685440705,
+      "learning_rate": 4.574378201281857e-05,
+      "loss": 1.1929,
+      "step": 15023
+    },
+    {
+      "epoch": 0.6921269636522781,
+      "grad_norm": 1.1489941745420038,
+      "learning_rate": 4.5731248838750054e-05,
+      "loss": 1.2489,
+      "step": 15024
+    },
+    {
+      "epoch": 0.692173031740913,
+      "grad_norm": 1.3885876151886178,
+      "learning_rate": 4.5718716872882296e-05,
+      "loss": 1.263,
+      "step": 15025
+    },
+    {
+      "epoch": 0.6922190998295481,
+      "grad_norm": 1.08318413223063,
+      "learning_rate": 4.570618611549421e-05,
+      "loss": 1.1396,
+      "step": 15026
+    },
+    {
+      "epoch": 0.692265167918183,
+      "grad_norm": 1.3285101802175006,
+      "learning_rate": 4.5693656566864785e-05,
+      "loss": 1.2331,
+      "step": 15027
+    },
+    {
+      "epoch": 0.6923112360068181,
+      "grad_norm": 1.023293800637201,
+      "learning_rate": 4.568112822727303e-05,
+      "loss": 1.1213,
+      "step": 15028
+    },
+    {
+      "epoch": 0.692357304095453,
+      "grad_norm": 0.8687357515168828,
+      "learning_rate": 4.5668601096997776e-05,
+      "loss": 1.4001,
+      "step": 15029
+    },
+    {
+      "epoch": 0.6924033721840881,
+      "grad_norm": 1.1518830019329047,
+      "learning_rate": 4.565607517631799e-05,
+      "loss": 1.2081,
+      "step": 15030
+    },
+    {
+      "epoch": 0.692449440272723,
+      "grad_norm": 0.70836188550128,
+      "learning_rate": 4.564355046551244e-05,
+      "loss": 1.3431,
+      "step": 15031
+    },
+    {
+      "epoch": 0.6924955083613581,
+      "grad_norm": 0.9067421666246921,
+      "learning_rate": 4.563102696486011e-05,
+      "loss": 1.3227,
+      "step": 15032
+    },
+    {
+      "epoch": 0.692541576449993,
+      "grad_norm": 1.2060230222808836,
+      "learning_rate": 4.5618504674639714e-05,
+      "loss": 0.9939,
+      "step": 15033
+    },
+    {
+      "epoch": 0.6925876445386281,
+      "grad_norm": 1.1693195786724848,
+      "learning_rate": 4.5605983595130096e-05,
+      "loss": 1.3887,
+      "step": 15034
+    },
+    {
+      "epoch": 0.6926337126272631,
+      "grad_norm": 1.301573852119809,
+      "learning_rate": 4.559346372660992e-05,
+      "loss": 1.3683,
+      "step": 15035
+    },
+    {
+      "epoch": 0.6926797807158981,
+      "grad_norm": 1.0604692834010125,
+      "learning_rate": 4.558094506935806e-05,
+      "loss": 1.4602,
+      "step": 15036
+    },
+    {
+      "epoch": 0.6927258488045331,
+      "grad_norm": 1.1254949885563956,
+      "learning_rate": 4.5568427623653143e-05,
+      "loss": 1.0748,
+      "step": 15037
+    },
+    {
+      "epoch": 0.6927719168931681,
+      "grad_norm": 0.876840871794262,
+      "learning_rate": 4.555591138977378e-05,
+      "loss": 1.1262,
+      "step": 15038
+    },
+    {
+      "epoch": 0.6928179849818031,
+      "grad_norm": 0.9836071965308215,
+      "learning_rate": 4.554339636799878e-05,
+      "loss": 1.0728,
+      "step": 15039
+    },
+    {
+      "epoch": 0.6928640530704381,
+      "grad_norm": 1.0261619507291504,
+      "learning_rate": 4.553088255860665e-05,
+      "loss": 1.6068,
+      "step": 15040
+    },
+    {
+      "epoch": 0.6929101211590731,
+      "grad_norm": 1.7610092094160334,
+      "learning_rate": 4.551836996187606e-05,
+      "loss": 1.5059,
+      "step": 15041
+    },
+    {
+      "epoch": 0.6929561892477081,
+      "grad_norm": 1.4907330097264944,
+      "learning_rate": 4.550585857808547e-05,
+      "loss": 1.3179,
+      "step": 15042
+    },
+    {
+      "epoch": 0.6930022573363431,
+      "grad_norm": 1.210866918342224,
+      "learning_rate": 4.549334840751358e-05,
+      "loss": 1.07,
+      "step": 15043
+    },
+    {
+      "epoch": 0.6930483254249781,
+      "grad_norm": 1.0898325147563384,
+      "learning_rate": 4.5480839450438786e-05,
+      "loss": 1.2267,
+      "step": 15044
+    },
+    {
+      "epoch": 0.6930943935136131,
+      "grad_norm": 1.2148593518254682,
+      "learning_rate": 4.546833170713965e-05,
+      "loss": 1.4856,
+      "step": 15045
+    },
+    {
+      "epoch": 0.6931404616022481,
+      "grad_norm": 1.9427976426595408,
+      "learning_rate": 4.545582517789459e-05,
+      "loss": 1.1342,
+      "step": 15046
+    },
+    {
+      "epoch": 0.6931865296908831,
+      "grad_norm": 1.1090137929213828,
+      "learning_rate": 4.544331986298205e-05,
+      "loss": 1.494,
+      "step": 15047
+    },
+    {
+      "epoch": 0.6932325977795182,
+      "grad_norm": 0.8571997388075377,
+      "learning_rate": 4.543081576268049e-05,
+      "loss": 1.5297,
+      "step": 15048
+    },
+    {
+      "epoch": 0.6932786658681531,
+      "grad_norm": 1.1249072705602103,
+      "learning_rate": 4.541831287726821e-05,
+      "loss": 1.3422,
+      "step": 15049
+    },
+    {
+      "epoch": 0.6933247339567882,
+      "grad_norm": 0.9013975681061943,
+      "learning_rate": 4.540581120702361e-05,
+      "loss": 1.4526,
+      "step": 15050
+    },
+    {
+      "epoch": 0.6933708020454231,
+      "grad_norm": 0.8550731940678378,
+      "learning_rate": 4.5393310752225e-05,
+      "loss": 1.1658,
+      "step": 15051
+    },
+    {
+      "epoch": 0.6934168701340582,
+      "grad_norm": 0.7553775724137188,
+      "learning_rate": 4.538081151315075e-05,
+      "loss": 1.4209,
+      "step": 15052
+    },
+    {
+      "epoch": 0.6934629382226931,
+      "grad_norm": 0.7480601013479894,
+      "learning_rate": 4.536831349007899e-05,
+      "loss": 1.2829,
+      "step": 15053
+    },
+    {
+      "epoch": 0.6935090063113282,
+      "grad_norm": 0.9650783419864859,
+      "learning_rate": 4.5355816683288144e-05,
+      "loss": 1.2864,
+      "step": 15054
+    },
+    {
+      "epoch": 0.6935550743999631,
+      "grad_norm": 5.380174032588154,
+      "learning_rate": 4.534332109305635e-05,
+      "loss": 1.1111,
+      "step": 15055
+    },
+    {
+      "epoch": 0.6936011424885982,
+      "grad_norm": 0.7126112662614347,
+      "learning_rate": 4.5330826719661744e-05,
+      "loss": 1.1347,
+      "step": 15056
+    },
+    {
+      "epoch": 0.6936472105772331,
+      "grad_norm": 0.9929985558492995,
+      "learning_rate": 4.5318333563382545e-05,
+      "loss": 1.4074,
+      "step": 15057
+    },
+    {
+      "epoch": 0.6936932786658682,
+      "grad_norm": 1.049659071858152,
+      "learning_rate": 4.5305841624496895e-05,
+      "loss": 1.4817,
+      "step": 15058
+    },
+    {
+      "epoch": 0.6937393467545031,
+      "grad_norm": 0.8517639060860585,
+      "learning_rate": 4.529335090328294e-05,
+      "loss": 1.2768,
+      "step": 15059
+    },
+    {
+      "epoch": 0.6937854148431382,
+      "grad_norm": 1.2010308436580146,
+      "learning_rate": 4.5280861400018684e-05,
+      "loss": 1.5099,
+      "step": 15060
+    },
+    {
+      "epoch": 0.6938314829317732,
+      "grad_norm": 1.4446508921207284,
+      "learning_rate": 4.526837311498223e-05,
+      "loss": 1.4743,
+      "step": 15061
+    },
+    {
+      "epoch": 0.6938775510204082,
+      "grad_norm": 1.0698362474014627,
+      "learning_rate": 4.52558860484516e-05,
+      "loss": 1.4036,
+      "step": 15062
+    },
+    {
+      "epoch": 0.6939236191090432,
+      "grad_norm": 5.156641620804453,
+      "learning_rate": 4.524340020070483e-05,
+      "loss": 0.9915,
+      "step": 15063
+    },
+    {
+      "epoch": 0.6939696871976782,
+      "grad_norm": 1.0502512481684487,
+      "learning_rate": 4.523091557201985e-05,
+      "loss": 1.0875,
+      "step": 15064
+    },
+    {
+      "epoch": 0.6940157552863132,
+      "grad_norm": 0.8005858482036955,
+      "learning_rate": 4.5218432162674603e-05,
+      "loss": 1.4082,
+      "step": 15065
+    },
+    {
+      "epoch": 0.6940618233749482,
+      "grad_norm": 0.9274972954627185,
+      "learning_rate": 4.520594997294709e-05,
+      "loss": 1.4399,
+      "step": 15066
+    },
+    {
+      "epoch": 0.6941078914635832,
+      "grad_norm": 1.0093961523221004,
+      "learning_rate": 4.5193469003115105e-05,
+      "loss": 1.3239,
+      "step": 15067
+    },
+    {
+      "epoch": 0.6941539595522181,
+      "grad_norm": 0.778162559462747,
+      "learning_rate": 4.518098925345656e-05,
+      "loss": 1.2498,
+      "step": 15068
+    },
+    {
+      "epoch": 0.6942000276408532,
+      "grad_norm": 1.1208043965067198,
+      "learning_rate": 4.5168510724249294e-05,
+      "loss": 1.2436,
+      "step": 15069
+    },
+    {
+      "epoch": 0.6942460957294881,
+      "grad_norm": 0.9775258988908198,
+      "learning_rate": 4.515603341577116e-05,
+      "loss": 1.4197,
+      "step": 15070
+    },
+    {
+      "epoch": 0.6942921638181232,
+      "grad_norm": 1.9818719885055915,
+      "learning_rate": 4.514355732829987e-05,
+      "loss": 1.4732,
+      "step": 15071
+    },
+    {
+      "epoch": 0.6943382319067581,
+      "grad_norm": 0.754428157482089,
+      "learning_rate": 4.513108246211321e-05,
+      "loss": 1.4417,
+      "step": 15072
+    },
+    {
+      "epoch": 0.6943842999953932,
+      "grad_norm": 0.7277845764030608,
+      "learning_rate": 4.511860881748896e-05,
+      "loss": 1.224,
+      "step": 15073
+    },
+    {
+      "epoch": 0.6944303680840282,
+      "grad_norm": 1.1606429031612768,
+      "learning_rate": 4.5106136394704746e-05,
+      "loss": 1.0371,
+      "step": 15074
+    },
+    {
+      "epoch": 0.6944764361726632,
+      "grad_norm": 1.654763766604842,
+      "learning_rate": 4.509366519403828e-05,
+      "loss": 1.246,
+      "step": 15075
+    },
+    {
+      "epoch": 0.6945225042612982,
+      "grad_norm": 1.6388184262374257,
+      "learning_rate": 4.508119521576723e-05,
+      "loss": 0.9643,
+      "step": 15076
+    },
+    {
+      "epoch": 0.6945685723499332,
+      "grad_norm": 0.7547843645589758,
+      "learning_rate": 4.506872646016922e-05,
+      "loss": 1.3812,
+      "step": 15077
+    },
+    {
+      "epoch": 0.6946146404385682,
+      "grad_norm": 1.2534783264873972,
+      "learning_rate": 4.50562589275218e-05,
+      "loss": 1.363,
+      "step": 15078
+    },
+    {
+      "epoch": 0.6946607085272032,
+      "grad_norm": 7.383650987101674,
+      "learning_rate": 4.5043792618102556e-05,
+      "loss": 1.2036,
+      "step": 15079
+    },
+    {
+      "epoch": 0.6947067766158382,
+      "grad_norm": 1.099362680432036,
+      "learning_rate": 4.503132753218905e-05,
+      "loss": 1.1403,
+      "step": 15080
+    },
+    {
+      "epoch": 0.6947528447044732,
+      "grad_norm": 1.536703678448733,
+      "learning_rate": 4.501886367005882e-05,
+      "loss": 1.1248,
+      "step": 15081
+    },
+    {
+      "epoch": 0.6947989127931082,
+      "grad_norm": 0.9908349483636167,
+      "learning_rate": 4.500640103198929e-05,
+      "loss": 1.3295,
+      "step": 15082
+    },
+    {
+      "epoch": 0.6948449808817432,
+      "grad_norm": 0.8363745942250629,
+      "learning_rate": 4.499393961825793e-05,
+      "loss": 1.309,
+      "step": 15083
+    },
+    {
+      "epoch": 0.6948910489703782,
+      "grad_norm": 0.6968923236720215,
+      "learning_rate": 4.4981479429142235e-05,
+      "loss": 1.1558,
+      "step": 15084
+    },
+    {
+      "epoch": 0.6949371170590132,
+      "grad_norm": 0.7154802327736407,
+      "learning_rate": 4.496902046491953e-05,
+      "loss": 1.5227,
+      "step": 15085
+    },
+    {
+      "epoch": 0.6949831851476482,
+      "grad_norm": 0.8946830174527918,
+      "learning_rate": 4.495656272586723e-05,
+      "loss": 1.3239,
+      "step": 15086
+    },
+    {
+      "epoch": 0.6950292532362832,
+      "grad_norm": 0.9331932236618353,
+      "learning_rate": 4.4944106212262674e-05,
+      "loss": 1.2325,
+      "step": 15087
+    },
+    {
+      "epoch": 0.6950753213249182,
+      "grad_norm": 0.9053439615029948,
+      "learning_rate": 4.493165092438323e-05,
+      "loss": 1.3599,
+      "step": 15088
+    },
+    {
+      "epoch": 0.6951213894135533,
+      "grad_norm": 1.0433064098834384,
+      "learning_rate": 4.4919196862506116e-05,
+      "loss": 1.2306,
+      "step": 15089
+    },
+    {
+      "epoch": 0.6951674575021882,
+      "grad_norm": 1.1894811526040832,
+      "learning_rate": 4.490674402690864e-05,
+      "loss": 1.3016,
+      "step": 15090
+    },
+    {
+      "epoch": 0.6952135255908233,
+      "grad_norm": 1.103472645839164,
+      "learning_rate": 4.489429241786806e-05,
+      "loss": 1.1015,
+      "step": 15091
+    },
+    {
+      "epoch": 0.6952595936794582,
+      "grad_norm": 0.7453997695789623,
+      "learning_rate": 4.4881842035661594e-05,
+      "loss": 1.4071,
+      "step": 15092
+    },
+    {
+      "epoch": 0.6953056617680933,
+      "grad_norm": 0.9348558295323627,
+      "learning_rate": 4.486939288056638e-05,
+      "loss": 1.1813,
+      "step": 15093
+    },
+    {
+      "epoch": 0.6953517298567282,
+      "grad_norm": 3.474341869253531,
+      "learning_rate": 4.485694495285959e-05,
+      "loss": 1.2279,
+      "step": 15094
+    },
+    {
+      "epoch": 0.6953977979453633,
+      "grad_norm": 1.394375979991196,
+      "learning_rate": 4.4844498252818425e-05,
+      "loss": 1.2238,
+      "step": 15095
+    },
+    {
+      "epoch": 0.6954438660339982,
+      "grad_norm": 1.660692477951301,
+      "learning_rate": 4.483205278071988e-05,
+      "loss": 1.1056,
+      "step": 15096
+    },
+    {
+      "epoch": 0.6954899341226333,
+      "grad_norm": 0.9015743640162365,
+      "learning_rate": 4.481960853684111e-05,
+      "loss": 1.2242,
+      "step": 15097
+    },
+    {
+      "epoch": 0.6955360022112682,
+      "grad_norm": 0.7964004880855288,
+      "learning_rate": 4.480716552145913e-05,
+      "loss": 1.3573,
+      "step": 15098
+    },
+    {
+      "epoch": 0.6955820702999033,
+      "grad_norm": 0.8072670248370203,
+      "learning_rate": 4.4794723734851005e-05,
+      "loss": 1.3396,
+      "step": 15099
+    },
+    {
+      "epoch": 0.6956281383885382,
+      "grad_norm": 1.3340632507397343,
+      "learning_rate": 4.478228317729366e-05,
+      "loss": 1.4199,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6956742064771733,
+      "grad_norm": 1.0618884618792848,
+      "learning_rate": 4.476984384906412e-05,
+      "loss": 1.2097,
+      "step": 15101
+    },
+    {
+      "epoch": 0.6957202745658083,
+      "grad_norm": 1.1637488491250771,
+      "learning_rate": 4.475740575043933e-05,
+      "loss": 1.7279,
+      "step": 15102
+    },
+    {
+      "epoch": 0.6957663426544433,
+      "grad_norm": 0.883715435537316,
+      "learning_rate": 4.474496888169615e-05,
+      "loss": 1.3069,
+      "step": 15103
+    },
+    {
+      "epoch": 0.6958124107430783,
+      "grad_norm": 1.043642653584548,
+      "learning_rate": 4.473253324311148e-05,
+      "loss": 1.2237,
+      "step": 15104
+    },
+    {
+      "epoch": 0.6958584788317133,
+      "grad_norm": 1.5201113100892145,
+      "learning_rate": 4.472009883496222e-05,
+      "loss": 1.3056,
+      "step": 15105
+    },
+    {
+      "epoch": 0.6959045469203483,
+      "grad_norm": 1.076241094196365,
+      "learning_rate": 4.47076656575252e-05,
+      "loss": 1.1711,
+      "step": 15106
+    },
+    {
+      "epoch": 0.6959506150089833,
+      "grad_norm": 1.067799260003625,
+      "learning_rate": 4.4695233711077155e-05,
+      "loss": 1.0805,
+      "step": 15107
+    },
+    {
+      "epoch": 0.6959966830976183,
+      "grad_norm": 0.7501917556946773,
+      "learning_rate": 4.468280299589491e-05,
+      "loss": 1.4107,
+      "step": 15108
+    },
+    {
+      "epoch": 0.6960427511862533,
+      "grad_norm": 0.9510445145646609,
+      "learning_rate": 4.46703735122552e-05,
+      "loss": 1.0549,
+      "step": 15109
+    },
+    {
+      "epoch": 0.6960888192748883,
+      "grad_norm": 1.254278390622919,
+      "learning_rate": 4.465794526043481e-05,
+      "loss": 1.0414,
+      "step": 15110
+    },
+    {
+      "epoch": 0.6961348873635232,
+      "grad_norm": 0.7396442016248119,
+      "learning_rate": 4.4645518240710325e-05,
+      "loss": 1.4438,
+      "step": 15111
+    },
+    {
+      "epoch": 0.6961809554521583,
+      "grad_norm": 1.10476503079977,
+      "learning_rate": 4.4633092453358464e-05,
+      "loss": 1.1295,
+      "step": 15112
+    },
+    {
+      "epoch": 0.6962270235407932,
+      "grad_norm": 3.849722503101912,
+      "learning_rate": 4.462066789865591e-05,
+      "loss": 1.0755,
+      "step": 15113
+    },
+    {
+      "epoch": 0.6962730916294283,
+      "grad_norm": 0.9856248307340166,
+      "learning_rate": 4.4608244576879186e-05,
+      "loss": 1.2597,
+      "step": 15114
+    },
+    {
+      "epoch": 0.6963191597180634,
+      "grad_norm": 0.7811358391636213,
+      "learning_rate": 4.459582248830492e-05,
+      "loss": 1.5277,
+      "step": 15115
+    },
+    {
+      "epoch": 0.6963652278066983,
+      "grad_norm": 0.732215205815707,
+      "learning_rate": 4.458340163320968e-05,
+      "loss": 0.9914,
+      "step": 15116
+    },
+    {
+      "epoch": 0.6964112958953333,
+      "grad_norm": 1.368956487089115,
+      "learning_rate": 4.457098201187e-05,
+      "loss": 1.2363,
+      "step": 15117
+    },
+    {
+      "epoch": 0.6964573639839683,
+      "grad_norm": 0.807326358720916,
+      "learning_rate": 4.455856362456234e-05,
+      "loss": 1.2256,
+      "step": 15118
+    },
+    {
+      "epoch": 0.6965034320726033,
+      "grad_norm": 0.8492402136134546,
+      "learning_rate": 4.45461464715632e-05,
+      "loss": 1.4043,
+      "step": 15119
+    },
+    {
+      "epoch": 0.6965495001612383,
+      "grad_norm": 1.1622548296635193,
+      "learning_rate": 4.4533730553149065e-05,
+      "loss": 1.25,
+      "step": 15120
+    },
+    {
+      "epoch": 0.6965955682498733,
+      "grad_norm": 0.7379713413074851,
+      "learning_rate": 4.452131586959627e-05,
+      "loss": 1.6095,
+      "step": 15121
+    },
+    {
+      "epoch": 0.6966416363385083,
+      "grad_norm": 1.0819930369361765,
+      "learning_rate": 4.45089024211813e-05,
+      "loss": 0.9241,
+      "step": 15122
+    },
+    {
+      "epoch": 0.6966877044271433,
+      "grad_norm": 0.735959414116595,
+      "learning_rate": 4.4496490208180386e-05,
+      "loss": 1.2893,
+      "step": 15123
+    },
+    {
+      "epoch": 0.6967337725157783,
+      "grad_norm": 0.9681901312216981,
+      "learning_rate": 4.448407923087002e-05,
+      "loss": 1.1813,
+      "step": 15124
+    },
+    {
+      "epoch": 0.6967798406044133,
+      "grad_norm": 0.9952575845075787,
+      "learning_rate": 4.4471669489526414e-05,
+      "loss": 1.0416,
+      "step": 15125
+    },
+    {
+      "epoch": 0.6968259086930483,
+      "grad_norm": 0.8287878864506802,
+      "learning_rate": 4.445926098442588e-05,
+      "loss": 1.603,
+      "step": 15126
+    },
+    {
+      "epoch": 0.6968719767816833,
+      "grad_norm": 1.0018830566088515,
+      "learning_rate": 4.444685371584466e-05,
+      "loss": 1.3618,
+      "step": 15127
+    },
+    {
+      "epoch": 0.6969180448703183,
+      "grad_norm": 1.242378164917797,
+      "learning_rate": 4.443444768405902e-05,
+      "loss": 1.3617,
+      "step": 15128
+    },
+    {
+      "epoch": 0.6969641129589533,
+      "grad_norm": 0.80434908413368,
+      "learning_rate": 4.442204288934514e-05,
+      "loss": 1.3182,
+      "step": 15129
+    },
+    {
+      "epoch": 0.6970101810475884,
+      "grad_norm": 0.8850285765676326,
+      "learning_rate": 4.44096393319791e-05,
+      "loss": 1.1852,
+      "step": 15130
+    },
+    {
+      "epoch": 0.6970562491362233,
+      "grad_norm": 2.182513775932022,
+      "learning_rate": 4.4397237012237204e-05,
+      "loss": 1.2699,
+      "step": 15131
+    },
+    {
+      "epoch": 0.6971023172248584,
+      "grad_norm": 1.3285117083393885,
+      "learning_rate": 4.438483593039544e-05,
+      "loss": 1.2874,
+      "step": 15132
+    },
+    {
+      "epoch": 0.6971483853134933,
+      "grad_norm": 1.0605511376434722,
+      "learning_rate": 4.437243608672999e-05,
+      "loss": 1.1727,
+      "step": 15133
+    },
+    {
+      "epoch": 0.6971944534021284,
+      "grad_norm": 0.8727562375268937,
+      "learning_rate": 4.436003748151678e-05,
+      "loss": 1.6006,
+      "step": 15134
+    },
+    {
+      "epoch": 0.6972405214907633,
+      "grad_norm": 1.3501607434435015,
+      "learning_rate": 4.434764011503203e-05,
+      "loss": 1.329,
+      "step": 15135
+    },
+    {
+      "epoch": 0.6972865895793984,
+      "grad_norm": 1.0669058113162058,
+      "learning_rate": 4.43352439875516e-05,
+      "loss": 1.2262,
+      "step": 15136
+    },
+    {
+      "epoch": 0.6973326576680333,
+      "grad_norm": 0.8205356436350651,
+      "learning_rate": 4.432284909935151e-05,
+      "loss": 1.4918,
+      "step": 15137
+    },
+    {
+      "epoch": 0.6973787257566684,
+      "grad_norm": 1.1171868932099718,
+      "learning_rate": 4.431045545070777e-05,
+      "loss": 0.9507,
+      "step": 15138
+    },
+    {
+      "epoch": 0.6974247938453033,
+      "grad_norm": 1.005033544161071,
+      "learning_rate": 4.429806304189621e-05,
+      "loss": 1.451,
+      "step": 15139
+    },
+    {
+      "epoch": 0.6974708619339384,
+      "grad_norm": 1.5103431956800821,
+      "learning_rate": 4.4285671873192803e-05,
+      "loss": 1.2067,
+      "step": 15140
+    },
+    {
+      "epoch": 0.6975169300225733,
+      "grad_norm": 0.7571614147410545,
+      "learning_rate": 4.427328194487331e-05,
+      "loss": 1.4224,
+      "step": 15141
+    },
+    {
+      "epoch": 0.6975629981112084,
+      "grad_norm": 0.8260667152805149,
+      "learning_rate": 4.426089325721372e-05,
+      "loss": 1.4147,
+      "step": 15142
+    },
+    {
+      "epoch": 0.6976090661998434,
+      "grad_norm": 1.244165274187885,
+      "learning_rate": 4.424850581048974e-05,
+      "loss": 1.01,
+      "step": 15143
+    },
+    {
+      "epoch": 0.6976551342884784,
+      "grad_norm": 1.2265071030351073,
+      "learning_rate": 4.423611960497721e-05,
+      "loss": 1.3483,
+      "step": 15144
+    },
+    {
+      "epoch": 0.6977012023771134,
+      "grad_norm": 1.0557320189632227,
+      "learning_rate": 4.4223734640951784e-05,
+      "loss": 1.1683,
+      "step": 15145
+    },
+    {
+      "epoch": 0.6977472704657484,
+      "grad_norm": 0.8057443319929126,
+      "learning_rate": 4.421135091868935e-05,
+      "loss": 1.434,
+      "step": 15146
+    },
+    {
+      "epoch": 0.6977933385543834,
+      "grad_norm": 1.438733710085949,
+      "learning_rate": 4.4198968438465535e-05,
+      "loss": 1.4033,
+      "step": 15147
+    },
+    {
+      "epoch": 0.6978394066430184,
+      "grad_norm": 1.2208130465332252,
+      "learning_rate": 4.4186587200555975e-05,
+      "loss": 1.2471,
+      "step": 15148
+    },
+    {
+      "epoch": 0.6978854747316534,
+      "grad_norm": 0.9894978546285944,
+      "learning_rate": 4.417420720523634e-05,
+      "loss": 1.2988,
+      "step": 15149
+    },
+    {
+      "epoch": 0.6979315428202884,
+      "grad_norm": 0.8310001979689292,
+      "learning_rate": 4.416182845278227e-05,
+      "loss": 1.2233,
+      "step": 15150
+    },
+    {
+      "epoch": 0.6979776109089234,
+      "grad_norm": 0.8206197136000399,
+      "learning_rate": 4.414945094346938e-05,
+      "loss": 1.4134,
+      "step": 15151
+    },
+    {
+      "epoch": 0.6980236789975584,
+      "grad_norm": 1.3221497345494442,
+      "learning_rate": 4.4137074677573134e-05,
+      "loss": 0.9293,
+      "step": 15152
+    },
+    {
+      "epoch": 0.6980697470861934,
+      "grad_norm": 0.6305350867452199,
+      "learning_rate": 4.4124699655369204e-05,
+      "loss": 1.5155,
+      "step": 15153
+    },
+    {
+      "epoch": 0.6981158151748283,
+      "grad_norm": 0.7922437758864739,
+      "learning_rate": 4.4112325877133e-05,
+      "loss": 1.0842,
+      "step": 15154
+    },
+    {
+      "epoch": 0.6981618832634634,
+      "grad_norm": 0.9744234996024427,
+      "learning_rate": 4.409995334314005e-05,
+      "loss": 1.0656,
+      "step": 15155
+    },
+    {
+      "epoch": 0.6982079513520985,
+      "grad_norm": 1.122226121565158,
+      "learning_rate": 4.408758205366578e-05,
+      "loss": 1.1675,
+      "step": 15156
+    },
+    {
+      "epoch": 0.6982540194407334,
+      "grad_norm": 0.8674877760580848,
+      "learning_rate": 4.407521200898561e-05,
+      "loss": 1.4638,
+      "step": 15157
+    },
+    {
+      "epoch": 0.6983000875293685,
+      "grad_norm": 0.7307617188347207,
+      "learning_rate": 4.4062843209375e-05,
+      "loss": 1.0678,
+      "step": 15158
+    },
+    {
+      "epoch": 0.6983461556180034,
+      "grad_norm": 0.9213476463875148,
+      "learning_rate": 4.405047565510923e-05,
+      "loss": 1.2821,
+      "step": 15159
+    },
+    {
+      "epoch": 0.6983922237066384,
+      "grad_norm": 2.078360918847771,
+      "learning_rate": 4.4038109346463695e-05,
+      "loss": 1.047,
+      "step": 15160
+    },
+    {
+      "epoch": 0.6984382917952734,
+      "grad_norm": 0.8319895621173965,
+      "learning_rate": 4.40257442837137e-05,
+      "loss": 1.1029,
+      "step": 15161
+    },
+    {
+      "epoch": 0.6984843598839084,
+      "grad_norm": 0.9260557831197255,
+      "learning_rate": 4.401338046713457e-05,
+      "loss": 1.1901,
+      "step": 15162
+    },
+    {
+      "epoch": 0.6985304279725434,
+      "grad_norm": 1.2853142809838343,
+      "learning_rate": 4.400101789700144e-05,
+      "loss": 1.1972,
+      "step": 15163
+    },
+    {
+      "epoch": 0.6985764960611784,
+      "grad_norm": 0.8942149328017607,
+      "learning_rate": 4.398865657358971e-05,
+      "loss": 1.2557,
+      "step": 15164
+    },
+    {
+      "epoch": 0.6986225641498134,
+      "grad_norm": 0.937148787205365,
+      "learning_rate": 4.3976296497174516e-05,
+      "loss": 0.9422,
+      "step": 15165
+    },
+    {
+      "epoch": 0.6986686322384484,
+      "grad_norm": 1.0033222196283962,
+      "learning_rate": 4.3963937668030974e-05,
+      "loss": 1.1592,
+      "step": 15166
+    },
+    {
+      "epoch": 0.6987147003270834,
+      "grad_norm": 1.002653818830221,
+      "learning_rate": 4.395158008643428e-05,
+      "loss": 1.213,
+      "step": 15167
+    },
+    {
+      "epoch": 0.6987607684157184,
+      "grad_norm": 1.4318271750564615,
+      "learning_rate": 4.393922375265954e-05,
+      "loss": 1.4104,
+      "step": 15168
+    },
+    {
+      "epoch": 0.6988068365043534,
+      "grad_norm": 0.918974008629844,
+      "learning_rate": 4.392686866698191e-05,
+      "loss": 1.3128,
+      "step": 15169
+    },
+    {
+      "epoch": 0.6988529045929884,
+      "grad_norm": 1.6686397086865656,
+      "learning_rate": 4.391451482967637e-05,
+      "loss": 1.3226,
+      "step": 15170
+    },
+    {
+      "epoch": 0.6988989726816235,
+      "grad_norm": 0.9053741842692279,
+      "learning_rate": 4.3902162241017984e-05,
+      "loss": 1.2114,
+      "step": 15171
+    },
+    {
+      "epoch": 0.6989450407702584,
+      "grad_norm": 0.9352378449686586,
+      "learning_rate": 4.388981090128177e-05,
+      "loss": 1.3752,
+      "step": 15172
+    },
+    {
+      "epoch": 0.6989911088588935,
+      "grad_norm": 1.5049805556019964,
+      "learning_rate": 4.387746081074274e-05,
+      "loss": 0.9854,
+      "step": 15173
+    },
+    {
+      "epoch": 0.6990371769475284,
+      "grad_norm": 0.73848879353008,
+      "learning_rate": 4.386511196967579e-05,
+      "loss": 1.1026,
+      "step": 15174
+    },
+    {
+      "epoch": 0.6990832450361635,
+      "grad_norm": 1.0672977302322701,
+      "learning_rate": 4.385276437835587e-05,
+      "loss": 1.4667,
+      "step": 15175
+    },
+    {
+      "epoch": 0.6991293131247984,
+      "grad_norm": 1.1298444075554255,
+      "learning_rate": 4.3840418037057905e-05,
+      "loss": 1.4755,
+      "step": 15176
+    },
+    {
+      "epoch": 0.6991753812134335,
+      "grad_norm": 0.8692427199368491,
+      "learning_rate": 4.382807294605671e-05,
+      "loss": 1.4192,
+      "step": 15177
+    },
+    {
+      "epoch": 0.6992214493020684,
+      "grad_norm": 1.0100260017567508,
+      "learning_rate": 4.381572910562716e-05,
+      "loss": 1.0908,
+      "step": 15178
+    },
+    {
+      "epoch": 0.6992675173907035,
+      "grad_norm": 0.9341253840280661,
+      "learning_rate": 4.3803386516044066e-05,
+      "loss": 1.3345,
+      "step": 15179
+    },
+    {
+      "epoch": 0.6993135854793384,
+      "grad_norm": 1.2824862948404074,
+      "learning_rate": 4.3791045177582244e-05,
+      "loss": 1.3313,
+      "step": 15180
+    },
+    {
+      "epoch": 0.6993596535679735,
+      "grad_norm": 0.988885881596777,
+      "learning_rate": 4.377870509051639e-05,
+      "loss": 1.2361,
+      "step": 15181
+    },
+    {
+      "epoch": 0.6994057216566084,
+      "grad_norm": 0.9755995466496674,
+      "learning_rate": 4.376636625512127e-05,
+      "loss": 1.1669,
+      "step": 15182
+    },
+    {
+      "epoch": 0.6994517897452435,
+      "grad_norm": 0.9985841802121969,
+      "learning_rate": 4.375402867167162e-05,
+      "loss": 1.3672,
+      "step": 15183
+    },
+    {
+      "epoch": 0.6994978578338785,
+      "grad_norm": 0.925188515113427,
+      "learning_rate": 4.374169234044204e-05,
+      "loss": 1.2993,
+      "step": 15184
+    },
+    {
+      "epoch": 0.6995439259225135,
+      "grad_norm": 0.7711256374189407,
+      "learning_rate": 4.3729357261707225e-05,
+      "loss": 1.2613,
+      "step": 15185
+    },
+    {
+      "epoch": 0.6995899940111485,
+      "grad_norm": 0.9257812691512912,
+      "learning_rate": 4.371702343574178e-05,
+      "loss": 0.7974,
+      "step": 15186
+    },
+    {
+      "epoch": 0.6996360620997835,
+      "grad_norm": 0.7724657908302593,
+      "learning_rate": 4.370469086282034e-05,
+      "loss": 1.2308,
+      "step": 15187
+    },
+    {
+      "epoch": 0.6996821301884185,
+      "grad_norm": 0.7394993050417734,
+      "learning_rate": 4.369235954321741e-05,
+      "loss": 1.2166,
+      "step": 15188
+    },
+    {
+      "epoch": 0.6997281982770535,
+      "grad_norm": 1.1670756639860402,
+      "learning_rate": 4.368002947720753e-05,
+      "loss": 1.4163,
+      "step": 15189
+    },
+    {
+      "epoch": 0.6997742663656885,
+      "grad_norm": 0.8734615985961552,
+      "learning_rate": 4.3667700665065234e-05,
+      "loss": 1.2917,
+      "step": 15190
+    },
+    {
+      "epoch": 0.6998203344543235,
+      "grad_norm": 0.7672332208417464,
+      "learning_rate": 4.3655373107065034e-05,
+      "loss": 1.2272,
+      "step": 15191
+    },
+    {
+      "epoch": 0.6998664025429585,
+      "grad_norm": 1.6167451725853879,
+      "learning_rate": 4.36430468034813e-05,
+      "loss": 1.2789,
+      "step": 15192
+    },
+    {
+      "epoch": 0.6999124706315935,
+      "grad_norm": 0.801402319447201,
+      "learning_rate": 4.3630721754588496e-05,
+      "loss": 1.2776,
+      "step": 15193
+    },
+    {
+      "epoch": 0.6999585387202285,
+      "grad_norm": 0.9220993458116917,
+      "learning_rate": 4.361839796066106e-05,
+      "loss": 1.0639,
+      "step": 15194
+    },
+    {
+      "epoch": 0.7000046068088634,
+      "grad_norm": 0.7296167437339554,
+      "learning_rate": 4.360607542197328e-05,
+      "loss": 1.1262,
+      "step": 15195
+    },
+    {
+      "epoch": 0.7000506748974985,
+      "grad_norm": 0.9561096793919392,
+      "learning_rate": 4.359375413879954e-05,
+      "loss": 1.2037,
+      "step": 15196
+    },
+    {
+      "epoch": 0.7000967429861336,
+      "grad_norm": 1.0244995674676436,
+      "learning_rate": 4.358143411141414e-05,
+      "loss": 1.2444,
+      "step": 15197
+    },
+    {
+      "epoch": 0.7001428110747685,
+      "grad_norm": 0.9373746256101058,
+      "learning_rate": 4.3569115340091416e-05,
+      "loss": 1.598,
+      "step": 15198
+    },
+    {
+      "epoch": 0.7001888791634036,
+      "grad_norm": 1.1315070298926908,
+      "learning_rate": 4.355679782510554e-05,
+      "loss": 1.0987,
+      "step": 15199
+    },
+    {
+      "epoch": 0.7002349472520385,
+      "grad_norm": 1.3228753175587626,
+      "learning_rate": 4.354448156673079e-05,
+      "loss": 1.2268,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7002810153406736,
+      "grad_norm": 1.5740243241700922,
+      "learning_rate": 4.353216656524135e-05,
+      "loss": 1.4755,
+      "step": 15201
+    },
+    {
+      "epoch": 0.7003270834293085,
+      "grad_norm": 0.7660703335214736,
+      "learning_rate": 4.351985282091143e-05,
+      "loss": 1.3421,
+      "step": 15202
+    },
+    {
+      "epoch": 0.7003731515179435,
+      "grad_norm": 0.8268397947618121,
+      "learning_rate": 4.3507540334015116e-05,
+      "loss": 1.2529,
+      "step": 15203
+    },
+    {
+      "epoch": 0.7004192196065785,
+      "grad_norm": 0.7796795339177919,
+      "learning_rate": 4.349522910482655e-05,
+      "loss": 1.1744,
+      "step": 15204
+    },
+    {
+      "epoch": 0.7004652876952135,
+      "grad_norm": 0.6478436858779676,
+      "learning_rate": 4.3482919133619856e-05,
+      "loss": 1.1698,
+      "step": 15205
+    },
+    {
+      "epoch": 0.7005113557838485,
+      "grad_norm": 1.0420638324383045,
+      "learning_rate": 4.347061042066903e-05,
+      "loss": 1.3706,
+      "step": 15206
+    },
+    {
+      "epoch": 0.7005574238724835,
+      "grad_norm": 1.4489813716691338,
+      "learning_rate": 4.345830296624814e-05,
+      "loss": 1.3897,
+      "step": 15207
+    },
+    {
+      "epoch": 0.7006034919611185,
+      "grad_norm": 3.8607645102797044,
+      "learning_rate": 4.344599677063118e-05,
+      "loss": 0.957,
+      "step": 15208
+    },
+    {
+      "epoch": 0.7006495600497535,
+      "grad_norm": 1.8634213281350782,
+      "learning_rate": 4.3433691834092174e-05,
+      "loss": 1.2109,
+      "step": 15209
+    },
+    {
+      "epoch": 0.7006956281383886,
+      "grad_norm": 0.7291261999151082,
+      "learning_rate": 4.3421388156905e-05,
+      "loss": 1.3903,
+      "step": 15210
+    },
+    {
+      "epoch": 0.7007416962270235,
+      "grad_norm": 1.0884334751435074,
+      "learning_rate": 4.3409085739343594e-05,
+      "loss": 1.2734,
+      "step": 15211
+    },
+    {
+      "epoch": 0.7007877643156586,
+      "grad_norm": 1.3709159944461182,
+      "learning_rate": 4.339678458168189e-05,
+      "loss": 0.9876,
+      "step": 15212
+    },
+    {
+      "epoch": 0.7008338324042935,
+      "grad_norm": 1.3993702232993224,
+      "learning_rate": 4.33844846841937e-05,
+      "loss": 1.3489,
+      "step": 15213
+    },
+    {
+      "epoch": 0.7008799004929286,
+      "grad_norm": 0.8184149710255796,
+      "learning_rate": 4.337218604715289e-05,
+      "loss": 1.3477,
+      "step": 15214
+    },
+    {
+      "epoch": 0.7009259685815635,
+      "grad_norm": 1.6000264301707234,
+      "learning_rate": 4.335988867083325e-05,
+      "loss": 1.2877,
+      "step": 15215
+    },
+    {
+      "epoch": 0.7009720366701986,
+      "grad_norm": 1.600917860939316,
+      "learning_rate": 4.334759255550861e-05,
+      "loss": 1.1441,
+      "step": 15216
+    },
+    {
+      "epoch": 0.7010181047588335,
+      "grad_norm": 0.9612134295205033,
+      "learning_rate": 4.3335297701452646e-05,
+      "loss": 1.3672,
+      "step": 15217
+    },
+    {
+      "epoch": 0.7010641728474686,
+      "grad_norm": 1.6630981056649163,
+      "learning_rate": 4.3323004108939126e-05,
+      "loss": 1.1418,
+      "step": 15218
+    },
+    {
+      "epoch": 0.7011102409361035,
+      "grad_norm": 0.7473879297049596,
+      "learning_rate": 4.331071177824173e-05,
+      "loss": 1.3756,
+      "step": 15219
+    },
+    {
+      "epoch": 0.7011563090247386,
+      "grad_norm": 3.58349107303801,
+      "learning_rate": 4.329842070963418e-05,
+      "loss": 1.2535,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7012023771133735,
+      "grad_norm": 1.037584109454624,
+      "learning_rate": 4.328613090339007e-05,
+      "loss": 1.23,
+      "step": 15221
+    },
+    {
+      "epoch": 0.7012484452020086,
+      "grad_norm": 1.0823479001209968,
+      "learning_rate": 4.327384235978294e-05,
+      "loss": 1.0136,
+      "step": 15222
+    },
+    {
+      "epoch": 0.7012945132906435,
+      "grad_norm": 1.0841523451130648,
+      "learning_rate": 4.3261555079086514e-05,
+      "loss": 1.4051,
+      "step": 15223
+    },
+    {
+      "epoch": 0.7013405813792786,
+      "grad_norm": 0.7906192591733462,
+      "learning_rate": 4.3249269061574235e-05,
+      "loss": 0.849,
+      "step": 15224
+    },
+    {
+      "epoch": 0.7013866494679136,
+      "grad_norm": 1.1068069351653806,
+      "learning_rate": 4.323698430751969e-05,
+      "loss": 1.2406,
+      "step": 15225
+    },
+    {
+      "epoch": 0.7014327175565486,
+      "grad_norm": 1.0762892383459224,
+      "learning_rate": 4.3224700817196354e-05,
+      "loss": 1.0805,
+      "step": 15226
+    },
+    {
+      "epoch": 0.7014787856451836,
+      "grad_norm": 0.7228470597656813,
+      "learning_rate": 4.321241859087773e-05,
+      "loss": 1.4731,
+      "step": 15227
+    },
+    {
+      "epoch": 0.7015248537338186,
+      "grad_norm": 1.1519763072915656,
+      "learning_rate": 4.320013762883722e-05,
+      "loss": 1.41,
+      "step": 15228
+    },
+    {
+      "epoch": 0.7015709218224536,
+      "grad_norm": 0.9756562251075911,
+      "learning_rate": 4.318785793134824e-05,
+      "loss": 1.4149,
+      "step": 15229
+    },
+    {
+      "epoch": 0.7016169899110886,
+      "grad_norm": 0.755068104158533,
+      "learning_rate": 4.317557949868424e-05,
+      "loss": 1.3692,
+      "step": 15230
+    },
+    {
+      "epoch": 0.7016630579997236,
+      "grad_norm": 1.284872258126062,
+      "learning_rate": 4.316330233111848e-05,
+      "loss": 1.1248,
+      "step": 15231
+    },
+    {
+      "epoch": 0.7017091260883586,
+      "grad_norm": 1.1560088754298454,
+      "learning_rate": 4.315102642892439e-05,
+      "loss": 1.0999,
+      "step": 15232
+    },
+    {
+      "epoch": 0.7017551941769936,
+      "grad_norm": 0.901156360637711,
+      "learning_rate": 4.3138751792375145e-05,
+      "loss": 1.1745,
+      "step": 15233
+    },
+    {
+      "epoch": 0.7018012622656286,
+      "grad_norm": 1.4012272394770793,
+      "learning_rate": 4.3126478421744156e-05,
+      "loss": 1.3719,
+      "step": 15234
+    },
+    {
+      "epoch": 0.7018473303542636,
+      "grad_norm": 2.216790807861939,
+      "learning_rate": 4.3114206317304584e-05,
+      "loss": 1.2585,
+      "step": 15235
+    },
+    {
+      "epoch": 0.7018933984428986,
+      "grad_norm": 0.7455876146412035,
+      "learning_rate": 4.310193547932967e-05,
+      "loss": 1.3032,
+      "step": 15236
+    },
+    {
+      "epoch": 0.7019394665315336,
+      "grad_norm": 0.9275260684245017,
+      "learning_rate": 4.308966590809259e-05,
+      "loss": 1.1671,
+      "step": 15237
+    },
+    {
+      "epoch": 0.7019855346201687,
+      "grad_norm": 1.1188090295415376,
+      "learning_rate": 4.3077397603866567e-05,
+      "loss": 1.4026,
+      "step": 15238
+    },
+    {
+      "epoch": 0.7020316027088036,
+      "grad_norm": 0.9940829722651137,
+      "learning_rate": 4.3065130566924685e-05,
+      "loss": 1.231,
+      "step": 15239
+    },
+    {
+      "epoch": 0.7020776707974387,
+      "grad_norm": 2.9495439487574355,
+      "learning_rate": 4.305286479753997e-05,
+      "loss": 1.0075,
+      "step": 15240
+    },
+    {
+      "epoch": 0.7021237388860736,
+      "grad_norm": 1.0348685631844132,
+      "learning_rate": 4.304060029598566e-05,
+      "loss": 1.1952,
+      "step": 15241
+    },
+    {
+      "epoch": 0.7021698069747087,
+      "grad_norm": 1.1710664328762057,
+      "learning_rate": 4.3028337062534664e-05,
+      "loss": 0.8937,
+      "step": 15242
+    },
+    {
+      "epoch": 0.7022158750633436,
+      "grad_norm": 0.9478109502026547,
+      "learning_rate": 4.301607509746011e-05,
+      "loss": 1.0662,
+      "step": 15243
+    },
+    {
+      "epoch": 0.7022619431519787,
+      "grad_norm": 1.1590266612012496,
+      "learning_rate": 4.300381440103486e-05,
+      "loss": 1.338,
+      "step": 15244
+    },
+    {
+      "epoch": 0.7023080112406136,
+      "grad_norm": 1.069013183095311,
+      "learning_rate": 4.299155497353203e-05,
+      "loss": 1.033,
+      "step": 15245
+    },
+    {
+      "epoch": 0.7023540793292486,
+      "grad_norm": 0.7810469147836133,
+      "learning_rate": 4.297929681522445e-05,
+      "loss": 1.2975,
+      "step": 15246
+    },
+    {
+      "epoch": 0.7024001474178836,
+      "grad_norm": 0.9337226245454574,
+      "learning_rate": 4.296703992638508e-05,
+      "loss": 1.4249,
+      "step": 15247
+    },
+    {
+      "epoch": 0.7024462155065186,
+      "grad_norm": 0.881657139839631,
+      "learning_rate": 4.295478430728674e-05,
+      "loss": 1.3819,
+      "step": 15248
+    },
+    {
+      "epoch": 0.7024922835951536,
+      "grad_norm": 0.9268830191852567,
+      "learning_rate": 4.294252995820232e-05,
+      "loss": 1.4579,
+      "step": 15249
+    },
+    {
+      "epoch": 0.7025383516837886,
+      "grad_norm": 0.9505296582026823,
+      "learning_rate": 4.2930276879404675e-05,
+      "loss": 1.4864,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7025844197724237,
+      "grad_norm": 1.7458868672742756,
+      "learning_rate": 4.291802507116649e-05,
+      "loss": 1.1366,
+      "step": 15251
+    },
+    {
+      "epoch": 0.7026304878610586,
+      "grad_norm": 0.5976670252359186,
+      "learning_rate": 4.2905774533760665e-05,
+      "loss": 1.1146,
+      "step": 15252
+    },
+    {
+      "epoch": 0.7026765559496937,
+      "grad_norm": 2.018981476470569,
+      "learning_rate": 4.2893525267459856e-05,
+      "loss": 1.3823,
+      "step": 15253
+    },
+    {
+      "epoch": 0.7027226240383286,
+      "grad_norm": 4.347318811994156,
+      "learning_rate": 4.288127727253681e-05,
+      "loss": 1.3042,
+      "step": 15254
+    },
+    {
+      "epoch": 0.7027686921269637,
+      "grad_norm": 1.2167912051756988,
+      "learning_rate": 4.286903054926412e-05,
+      "loss": 1.016,
+      "step": 15255
+    },
+    {
+      "epoch": 0.7028147602155986,
+      "grad_norm": 0.8498358367393596,
+      "learning_rate": 4.285678509791459e-05,
+      "loss": 1.134,
+      "step": 15256
+    },
+    {
+      "epoch": 0.7028608283042337,
+      "grad_norm": 1.2917248917454596,
+      "learning_rate": 4.284454091876074e-05,
+      "loss": 1.2676,
+      "step": 15257
+    },
+    {
+      "epoch": 0.7029068963928686,
+      "grad_norm": 1.2666574368163073,
+      "learning_rate": 4.283229801207517e-05,
+      "loss": 1.3349,
+      "step": 15258
+    },
+    {
+      "epoch": 0.7029529644815037,
+      "grad_norm": 1.5767977968921718,
+      "learning_rate": 4.2820056378130456e-05,
+      "loss": 1.3036,
+      "step": 15259
+    },
+    {
+      "epoch": 0.7029990325701386,
+      "grad_norm": 0.8656105666183742,
+      "learning_rate": 4.2807816017199144e-05,
+      "loss": 1.173,
+      "step": 15260
+    },
+    {
+      "epoch": 0.7030451006587737,
+      "grad_norm": 0.874955290979156,
+      "learning_rate": 4.279557692955377e-05,
+      "loss": 1.3591,
+      "step": 15261
+    },
+    {
+      "epoch": 0.7030911687474086,
+      "grad_norm": 1.0167494313141328,
+      "learning_rate": 4.278333911546674e-05,
+      "loss": 1.1371,
+      "step": 15262
+    },
+    {
+      "epoch": 0.7031372368360437,
+      "grad_norm": 1.123625958664231,
+      "learning_rate": 4.277110257521061e-05,
+      "loss": 1.5317,
+      "step": 15263
+    },
+    {
+      "epoch": 0.7031833049246786,
+      "grad_norm": 0.7841069109653075,
+      "learning_rate": 4.275886730905773e-05,
+      "loss": 1.3327,
+      "step": 15264
+    },
+    {
+      "epoch": 0.7032293730133137,
+      "grad_norm": 1.3939146273662348,
+      "learning_rate": 4.274663331728056e-05,
+      "loss": 1.0252,
+      "step": 15265
+    },
+    {
+      "epoch": 0.7032754411019487,
+      "grad_norm": 1.5534276246207808,
+      "learning_rate": 4.2734400600151394e-05,
+      "loss": 1.3899,
+      "step": 15266
+    },
+    {
+      "epoch": 0.7033215091905837,
+      "grad_norm": 0.8292041494258057,
+      "learning_rate": 4.2722169157942606e-05,
+      "loss": 1.4702,
+      "step": 15267
+    },
+    {
+      "epoch": 0.7033675772792187,
+      "grad_norm": 1.1371613539188592,
+      "learning_rate": 4.2709938990926545e-05,
+      "loss": 1.3113,
+      "step": 15268
+    },
+    {
+      "epoch": 0.7034136453678537,
+      "grad_norm": 0.7630104483205394,
+      "learning_rate": 4.269771009937542e-05,
+      "loss": 1.4123,
+      "step": 15269
+    },
+    {
+      "epoch": 0.7034597134564887,
+      "grad_norm": 1.169749028056961,
+      "learning_rate": 4.2685482483561535e-05,
+      "loss": 0.9956,
+      "step": 15270
+    },
+    {
+      "epoch": 0.7035057815451237,
+      "grad_norm": 1.1330135337911449,
+      "learning_rate": 4.2673256143757114e-05,
+      "loss": 1.3346,
+      "step": 15271
+    },
+    {
+      "epoch": 0.7035518496337587,
+      "grad_norm": 0.9304109786514179,
+      "learning_rate": 4.2661031080234385e-05,
+      "loss": 1.3816,
+      "step": 15272
+    },
+    {
+      "epoch": 0.7035979177223937,
+      "grad_norm": 1.2697070520876295,
+      "learning_rate": 4.2648807293265435e-05,
+      "loss": 1.4497,
+      "step": 15273
+    },
+    {
+      "epoch": 0.7036439858110287,
+      "grad_norm": 0.8023136238425369,
+      "learning_rate": 4.263658478312247e-05,
+      "loss": 1.5187,
+      "step": 15274
+    },
+    {
+      "epoch": 0.7036900538996637,
+      "grad_norm": 2.7936672591772598,
+      "learning_rate": 4.2624363550077606e-05,
+      "loss": 1.4631,
+      "step": 15275
+    },
+    {
+      "epoch": 0.7037361219882987,
+      "grad_norm": 0.7428216644560298,
+      "learning_rate": 4.2612143594402886e-05,
+      "loss": 1.4428,
+      "step": 15276
+    },
+    {
+      "epoch": 0.7037821900769337,
+      "grad_norm": 1.0837679787663748,
+      "learning_rate": 4.259992491637039e-05,
+      "loss": 1.0778,
+      "step": 15277
+    },
+    {
+      "epoch": 0.7038282581655687,
+      "grad_norm": 0.8062942906196875,
+      "learning_rate": 4.258770751625214e-05,
+      "loss": 1.4184,
+      "step": 15278
+    },
+    {
+      "epoch": 0.7038743262542038,
+      "grad_norm": 0.9487873924029037,
+      "learning_rate": 4.257549139432019e-05,
+      "loss": 1.4006,
+      "step": 15279
+    },
+    {
+      "epoch": 0.7039203943428387,
+      "grad_norm": 1.2308626687190922,
+      "learning_rate": 4.256327655084641e-05,
+      "loss": 1.4097,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7039664624314738,
+      "grad_norm": 0.8931720590879034,
+      "learning_rate": 4.255106298610281e-05,
+      "loss": 1.3506,
+      "step": 15281
+    },
+    {
+      "epoch": 0.7040125305201087,
+      "grad_norm": 1.1946116129146824,
+      "learning_rate": 4.2538850700361275e-05,
+      "loss": 1.0032,
+      "step": 15282
+    },
+    {
+      "epoch": 0.7040585986087438,
+      "grad_norm": 0.9479734485935237,
+      "learning_rate": 4.252663969389374e-05,
+      "loss": 1.0947,
+      "step": 15283
+    },
+    {
+      "epoch": 0.7041046666973787,
+      "grad_norm": 1.3302838576939588,
+      "learning_rate": 4.2514429966972e-05,
+      "loss": 1.0967,
+      "step": 15284
+    },
+    {
+      "epoch": 0.7041507347860138,
+      "grad_norm": 1.5365557864002217,
+      "learning_rate": 4.250222151986791e-05,
+      "loss": 0.9347,
+      "step": 15285
+    },
+    {
+      "epoch": 0.7041968028746487,
+      "grad_norm": 1.655411774872266,
+      "learning_rate": 4.2490014352853305e-05,
+      "loss": 1.3023,
+      "step": 15286
+    },
+    {
+      "epoch": 0.7042428709632838,
+      "grad_norm": 0.8554055914405363,
+      "learning_rate": 4.2477808466199884e-05,
+      "loss": 1.643,
+      "step": 15287
+    },
+    {
+      "epoch": 0.7042889390519187,
+      "grad_norm": 1.0587534721129792,
+      "learning_rate": 4.2465603860179424e-05,
+      "loss": 1.3167,
+      "step": 15288
+    },
+    {
+      "epoch": 0.7043350071405537,
+      "grad_norm": 1.7300960085341008,
+      "learning_rate": 4.245340053506365e-05,
+      "loss": 1.3579,
+      "step": 15289
+    },
+    {
+      "epoch": 0.7043810752291887,
+      "grad_norm": 0.9260017182832416,
+      "learning_rate": 4.2441198491124276e-05,
+      "loss": 1.3222,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7044271433178237,
+      "grad_norm": 1.2215691768578532,
+      "learning_rate": 4.242899772863289e-05,
+      "loss": 1.2282,
+      "step": 15291
+    },
+    {
+      "epoch": 0.7044732114064588,
+      "grad_norm": 1.059183358060652,
+      "learning_rate": 4.241679824786116e-05,
+      "loss": 1.3255,
+      "step": 15292
+    },
+    {
+      "epoch": 0.7045192794950937,
+      "grad_norm": 0.8463160504770906,
+      "learning_rate": 4.240460004908071e-05,
+      "loss": 1.2886,
+      "step": 15293
+    },
+    {
+      "epoch": 0.7045653475837288,
+      "grad_norm": 0.9710136686813162,
+      "learning_rate": 4.2392403132563054e-05,
+      "loss": 0.779,
+      "step": 15294
+    },
+    {
+      "epoch": 0.7046114156723637,
+      "grad_norm": 0.7348258001602038,
+      "learning_rate": 4.238020749857977e-05,
+      "loss": 1.1998,
+      "step": 15295
+    },
+    {
+      "epoch": 0.7046574837609988,
+      "grad_norm": 0.9199427267160503,
+      "learning_rate": 4.236801314740236e-05,
+      "loss": 1.3002,
+      "step": 15296
+    },
+    {
+      "epoch": 0.7047035518496337,
+      "grad_norm": 1.044482113120663,
+      "learning_rate": 4.235582007930237e-05,
+      "loss": 1.0957,
+      "step": 15297
+    },
+    {
+      "epoch": 0.7047496199382688,
+      "grad_norm": 1.2905140650605895,
+      "learning_rate": 4.234362829455115e-05,
+      "loss": 0.8982,
+      "step": 15298
+    },
+    {
+      "epoch": 0.7047956880269037,
+      "grad_norm": 0.8184301556029977,
+      "learning_rate": 4.23314377934202e-05,
+      "loss": 1.5453,
+      "step": 15299
+    },
+    {
+      "epoch": 0.7048417561155388,
+      "grad_norm": 0.9309286952598647,
+      "learning_rate": 4.231924857618091e-05,
+      "loss": 1.22,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7048878242041737,
+      "grad_norm": 1.060604927031903,
+      "learning_rate": 4.230706064310468e-05,
+      "loss": 1.0998,
+      "step": 15301
+    },
+    {
+      "epoch": 0.7049338922928088,
+      "grad_norm": 0.795527955499317,
+      "learning_rate": 4.229487399446279e-05,
+      "loss": 1.414,
+      "step": 15302
+    },
+    {
+      "epoch": 0.7049799603814437,
+      "grad_norm": 2.0407548475981145,
+      "learning_rate": 4.228268863052658e-05,
+      "loss": 1.3289,
+      "step": 15303
+    },
+    {
+      "epoch": 0.7050260284700788,
+      "grad_norm": 16.70541100781603,
+      "learning_rate": 4.227050455156739e-05,
+      "loss": 1.2368,
+      "step": 15304
+    },
+    {
+      "epoch": 0.7050720965587138,
+      "grad_norm": 1.6585758326276578,
+      "learning_rate": 4.225832175785639e-05,
+      "loss": 1.3126,
+      "step": 15305
+    },
+    {
+      "epoch": 0.7051181646473488,
+      "grad_norm": 7.166646014060731,
+      "learning_rate": 4.224614024966484e-05,
+      "loss": 1.0349,
+      "step": 15306
+    },
+    {
+      "epoch": 0.7051642327359838,
+      "grad_norm": 0.8945646660524029,
+      "learning_rate": 4.223396002726398e-05,
+      "loss": 1.7186,
+      "step": 15307
+    },
+    {
+      "epoch": 0.7052103008246188,
+      "grad_norm": 1.0547082508081973,
+      "learning_rate": 4.222178109092496e-05,
+      "loss": 1.09,
+      "step": 15308
+    },
+    {
+      "epoch": 0.7052563689132538,
+      "grad_norm": 1.8258030698672862,
+      "learning_rate": 4.2209603440918886e-05,
+      "loss": 1.334,
+      "step": 15309
+    },
+    {
+      "epoch": 0.7053024370018888,
+      "grad_norm": 0.7413099834235575,
+      "learning_rate": 4.21974270775169e-05,
+      "loss": 1.385,
+      "step": 15310
+    },
+    {
+      "epoch": 0.7053485050905238,
+      "grad_norm": 1.0201430697215843,
+      "learning_rate": 4.218525200099014e-05,
+      "loss": 1.1397,
+      "step": 15311
+    },
+    {
+      "epoch": 0.7053945731791588,
+      "grad_norm": 1.5910876946657033,
+      "learning_rate": 4.217307821160956e-05,
+      "loss": 1.3444,
+      "step": 15312
+    },
+    {
+      "epoch": 0.7054406412677938,
+      "grad_norm": 1.120423363444692,
+      "learning_rate": 4.216090570964626e-05,
+      "loss": 1.09,
+      "step": 15313
+    },
+    {
+      "epoch": 0.7054867093564288,
+      "grad_norm": 1.0348210625714802,
+      "learning_rate": 4.214873449537121e-05,
+      "loss": 1.4085,
+      "step": 15314
+    },
+    {
+      "epoch": 0.7055327774450638,
+      "grad_norm": 1.829786499661859,
+      "learning_rate": 4.2136564569055423e-05,
+      "loss": 1.1931,
+      "step": 15315
+    },
+    {
+      "epoch": 0.7055788455336988,
+      "grad_norm": 0.8104728248918425,
+      "learning_rate": 4.21243959309698e-05,
+      "loss": 0.9592,
+      "step": 15316
+    },
+    {
+      "epoch": 0.7056249136223338,
+      "grad_norm": 1.6042064854376628,
+      "learning_rate": 4.211222858138524e-05,
+      "loss": 1.1423,
+      "step": 15317
+    },
+    {
+      "epoch": 0.7056709817109688,
+      "grad_norm": 1.3400022360341466,
+      "learning_rate": 4.2100062520572656e-05,
+      "loss": 1.229,
+      "step": 15318
+    },
+    {
+      "epoch": 0.7057170497996038,
+      "grad_norm": 1.2370340394717954,
+      "learning_rate": 4.208789774880295e-05,
+      "loss": 1.6376,
+      "step": 15319
+    },
+    {
+      "epoch": 0.7057631178882389,
+      "grad_norm": 0.8670758815170995,
+      "learning_rate": 4.207573426634687e-05,
+      "loss": 1.2333,
+      "step": 15320
+    },
+    {
+      "epoch": 0.7058091859768738,
+      "grad_norm": 1.9011100391408842,
+      "learning_rate": 4.2063572073475234e-05,
+      "loss": 1.3448,
+      "step": 15321
+    },
+    {
+      "epoch": 0.7058552540655089,
+      "grad_norm": 0.9508425567540604,
+      "learning_rate": 4.2051411170458874e-05,
+      "loss": 1.3873,
+      "step": 15322
+    },
+    {
+      "epoch": 0.7059013221541438,
+      "grad_norm": 1.4046702938275344,
+      "learning_rate": 4.203925155756844e-05,
+      "loss": 1.7447,
+      "step": 15323
+    },
+    {
+      "epoch": 0.7059473902427789,
+      "grad_norm": 1.0539585683421706,
+      "learning_rate": 4.2027093235074686e-05,
+      "loss": 1.1792,
+      "step": 15324
+    },
+    {
+      "epoch": 0.7059934583314138,
+      "grad_norm": 1.3225727035743198,
+      "learning_rate": 4.201493620324831e-05,
+      "loss": 1.3038,
+      "step": 15325
+    },
+    {
+      "epoch": 0.7060395264200489,
+      "grad_norm": 1.478243781251554,
+      "learning_rate": 4.200278046235999e-05,
+      "loss": 1.2263,
+      "step": 15326
+    },
+    {
+      "epoch": 0.7060855945086838,
+      "grad_norm": 1.0233689071292145,
+      "learning_rate": 4.199062601268028e-05,
+      "loss": 1.2779,
+      "step": 15327
+    },
+    {
+      "epoch": 0.7061316625973189,
+      "grad_norm": 1.193633095157574,
+      "learning_rate": 4.197847285447981e-05,
+      "loss": 1.2392,
+      "step": 15328
+    },
+    {
+      "epoch": 0.7061777306859538,
+      "grad_norm": 0.9124776505870881,
+      "learning_rate": 4.1966320988029164e-05,
+      "loss": 1.2079,
+      "step": 15329
+    },
+    {
+      "epoch": 0.7062237987745889,
+      "grad_norm": 1.2903273184644826,
+      "learning_rate": 4.1954170413598914e-05,
+      "loss": 1.3346,
+      "step": 15330
+    },
+    {
+      "epoch": 0.7062698668632238,
+      "grad_norm": 1.2390047671264703,
+      "learning_rate": 4.194202113145953e-05,
+      "loss": 1.1262,
+      "step": 15331
+    },
+    {
+      "epoch": 0.7063159349518588,
+      "grad_norm": 2.959993395945273,
+      "learning_rate": 4.1929873141881426e-05,
+      "loss": 1.2789,
+      "step": 15332
+    },
+    {
+      "epoch": 0.7063620030404939,
+      "grad_norm": 0.9800150285447697,
+      "learning_rate": 4.19177264451352e-05,
+      "loss": 1.195,
+      "step": 15333
+    },
+    {
+      "epoch": 0.7064080711291288,
+      "grad_norm": 2.0858261777126867,
+      "learning_rate": 4.190558104149116e-05,
+      "loss": 1.2906,
+      "step": 15334
+    },
+    {
+      "epoch": 0.7064541392177639,
+      "grad_norm": 6.85996026733861,
+      "learning_rate": 4.1893436931219755e-05,
+      "loss": 1.3549,
+      "step": 15335
+    },
+    {
+      "epoch": 0.7065002073063988,
+      "grad_norm": 0.909998595081604,
+      "learning_rate": 4.188129411459134e-05,
+      "loss": 1.118,
+      "step": 15336
+    },
+    {
+      "epoch": 0.7065462753950339,
+      "grad_norm": 3.4017214645501723,
+      "learning_rate": 4.1869152591876304e-05,
+      "loss": 1.3243,
+      "step": 15337
+    },
+    {
+      "epoch": 0.7065923434836688,
+      "grad_norm": 1.1597371240512304,
+      "learning_rate": 4.1857012363344875e-05,
+      "loss": 1.2281,
+      "step": 15338
+    },
+    {
+      "epoch": 0.7066384115723039,
+      "grad_norm": 1.1752928490634615,
+      "learning_rate": 4.184487342926736e-05,
+      "loss": 1.2254,
+      "step": 15339
+    },
+    {
+      "epoch": 0.7066844796609388,
+      "grad_norm": 1.0324095354900977,
+      "learning_rate": 4.1832735789914066e-05,
+      "loss": 1.3477,
+      "step": 15340
+    },
+    {
+      "epoch": 0.7067305477495739,
+      "grad_norm": 3.098250567341558,
+      "learning_rate": 4.182059944555514e-05,
+      "loss": 1.2561,
+      "step": 15341
+    },
+    {
+      "epoch": 0.7067766158382088,
+      "grad_norm": 11.784683740737945,
+      "learning_rate": 4.180846439646086e-05,
+      "loss": 1.1253,
+      "step": 15342
+    },
+    {
+      "epoch": 0.7068226839268439,
+      "grad_norm": 0.95835105925872,
+      "learning_rate": 4.1796330642901246e-05,
+      "loss": 1.4093,
+      "step": 15343
+    },
+    {
+      "epoch": 0.7068687520154788,
+      "grad_norm": 2.2932214709416145,
+      "learning_rate": 4.178419818514663e-05,
+      "loss": 1.2136,
+      "step": 15344
+    },
+    {
+      "epoch": 0.7069148201041139,
+      "grad_norm": 1.1551359032963027,
+      "learning_rate": 4.177206702346696e-05,
+      "loss": 1.3405,
+      "step": 15345
+    },
+    {
+      "epoch": 0.7069608881927489,
+      "grad_norm": 0.9035036551808433,
+      "learning_rate": 4.175993715813243e-05,
+      "loss": 1.2273,
+      "step": 15346
+    },
+    {
+      "epoch": 0.7070069562813839,
+      "grad_norm": 1.0356403044955873,
+      "learning_rate": 4.1747808589412974e-05,
+      "loss": 1.5081,
+      "step": 15347
+    },
+    {
+      "epoch": 0.7070530243700189,
+      "grad_norm": 0.9649453064693185,
+      "learning_rate": 4.1735681317578736e-05,
+      "loss": 1.0634,
+      "step": 15348
+    },
+    {
+      "epoch": 0.7070990924586539,
+      "grad_norm": 2.347262670685033,
+      "learning_rate": 4.172355534289966e-05,
+      "loss": 1.3227,
+      "step": 15349
+    },
+    {
+      "epoch": 0.7071451605472889,
+      "grad_norm": 1.355722818151245,
+      "learning_rate": 4.171143066564562e-05,
+      "loss": 1.2999,
+      "step": 15350
+    },
+    {
+      "epoch": 0.7071912286359239,
+      "grad_norm": 1.7656763993074798,
+      "learning_rate": 4.169930728608672e-05,
+      "loss": 1.0832,
+      "step": 15351
+    },
+    {
+      "epoch": 0.7072372967245589,
+      "grad_norm": 1.1114683206789597,
+      "learning_rate": 4.168718520449273e-05,
+      "loss": 1.4372,
+      "step": 15352
+    },
+    {
+      "epoch": 0.7072833648131939,
+      "grad_norm": 0.8186462676702487,
+      "learning_rate": 4.167506442113361e-05,
+      "loss": 1.3817,
+      "step": 15353
+    },
+    {
+      "epoch": 0.7073294329018289,
+      "grad_norm": 1.0823967902903155,
+      "learning_rate": 4.1662944936279114e-05,
+      "loss": 1.1872,
+      "step": 15354
+    },
+    {
+      "epoch": 0.7073755009904639,
+      "grad_norm": 0.8792458898486905,
+      "learning_rate": 4.165082675019919e-05,
+      "loss": 1.2292,
+      "step": 15355
+    },
+    {
+      "epoch": 0.7074215690790989,
+      "grad_norm": 0.9486065221892936,
+      "learning_rate": 4.163870986316353e-05,
+      "loss": 1.2429,
+      "step": 15356
+    },
+    {
+      "epoch": 0.7074676371677339,
+      "grad_norm": 1.2225664693524005,
+      "learning_rate": 4.162659427544196e-05,
+      "loss": 1.1989,
+      "step": 15357
+    },
+    {
+      "epoch": 0.7075137052563689,
+      "grad_norm": 1.3561156966193342,
+      "learning_rate": 4.1614479987304146e-05,
+      "loss": 0.9675,
+      "step": 15358
+    },
+    {
+      "epoch": 0.7075597733450039,
+      "grad_norm": 3.711162245071803,
+      "learning_rate": 4.1602366999019817e-05,
+      "loss": 1.3603,
+      "step": 15359
+    },
+    {
+      "epoch": 0.7076058414336389,
+      "grad_norm": 14.574092028636793,
+      "learning_rate": 4.15902553108587e-05,
+      "loss": 1.0306,
+      "step": 15360
+    },
+    {
+      "epoch": 0.707651909522274,
+      "grad_norm": 1.753776928355092,
+      "learning_rate": 4.1578144923090314e-05,
+      "loss": 1.4543,
+      "step": 15361
+    },
+    {
+      "epoch": 0.7076979776109089,
+      "grad_norm": 1.1104582835572951,
+      "learning_rate": 4.156603583598444e-05,
+      "loss": 1.2589,
+      "step": 15362
+    },
+    {
+      "epoch": 0.707744045699544,
+      "grad_norm": 1.1321958195623751,
+      "learning_rate": 4.1553928049810544e-05,
+      "loss": 1.3109,
+      "step": 15363
+    },
+    {
+      "epoch": 0.7077901137881789,
+      "grad_norm": 9.227908347447542,
+      "learning_rate": 4.154182156483827e-05,
+      "loss": 1.286,
+      "step": 15364
+    },
+    {
+      "epoch": 0.707836181876814,
+      "grad_norm": 1.2602189865618523,
+      "learning_rate": 4.1529716381337025e-05,
+      "loss": 1.1374,
+      "step": 15365
+    },
+    {
+      "epoch": 0.7078822499654489,
+      "grad_norm": 1.7745117577076237,
+      "learning_rate": 4.151761249957648e-05,
+      "loss": 1.0973,
+      "step": 15366
+    },
+    {
+      "epoch": 0.707928318054084,
+      "grad_norm": 0.7668695896336629,
+      "learning_rate": 4.150550991982599e-05,
+      "loss": 1.3006,
+      "step": 15367
+    },
+    {
+      "epoch": 0.7079743861427189,
+      "grad_norm": 2.5449908135603727,
+      "learning_rate": 4.1493408642355005e-05,
+      "loss": 1.3251,
+      "step": 15368
+    },
+    {
+      "epoch": 0.708020454231354,
+      "grad_norm": 0.8358334999510928,
+      "learning_rate": 4.148130866743295e-05,
+      "loss": 1.2649,
+      "step": 15369
+    },
+    {
+      "epoch": 0.7080665223199889,
+      "grad_norm": 1.0799703993691039,
+      "learning_rate": 4.146920999532923e-05,
+      "loss": 1.3117,
+      "step": 15370
+    },
+    {
+      "epoch": 0.708112590408624,
+      "grad_norm": 8.97924099211059,
+      "learning_rate": 4.145711262631322e-05,
+      "loss": 1.3552,
+      "step": 15371
+    },
+    {
+      "epoch": 0.7081586584972589,
+      "grad_norm": 0.8960339418297127,
+      "learning_rate": 4.144501656065418e-05,
+      "loss": 1.3676,
+      "step": 15372
+    },
+    {
+      "epoch": 0.708204726585894,
+      "grad_norm": 1.0301475317662137,
+      "learning_rate": 4.143292179862146e-05,
+      "loss": 1.3544,
+      "step": 15373
+    },
+    {
+      "epoch": 0.708250794674529,
+      "grad_norm": 0.8451385888652073,
+      "learning_rate": 4.1420828340484305e-05,
+      "loss": 1.4867,
+      "step": 15374
+    },
+    {
+      "epoch": 0.708296862763164,
+      "grad_norm": 1.0560605593375405,
+      "learning_rate": 4.1408736186512e-05,
+      "loss": 1.3547,
+      "step": 15375
+    },
+    {
+      "epoch": 0.708342930851799,
+      "grad_norm": 4.142780822373453,
+      "learning_rate": 4.139664533697368e-05,
+      "loss": 1.2003,
+      "step": 15376
+    },
+    {
+      "epoch": 0.708388998940434,
+      "grad_norm": 1.4168924859649765,
+      "learning_rate": 4.138455579213858e-05,
+      "loss": 1.2897,
+      "step": 15377
+    },
+    {
+      "epoch": 0.708435067029069,
+      "grad_norm": 1.7241169589117986,
+      "learning_rate": 4.137246755227587e-05,
+      "loss": 1.1613,
+      "step": 15378
+    },
+    {
+      "epoch": 0.7084811351177039,
+      "grad_norm": 1.168139751978962,
+      "learning_rate": 4.136038061765461e-05,
+      "loss": 1.3448,
+      "step": 15379
+    },
+    {
+      "epoch": 0.708527203206339,
+      "grad_norm": 0.8329661278187604,
+      "learning_rate": 4.134829498854393e-05,
+      "loss": 1.637,
+      "step": 15380
+    },
+    {
+      "epoch": 0.7085732712949739,
+      "grad_norm": 1.6422411007016606,
+      "learning_rate": 4.133621066521291e-05,
+      "loss": 1.1903,
+      "step": 15381
+    },
+    {
+      "epoch": 0.708619339383609,
+      "grad_norm": 1.6146142279117575,
+      "learning_rate": 4.132412764793059e-05,
+      "loss": 1.325,
+      "step": 15382
+    },
+    {
+      "epoch": 0.7086654074722439,
+      "grad_norm": 0.892998101463588,
+      "learning_rate": 4.1312045936965935e-05,
+      "loss": 1.4283,
+      "step": 15383
+    },
+    {
+      "epoch": 0.708711475560879,
+      "grad_norm": 0.9950458645174677,
+      "learning_rate": 4.129996553258795e-05,
+      "loss": 1.5448,
+      "step": 15384
+    },
+    {
+      "epoch": 0.7087575436495139,
+      "grad_norm": 0.775601563386883,
+      "learning_rate": 4.12878864350656e-05,
+      "loss": 1.2461,
+      "step": 15385
+    },
+    {
+      "epoch": 0.708803611738149,
+      "grad_norm": 1.3251456522934701,
+      "learning_rate": 4.127580864466778e-05,
+      "loss": 1.047,
+      "step": 15386
+    },
+    {
+      "epoch": 0.708849679826784,
+      "grad_norm": 0.7631700881172685,
+      "learning_rate": 4.126373216166337e-05,
+      "loss": 1.5061,
+      "step": 15387
+    },
+    {
+      "epoch": 0.708895747915419,
+      "grad_norm": 0.9183872830575186,
+      "learning_rate": 4.1251656986321264e-05,
+      "loss": 1.547,
+      "step": 15388
+    },
+    {
+      "epoch": 0.708941816004054,
+      "grad_norm": 1.2262816369574803,
+      "learning_rate": 4.1239583118910304e-05,
+      "loss": 1.3313,
+      "step": 15389
+    },
+    {
+      "epoch": 0.708987884092689,
+      "grad_norm": 22.453678811437204,
+      "learning_rate": 4.122751055969926e-05,
+      "loss": 1.1655,
+      "step": 15390
+    },
+    {
+      "epoch": 0.709033952181324,
+      "grad_norm": 1.0777376214188112,
+      "learning_rate": 4.1215439308956906e-05,
+      "loss": 1.4241,
+      "step": 15391
+    },
+    {
+      "epoch": 0.709080020269959,
+      "grad_norm": 0.8838383755802496,
+      "learning_rate": 4.120336936695201e-05,
+      "loss": 1.321,
+      "step": 15392
+    },
+    {
+      "epoch": 0.709126088358594,
+      "grad_norm": 0.7557907918427601,
+      "learning_rate": 4.119130073395331e-05,
+      "loss": 1.2458,
+      "step": 15393
+    },
+    {
+      "epoch": 0.709172156447229,
+      "grad_norm": 1.6788765622549424,
+      "learning_rate": 4.1179233410229424e-05,
+      "loss": 1.1304,
+      "step": 15394
+    },
+    {
+      "epoch": 0.709218224535864,
+      "grad_norm": 1.1148313959678244,
+      "learning_rate": 4.1167167396049056e-05,
+      "loss": 1.3661,
+      "step": 15395
+    },
+    {
+      "epoch": 0.709264292624499,
+      "grad_norm": 0.9512870992902591,
+      "learning_rate": 4.115510269168086e-05,
+      "loss": 1.1615,
+      "step": 15396
+    },
+    {
+      "epoch": 0.709310360713134,
+      "grad_norm": 1.1497763718010203,
+      "learning_rate": 4.114303929739337e-05,
+      "loss": 1.1225,
+      "step": 15397
+    },
+    {
+      "epoch": 0.709356428801769,
+      "grad_norm": 4.218647247065204,
+      "learning_rate": 4.11309772134552e-05,
+      "loss": 1.0671,
+      "step": 15398
+    },
+    {
+      "epoch": 0.709402496890404,
+      "grad_norm": 1.0773887713392412,
+      "learning_rate": 4.111891644013488e-05,
+      "loss": 1.564,
+      "step": 15399
+    },
+    {
+      "epoch": 0.709448564979039,
+      "grad_norm": 1.5986829062913812,
+      "learning_rate": 4.1106856977700957e-05,
+      "loss": 0.9247,
+      "step": 15400
+    },
+    {
+      "epoch": 0.709494633067674,
+      "grad_norm": 1.1430942305975225,
+      "learning_rate": 4.109479882642186e-05,
+      "loss": 1.2133,
+      "step": 15401
+    },
+    {
+      "epoch": 0.7095407011563091,
+      "grad_norm": 1.3704583398929313,
+      "learning_rate": 4.108274198656605e-05,
+      "loss": 1.093,
+      "step": 15402
+    },
+    {
+      "epoch": 0.709586769244944,
+      "grad_norm": 1.0993782051576606,
+      "learning_rate": 4.107068645840202e-05,
+      "loss": 1.2145,
+      "step": 15403
+    },
+    {
+      "epoch": 0.7096328373335791,
+      "grad_norm": 0.9030995964920164,
+      "learning_rate": 4.105863224219807e-05,
+      "loss": 1.4177,
+      "step": 15404
+    },
+    {
+      "epoch": 0.709678905422214,
+      "grad_norm": 2.9308274918406254,
+      "learning_rate": 4.1046579338222604e-05,
+      "loss": 1.2961,
+      "step": 15405
+    },
+    {
+      "epoch": 0.7097249735108491,
+      "grad_norm": 1.0214884564032927,
+      "learning_rate": 4.103452774674398e-05,
+      "loss": 1.5633,
+      "step": 15406
+    },
+    {
+      "epoch": 0.709771041599484,
+      "grad_norm": 1.0296235924750348,
+      "learning_rate": 4.102247746803052e-05,
+      "loss": 1.3812,
+      "step": 15407
+    },
+    {
+      "epoch": 0.7098171096881191,
+      "grad_norm": 0.8332098402977541,
+      "learning_rate": 4.1010428502350426e-05,
+      "loss": 1.2109,
+      "step": 15408
+    },
+    {
+      "epoch": 0.709863177776754,
+      "grad_norm": 1.3049512952258693,
+      "learning_rate": 4.099838084997202e-05,
+      "loss": 1.0823,
+      "step": 15409
+    },
+    {
+      "epoch": 0.7099092458653891,
+      "grad_norm": 1.3365554473031152,
+      "learning_rate": 4.098633451116348e-05,
+      "loss": 1.4156,
+      "step": 15410
+    },
+    {
+      "epoch": 0.709955313954024,
+      "grad_norm": 4.74855756342577,
+      "learning_rate": 4.097428948619305e-05,
+      "loss": 1.2288,
+      "step": 15411
+    },
+    {
+      "epoch": 0.7100013820426591,
+      "grad_norm": 0.8427050426644027,
+      "learning_rate": 4.096224577532883e-05,
+      "loss": 1.3813,
+      "step": 15412
+    },
+    {
+      "epoch": 0.710047450131294,
+      "grad_norm": 0.9804183011186286,
+      "learning_rate": 4.0950203378838983e-05,
+      "loss": 1.4587,
+      "step": 15413
+    },
+    {
+      "epoch": 0.710093518219929,
+      "grad_norm": 1.3764927760393697,
+      "learning_rate": 4.093816229699164e-05,
+      "loss": 1.3492,
+      "step": 15414
+    },
+    {
+      "epoch": 0.7101395863085641,
+      "grad_norm": 0.889651931358464,
+      "learning_rate": 4.092612253005481e-05,
+      "loss": 1.2909,
+      "step": 15415
+    },
+    {
+      "epoch": 0.710185654397199,
+      "grad_norm": 0.7873494535784609,
+      "learning_rate": 4.0914084078296566e-05,
+      "loss": 1.4896,
+      "step": 15416
+    },
+    {
+      "epoch": 0.7102317224858341,
+      "grad_norm": 2.663732205011329,
+      "learning_rate": 4.090204694198494e-05,
+      "loss": 1.0801,
+      "step": 15417
+    },
+    {
+      "epoch": 0.710277790574469,
+      "grad_norm": 0.9760597834824201,
+      "learning_rate": 4.089001112138794e-05,
+      "loss": 1.5884,
+      "step": 15418
+    },
+    {
+      "epoch": 0.7103238586631041,
+      "grad_norm": 0.8511151114429886,
+      "learning_rate": 4.087797661677345e-05,
+      "loss": 1.549,
+      "step": 15419
+    },
+    {
+      "epoch": 0.710369926751739,
+      "grad_norm": 0.7917078965128485,
+      "learning_rate": 4.086594342840944e-05,
+      "loss": 1.3419,
+      "step": 15420
+    },
+    {
+      "epoch": 0.7104159948403741,
+      "grad_norm": 0.9278559255765679,
+      "learning_rate": 4.085391155656384e-05,
+      "loss": 1.3513,
+      "step": 15421
+    },
+    {
+      "epoch": 0.710462062929009,
+      "grad_norm": 0.8496811123369602,
+      "learning_rate": 4.084188100150445e-05,
+      "loss": 1.5029,
+      "step": 15422
+    },
+    {
+      "epoch": 0.7105081310176441,
+      "grad_norm": 1.5649963448614042,
+      "learning_rate": 4.082985176349915e-05,
+      "loss": 0.9124,
+      "step": 15423
+    },
+    {
+      "epoch": 0.710554199106279,
+      "grad_norm": 0.9897423126584395,
+      "learning_rate": 4.0817823842815737e-05,
+      "loss": 1.3079,
+      "step": 15424
+    },
+    {
+      "epoch": 0.7106002671949141,
+      "grad_norm": 1.9705682657082262,
+      "learning_rate": 4.080579723972204e-05,
+      "loss": 1.2602,
+      "step": 15425
+    },
+    {
+      "epoch": 0.710646335283549,
+      "grad_norm": 1.1086280187565951,
+      "learning_rate": 4.079377195448574e-05,
+      "loss": 1.4184,
+      "step": 15426
+    },
+    {
+      "epoch": 0.7106924033721841,
+      "grad_norm": 0.9237707288353607,
+      "learning_rate": 4.078174798737459e-05,
+      "loss": 1.1597,
+      "step": 15427
+    },
+    {
+      "epoch": 0.7107384714608191,
+      "grad_norm": 1.1288421900955221,
+      "learning_rate": 4.076972533865629e-05,
+      "loss": 1.1201,
+      "step": 15428
+    },
+    {
+      "epoch": 0.7107845395494541,
+      "grad_norm": 1.159790164860301,
+      "learning_rate": 4.075770400859852e-05,
+      "loss": 1.4993,
+      "step": 15429
+    },
+    {
+      "epoch": 0.7108306076380891,
+      "grad_norm": 1.8938442187685287,
+      "learning_rate": 4.0745683997468884e-05,
+      "loss": 1.3851,
+      "step": 15430
+    },
+    {
+      "epoch": 0.7108766757267241,
+      "grad_norm": 0.8997340495793452,
+      "learning_rate": 4.073366530553494e-05,
+      "loss": 1.5906,
+      "step": 15431
+    },
+    {
+      "epoch": 0.7109227438153591,
+      "grad_norm": 1.0732592605197566,
+      "learning_rate": 4.072164793306439e-05,
+      "loss": 0.9915,
+      "step": 15432
+    },
+    {
+      "epoch": 0.7109688119039941,
+      "grad_norm": 2.1032893373799944,
+      "learning_rate": 4.070963188032466e-05,
+      "loss": 1.5628,
+      "step": 15433
+    },
+    {
+      "epoch": 0.7110148799926291,
+      "grad_norm": 0.813345693456065,
+      "learning_rate": 4.069761714758332e-05,
+      "loss": 1.3532,
+      "step": 15434
+    },
+    {
+      "epoch": 0.7110609480812641,
+      "grad_norm": 0.9602188340012336,
+      "learning_rate": 4.068560373510786e-05,
+      "loss": 1.4234,
+      "step": 15435
+    },
+    {
+      "epoch": 0.7111070161698991,
+      "grad_norm": 0.7456132044612336,
+      "learning_rate": 4.0673591643165756e-05,
+      "loss": 1.1991,
+      "step": 15436
+    },
+    {
+      "epoch": 0.7111530842585341,
+      "grad_norm": 1.1171346983659993,
+      "learning_rate": 4.0661580872024386e-05,
+      "loss": 1.2538,
+      "step": 15437
+    },
+    {
+      "epoch": 0.7111991523471691,
+      "grad_norm": 0.8746149003366516,
+      "learning_rate": 4.064957142195116e-05,
+      "loss": 1.2251,
+      "step": 15438
+    },
+    {
+      "epoch": 0.7112452204358041,
+      "grad_norm": 2.761879038035928,
+      "learning_rate": 4.063756329321351e-05,
+      "loss": 1.1173,
+      "step": 15439
+    },
+    {
+      "epoch": 0.7112912885244391,
+      "grad_norm": 1.4277466889587145,
+      "learning_rate": 4.0625556486078685e-05,
+      "loss": 1.2902,
+      "step": 15440
+    },
+    {
+      "epoch": 0.7113373566130742,
+      "grad_norm": 0.9654651644943871,
+      "learning_rate": 4.061355100081408e-05,
+      "loss": 1.0948,
+      "step": 15441
+    },
+    {
+      "epoch": 0.7113834247017091,
+      "grad_norm": 2.2915320978329516,
+      "learning_rate": 4.060154683768687e-05,
+      "loss": 1.1752,
+      "step": 15442
+    },
+    {
+      "epoch": 0.7114294927903442,
+      "grad_norm": 1.6173595544542028,
+      "learning_rate": 4.058954399696443e-05,
+      "loss": 1.2528,
+      "step": 15443
+    },
+    {
+      "epoch": 0.7114755608789791,
+      "grad_norm": 1.2070633049132575,
+      "learning_rate": 4.0577542478913896e-05,
+      "loss": 1.1362,
+      "step": 15444
+    },
+    {
+      "epoch": 0.7115216289676142,
+      "grad_norm": 3.596163698508231,
+      "learning_rate": 4.056554228380253e-05,
+      "loss": 1.1191,
+      "step": 15445
+    },
+    {
+      "epoch": 0.7115676970562491,
+      "grad_norm": 3.3638191246664504,
+      "learning_rate": 4.0553543411897385e-05,
+      "loss": 1.2232,
+      "step": 15446
+    },
+    {
+      "epoch": 0.7116137651448842,
+      "grad_norm": 0.9010683469318012,
+      "learning_rate": 4.054154586346573e-05,
+      "loss": 1.47,
+      "step": 15447
+    },
+    {
+      "epoch": 0.7116598332335191,
+      "grad_norm": 1.1554831739000597,
+      "learning_rate": 4.052954963877459e-05,
+      "loss": 1.583,
+      "step": 15448
+    },
+    {
+      "epoch": 0.7117059013221542,
+      "grad_norm": 0.9897302282191677,
+      "learning_rate": 4.0517554738091046e-05,
+      "loss": 1.3541,
+      "step": 15449
+    },
+    {
+      "epoch": 0.7117519694107891,
+      "grad_norm": 1.5284195569146637,
+      "learning_rate": 4.0505561161682185e-05,
+      "loss": 1.0054,
+      "step": 15450
+    },
+    {
+      "epoch": 0.7117980374994242,
+      "grad_norm": 2.40121865530174,
+      "learning_rate": 4.049356890981496e-05,
+      "loss": 1.37,
+      "step": 15451
+    },
+    {
+      "epoch": 0.7118441055880591,
+      "grad_norm": 0.9961661908689857,
+      "learning_rate": 4.048157798275644e-05,
+      "loss": 1.4409,
+      "step": 15452
+    },
+    {
+      "epoch": 0.7118901736766942,
+      "grad_norm": 0.7211316728854273,
+      "learning_rate": 4.0469588380773446e-05,
+      "loss": 1.3902,
+      "step": 15453
+    },
+    {
+      "epoch": 0.7119362417653291,
+      "grad_norm": 1.7723207389303937,
+      "learning_rate": 4.0457600104133076e-05,
+      "loss": 1.3046,
+      "step": 15454
+    },
+    {
+      "epoch": 0.7119823098539642,
+      "grad_norm": 0.8439806472475827,
+      "learning_rate": 4.044561315310211e-05,
+      "loss": 1.2991,
+      "step": 15455
+    },
+    {
+      "epoch": 0.7120283779425992,
+      "grad_norm": 0.7183617351830729,
+      "learning_rate": 4.043362752794748e-05,
+      "loss": 1.3321,
+      "step": 15456
+    },
+    {
+      "epoch": 0.7120744460312342,
+      "grad_norm": 0.7650119455272095,
+      "learning_rate": 4.042164322893595e-05,
+      "loss": 1.4923,
+      "step": 15457
+    },
+    {
+      "epoch": 0.7121205141198692,
+      "grad_norm": 1.3610061147999368,
+      "learning_rate": 4.040966025633443e-05,
+      "loss": 1.1162,
+      "step": 15458
+    },
+    {
+      "epoch": 0.7121665822085042,
+      "grad_norm": 0.9874585340881006,
+      "learning_rate": 4.039767861040965e-05,
+      "loss": 1.317,
+      "step": 15459
+    },
+    {
+      "epoch": 0.7122126502971392,
+      "grad_norm": 0.7802939072815496,
+      "learning_rate": 4.038569829142829e-05,
+      "loss": 1.3033,
+      "step": 15460
+    },
+    {
+      "epoch": 0.7122587183857741,
+      "grad_norm": 0.732863224801346,
+      "learning_rate": 4.037371929965721e-05,
+      "loss": 1.4522,
+      "step": 15461
+    },
+    {
+      "epoch": 0.7123047864744092,
+      "grad_norm": 1.403700167838238,
+      "learning_rate": 4.0361741635363e-05,
+      "loss": 1.2448,
+      "step": 15462
+    },
+    {
+      "epoch": 0.7123508545630441,
+      "grad_norm": 0.8481315062689807,
+      "learning_rate": 4.034976529881239e-05,
+      "loss": 1.2426,
+      "step": 15463
+    },
+    {
+      "epoch": 0.7123969226516792,
+      "grad_norm": 1.0587208631089025,
+      "learning_rate": 4.0337790290271914e-05,
+      "loss": 1.3626,
+      "step": 15464
+    },
+    {
+      "epoch": 0.7124429907403141,
+      "grad_norm": 1.1267759539461657,
+      "learning_rate": 4.0325816610008304e-05,
+      "loss": 0.9404,
+      "step": 15465
+    },
+    {
+      "epoch": 0.7124890588289492,
+      "grad_norm": 0.8445725240414069,
+      "learning_rate": 4.031384425828803e-05,
+      "loss": 1.1531,
+      "step": 15466
+    },
+    {
+      "epoch": 0.7125351269175841,
+      "grad_norm": 1.7558758916708934,
+      "learning_rate": 4.030187323537771e-05,
+      "loss": 1.4186,
+      "step": 15467
+    },
+    {
+      "epoch": 0.7125811950062192,
+      "grad_norm": 0.8641011494709067,
+      "learning_rate": 4.028990354154378e-05,
+      "loss": 1.3673,
+      "step": 15468
+    },
+    {
+      "epoch": 0.7126272630948542,
+      "grad_norm": 0.7890408620738316,
+      "learning_rate": 4.0277935177052775e-05,
+      "loss": 1.1304,
+      "step": 15469
+    },
+    {
+      "epoch": 0.7126733311834892,
+      "grad_norm": 0.7759063390092578,
+      "learning_rate": 4.0265968142171176e-05,
+      "loss": 1.3312,
+      "step": 15470
+    },
+    {
+      "epoch": 0.7127193992721242,
+      "grad_norm": 1.6725046566829262,
+      "learning_rate": 4.025400243716534e-05,
+      "loss": 1.188,
+      "step": 15471
+    },
+    {
+      "epoch": 0.7127654673607592,
+      "grad_norm": 2.1681343383740326,
+      "learning_rate": 4.02420380623017e-05,
+      "loss": 1.2978,
+      "step": 15472
+    },
+    {
+      "epoch": 0.7128115354493942,
+      "grad_norm": 0.77460899848531,
+      "learning_rate": 4.0230075017846616e-05,
+      "loss": 1.3878,
+      "step": 15473
+    },
+    {
+      "epoch": 0.7128576035380292,
+      "grad_norm": 0.9329605585815143,
+      "learning_rate": 4.021811330406646e-05,
+      "loss": 1.2852,
+      "step": 15474
+    },
+    {
+      "epoch": 0.7129036716266642,
+      "grad_norm": 0.9257553657997158,
+      "learning_rate": 4.020615292122743e-05,
+      "loss": 1.3002,
+      "step": 15475
+    },
+    {
+      "epoch": 0.7129497397152992,
+      "grad_norm": 0.6432806463691041,
+      "learning_rate": 4.019419386959598e-05,
+      "loss": 1.2604,
+      "step": 15476
+    },
+    {
+      "epoch": 0.7129958078039342,
+      "grad_norm": 0.735135916724903,
+      "learning_rate": 4.0182236149438236e-05,
+      "loss": 1.3178,
+      "step": 15477
+    },
+    {
+      "epoch": 0.7130418758925692,
+      "grad_norm": 27.74661857260439,
+      "learning_rate": 4.0170279761020404e-05,
+      "loss": 1.6007,
+      "step": 15478
+    },
+    {
+      "epoch": 0.7130879439812042,
+      "grad_norm": 0.9746946347853518,
+      "learning_rate": 4.0158324704608705e-05,
+      "loss": 1.1604,
+      "step": 15479
+    },
+    {
+      "epoch": 0.7131340120698392,
+      "grad_norm": 1.0152532040930777,
+      "learning_rate": 4.014637098046932e-05,
+      "loss": 1.2249,
+      "step": 15480
+    },
+    {
+      "epoch": 0.7131800801584742,
+      "grad_norm": 0.7549858996934481,
+      "learning_rate": 4.013441858886837e-05,
+      "loss": 1.475,
+      "step": 15481
+    },
+    {
+      "epoch": 0.7132261482471093,
+      "grad_norm": 0.887643517265509,
+      "learning_rate": 4.0122467530071915e-05,
+      "loss": 1.6054,
+      "step": 15482
+    },
+    {
+      "epoch": 0.7132722163357442,
+      "grad_norm": 0.7930569399227919,
+      "learning_rate": 4.011051780434606e-05,
+      "loss": 1.0604,
+      "step": 15483
+    },
+    {
+      "epoch": 0.7133182844243793,
+      "grad_norm": 1.2005938872622126,
+      "learning_rate": 4.009856941195683e-05,
+      "loss": 1.3339,
+      "step": 15484
+    },
+    {
+      "epoch": 0.7133643525130142,
+      "grad_norm": 0.7400425428059157,
+      "learning_rate": 4.0086622353170276e-05,
+      "loss": 1.3894,
+      "step": 15485
+    },
+    {
+      "epoch": 0.7134104206016493,
+      "grad_norm": 0.9380854589406648,
+      "learning_rate": 4.007467662825232e-05,
+      "loss": 1.5236,
+      "step": 15486
+    },
+    {
+      "epoch": 0.7134564886902842,
+      "grad_norm": 1.0417267096763498,
+      "learning_rate": 4.0062732237468925e-05,
+      "loss": 1.4502,
+      "step": 15487
+    },
+    {
+      "epoch": 0.7135025567789193,
+      "grad_norm": 0.7797181740675698,
+      "learning_rate": 4.005078918108608e-05,
+      "loss": 1.1929,
+      "step": 15488
+    },
+    {
+      "epoch": 0.7135486248675542,
+      "grad_norm": 0.845982693643884,
+      "learning_rate": 4.003884745936958e-05,
+      "loss": 1.2639,
+      "step": 15489
+    },
+    {
+      "epoch": 0.7135946929561893,
+      "grad_norm": 0.7992962936438784,
+      "learning_rate": 4.0026907072585317e-05,
+      "loss": 1.4104,
+      "step": 15490
+    },
+    {
+      "epoch": 0.7136407610448242,
+      "grad_norm": 1.0428364580549934,
+      "learning_rate": 4.0014968020999154e-05,
+      "loss": 1.5663,
+      "step": 15491
+    },
+    {
+      "epoch": 0.7136868291334593,
+      "grad_norm": 0.7595978394979818,
+      "learning_rate": 4.00030303048769e-05,
+      "loss": 1.1144,
+      "step": 15492
+    },
+    {
+      "epoch": 0.7137328972220942,
+      "grad_norm": 0.8495426691308945,
+      "learning_rate": 3.999109392448427e-05,
+      "loss": 1.2231,
+      "step": 15493
+    },
+    {
+      "epoch": 0.7137789653107293,
+      "grad_norm": 0.8895304295225601,
+      "learning_rate": 3.997915888008704e-05,
+      "loss": 1.5059,
+      "step": 15494
+    },
+    {
+      "epoch": 0.7138250333993642,
+      "grad_norm": 1.2898875105225451,
+      "learning_rate": 3.9967225171950976e-05,
+      "loss": 1.257,
+      "step": 15495
+    },
+    {
+      "epoch": 0.7138711014879993,
+      "grad_norm": 1.0552334943513173,
+      "learning_rate": 3.995529280034165e-05,
+      "loss": 1.05,
+      "step": 15496
+    },
+    {
+      "epoch": 0.7139171695766343,
+      "grad_norm": 1.0378804339300924,
+      "learning_rate": 3.994336176552479e-05,
+      "loss": 0.9821,
+      "step": 15497
+    },
+    {
+      "epoch": 0.7139632376652693,
+      "grad_norm": 1.12307323928894,
+      "learning_rate": 3.993143206776599e-05,
+      "loss": 1.1648,
+      "step": 15498
+    },
+    {
+      "epoch": 0.7140093057539043,
+      "grad_norm": 1.389652052184593,
+      "learning_rate": 3.991950370733092e-05,
+      "loss": 1.2838,
+      "step": 15499
+    },
+    {
+      "epoch": 0.7140553738425393,
+      "grad_norm": 0.7390978761512697,
+      "learning_rate": 3.9907576684485026e-05,
+      "loss": 1.2116,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7141014419311743,
+      "grad_norm": 1.1154614799508547,
+      "learning_rate": 3.989565099949392e-05,
+      "loss": 1.5028,
+      "step": 15501
+    },
+    {
+      "epoch": 0.7141475100198093,
+      "grad_norm": 2.025247342582644,
+      "learning_rate": 3.988372665262307e-05,
+      "loss": 1.24,
+      "step": 15502
+    },
+    {
+      "epoch": 0.7141935781084443,
+      "grad_norm": 1.0235778436462195,
+      "learning_rate": 3.9871803644138004e-05,
+      "loss": 1.2792,
+      "step": 15503
+    },
+    {
+      "epoch": 0.7142396461970792,
+      "grad_norm": 0.8281320697430347,
+      "learning_rate": 3.98598819743041e-05,
+      "loss": 1.4718,
+      "step": 15504
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.8791771010214764,
+      "learning_rate": 3.984796164338682e-05,
+      "loss": 1.0701,
+      "step": 15505
+    },
+    {
+      "epoch": 0.7143317823743492,
+      "grad_norm": 0.7045215267532328,
+      "learning_rate": 3.983604265165156e-05,
+      "loss": 1.3412,
+      "step": 15506
+    },
+    {
+      "epoch": 0.7143778504629843,
+      "grad_norm": 0.8237472191437184,
+      "learning_rate": 3.982412499936361e-05,
+      "loss": 1.3533,
+      "step": 15507
+    },
+    {
+      "epoch": 0.7144239185516192,
+      "grad_norm": 0.8334286099037759,
+      "learning_rate": 3.981220868678834e-05,
+      "loss": 1.5793,
+      "step": 15508
+    },
+    {
+      "epoch": 0.7144699866402543,
+      "grad_norm": 0.8585205386308696,
+      "learning_rate": 3.980029371419105e-05,
+      "loss": 1.326,
+      "step": 15509
+    },
+    {
+      "epoch": 0.7145160547288893,
+      "grad_norm": 1.0763905135157599,
+      "learning_rate": 3.978838008183704e-05,
+      "loss": 1.4062,
+      "step": 15510
+    },
+    {
+      "epoch": 0.7145621228175243,
+      "grad_norm": 1.0055502642083793,
+      "learning_rate": 3.9776467789991456e-05,
+      "loss": 1.5626,
+      "step": 15511
+    },
+    {
+      "epoch": 0.7146081909061593,
+      "grad_norm": 0.6171954134312454,
+      "learning_rate": 3.9764556838919566e-05,
+      "loss": 1.1654,
+      "step": 15512
+    },
+    {
+      "epoch": 0.7146542589947943,
+      "grad_norm": 0.9599549088626759,
+      "learning_rate": 3.975264722888656e-05,
+      "loss": 1.2013,
+      "step": 15513
+    },
+    {
+      "epoch": 0.7147003270834293,
+      "grad_norm": 1.2010697689015837,
+      "learning_rate": 3.974073896015754e-05,
+      "loss": 1.3081,
+      "step": 15514
+    },
+    {
+      "epoch": 0.7147463951720643,
+      "grad_norm": 0.7753739324487702,
+      "learning_rate": 3.9728832032997645e-05,
+      "loss": 1.2908,
+      "step": 15515
+    },
+    {
+      "epoch": 0.7147924632606993,
+      "grad_norm": 0.7523423666753212,
+      "learning_rate": 3.971692644767195e-05,
+      "loss": 1.2326,
+      "step": 15516
+    },
+    {
+      "epoch": 0.7148385313493343,
+      "grad_norm": 0.923953910527598,
+      "learning_rate": 3.9705022204445565e-05,
+      "loss": 1.1629,
+      "step": 15517
+    },
+    {
+      "epoch": 0.7148845994379693,
+      "grad_norm": 0.9379510591072203,
+      "learning_rate": 3.9693119303583445e-05,
+      "loss": 1.2476,
+      "step": 15518
+    },
+    {
+      "epoch": 0.7149306675266043,
+      "grad_norm": 0.6917709743567549,
+      "learning_rate": 3.968121774535062e-05,
+      "loss": 1.2868,
+      "step": 15519
+    },
+    {
+      "epoch": 0.7149767356152393,
+      "grad_norm": 0.6967358660416532,
+      "learning_rate": 3.966931753001205e-05,
+      "loss": 1.5627,
+      "step": 15520
+    },
+    {
+      "epoch": 0.7150228037038743,
+      "grad_norm": 0.9484866302640235,
+      "learning_rate": 3.965741865783272e-05,
+      "loss": 1.2024,
+      "step": 15521
+    },
+    {
+      "epoch": 0.7150688717925093,
+      "grad_norm": 0.6840637935973324,
+      "learning_rate": 3.964552112907747e-05,
+      "loss": 1.3167,
+      "step": 15522
+    },
+    {
+      "epoch": 0.7151149398811444,
+      "grad_norm": 0.7798969735309338,
+      "learning_rate": 3.963362494401119e-05,
+      "loss": 1.0915,
+      "step": 15523
+    },
+    {
+      "epoch": 0.7151610079697793,
+      "grad_norm": 0.7241770251945826,
+      "learning_rate": 3.962173010289879e-05,
+      "loss": 0.9662,
+      "step": 15524
+    },
+    {
+      "epoch": 0.7152070760584144,
+      "grad_norm": 1.2239749720533097,
+      "learning_rate": 3.960983660600501e-05,
+      "loss": 1.4039,
+      "step": 15525
+    },
+    {
+      "epoch": 0.7152531441470493,
+      "grad_norm": 1.0695878815848079,
+      "learning_rate": 3.959794445359467e-05,
+      "loss": 1.4076,
+      "step": 15526
+    },
+    {
+      "epoch": 0.7152992122356844,
+      "grad_norm": 0.7597793726606138,
+      "learning_rate": 3.958605364593252e-05,
+      "loss": 1.1336,
+      "step": 15527
+    },
+    {
+      "epoch": 0.7153452803243193,
+      "grad_norm": 1.287379346425624,
+      "learning_rate": 3.957416418328335e-05,
+      "loss": 1.0013,
+      "step": 15528
+    },
+    {
+      "epoch": 0.7153913484129544,
+      "grad_norm": 1.4040543301524093,
+      "learning_rate": 3.9562276065911754e-05,
+      "loss": 1.2114,
+      "step": 15529
+    },
+    {
+      "epoch": 0.7154374165015893,
+      "grad_norm": 65.67600567895478,
+      "learning_rate": 3.9550389294082456e-05,
+      "loss": 1.5349,
+      "step": 15530
+    },
+    {
+      "epoch": 0.7154834845902244,
+      "grad_norm": 1.0856319135755332,
+      "learning_rate": 3.9538503868060136e-05,
+      "loss": 1.5208,
+      "step": 15531
+    },
+    {
+      "epoch": 0.7155295526788593,
+      "grad_norm": 1.0382178497617836,
+      "learning_rate": 3.952661978810932e-05,
+      "loss": 1.4926,
+      "step": 15532
+    },
+    {
+      "epoch": 0.7155756207674944,
+      "grad_norm": 1.3161979947736306,
+      "learning_rate": 3.951473705449462e-05,
+      "loss": 1.5662,
+      "step": 15533
+    },
+    {
+      "epoch": 0.7156216888561293,
+      "grad_norm": 0.8578841069507341,
+      "learning_rate": 3.9502855667480585e-05,
+      "loss": 1.2933,
+      "step": 15534
+    },
+    {
+      "epoch": 0.7156677569447644,
+      "grad_norm": 0.8637538150953966,
+      "learning_rate": 3.949097562733178e-05,
+      "loss": 1.2845,
+      "step": 15535
+    },
+    {
+      "epoch": 0.7157138250333993,
+      "grad_norm": 0.8587289225152089,
+      "learning_rate": 3.947909693431262e-05,
+      "loss": 1.1456,
+      "step": 15536
+    },
+    {
+      "epoch": 0.7157598931220344,
+      "grad_norm": 1.2011435580102177,
+      "learning_rate": 3.94672195886876e-05,
+      "loss": 1.2522,
+      "step": 15537
+    },
+    {
+      "epoch": 0.7158059612106694,
+      "grad_norm": 0.8942488982214999,
+      "learning_rate": 3.9455343590721136e-05,
+      "loss": 1.7882,
+      "step": 15538
+    },
+    {
+      "epoch": 0.7158520292993044,
+      "grad_norm": 0.9502850959987974,
+      "learning_rate": 3.944346894067767e-05,
+      "loss": 1.4877,
+      "step": 15539
+    },
+    {
+      "epoch": 0.7158980973879394,
+      "grad_norm": 0.9153061458625468,
+      "learning_rate": 3.943159563882154e-05,
+      "loss": 1.2662,
+      "step": 15540
+    },
+    {
+      "epoch": 0.7159441654765744,
+      "grad_norm": 0.9906502524016302,
+      "learning_rate": 3.941972368541701e-05,
+      "loss": 1.4212,
+      "step": 15541
+    },
+    {
+      "epoch": 0.7159902335652094,
+      "grad_norm": 1.2052835758961409,
+      "learning_rate": 3.940785308072852e-05,
+      "loss": 1.1759,
+      "step": 15542
+    },
+    {
+      "epoch": 0.7160363016538444,
+      "grad_norm": 1.1563245880371633,
+      "learning_rate": 3.939598382502027e-05,
+      "loss": 1.5734,
+      "step": 15543
+    },
+    {
+      "epoch": 0.7160823697424794,
+      "grad_norm": 1.008735426026556,
+      "learning_rate": 3.938411591855654e-05,
+      "loss": 1.4198,
+      "step": 15544
+    },
+    {
+      "epoch": 0.7161284378311144,
+      "grad_norm": 1.1034323135952815,
+      "learning_rate": 3.937224936160148e-05,
+      "loss": 1.276,
+      "step": 15545
+    },
+    {
+      "epoch": 0.7161745059197494,
+      "grad_norm": 1.0316639573936093,
+      "learning_rate": 3.9360384154419405e-05,
+      "loss": 1.2253,
+      "step": 15546
+    },
+    {
+      "epoch": 0.7162205740083843,
+      "grad_norm": 0.9294537803533621,
+      "learning_rate": 3.9348520297274364e-05,
+      "loss": 1.3915,
+      "step": 15547
+    },
+    {
+      "epoch": 0.7162666420970194,
+      "grad_norm": 0.8598734275401986,
+      "learning_rate": 3.933665779043052e-05,
+      "loss": 1.3489,
+      "step": 15548
+    },
+    {
+      "epoch": 0.7163127101856543,
+      "grad_norm": 0.8668861470163707,
+      "learning_rate": 3.932479663415201e-05,
+      "loss": 1.5803,
+      "step": 15549
+    },
+    {
+      "epoch": 0.7163587782742894,
+      "grad_norm": 0.8782792160349194,
+      "learning_rate": 3.931293682870283e-05,
+      "loss": 1.38,
+      "step": 15550
+    },
+    {
+      "epoch": 0.7164048463629245,
+      "grad_norm": 0.8395000502367443,
+      "learning_rate": 3.93010783743471e-05,
+      "loss": 1.6157,
+      "step": 15551
+    },
+    {
+      "epoch": 0.7164509144515594,
+      "grad_norm": 1.1334585117757412,
+      "learning_rate": 3.9289221271348706e-05,
+      "loss": 0.8987,
+      "step": 15552
+    },
+    {
+      "epoch": 0.7164969825401944,
+      "grad_norm": 0.8865180468110943,
+      "learning_rate": 3.9277365519971777e-05,
+      "loss": 1.427,
+      "step": 15553
+    },
+    {
+      "epoch": 0.7165430506288294,
+      "grad_norm": 1.0865879439899677,
+      "learning_rate": 3.9265511120480146e-05,
+      "loss": 1.5701,
+      "step": 15554
+    },
+    {
+      "epoch": 0.7165891187174644,
+      "grad_norm": 0.8258480998479317,
+      "learning_rate": 3.9253658073137824e-05,
+      "loss": 1.5953,
+      "step": 15555
+    },
+    {
+      "epoch": 0.7166351868060994,
+      "grad_norm": 1.095764210885297,
+      "learning_rate": 3.9241806378208557e-05,
+      "loss": 1.6202,
+      "step": 15556
+    },
+    {
+      "epoch": 0.7166812548947344,
+      "grad_norm": 1.0529945917848116,
+      "learning_rate": 3.9229956035956374e-05,
+      "loss": 1.4789,
+      "step": 15557
+    },
+    {
+      "epoch": 0.7167273229833694,
+      "grad_norm": 1.156970054864331,
+      "learning_rate": 3.921810704664502e-05,
+      "loss": 1.2093,
+      "step": 15558
+    },
+    {
+      "epoch": 0.7167733910720044,
+      "grad_norm": 0.792786634592525,
+      "learning_rate": 3.920625941053823e-05,
+      "loss": 1.2494,
+      "step": 15559
+    },
+    {
+      "epoch": 0.7168194591606394,
+      "grad_norm": 0.8356475234565458,
+      "learning_rate": 3.9194413127899896e-05,
+      "loss": 1.4913,
+      "step": 15560
+    },
+    {
+      "epoch": 0.7168655272492744,
+      "grad_norm": 1.2013447104155826,
+      "learning_rate": 3.9182568198993666e-05,
+      "loss": 1.3599,
+      "step": 15561
+    },
+    {
+      "epoch": 0.7169115953379094,
+      "grad_norm": 0.8981831444915264,
+      "learning_rate": 3.917072462408331e-05,
+      "loss": 1.1143,
+      "step": 15562
+    },
+    {
+      "epoch": 0.7169576634265444,
+      "grad_norm": 0.9576304388976751,
+      "learning_rate": 3.9158882403432394e-05,
+      "loss": 1.2104,
+      "step": 15563
+    },
+    {
+      "epoch": 0.7170037315151795,
+      "grad_norm": 0.697836580195668,
+      "learning_rate": 3.914704153730472e-05,
+      "loss": 1.5719,
+      "step": 15564
+    },
+    {
+      "epoch": 0.7170497996038144,
+      "grad_norm": 0.8243509360220955,
+      "learning_rate": 3.913520202596378e-05,
+      "loss": 0.9327,
+      "step": 15565
+    },
+    {
+      "epoch": 0.7170958676924495,
+      "grad_norm": 0.7585574749112297,
+      "learning_rate": 3.912336386967325e-05,
+      "loss": 1.3607,
+      "step": 15566
+    },
+    {
+      "epoch": 0.7171419357810844,
+      "grad_norm": 0.9781527510432279,
+      "learning_rate": 3.911152706869656e-05,
+      "loss": 1.3166,
+      "step": 15567
+    },
+    {
+      "epoch": 0.7171880038697195,
+      "grad_norm": 0.8611090433259689,
+      "learning_rate": 3.909969162329741e-05,
+      "loss": 1.1954,
+      "step": 15568
+    },
+    {
+      "epoch": 0.7172340719583544,
+      "grad_norm": 0.8178025093530294,
+      "learning_rate": 3.908785753373919e-05,
+      "loss": 1.4243,
+      "step": 15569
+    },
+    {
+      "epoch": 0.7172801400469895,
+      "grad_norm": 0.8865330616781586,
+      "learning_rate": 3.907602480028535e-05,
+      "loss": 1.2551,
+      "step": 15570
+    },
+    {
+      "epoch": 0.7173262081356244,
+      "grad_norm": 0.8266800451409144,
+      "learning_rate": 3.906419342319935e-05,
+      "loss": 1.4516,
+      "step": 15571
+    },
+    {
+      "epoch": 0.7173722762242595,
+      "grad_norm": 0.9497704083352759,
+      "learning_rate": 3.9052363402744594e-05,
+      "loss": 1.216,
+      "step": 15572
+    },
+    {
+      "epoch": 0.7174183443128944,
+      "grad_norm": 0.8897675919698386,
+      "learning_rate": 3.904053473918452e-05,
+      "loss": 1.3333,
+      "step": 15573
+    },
+    {
+      "epoch": 0.7174644124015295,
+      "grad_norm": 0.8373365320180146,
+      "learning_rate": 3.902870743278233e-05,
+      "loss": 1.0352,
+      "step": 15574
+    },
+    {
+      "epoch": 0.7175104804901644,
+      "grad_norm": 0.7382426939369554,
+      "learning_rate": 3.9016881483801506e-05,
+      "loss": 1.4625,
+      "step": 15575
+    },
+    {
+      "epoch": 0.7175565485787995,
+      "grad_norm": 1.267872866171311,
+      "learning_rate": 3.900505689250521e-05,
+      "loss": 1.3837,
+      "step": 15576
+    },
+    {
+      "epoch": 0.7176026166674345,
+      "grad_norm": 0.6573793746497733,
+      "learning_rate": 3.899323365915678e-05,
+      "loss": 0.9134,
+      "step": 15577
+    },
+    {
+      "epoch": 0.7176486847560695,
+      "grad_norm": 1.3155711650430815,
+      "learning_rate": 3.8981411784019364e-05,
+      "loss": 1.2778,
+      "step": 15578
+    },
+    {
+      "epoch": 0.7176947528447045,
+      "grad_norm": 0.9142199029258476,
+      "learning_rate": 3.8969591267356174e-05,
+      "loss": 1.1695,
+      "step": 15579
+    },
+    {
+      "epoch": 0.7177408209333395,
+      "grad_norm": 2.1786018031790033,
+      "learning_rate": 3.895777210943045e-05,
+      "loss": 1.3411,
+      "step": 15580
+    },
+    {
+      "epoch": 0.7177868890219745,
+      "grad_norm": 0.9708790463927032,
+      "learning_rate": 3.8945954310505206e-05,
+      "loss": 1.5187,
+      "step": 15581
+    },
+    {
+      "epoch": 0.7178329571106095,
+      "grad_norm": 1.0639644393765444,
+      "learning_rate": 3.893413787084361e-05,
+      "loss": 1.5556,
+      "step": 15582
+    },
+    {
+      "epoch": 0.7178790251992445,
+      "grad_norm": 0.903801948576063,
+      "learning_rate": 3.8922322790708745e-05,
+      "loss": 1.4743,
+      "step": 15583
+    },
+    {
+      "epoch": 0.7179250932878795,
+      "grad_norm": 0.9410381330339126,
+      "learning_rate": 3.8910509070363655e-05,
+      "loss": 1.3171,
+      "step": 15584
+    },
+    {
+      "epoch": 0.7179711613765145,
+      "grad_norm": 0.7294432879421517,
+      "learning_rate": 3.889869671007127e-05,
+      "loss": 1.2526,
+      "step": 15585
+    },
+    {
+      "epoch": 0.7180172294651495,
+      "grad_norm": 0.7489949913152132,
+      "learning_rate": 3.8886885710094704e-05,
+      "loss": 1.0697,
+      "step": 15586
+    },
+    {
+      "epoch": 0.7180632975537845,
+      "grad_norm": 0.9594481425159136,
+      "learning_rate": 3.887507607069685e-05,
+      "loss": 1.033,
+      "step": 15587
+    },
+    {
+      "epoch": 0.7181093656424195,
+      "grad_norm": 0.7258952932721144,
+      "learning_rate": 3.886326779214058e-05,
+      "loss": 1.4237,
+      "step": 15588
+    },
+    {
+      "epoch": 0.7181554337310545,
+      "grad_norm": 0.8106481088727495,
+      "learning_rate": 3.885146087468882e-05,
+      "loss": 1.1143,
+      "step": 15589
+    },
+    {
+      "epoch": 0.7182015018196894,
+      "grad_norm": 0.6131518261405919,
+      "learning_rate": 3.883965531860444e-05,
+      "loss": 1.689,
+      "step": 15590
+    },
+    {
+      "epoch": 0.7182475699083245,
+      "grad_norm": 0.9258877910822785,
+      "learning_rate": 3.8827851124150304e-05,
+      "loss": 1.6239,
+      "step": 15591
+    },
+    {
+      "epoch": 0.7182936379969596,
+      "grad_norm": 1.2136626919501081,
+      "learning_rate": 3.8816048291589134e-05,
+      "loss": 1.1924,
+      "step": 15592
+    },
+    {
+      "epoch": 0.7183397060855945,
+      "grad_norm": 0.8041039311197457,
+      "learning_rate": 3.8804246821183754e-05,
+      "loss": 1.29,
+      "step": 15593
+    },
+    {
+      "epoch": 0.7183857741742296,
+      "grad_norm": 0.9506147157721346,
+      "learning_rate": 3.879244671319689e-05,
+      "loss": 1.7656,
+      "step": 15594
+    },
+    {
+      "epoch": 0.7184318422628645,
+      "grad_norm": 0.823546081403264,
+      "learning_rate": 3.878064796789127e-05,
+      "loss": 1.4958,
+      "step": 15595
+    },
+    {
+      "epoch": 0.7184779103514995,
+      "grad_norm": 0.8889040774896592,
+      "learning_rate": 3.8768850585529527e-05,
+      "loss": 0.975,
+      "step": 15596
+    },
+    {
+      "epoch": 0.7185239784401345,
+      "grad_norm": 0.8373084019439803,
+      "learning_rate": 3.875705456637434e-05,
+      "loss": 1.4592,
+      "step": 15597
+    },
+    {
+      "epoch": 0.7185700465287695,
+      "grad_norm": 1.0420898834048336,
+      "learning_rate": 3.874525991068836e-05,
+      "loss": 1.2481,
+      "step": 15598
+    },
+    {
+      "epoch": 0.7186161146174045,
+      "grad_norm": 0.93590801053473,
+      "learning_rate": 3.873346661873409e-05,
+      "loss": 1.4212,
+      "step": 15599
+    },
+    {
+      "epoch": 0.7186621827060395,
+      "grad_norm": 0.8487220944482404,
+      "learning_rate": 3.872167469077415e-05,
+      "loss": 1.3108,
+      "step": 15600
+    },
+    {
+      "epoch": 0.7187082507946745,
+      "grad_norm": 1.853689724990382,
+      "learning_rate": 3.870988412707105e-05,
+      "loss": 1.2727,
+      "step": 15601
+    },
+    {
+      "epoch": 0.7187543188833095,
+      "grad_norm": 0.8688024797190598,
+      "learning_rate": 3.869809492788733e-05,
+      "loss": 0.9518,
+      "step": 15602
+    },
+    {
+      "epoch": 0.7188003869719445,
+      "grad_norm": 0.8946051519824354,
+      "learning_rate": 3.8686307093485375e-05,
+      "loss": 1.4631,
+      "step": 15603
+    },
+    {
+      "epoch": 0.7188464550605795,
+      "grad_norm": 0.778983360204137,
+      "learning_rate": 3.867452062412766e-05,
+      "loss": 1.3658,
+      "step": 15604
+    },
+    {
+      "epoch": 0.7188925231492146,
+      "grad_norm": 3.3937606773258544,
+      "learning_rate": 3.866273552007663e-05,
+      "loss": 1.1393,
+      "step": 15605
+    },
+    {
+      "epoch": 0.7189385912378495,
+      "grad_norm": 1.0152828974754875,
+      "learning_rate": 3.865095178159459e-05,
+      "loss": 1.1231,
+      "step": 15606
+    },
+    {
+      "epoch": 0.7189846593264846,
+      "grad_norm": 1.2980244229653566,
+      "learning_rate": 3.863916940894391e-05,
+      "loss": 1.4139,
+      "step": 15607
+    },
+    {
+      "epoch": 0.7190307274151195,
+      "grad_norm": 0.8434744710112282,
+      "learning_rate": 3.862738840238692e-05,
+      "loss": 1.3179,
+      "step": 15608
+    },
+    {
+      "epoch": 0.7190767955037546,
+      "grad_norm": 0.7364549653712099,
+      "learning_rate": 3.8615608762185927e-05,
+      "loss": 1.2843,
+      "step": 15609
+    },
+    {
+      "epoch": 0.7191228635923895,
+      "grad_norm": 1.1851067317178452,
+      "learning_rate": 3.860383048860312e-05,
+      "loss": 1.3809,
+      "step": 15610
+    },
+    {
+      "epoch": 0.7191689316810246,
+      "grad_norm": 1.0784435049040737,
+      "learning_rate": 3.8592053581900756e-05,
+      "loss": 0.8774,
+      "step": 15611
+    },
+    {
+      "epoch": 0.7192149997696595,
+      "grad_norm": 1.0622962132825213,
+      "learning_rate": 3.858027804234102e-05,
+      "loss": 1.2868,
+      "step": 15612
+    },
+    {
+      "epoch": 0.7192610678582946,
+      "grad_norm": 0.9612742513900518,
+      "learning_rate": 3.8568503870186135e-05,
+      "loss": 1.2643,
+      "step": 15613
+    },
+    {
+      "epoch": 0.7193071359469295,
+      "grad_norm": 1.2368888581865254,
+      "learning_rate": 3.8556731065698126e-05,
+      "loss": 1.1723,
+      "step": 15614
+    },
+    {
+      "epoch": 0.7193532040355646,
+      "grad_norm": 0.673400693760901,
+      "learning_rate": 3.8544959629139155e-05,
+      "loss": 1.1794,
+      "step": 15615
+    },
+    {
+      "epoch": 0.7193992721241995,
+      "grad_norm": 0.7222305927360181,
+      "learning_rate": 3.853318956077131e-05,
+      "loss": 1.327,
+      "step": 15616
+    },
+    {
+      "epoch": 0.7194453402128346,
+      "grad_norm": 1.660317944738075,
+      "learning_rate": 3.852142086085658e-05,
+      "loss": 1.4261,
+      "step": 15617
+    },
+    {
+      "epoch": 0.7194914083014696,
+      "grad_norm": 0.8517300072559005,
+      "learning_rate": 3.8509653529657006e-05,
+      "loss": 1.4311,
+      "step": 15618
+    },
+    {
+      "epoch": 0.7195374763901046,
+      "grad_norm": 1.6214315927537697,
+      "learning_rate": 3.8497887567434555e-05,
+      "loss": 1.1421,
+      "step": 15619
+    },
+    {
+      "epoch": 0.7195835444787396,
+      "grad_norm": 0.7906124542335656,
+      "learning_rate": 3.848612297445122e-05,
+      "loss": 1.3776,
+      "step": 15620
+    },
+    {
+      "epoch": 0.7196296125673746,
+      "grad_norm": 1.154504304509917,
+      "learning_rate": 3.847435975096885e-05,
+      "loss": 1.15,
+      "step": 15621
+    },
+    {
+      "epoch": 0.7196756806560096,
+      "grad_norm": 0.767345345126558,
+      "learning_rate": 3.846259789724936e-05,
+      "loss": 1.3912,
+      "step": 15622
+    },
+    {
+      "epoch": 0.7197217487446446,
+      "grad_norm": 0.8073663502516462,
+      "learning_rate": 3.8450837413554664e-05,
+      "loss": 1.301,
+      "step": 15623
+    },
+    {
+      "epoch": 0.7197678168332796,
+      "grad_norm": 0.9162797094847329,
+      "learning_rate": 3.843907830014649e-05,
+      "loss": 1.0957,
+      "step": 15624
+    },
+    {
+      "epoch": 0.7198138849219146,
+      "grad_norm": 0.7887433257259462,
+      "learning_rate": 3.8427320557286685e-05,
+      "loss": 1.168,
+      "step": 15625
+    },
+    {
+      "epoch": 0.7198599530105496,
+      "grad_norm": 0.9096888300113505,
+      "learning_rate": 3.841556418523702e-05,
+      "loss": 1.2966,
+      "step": 15626
+    },
+    {
+      "epoch": 0.7199060210991846,
+      "grad_norm": 0.9457122233575205,
+      "learning_rate": 3.8403809184259256e-05,
+      "loss": 1.1676,
+      "step": 15627
+    },
+    {
+      "epoch": 0.7199520891878196,
+      "grad_norm": 0.7228577009956894,
+      "learning_rate": 3.839205555461503e-05,
+      "loss": 1.2171,
+      "step": 15628
+    },
+    {
+      "epoch": 0.7199981572764546,
+      "grad_norm": 0.918799578612656,
+      "learning_rate": 3.8380303296566046e-05,
+      "loss": 1.1976,
+      "step": 15629
+    },
+    {
+      "epoch": 0.7200442253650896,
+      "grad_norm": 0.9271854039258598,
+      "learning_rate": 3.836855241037396e-05,
+      "loss": 1.0824,
+      "step": 15630
+    },
+    {
+      "epoch": 0.7200902934537246,
+      "grad_norm": 0.9638727997747222,
+      "learning_rate": 3.835680289630041e-05,
+      "loss": 1.3158,
+      "step": 15631
+    },
+    {
+      "epoch": 0.7201363615423596,
+      "grad_norm": 1.071440887122076,
+      "learning_rate": 3.834505475460692e-05,
+      "loss": 1.2825,
+      "step": 15632
+    },
+    {
+      "epoch": 0.7201824296309947,
+      "grad_norm": 0.6599192913478832,
+      "learning_rate": 3.833330798555507e-05,
+      "loss": 1.2195,
+      "step": 15633
+    },
+    {
+      "epoch": 0.7202284977196296,
+      "grad_norm": 0.7855011812164827,
+      "learning_rate": 3.832156258940641e-05,
+      "loss": 1.2631,
+      "step": 15634
+    },
+    {
+      "epoch": 0.7202745658082647,
+      "grad_norm": 0.9277544966337022,
+      "learning_rate": 3.830981856642237e-05,
+      "loss": 1.3866,
+      "step": 15635
+    },
+    {
+      "epoch": 0.7203206338968996,
+      "grad_norm": 0.8150030869308073,
+      "learning_rate": 3.8298075916864454e-05,
+      "loss": 1.4842,
+      "step": 15636
+    },
+    {
+      "epoch": 0.7203667019855347,
+      "grad_norm": 0.7393597514654191,
+      "learning_rate": 3.828633464099407e-05,
+      "loss": 1.3947,
+      "step": 15637
+    },
+    {
+      "epoch": 0.7204127700741696,
+      "grad_norm": 0.6724959596103086,
+      "learning_rate": 3.827459473907267e-05,
+      "loss": 1.0765,
+      "step": 15638
+    },
+    {
+      "epoch": 0.7204588381628046,
+      "grad_norm": 1.3315880093539434,
+      "learning_rate": 3.826285621136154e-05,
+      "loss": 1.1873,
+      "step": 15639
+    },
+    {
+      "epoch": 0.7205049062514396,
+      "grad_norm": 0.9178927242646996,
+      "learning_rate": 3.825111905812206e-05,
+      "loss": 1.3399,
+      "step": 15640
+    },
+    {
+      "epoch": 0.7205509743400746,
+      "grad_norm": 1.2192379718236421,
+      "learning_rate": 3.823938327961558e-05,
+      "loss": 1.1966,
+      "step": 15641
+    },
+    {
+      "epoch": 0.7205970424287096,
+      "grad_norm": 0.8848896794375436,
+      "learning_rate": 3.822764887610329e-05,
+      "loss": 1.105,
+      "step": 15642
+    },
+    {
+      "epoch": 0.7206431105173446,
+      "grad_norm": 0.8467526499194069,
+      "learning_rate": 3.821591584784653e-05,
+      "loss": 1.0792,
+      "step": 15643
+    },
+    {
+      "epoch": 0.7206891786059796,
+      "grad_norm": 0.8707943115858945,
+      "learning_rate": 3.820418419510639e-05,
+      "loss": 1.2354,
+      "step": 15644
+    },
+    {
+      "epoch": 0.7207352466946146,
+      "grad_norm": 0.6893843271288022,
+      "learning_rate": 3.819245391814419e-05,
+      "loss": 1.0152,
+      "step": 15645
+    },
+    {
+      "epoch": 0.7207813147832497,
+      "grad_norm": 0.7597904660418746,
+      "learning_rate": 3.818072501722101e-05,
+      "loss": 1.3995,
+      "step": 15646
+    },
+    {
+      "epoch": 0.7208273828718846,
+      "grad_norm": 0.9687446873578325,
+      "learning_rate": 3.816899749259798e-05,
+      "loss": 1.4192,
+      "step": 15647
+    },
+    {
+      "epoch": 0.7208734509605197,
+      "grad_norm": 0.7221936825156412,
+      "learning_rate": 3.81572713445362e-05,
+      "loss": 1.5152,
+      "step": 15648
+    },
+    {
+      "epoch": 0.7209195190491546,
+      "grad_norm": 0.7843627284711054,
+      "learning_rate": 3.8145546573296776e-05,
+      "loss": 1.2934,
+      "step": 15649
+    },
+    {
+      "epoch": 0.7209655871377897,
+      "grad_norm": 0.7313646486222851,
+      "learning_rate": 3.8133823179140706e-05,
+      "loss": 1.1493,
+      "step": 15650
+    },
+    {
+      "epoch": 0.7210116552264246,
+      "grad_norm": 0.8521854001684628,
+      "learning_rate": 3.8122101162328906e-05,
+      "loss": 1.426,
+      "step": 15651
+    },
+    {
+      "epoch": 0.7210577233150597,
+      "grad_norm": 0.7882681285734727,
+      "learning_rate": 3.81103805231225e-05,
+      "loss": 1.3799,
+      "step": 15652
+    },
+    {
+      "epoch": 0.7211037914036946,
+      "grad_norm": 0.7385292326629694,
+      "learning_rate": 3.809866126178232e-05,
+      "loss": 1.4119,
+      "step": 15653
+    },
+    {
+      "epoch": 0.7211498594923297,
+      "grad_norm": 1.1679066425038434,
+      "learning_rate": 3.808694337856934e-05,
+      "loss": 1.3375,
+      "step": 15654
+    },
+    {
+      "epoch": 0.7211959275809646,
+      "grad_norm": 0.668439219672547,
+      "learning_rate": 3.8075226873744343e-05,
+      "loss": 1.3144,
+      "step": 15655
+    },
+    {
+      "epoch": 0.7212419956695997,
+      "grad_norm": 0.6473341537871318,
+      "learning_rate": 3.8063511747568303e-05,
+      "loss": 1.3099,
+      "step": 15656
+    },
+    {
+      "epoch": 0.7212880637582346,
+      "grad_norm": 1.084603410699328,
+      "learning_rate": 3.805179800030194e-05,
+      "loss": 1.1314,
+      "step": 15657
+    },
+    {
+      "epoch": 0.7213341318468697,
+      "grad_norm": 1.2735874556158127,
+      "learning_rate": 3.8040085632206094e-05,
+      "loss": 1.1437,
+      "step": 15658
+    },
+    {
+      "epoch": 0.7213801999355047,
+      "grad_norm": 0.6900455879592114,
+      "learning_rate": 3.802837464354152e-05,
+      "loss": 1.5526,
+      "step": 15659
+    },
+    {
+      "epoch": 0.7214262680241397,
+      "grad_norm": 1.0122246273000661,
+      "learning_rate": 3.801666503456889e-05,
+      "loss": 1.337,
+      "step": 15660
+    },
+    {
+      "epoch": 0.7214723361127747,
+      "grad_norm": 0.9710521879657429,
+      "learning_rate": 3.800495680554897e-05,
+      "loss": 1.3676,
+      "step": 15661
+    },
+    {
+      "epoch": 0.7215184042014097,
+      "grad_norm": 0.6437040321612492,
+      "learning_rate": 3.7993249956742337e-05,
+      "loss": 1.1433,
+      "step": 15662
+    },
+    {
+      "epoch": 0.7215644722900447,
+      "grad_norm": 0.8520632053930798,
+      "learning_rate": 3.7981544488409723e-05,
+      "loss": 1.1951,
+      "step": 15663
+    },
+    {
+      "epoch": 0.7216105403786797,
+      "grad_norm": 0.7814867659316698,
+      "learning_rate": 3.796984040081166e-05,
+      "loss": 1.1526,
+      "step": 15664
+    },
+    {
+      "epoch": 0.7216566084673147,
+      "grad_norm": 0.9810496649356847,
+      "learning_rate": 3.7958137694208776e-05,
+      "loss": 1.4112,
+      "step": 15665
+    },
+    {
+      "epoch": 0.7217026765559497,
+      "grad_norm": 0.6473419894962141,
+      "learning_rate": 3.79464363688615e-05,
+      "loss": 1.3641,
+      "step": 15666
+    },
+    {
+      "epoch": 0.7217487446445847,
+      "grad_norm": 0.6707910128377336,
+      "learning_rate": 3.7934736425030494e-05,
+      "loss": 1.3522,
+      "step": 15667
+    },
+    {
+      "epoch": 0.7217948127332197,
+      "grad_norm": 0.9566528052157811,
+      "learning_rate": 3.7923037862976164e-05,
+      "loss": 1.2263,
+      "step": 15668
+    },
+    {
+      "epoch": 0.7218408808218547,
+      "grad_norm": 0.7328361916247278,
+      "learning_rate": 3.791134068295892e-05,
+      "loss": 1.451,
+      "step": 15669
+    },
+    {
+      "epoch": 0.7218869489104897,
+      "grad_norm": 0.8938559374432022,
+      "learning_rate": 3.789964488523921e-05,
+      "loss": 1.2781,
+      "step": 15670
+    },
+    {
+      "epoch": 0.7219330169991247,
+      "grad_norm": 0.8286529651521946,
+      "learning_rate": 3.788795047007744e-05,
+      "loss": 1.0541,
+      "step": 15671
+    },
+    {
+      "epoch": 0.7219790850877598,
+      "grad_norm": 0.9022127897775423,
+      "learning_rate": 3.7876257437733973e-05,
+      "loss": 1.132,
+      "step": 15672
+    },
+    {
+      "epoch": 0.7220251531763947,
+      "grad_norm": 1.1268016378701586,
+      "learning_rate": 3.7864565788469056e-05,
+      "loss": 1.4154,
+      "step": 15673
+    },
+    {
+      "epoch": 0.7220712212650298,
+      "grad_norm": 0.9340349331846947,
+      "learning_rate": 3.7852875522543105e-05,
+      "loss": 1.0223,
+      "step": 15674
+    },
+    {
+      "epoch": 0.7221172893536647,
+      "grad_norm": 0.8295256144544739,
+      "learning_rate": 3.7841186640216295e-05,
+      "loss": 1.3215,
+      "step": 15675
+    },
+    {
+      "epoch": 0.7221633574422998,
+      "grad_norm": 0.7681502495905272,
+      "learning_rate": 3.7829499141748915e-05,
+      "loss": 0.9997,
+      "step": 15676
+    },
+    {
+      "epoch": 0.7222094255309347,
+      "grad_norm": 0.6722638999542502,
+      "learning_rate": 3.781781302740111e-05,
+      "loss": 1.277,
+      "step": 15677
+    },
+    {
+      "epoch": 0.7222554936195698,
+      "grad_norm": 0.8159627438994457,
+      "learning_rate": 3.7806128297433065e-05,
+      "loss": 1.4884,
+      "step": 15678
+    },
+    {
+      "epoch": 0.7223015617082047,
+      "grad_norm": 0.9250135093108044,
+      "learning_rate": 3.779444495210497e-05,
+      "loss": 1.335,
+      "step": 15679
+    },
+    {
+      "epoch": 0.7223476297968398,
+      "grad_norm": 0.9059415797807189,
+      "learning_rate": 3.778276299167687e-05,
+      "loss": 1.2208,
+      "step": 15680
+    },
+    {
+      "epoch": 0.7223936978854747,
+      "grad_norm": 0.6400101475738492,
+      "learning_rate": 3.777108241640886e-05,
+      "loss": 1.2566,
+      "step": 15681
+    },
+    {
+      "epoch": 0.7224397659741097,
+      "grad_norm": 0.6928596121163886,
+      "learning_rate": 3.7759403226561016e-05,
+      "loss": 1.3468,
+      "step": 15682
+    },
+    {
+      "epoch": 0.7224858340627447,
+      "grad_norm": 0.8500047077116871,
+      "learning_rate": 3.7747725422393375e-05,
+      "loss": 1.2675,
+      "step": 15683
+    },
+    {
+      "epoch": 0.7225319021513797,
+      "grad_norm": 0.817172252710766,
+      "learning_rate": 3.773604900416581e-05,
+      "loss": 1.201,
+      "step": 15684
+    },
+    {
+      "epoch": 0.7225779702400147,
+      "grad_norm": 0.8434732323030387,
+      "learning_rate": 3.7724373972138426e-05,
+      "loss": 1.2799,
+      "step": 15685
+    },
+    {
+      "epoch": 0.7226240383286497,
+      "grad_norm": 1.0968907249698285,
+      "learning_rate": 3.7712700326571034e-05,
+      "loss": 1.4374,
+      "step": 15686
+    },
+    {
+      "epoch": 0.7226701064172848,
+      "grad_norm": 0.7591702395678729,
+      "learning_rate": 3.770102806772362e-05,
+      "loss": 1.315,
+      "step": 15687
+    },
+    {
+      "epoch": 0.7227161745059197,
+      "grad_norm": 0.6963045864304939,
+      "learning_rate": 3.7689357195855946e-05,
+      "loss": 1.3679,
+      "step": 15688
+    },
+    {
+      "epoch": 0.7227622425945548,
+      "grad_norm": 0.9180991057101922,
+      "learning_rate": 3.76776877112279e-05,
+      "loss": 1.1439,
+      "step": 15689
+    },
+    {
+      "epoch": 0.7228083106831897,
+      "grad_norm": 0.8248547975686525,
+      "learning_rate": 3.7666019614099314e-05,
+      "loss": 1.5885,
+      "step": 15690
+    },
+    {
+      "epoch": 0.7228543787718248,
+      "grad_norm": 0.8231350606362279,
+      "learning_rate": 3.765435290472988e-05,
+      "loss": 1.0775,
+      "step": 15691
+    },
+    {
+      "epoch": 0.7229004468604597,
+      "grad_norm": 1.5828246571528797,
+      "learning_rate": 3.764268758337939e-05,
+      "loss": 1.3745,
+      "step": 15692
+    },
+    {
+      "epoch": 0.7229465149490948,
+      "grad_norm": 0.8349914005225606,
+      "learning_rate": 3.763102365030754e-05,
+      "loss": 1.4334,
+      "step": 15693
+    },
+    {
+      "epoch": 0.7229925830377297,
+      "grad_norm": 1.1786344120750285,
+      "learning_rate": 3.761936110577403e-05,
+      "loss": 0.8867,
+      "step": 15694
+    },
+    {
+      "epoch": 0.7230386511263648,
+      "grad_norm": 2.0795303581985687,
+      "learning_rate": 3.760769995003847e-05,
+      "loss": 1.0971,
+      "step": 15695
+    },
+    {
+      "epoch": 0.7230847192149997,
+      "grad_norm": 0.6831162446693665,
+      "learning_rate": 3.759604018336047e-05,
+      "loss": 1.2558,
+      "step": 15696
+    },
+    {
+      "epoch": 0.7231307873036348,
+      "grad_norm": 1.0527734881668862,
+      "learning_rate": 3.758438180599968e-05,
+      "loss": 1.3057,
+      "step": 15697
+    },
+    {
+      "epoch": 0.7231768553922697,
+      "grad_norm": 0.7972968080367687,
+      "learning_rate": 3.757272481821558e-05,
+      "loss": 1.4615,
+      "step": 15698
+    },
+    {
+      "epoch": 0.7232229234809048,
+      "grad_norm": 0.624184785654881,
+      "learning_rate": 3.756106922026771e-05,
+      "loss": 1.311,
+      "step": 15699
+    },
+    {
+      "epoch": 0.7232689915695398,
+      "grad_norm": 0.9239204298727103,
+      "learning_rate": 3.7549415012415576e-05,
+      "loss": 1.0654,
+      "step": 15700
+    },
+    {
+      "epoch": 0.7233150596581748,
+      "grad_norm": 0.8160462069888276,
+      "learning_rate": 3.753776219491868e-05,
+      "loss": 1.2318,
+      "step": 15701
+    },
+    {
+      "epoch": 0.7233611277468098,
+      "grad_norm": 1.21655283740174,
+      "learning_rate": 3.752611076803637e-05,
+      "loss": 1.2189,
+      "step": 15702
+    },
+    {
+      "epoch": 0.7234071958354448,
+      "grad_norm": 0.7966321942312949,
+      "learning_rate": 3.751446073202809e-05,
+      "loss": 1.2088,
+      "step": 15703
+    },
+    {
+      "epoch": 0.7234532639240798,
+      "grad_norm": 0.9404039106123155,
+      "learning_rate": 3.7502812087153194e-05,
+      "loss": 1.2999,
+      "step": 15704
+    },
+    {
+      "epoch": 0.7234993320127148,
+      "grad_norm": 1.3044020357757005,
+      "learning_rate": 3.749116483367107e-05,
+      "loss": 1.0127,
+      "step": 15705
+    },
+    {
+      "epoch": 0.7235454001013498,
+      "grad_norm": 0.8368569685898575,
+      "learning_rate": 3.747951897184093e-05,
+      "loss": 1.3878,
+      "step": 15706
+    },
+    {
+      "epoch": 0.7235914681899848,
+      "grad_norm": 0.7035994030299876,
+      "learning_rate": 3.7467874501922115e-05,
+      "loss": 1.2958,
+      "step": 15707
+    },
+    {
+      "epoch": 0.7236375362786198,
+      "grad_norm": 1.1396977919886835,
+      "learning_rate": 3.745623142417388e-05,
+      "loss": 1.4539,
+      "step": 15708
+    },
+    {
+      "epoch": 0.7236836043672548,
+      "grad_norm": 1.816080381625097,
+      "learning_rate": 3.7444589738855395e-05,
+      "loss": 1.1543,
+      "step": 15709
+    },
+    {
+      "epoch": 0.7237296724558898,
+      "grad_norm": 0.7909425946992397,
+      "learning_rate": 3.743294944622584e-05,
+      "loss": 1.4356,
+      "step": 15710
+    },
+    {
+      "epoch": 0.7237757405445248,
+      "grad_norm": 0.7084458462890552,
+      "learning_rate": 3.74213105465444e-05,
+      "loss": 1.5289,
+      "step": 15711
+    },
+    {
+      "epoch": 0.7238218086331598,
+      "grad_norm": 0.8043756093404476,
+      "learning_rate": 3.7409673040070216e-05,
+      "loss": 1.1183,
+      "step": 15712
+    },
+    {
+      "epoch": 0.7238678767217949,
+      "grad_norm": 0.7370656763650612,
+      "learning_rate": 3.7398036927062296e-05,
+      "loss": 1.2907,
+      "step": 15713
+    },
+    {
+      "epoch": 0.7239139448104298,
+      "grad_norm": 1.316445741806515,
+      "learning_rate": 3.738640220777975e-05,
+      "loss": 1.4357,
+      "step": 15714
+    },
+    {
+      "epoch": 0.7239600128990649,
+      "grad_norm": 0.8327278343939287,
+      "learning_rate": 3.7374768882481634e-05,
+      "loss": 1.2089,
+      "step": 15715
+    },
+    {
+      "epoch": 0.7240060809876998,
+      "grad_norm": 7.110349176814641,
+      "learning_rate": 3.736313695142688e-05,
+      "loss": 1.2871,
+      "step": 15716
+    },
+    {
+      "epoch": 0.7240521490763349,
+      "grad_norm": 0.6575381253598945,
+      "learning_rate": 3.7351506414874474e-05,
+      "loss": 1.5928,
+      "step": 15717
+    },
+    {
+      "epoch": 0.7240982171649698,
+      "grad_norm": 0.6310769531853472,
+      "learning_rate": 3.733987727308337e-05,
+      "loss": 1.3595,
+      "step": 15718
+    },
+    {
+      "epoch": 0.7241442852536049,
+      "grad_norm": 0.8370937780613282,
+      "learning_rate": 3.732824952631249e-05,
+      "loss": 1.1049,
+      "step": 15719
+    },
+    {
+      "epoch": 0.7241903533422398,
+      "grad_norm": 0.928122680584179,
+      "learning_rate": 3.731662317482063e-05,
+      "loss": 1.1409,
+      "step": 15720
+    },
+    {
+      "epoch": 0.7242364214308749,
+      "grad_norm": 0.9295864893950749,
+      "learning_rate": 3.7304998218866684e-05,
+      "loss": 1.2303,
+      "step": 15721
+    },
+    {
+      "epoch": 0.7242824895195098,
+      "grad_norm": 1.134021679523039,
+      "learning_rate": 3.7293374658709456e-05,
+      "loss": 1.2955,
+      "step": 15722
+    },
+    {
+      "epoch": 0.7243285576081449,
+      "grad_norm": 0.84637971061117,
+      "learning_rate": 3.728175249460775e-05,
+      "loss": 1.1403,
+      "step": 15723
+    },
+    {
+      "epoch": 0.7243746256967798,
+      "grad_norm": 1.0018650115464671,
+      "learning_rate": 3.7270131726820255e-05,
+      "loss": 1.3033,
+      "step": 15724
+    },
+    {
+      "epoch": 0.7244206937854148,
+      "grad_norm": 1.0190730632036764,
+      "learning_rate": 3.7258512355605724e-05,
+      "loss": 1.1264,
+      "step": 15725
+    },
+    {
+      "epoch": 0.7244667618740498,
+      "grad_norm": 1.02494464278809,
+      "learning_rate": 3.7246894381222866e-05,
+      "loss": 1.3876,
+      "step": 15726
+    },
+    {
+      "epoch": 0.7245128299626848,
+      "grad_norm": 0.7023357059400144,
+      "learning_rate": 3.723527780393027e-05,
+      "loss": 1.4192,
+      "step": 15727
+    },
+    {
+      "epoch": 0.7245588980513199,
+      "grad_norm": 2.8305663972406068,
+      "learning_rate": 3.72236626239866e-05,
+      "loss": 1.268,
+      "step": 15728
+    },
+    {
+      "epoch": 0.7246049661399548,
+      "grad_norm": 0.759494442206147,
+      "learning_rate": 3.721204884165044e-05,
+      "loss": 1.1469,
+      "step": 15729
+    },
+    {
+      "epoch": 0.7246510342285899,
+      "grad_norm": 1.4997620493718185,
+      "learning_rate": 3.720043645718039e-05,
+      "loss": 1.2864,
+      "step": 15730
+    },
+    {
+      "epoch": 0.7246971023172248,
+      "grad_norm": 0.8713628152346472,
+      "learning_rate": 3.718882547083491e-05,
+      "loss": 1.2758,
+      "step": 15731
+    },
+    {
+      "epoch": 0.7247431704058599,
+      "grad_norm": 0.6703251762889844,
+      "learning_rate": 3.717721588287253e-05,
+      "loss": 1.3521,
+      "step": 15732
+    },
+    {
+      "epoch": 0.7247892384944948,
+      "grad_norm": 0.7461635680844985,
+      "learning_rate": 3.7165607693551764e-05,
+      "loss": 1.0249,
+      "step": 15733
+    },
+    {
+      "epoch": 0.7248353065831299,
+      "grad_norm": 0.8181396730002305,
+      "learning_rate": 3.715400090313096e-05,
+      "loss": 1.4552,
+      "step": 15734
+    },
+    {
+      "epoch": 0.7248813746717648,
+      "grad_norm": 0.7262700821563425,
+      "learning_rate": 3.714239551186857e-05,
+      "loss": 1.4057,
+      "step": 15735
+    },
+    {
+      "epoch": 0.7249274427603999,
+      "grad_norm": 0.6689883647992907,
+      "learning_rate": 3.7130791520022965e-05,
+      "loss": 1.1985,
+      "step": 15736
+    },
+    {
+      "epoch": 0.7249735108490348,
+      "grad_norm": 0.6349967955139595,
+      "learning_rate": 3.711918892785253e-05,
+      "loss": 1.4055,
+      "step": 15737
+    },
+    {
+      "epoch": 0.7250195789376699,
+      "grad_norm": 1.1194047033421952,
+      "learning_rate": 3.71075877356155e-05,
+      "loss": 1.2344,
+      "step": 15738
+    },
+    {
+      "epoch": 0.7250656470263048,
+      "grad_norm": 1.6470443606876564,
+      "learning_rate": 3.709598794357019e-05,
+      "loss": 1.5128,
+      "step": 15739
+    },
+    {
+      "epoch": 0.7251117151149399,
+      "grad_norm": 0.8791269673357729,
+      "learning_rate": 3.708438955197485e-05,
+      "loss": 1.3742,
+      "step": 15740
+    },
+    {
+      "epoch": 0.7251577832035749,
+      "grad_norm": 0.9348991227439779,
+      "learning_rate": 3.707279256108773e-05,
+      "loss": 1.3479,
+      "step": 15741
+    },
+    {
+      "epoch": 0.7252038512922099,
+      "grad_norm": 0.9928510540614309,
+      "learning_rate": 3.706119697116699e-05,
+      "loss": 1.347,
+      "step": 15742
+    },
+    {
+      "epoch": 0.7252499193808449,
+      "grad_norm": 4.609005021982667,
+      "learning_rate": 3.704960278247072e-05,
+      "loss": 1.1235,
+      "step": 15743
+    },
+    {
+      "epoch": 0.7252959874694799,
+      "grad_norm": 1.6100589947021955,
+      "learning_rate": 3.703800999525717e-05,
+      "loss": 1.1609,
+      "step": 15744
+    },
+    {
+      "epoch": 0.7253420555581149,
+      "grad_norm": 0.7420210563365894,
+      "learning_rate": 3.702641860978434e-05,
+      "loss": 0.8467,
+      "step": 15745
+    },
+    {
+      "epoch": 0.7253881236467499,
+      "grad_norm": 1.7387442491522063,
+      "learning_rate": 3.701482862631032e-05,
+      "loss": 1.1271,
+      "step": 15746
+    },
+    {
+      "epoch": 0.7254341917353849,
+      "grad_norm": 0.9423157990259367,
+      "learning_rate": 3.700324004509315e-05,
+      "loss": 1.3029,
+      "step": 15747
+    },
+    {
+      "epoch": 0.7254802598240199,
+      "grad_norm": 1.122553713870138,
+      "learning_rate": 3.6991652866390844e-05,
+      "loss": 1.4937,
+      "step": 15748
+    },
+    {
+      "epoch": 0.7255263279126549,
+      "grad_norm": 1.032708625950412,
+      "learning_rate": 3.698006709046132e-05,
+      "loss": 1.5048,
+      "step": 15749
+    },
+    {
+      "epoch": 0.7255723960012899,
+      "grad_norm": 0.7803858299413544,
+      "learning_rate": 3.696848271756256e-05,
+      "loss": 0.9427,
+      "step": 15750
+    },
+    {
+      "epoch": 0.7256184640899249,
+      "grad_norm": 1.1242723985581164,
+      "learning_rate": 3.695689974795248e-05,
+      "loss": 1.4347,
+      "step": 15751
+    },
+    {
+      "epoch": 0.7256645321785599,
+      "grad_norm": 2.050740007015069,
+      "learning_rate": 3.69453181818889e-05,
+      "loss": 0.9878,
+      "step": 15752
+    },
+    {
+      "epoch": 0.7257106002671949,
+      "grad_norm": 0.8862550771250776,
+      "learning_rate": 3.693373801962974e-05,
+      "loss": 1.4401,
+      "step": 15753
+    },
+    {
+      "epoch": 0.72575666835583,
+      "grad_norm": 0.9621130107254562,
+      "learning_rate": 3.692215926143268e-05,
+      "loss": 1.3063,
+      "step": 15754
+    },
+    {
+      "epoch": 0.7258027364444649,
+      "grad_norm": 0.8974350218088764,
+      "learning_rate": 3.691058190755566e-05,
+      "loss": 1.2496,
+      "step": 15755
+    },
+    {
+      "epoch": 0.7258488045331,
+      "grad_norm": 0.9720567096868062,
+      "learning_rate": 3.6899005958256336e-05,
+      "loss": 1.345,
+      "step": 15756
+    },
+    {
+      "epoch": 0.7258948726217349,
+      "grad_norm": 0.7476567435279808,
+      "learning_rate": 3.688743141379244e-05,
+      "loss": 1.4399,
+      "step": 15757
+    },
+    {
+      "epoch": 0.72594094071037,
+      "grad_norm": 1.1044228777513838,
+      "learning_rate": 3.687585827442167e-05,
+      "loss": 1.2542,
+      "step": 15758
+    },
+    {
+      "epoch": 0.7259870087990049,
+      "grad_norm": 1.2686014961564542,
+      "learning_rate": 3.686428654040173e-05,
+      "loss": 1.1974,
+      "step": 15759
+    },
+    {
+      "epoch": 0.72603307688764,
+      "grad_norm": 1.3686890919191048,
+      "learning_rate": 3.685271621199018e-05,
+      "loss": 1.1274,
+      "step": 15760
+    },
+    {
+      "epoch": 0.7260791449762749,
+      "grad_norm": 1.4025608003485293,
+      "learning_rate": 3.684114728944455e-05,
+      "loss": 1.294,
+      "step": 15761
+    },
+    {
+      "epoch": 0.72612521306491,
+      "grad_norm": 1.4968288155272467,
+      "learning_rate": 3.682957977302257e-05,
+      "loss": 1.2394,
+      "step": 15762
+    },
+    {
+      "epoch": 0.7261712811535449,
+      "grad_norm": 0.8803741469465544,
+      "learning_rate": 3.681801366298164e-05,
+      "loss": 1.3702,
+      "step": 15763
+    },
+    {
+      "epoch": 0.72621734924218,
+      "grad_norm": 0.7016650574530422,
+      "learning_rate": 3.6806448959579345e-05,
+      "loss": 1.533,
+      "step": 15764
+    },
+    {
+      "epoch": 0.7262634173308149,
+      "grad_norm": 0.8742811562206754,
+      "learning_rate": 3.679488566307303e-05,
+      "loss": 1.3768,
+      "step": 15765
+    },
+    {
+      "epoch": 0.72630948541945,
+      "grad_norm": 0.9349357301197935,
+      "learning_rate": 3.678332377372028e-05,
+      "loss": 1.3926,
+      "step": 15766
+    },
+    {
+      "epoch": 0.7263555535080849,
+      "grad_norm": 0.9101566345137501,
+      "learning_rate": 3.6771763291778406e-05,
+      "loss": 1.4103,
+      "step": 15767
+    },
+    {
+      "epoch": 0.72640162159672,
+      "grad_norm": 1.2376197695781241,
+      "learning_rate": 3.6760204217504846e-05,
+      "loss": 1.4995,
+      "step": 15768
+    },
+    {
+      "epoch": 0.726447689685355,
+      "grad_norm": 0.7442395836033606,
+      "learning_rate": 3.674864655115686e-05,
+      "loss": 1.1016,
+      "step": 15769
+    },
+    {
+      "epoch": 0.72649375777399,
+      "grad_norm": 0.781902206032534,
+      "learning_rate": 3.673709029299182e-05,
+      "loss": 1.2217,
+      "step": 15770
+    },
+    {
+      "epoch": 0.726539825862625,
+      "grad_norm": 0.7723125682260059,
+      "learning_rate": 3.6725535443267014e-05,
+      "loss": 1.0998,
+      "step": 15771
+    },
+    {
+      "epoch": 0.7265858939512599,
+      "grad_norm": 1.5736977395783436,
+      "learning_rate": 3.671398200223961e-05,
+      "loss": 1.0695,
+      "step": 15772
+    },
+    {
+      "epoch": 0.726631962039895,
+      "grad_norm": 1.0436481662790007,
+      "learning_rate": 3.670242997016694e-05,
+      "loss": 1.4452,
+      "step": 15773
+    },
+    {
+      "epoch": 0.7266780301285299,
+      "grad_norm": 1.1517453943767757,
+      "learning_rate": 3.6690879347306115e-05,
+      "loss": 1.1165,
+      "step": 15774
+    },
+    {
+      "epoch": 0.726724098217165,
+      "grad_norm": 0.7614829455091575,
+      "learning_rate": 3.667933013391434e-05,
+      "loss": 1.3175,
+      "step": 15775
+    },
+    {
+      "epoch": 0.7267701663057999,
+      "grad_norm": 1.5827352912941297,
+      "learning_rate": 3.666778233024863e-05,
+      "loss": 1.3129,
+      "step": 15776
+    },
+    {
+      "epoch": 0.726816234394435,
+      "grad_norm": 0.7455691021853,
+      "learning_rate": 3.665623593656623e-05,
+      "loss": 1.3962,
+      "step": 15777
+    },
+    {
+      "epoch": 0.7268623024830699,
+      "grad_norm": 0.8182337661462393,
+      "learning_rate": 3.6644690953124126e-05,
+      "loss": 1.3795,
+      "step": 15778
+    },
+    {
+      "epoch": 0.726908370571705,
+      "grad_norm": 1.1822916448244811,
+      "learning_rate": 3.663314738017931e-05,
+      "loss": 1.3737,
+      "step": 15779
+    },
+    {
+      "epoch": 0.7269544386603399,
+      "grad_norm": 1.0083288291289336,
+      "learning_rate": 3.662160521798882e-05,
+      "loss": 1.0346,
+      "step": 15780
+    },
+    {
+      "epoch": 0.727000506748975,
+      "grad_norm": 1.05727091486461,
+      "learning_rate": 3.661006446680963e-05,
+      "loss": 1.3674,
+      "step": 15781
+    },
+    {
+      "epoch": 0.72704657483761,
+      "grad_norm": 1.0040667680752258,
+      "learning_rate": 3.6598525126898675e-05,
+      "loss": 1.0319,
+      "step": 15782
+    },
+    {
+      "epoch": 0.727092642926245,
+      "grad_norm": 0.9147031402864985,
+      "learning_rate": 3.658698719851279e-05,
+      "loss": 1.2926,
+      "step": 15783
+    },
+    {
+      "epoch": 0.72713871101488,
+      "grad_norm": 0.9417031549706399,
+      "learning_rate": 3.6575450681908976e-05,
+      "loss": 1.3273,
+      "step": 15784
+    },
+    {
+      "epoch": 0.727184779103515,
+      "grad_norm": 0.8468353546863613,
+      "learning_rate": 3.6563915577343964e-05,
+      "loss": 1.0301,
+      "step": 15785
+    },
+    {
+      "epoch": 0.72723084719215,
+      "grad_norm": 1.3834056559750183,
+      "learning_rate": 3.655238188507464e-05,
+      "loss": 1.4182,
+      "step": 15786
+    },
+    {
+      "epoch": 0.727276915280785,
+      "grad_norm": 0.6631861483269684,
+      "learning_rate": 3.6540849605357705e-05,
+      "loss": 1.2648,
+      "step": 15787
+    },
+    {
+      "epoch": 0.72732298336942,
+      "grad_norm": 0.9364631149088628,
+      "learning_rate": 3.6529318738449956e-05,
+      "loss": 1.06,
+      "step": 15788
+    },
+    {
+      "epoch": 0.727369051458055,
+      "grad_norm": 0.6394946553123477,
+      "learning_rate": 3.651778928460812e-05,
+      "loss": 1.0761,
+      "step": 15789
+    },
+    {
+      "epoch": 0.72741511954669,
+      "grad_norm": 0.75087130922874,
+      "learning_rate": 3.650626124408883e-05,
+      "loss": 0.9834,
+      "step": 15790
+    },
+    {
+      "epoch": 0.727461187635325,
+      "grad_norm": 0.8780682701077163,
+      "learning_rate": 3.649473461714876e-05,
+      "loss": 1.3101,
+      "step": 15791
+    },
+    {
+      "epoch": 0.72750725572396,
+      "grad_norm": 1.3979825397928098,
+      "learning_rate": 3.6483209404044536e-05,
+      "loss": 1.2427,
+      "step": 15792
+    },
+    {
+      "epoch": 0.727553323812595,
+      "grad_norm": 1.154182025596911,
+      "learning_rate": 3.647168560503278e-05,
+      "loss": 1.0669,
+      "step": 15793
+    },
+    {
+      "epoch": 0.72759939190123,
+      "grad_norm": 2.571991958161749,
+      "learning_rate": 3.646016322036998e-05,
+      "loss": 1.5968,
+      "step": 15794
+    },
+    {
+      "epoch": 0.7276454599898651,
+      "grad_norm": 0.804910623751567,
+      "learning_rate": 3.64486422503127e-05,
+      "loss": 1.4882,
+      "step": 15795
+    },
+    {
+      "epoch": 0.7276915280785,
+      "grad_norm": 0.886540047046778,
+      "learning_rate": 3.643712269511748e-05,
+      "loss": 1.4492,
+      "step": 15796
+    },
+    {
+      "epoch": 0.7277375961671351,
+      "grad_norm": 0.8716908886644279,
+      "learning_rate": 3.642560455504068e-05,
+      "loss": 1.3329,
+      "step": 15797
+    },
+    {
+      "epoch": 0.72778366425577,
+      "grad_norm": 1.2246622704884147,
+      "learning_rate": 3.641408783033881e-05,
+      "loss": 1.2483,
+      "step": 15798
+    },
+    {
+      "epoch": 0.7278297323444051,
+      "grad_norm": 1.0489548738705323,
+      "learning_rate": 3.640257252126823e-05,
+      "loss": 1.1322,
+      "step": 15799
+    },
+    {
+      "epoch": 0.72787580043304,
+      "grad_norm": 0.6580992613641091,
+      "learning_rate": 3.6391058628085376e-05,
+      "loss": 0.9695,
+      "step": 15800
+    },
+    {
+      "epoch": 0.7279218685216751,
+      "grad_norm": 0.7447040854383435,
+      "learning_rate": 3.6379546151046485e-05,
+      "loss": 1.239,
+      "step": 15801
+    },
+    {
+      "epoch": 0.72796793661031,
+      "grad_norm": 0.6873144578642345,
+      "learning_rate": 3.636803509040792e-05,
+      "loss": 1.4794,
+      "step": 15802
+    },
+    {
+      "epoch": 0.7280140046989451,
+      "grad_norm": 0.9127095930361744,
+      "learning_rate": 3.635652544642595e-05,
+      "loss": 1.266,
+      "step": 15803
+    },
+    {
+      "epoch": 0.72806007278758,
+      "grad_norm": 0.8946322447031563,
+      "learning_rate": 3.634501721935686e-05,
+      "loss": 1.2157,
+      "step": 15804
+    },
+    {
+      "epoch": 0.7281061408762151,
+      "grad_norm": 1.008501744430875,
+      "learning_rate": 3.6333510409456765e-05,
+      "loss": 1.5228,
+      "step": 15805
+    },
+    {
+      "epoch": 0.72815220896485,
+      "grad_norm": 1.47663736048326,
+      "learning_rate": 3.632200501698191e-05,
+      "loss": 1.4067,
+      "step": 15806
+    },
+    {
+      "epoch": 0.728198277053485,
+      "grad_norm": 1.279233966722912,
+      "learning_rate": 3.631050104218845e-05,
+      "loss": 1.219,
+      "step": 15807
+    },
+    {
+      "epoch": 0.7282443451421201,
+      "grad_norm": 0.8565464967162156,
+      "learning_rate": 3.629899848533246e-05,
+      "loss": 1.2962,
+      "step": 15808
+    },
+    {
+      "epoch": 0.728290413230755,
+      "grad_norm": 0.674767807860052,
+      "learning_rate": 3.628749734667003e-05,
+      "loss": 1.4139,
+      "step": 15809
+    },
+    {
+      "epoch": 0.7283364813193901,
+      "grad_norm": 0.9321364108233228,
+      "learning_rate": 3.6275997626457234e-05,
+      "loss": 1.1838,
+      "step": 15810
+    },
+    {
+      "epoch": 0.728382549408025,
+      "grad_norm": 0.7752577457406093,
+      "learning_rate": 3.626449932495013e-05,
+      "loss": 1.301,
+      "step": 15811
+    },
+    {
+      "epoch": 0.7284286174966601,
+      "grad_norm": 0.9748469659188769,
+      "learning_rate": 3.6253002442404616e-05,
+      "loss": 1.0672,
+      "step": 15812
+    },
+    {
+      "epoch": 0.728474685585295,
+      "grad_norm": 0.7737874493721609,
+      "learning_rate": 3.624150697907671e-05,
+      "loss": 1.4031,
+      "step": 15813
+    },
+    {
+      "epoch": 0.7285207536739301,
+      "grad_norm": 1.0916368731476056,
+      "learning_rate": 3.6230012935222335e-05,
+      "loss": 1.3791,
+      "step": 15814
+    },
+    {
+      "epoch": 0.728566821762565,
+      "grad_norm": 0.8309580560679544,
+      "learning_rate": 3.62185203110974e-05,
+      "loss": 1.2844,
+      "step": 15815
+    },
+    {
+      "epoch": 0.7286128898512001,
+      "grad_norm": 0.6592226743132524,
+      "learning_rate": 3.6207029106957723e-05,
+      "loss": 1.0426,
+      "step": 15816
+    },
+    {
+      "epoch": 0.728658957939835,
+      "grad_norm": 0.7694145014670083,
+      "learning_rate": 3.6195539323059155e-05,
+      "loss": 1.1519,
+      "step": 15817
+    },
+    {
+      "epoch": 0.7287050260284701,
+      "grad_norm": 0.7106708735680115,
+      "learning_rate": 3.6184050959657546e-05,
+      "loss": 1.4172,
+      "step": 15818
+    },
+    {
+      "epoch": 0.728751094117105,
+      "grad_norm": 0.7761335423809288,
+      "learning_rate": 3.617256401700858e-05,
+      "loss": 1.3742,
+      "step": 15819
+    },
+    {
+      "epoch": 0.7287971622057401,
+      "grad_norm": 0.7650652256240327,
+      "learning_rate": 3.616107849536804e-05,
+      "loss": 1.1012,
+      "step": 15820
+    },
+    {
+      "epoch": 0.728843230294375,
+      "grad_norm": 0.9302244183940409,
+      "learning_rate": 3.614959439499162e-05,
+      "loss": 1.296,
+      "step": 15821
+    },
+    {
+      "epoch": 0.7288892983830101,
+      "grad_norm": 0.9601071918150746,
+      "learning_rate": 3.6138111716135046e-05,
+      "loss": 1.525,
+      "step": 15822
+    },
+    {
+      "epoch": 0.7289353664716451,
+      "grad_norm": 0.7461884994124821,
+      "learning_rate": 3.612663045905388e-05,
+      "loss": 1.3784,
+      "step": 15823
+    },
+    {
+      "epoch": 0.7289814345602801,
+      "grad_norm": 0.7382994851241403,
+      "learning_rate": 3.611515062400376e-05,
+      "loss": 1.3913,
+      "step": 15824
+    },
+    {
+      "epoch": 0.7290275026489151,
+      "grad_norm": 0.7985273788978399,
+      "learning_rate": 3.61036722112403e-05,
+      "loss": 1.3936,
+      "step": 15825
+    },
+    {
+      "epoch": 0.7290735707375501,
+      "grad_norm": 0.6972336299544512,
+      "learning_rate": 3.6092195221018997e-05,
+      "loss": 0.9741,
+      "step": 15826
+    },
+    {
+      "epoch": 0.7291196388261851,
+      "grad_norm": 0.8463284832029856,
+      "learning_rate": 3.6080719653595396e-05,
+      "loss": 1.1968,
+      "step": 15827
+    },
+    {
+      "epoch": 0.7291657069148201,
+      "grad_norm": 0.8903991284826052,
+      "learning_rate": 3.606924550922496e-05,
+      "loss": 0.9825,
+      "step": 15828
+    },
+    {
+      "epoch": 0.7292117750034551,
+      "grad_norm": 0.8056236955626651,
+      "learning_rate": 3.605777278816319e-05,
+      "loss": 1.3742,
+      "step": 15829
+    },
+    {
+      "epoch": 0.7292578430920901,
+      "grad_norm": 0.9397979758917068,
+      "learning_rate": 3.604630149066545e-05,
+      "loss": 1.2297,
+      "step": 15830
+    },
+    {
+      "epoch": 0.7293039111807251,
+      "grad_norm": 1.103445390669581,
+      "learning_rate": 3.603483161698713e-05,
+      "loss": 1.1913,
+      "step": 15831
+    },
+    {
+      "epoch": 0.7293499792693601,
+      "grad_norm": 0.8207620606756191,
+      "learning_rate": 3.6023363167383614e-05,
+      "loss": 1.1688,
+      "step": 15832
+    },
+    {
+      "epoch": 0.7293960473579951,
+      "grad_norm": 0.7265659574337262,
+      "learning_rate": 3.601189614211027e-05,
+      "loss": 1.4904,
+      "step": 15833
+    },
+    {
+      "epoch": 0.7294421154466301,
+      "grad_norm": 0.9703836214304806,
+      "learning_rate": 3.600043054142229e-05,
+      "loss": 1.101,
+      "step": 15834
+    },
+    {
+      "epoch": 0.7294881835352651,
+      "grad_norm": 4.511531653033189,
+      "learning_rate": 3.5988966365575006e-05,
+      "loss": 1.1978,
+      "step": 15835
+    },
+    {
+      "epoch": 0.7295342516239002,
+      "grad_norm": 0.7121021163583829,
+      "learning_rate": 3.597750361482365e-05,
+      "loss": 1.5338,
+      "step": 15836
+    },
+    {
+      "epoch": 0.7295803197125351,
+      "grad_norm": 0.8545963050907456,
+      "learning_rate": 3.5966042289423375e-05,
+      "loss": 1.3691,
+      "step": 15837
+    },
+    {
+      "epoch": 0.7296263878011702,
+      "grad_norm": 0.7186136779863214,
+      "learning_rate": 3.595458238962936e-05,
+      "loss": 1.0516,
+      "step": 15838
+    },
+    {
+      "epoch": 0.7296724558898051,
+      "grad_norm": 1.1047738163493175,
+      "learning_rate": 3.5943123915696766e-05,
+      "loss": 1.1325,
+      "step": 15839
+    },
+    {
+      "epoch": 0.7297185239784402,
+      "grad_norm": 0.7966803646213747,
+      "learning_rate": 3.593166686788072e-05,
+      "loss": 1.3251,
+      "step": 15840
+    },
+    {
+      "epoch": 0.7297645920670751,
+      "grad_norm": 1.1987627799965261,
+      "learning_rate": 3.592021124643622e-05,
+      "loss": 1.2307,
+      "step": 15841
+    },
+    {
+      "epoch": 0.7298106601557102,
+      "grad_norm": 0.753983933654933,
+      "learning_rate": 3.590875705161834e-05,
+      "loss": 1.2583,
+      "step": 15842
+    },
+    {
+      "epoch": 0.7298567282443451,
+      "grad_norm": 0.8278850485252058,
+      "learning_rate": 3.589730428368212e-05,
+      "loss": 0.9697,
+      "step": 15843
+    },
+    {
+      "epoch": 0.7299027963329802,
+      "grad_norm": 0.8483728776921007,
+      "learning_rate": 3.5885852942882494e-05,
+      "loss": 1.1254,
+      "step": 15844
+    },
+    {
+      "epoch": 0.7299488644216151,
+      "grad_norm": 1.2076936340391071,
+      "learning_rate": 3.58744030294744e-05,
+      "loss": 1.0584,
+      "step": 15845
+    },
+    {
+      "epoch": 0.7299949325102502,
+      "grad_norm": 0.6713502997772688,
+      "learning_rate": 3.586295454371278e-05,
+      "loss": 1.3051,
+      "step": 15846
+    },
+    {
+      "epoch": 0.7300410005988851,
+      "grad_norm": 2.077949060370077,
+      "learning_rate": 3.585150748585253e-05,
+      "loss": 1.2993,
+      "step": 15847
+    },
+    {
+      "epoch": 0.7300870686875202,
+      "grad_norm": 0.8523036970636292,
+      "learning_rate": 3.5840061856148454e-05,
+      "loss": 1.2847,
+      "step": 15848
+    },
+    {
+      "epoch": 0.7301331367761552,
+      "grad_norm": 0.7109633473801034,
+      "learning_rate": 3.582861765485539e-05,
+      "loss": 1.2039,
+      "step": 15849
+    },
+    {
+      "epoch": 0.7301792048647902,
+      "grad_norm": 0.7118230717624876,
+      "learning_rate": 3.581717488222811e-05,
+      "loss": 1.4686,
+      "step": 15850
+    },
+    {
+      "epoch": 0.7302252729534252,
+      "grad_norm": 0.7283850626362677,
+      "learning_rate": 3.580573353852142e-05,
+      "loss": 1.3997,
+      "step": 15851
+    },
+    {
+      "epoch": 0.7302713410420602,
+      "grad_norm": 0.8360256431479828,
+      "learning_rate": 3.5794293623990014e-05,
+      "loss": 1.4882,
+      "step": 15852
+    },
+    {
+      "epoch": 0.7303174091306952,
+      "grad_norm": 1.0233239404158305,
+      "learning_rate": 3.57828551388885e-05,
+      "loss": 1.1047,
+      "step": 15853
+    },
+    {
+      "epoch": 0.7303634772193301,
+      "grad_norm": 0.7573901118966638,
+      "learning_rate": 3.5771418083471665e-05,
+      "loss": 1.3229,
+      "step": 15854
+    },
+    {
+      "epoch": 0.7304095453079652,
+      "grad_norm": 0.6917588672697891,
+      "learning_rate": 3.5759982457994056e-05,
+      "loss": 1.2791,
+      "step": 15855
+    },
+    {
+      "epoch": 0.7304556133966001,
+      "grad_norm": 0.9993955801626727,
+      "learning_rate": 3.574854826271028e-05,
+      "loss": 1.1654,
+      "step": 15856
+    },
+    {
+      "epoch": 0.7305016814852352,
+      "grad_norm": 0.854637504623907,
+      "learning_rate": 3.573711549787491e-05,
+      "loss": 1.1462,
+      "step": 15857
+    },
+    {
+      "epoch": 0.7305477495738701,
+      "grad_norm": 0.8019628205257346,
+      "learning_rate": 3.572568416374251e-05,
+      "loss": 1.4521,
+      "step": 15858
+    },
+    {
+      "epoch": 0.7305938176625052,
+      "grad_norm": 0.9708672564205232,
+      "learning_rate": 3.5714254260567527e-05,
+      "loss": 1.249,
+      "step": 15859
+    },
+    {
+      "epoch": 0.7306398857511401,
+      "grad_norm": 0.7408369283858607,
+      "learning_rate": 3.570282578860443e-05,
+      "loss": 1.2118,
+      "step": 15860
+    },
+    {
+      "epoch": 0.7306859538397752,
+      "grad_norm": 1.703167520956309,
+      "learning_rate": 3.569139874810771e-05,
+      "loss": 1.518,
+      "step": 15861
+    },
+    {
+      "epoch": 0.7307320219284101,
+      "grad_norm": 0.773004104278561,
+      "learning_rate": 3.567997313933169e-05,
+      "loss": 1.28,
+      "step": 15862
+    },
+    {
+      "epoch": 0.7307780900170452,
+      "grad_norm": 0.6994059099452822,
+      "learning_rate": 3.5668548962530827e-05,
+      "loss": 0.9406,
+      "step": 15863
+    },
+    {
+      "epoch": 0.7308241581056802,
+      "grad_norm": 1.8291962113252502,
+      "learning_rate": 3.5657126217959344e-05,
+      "loss": 1.1303,
+      "step": 15864
+    },
+    {
+      "epoch": 0.7308702261943152,
+      "grad_norm": 2.2999345356640477,
+      "learning_rate": 3.564570490587169e-05,
+      "loss": 1.2537,
+      "step": 15865
+    },
+    {
+      "epoch": 0.7309162942829502,
+      "grad_norm": 0.8697970270047045,
+      "learning_rate": 3.563428502652204e-05,
+      "loss": 1.3387,
+      "step": 15866
+    },
+    {
+      "epoch": 0.7309623623715852,
+      "grad_norm": 0.6316638421881545,
+      "learning_rate": 3.56228665801647e-05,
+      "loss": 1.4514,
+      "step": 15867
+    },
+    {
+      "epoch": 0.7310084304602202,
+      "grad_norm": 0.8067550707448807,
+      "learning_rate": 3.561144956705379e-05,
+      "loss": 1.3323,
+      "step": 15868
+    },
+    {
+      "epoch": 0.7310544985488552,
+      "grad_norm": 0.9258777239695658,
+      "learning_rate": 3.5600033987443615e-05,
+      "loss": 1.394,
+      "step": 15869
+    },
+    {
+      "epoch": 0.7311005666374902,
+      "grad_norm": 0.8320713760684233,
+      "learning_rate": 3.5588619841588266e-05,
+      "loss": 1.3339,
+      "step": 15870
+    },
+    {
+      "epoch": 0.7311466347261252,
+      "grad_norm": 0.8189807356572112,
+      "learning_rate": 3.5577207129741786e-05,
+      "loss": 1.3175,
+      "step": 15871
+    },
+    {
+      "epoch": 0.7311927028147602,
+      "grad_norm": 0.7241378498867878,
+      "learning_rate": 3.556579585215839e-05,
+      "loss": 1.3302,
+      "step": 15872
+    },
+    {
+      "epoch": 0.7312387709033952,
+      "grad_norm": 0.8359204424753972,
+      "learning_rate": 3.5554386009092046e-05,
+      "loss": 1.428,
+      "step": 15873
+    },
+    {
+      "epoch": 0.7312848389920302,
+      "grad_norm": 1.1455375048675698,
+      "learning_rate": 3.554297760079681e-05,
+      "loss": 1.2237,
+      "step": 15874
+    },
+    {
+      "epoch": 0.7313309070806652,
+      "grad_norm": 1.0046094197851199,
+      "learning_rate": 3.553157062752661e-05,
+      "loss": 0.947,
+      "step": 15875
+    },
+    {
+      "epoch": 0.7313769751693002,
+      "grad_norm": 0.7970814494988329,
+      "learning_rate": 3.5520165089535506e-05,
+      "loss": 1.1898,
+      "step": 15876
+    },
+    {
+      "epoch": 0.7314230432579353,
+      "grad_norm": 1.0487954105188018,
+      "learning_rate": 3.550876098707735e-05,
+      "loss": 1.3804,
+      "step": 15877
+    },
+    {
+      "epoch": 0.7314691113465702,
+      "grad_norm": 0.829409876390876,
+      "learning_rate": 3.549735832040607e-05,
+      "loss": 1.0105,
+      "step": 15878
+    },
+    {
+      "epoch": 0.7315151794352053,
+      "grad_norm": 5.699972011701202,
+      "learning_rate": 3.548595708977548e-05,
+      "loss": 1.4674,
+      "step": 15879
+    },
+    {
+      "epoch": 0.7315612475238402,
+      "grad_norm": 0.8639192801279678,
+      "learning_rate": 3.547455729543944e-05,
+      "loss": 1.1696,
+      "step": 15880
+    },
+    {
+      "epoch": 0.7316073156124753,
+      "grad_norm": 1.1826308243140087,
+      "learning_rate": 3.5463158937651766e-05,
+      "loss": 1.2158,
+      "step": 15881
+    },
+    {
+      "epoch": 0.7316533837011102,
+      "grad_norm": 0.9410590490854215,
+      "learning_rate": 3.545176201666615e-05,
+      "loss": 1.4863,
+      "step": 15882
+    },
+    {
+      "epoch": 0.7316994517897453,
+      "grad_norm": 0.6654424763175629,
+      "learning_rate": 3.544036653273643e-05,
+      "loss": 0.8048,
+      "step": 15883
+    },
+    {
+      "epoch": 0.7317455198783802,
+      "grad_norm": 1.0092968672688394,
+      "learning_rate": 3.542897248611623e-05,
+      "loss": 1.2445,
+      "step": 15884
+    },
+    {
+      "epoch": 0.7317915879670153,
+      "grad_norm": 1.1744370078673465,
+      "learning_rate": 3.541757987705927e-05,
+      "loss": 1.357,
+      "step": 15885
+    },
+    {
+      "epoch": 0.7318376560556502,
+      "grad_norm": 0.8434124515424019,
+      "learning_rate": 3.540618870581909e-05,
+      "loss": 1.3189,
+      "step": 15886
+    },
+    {
+      "epoch": 0.7318837241442853,
+      "grad_norm": 0.7730271695239521,
+      "learning_rate": 3.5394798972649434e-05,
+      "loss": 1.3708,
+      "step": 15887
+    },
+    {
+      "epoch": 0.7319297922329202,
+      "grad_norm": 0.7879596699782565,
+      "learning_rate": 3.5383410677803806e-05,
+      "loss": 1.4745,
+      "step": 15888
+    },
+    {
+      "epoch": 0.7319758603215553,
+      "grad_norm": 0.8898579572658625,
+      "learning_rate": 3.537202382153571e-05,
+      "loss": 1.4787,
+      "step": 15889
+    },
+    {
+      "epoch": 0.7320219284101903,
+      "grad_norm": 0.8387095791787674,
+      "learning_rate": 3.536063840409869e-05,
+      "loss": 1.0073,
+      "step": 15890
+    },
+    {
+      "epoch": 0.7320679964988253,
+      "grad_norm": 0.949800580473846,
+      "learning_rate": 3.534925442574623e-05,
+      "loss": 1.0414,
+      "step": 15891
+    },
+    {
+      "epoch": 0.7321140645874603,
+      "grad_norm": 1.920678876241998,
+      "learning_rate": 3.533787188673179e-05,
+      "loss": 1.3482,
+      "step": 15892
+    },
+    {
+      "epoch": 0.7321601326760953,
+      "grad_norm": 1.258381531808586,
+      "learning_rate": 3.532649078730874e-05,
+      "loss": 1.1052,
+      "step": 15893
+    },
+    {
+      "epoch": 0.7322062007647303,
+      "grad_norm": 1.1676700455570728,
+      "learning_rate": 3.5315111127730475e-05,
+      "loss": 1.4518,
+      "step": 15894
+    },
+    {
+      "epoch": 0.7322522688533653,
+      "grad_norm": 1.1266381429400676,
+      "learning_rate": 3.530373290825034e-05,
+      "loss": 1.3305,
+      "step": 15895
+    },
+    {
+      "epoch": 0.7322983369420003,
+      "grad_norm": 0.94345197020939,
+      "learning_rate": 3.52923561291217e-05,
+      "loss": 1.23,
+      "step": 15896
+    },
+    {
+      "epoch": 0.7323444050306352,
+      "grad_norm": 0.8648511818216845,
+      "learning_rate": 3.5280980790597787e-05,
+      "loss": 1.174,
+      "step": 15897
+    },
+    {
+      "epoch": 0.7323904731192703,
+      "grad_norm": 3.735097060258656,
+      "learning_rate": 3.526960689293185e-05,
+      "loss": 1.4857,
+      "step": 15898
+    },
+    {
+      "epoch": 0.7324365412079052,
+      "grad_norm": 1.002635106824269,
+      "learning_rate": 3.525823443637716e-05,
+      "loss": 0.9336,
+      "step": 15899
+    },
+    {
+      "epoch": 0.7324826092965403,
+      "grad_norm": 0.7508011663917051,
+      "learning_rate": 3.524686342118684e-05,
+      "loss": 1.047,
+      "step": 15900
+    },
+    {
+      "epoch": 0.7325286773851752,
+      "grad_norm": 1.6235114982519183,
+      "learning_rate": 3.523549384761408e-05,
+      "loss": 1.3044,
+      "step": 15901
+    },
+    {
+      "epoch": 0.7325747454738103,
+      "grad_norm": 2.151251491620322,
+      "learning_rate": 3.5224125715912005e-05,
+      "loss": 1.1866,
+      "step": 15902
+    },
+    {
+      "epoch": 0.7326208135624452,
+      "grad_norm": 0.8675492190564468,
+      "learning_rate": 3.521275902633373e-05,
+      "loss": 1.0602,
+      "step": 15903
+    },
+    {
+      "epoch": 0.7326668816510803,
+      "grad_norm": 0.769385026703339,
+      "learning_rate": 3.520139377913225e-05,
+      "loss": 1.2262,
+      "step": 15904
+    },
+    {
+      "epoch": 0.7327129497397153,
+      "grad_norm": 1.6783068414589897,
+      "learning_rate": 3.519002997456065e-05,
+      "loss": 1.4066,
+      "step": 15905
+    },
+    {
+      "epoch": 0.7327590178283503,
+      "grad_norm": 0.733256904836827,
+      "learning_rate": 3.517866761287193e-05,
+      "loss": 1.1531,
+      "step": 15906
+    },
+    {
+      "epoch": 0.7328050859169853,
+      "grad_norm": 0.9731367443406573,
+      "learning_rate": 3.5167306694319e-05,
+      "loss": 1.1649,
+      "step": 15907
+    },
+    {
+      "epoch": 0.7328511540056203,
+      "grad_norm": 0.8394289741901062,
+      "learning_rate": 3.5155947219154815e-05,
+      "loss": 1.3272,
+      "step": 15908
+    },
+    {
+      "epoch": 0.7328972220942553,
+      "grad_norm": 1.207447000121228,
+      "learning_rate": 3.514458918763229e-05,
+      "loss": 0.9659,
+      "step": 15909
+    },
+    {
+      "epoch": 0.7329432901828903,
+      "grad_norm": 0.9337354202524009,
+      "learning_rate": 3.513323260000432e-05,
+      "loss": 1.2642,
+      "step": 15910
+    },
+    {
+      "epoch": 0.7329893582715253,
+      "grad_norm": 0.8995655693907425,
+      "learning_rate": 3.5121877456523664e-05,
+      "loss": 1.4485,
+      "step": 15911
+    },
+    {
+      "epoch": 0.7330354263601603,
+      "grad_norm": 2.45899841940527,
+      "learning_rate": 3.511052375744316e-05,
+      "loss": 1.2865,
+      "step": 15912
+    },
+    {
+      "epoch": 0.7330814944487953,
+      "grad_norm": 1.3334606694026936,
+      "learning_rate": 3.5099171503015596e-05,
+      "loss": 1.2326,
+      "step": 15913
+    },
+    {
+      "epoch": 0.7331275625374303,
+      "grad_norm": 0.8476899253812511,
+      "learning_rate": 3.5087820693493725e-05,
+      "loss": 1.1357,
+      "step": 15914
+    },
+    {
+      "epoch": 0.7331736306260653,
+      "grad_norm": 0.8217655946183952,
+      "learning_rate": 3.507647132913018e-05,
+      "loss": 1.2129,
+      "step": 15915
+    },
+    {
+      "epoch": 0.7332196987147003,
+      "grad_norm": 0.7715201488893559,
+      "learning_rate": 3.506512341017769e-05,
+      "loss": 1.4469,
+      "step": 15916
+    },
+    {
+      "epoch": 0.7332657668033353,
+      "grad_norm": 1.006426145169229,
+      "learning_rate": 3.505377693688893e-05,
+      "loss": 1.3119,
+      "step": 15917
+    },
+    {
+      "epoch": 0.7333118348919704,
+      "grad_norm": 1.1116999329587156,
+      "learning_rate": 3.5042431909516426e-05,
+      "loss": 1.3074,
+      "step": 15918
+    },
+    {
+      "epoch": 0.7333579029806053,
+      "grad_norm": 1.6597223113542445,
+      "learning_rate": 3.50310883283128e-05,
+      "loss": 1.251,
+      "step": 15919
+    },
+    {
+      "epoch": 0.7334039710692404,
+      "grad_norm": 0.80620166207515,
+      "learning_rate": 3.501974619353059e-05,
+      "loss": 1.5258,
+      "step": 15920
+    },
+    {
+      "epoch": 0.7334500391578753,
+      "grad_norm": 1.1049414749363353,
+      "learning_rate": 3.5008405505422346e-05,
+      "loss": 1.1948,
+      "step": 15921
+    },
+    {
+      "epoch": 0.7334961072465104,
+      "grad_norm": 10.629691863853138,
+      "learning_rate": 3.4997066264240484e-05,
+      "loss": 1.1984,
+      "step": 15922
+    },
+    {
+      "epoch": 0.7335421753351453,
+      "grad_norm": 0.8162198704488306,
+      "learning_rate": 3.4985728470237486e-05,
+      "loss": 1.3769,
+      "step": 15923
+    },
+    {
+      "epoch": 0.7335882434237804,
+      "grad_norm": 0.8460549221801912,
+      "learning_rate": 3.4974392123665814e-05,
+      "loss": 1.0004,
+      "step": 15924
+    },
+    {
+      "epoch": 0.7336343115124153,
+      "grad_norm": 0.7176986800797663,
+      "learning_rate": 3.496305722477776e-05,
+      "loss": 1.1975,
+      "step": 15925
+    },
+    {
+      "epoch": 0.7336803796010504,
+      "grad_norm": 0.9104505286778002,
+      "learning_rate": 3.495172377382573e-05,
+      "loss": 1.2233,
+      "step": 15926
+    },
+    {
+      "epoch": 0.7337264476896853,
+      "grad_norm": 1.2438569817496703,
+      "learning_rate": 3.494039177106202e-05,
+      "loss": 1.0088,
+      "step": 15927
+    },
+    {
+      "epoch": 0.7337725157783204,
+      "grad_norm": 0.7224044776362855,
+      "learning_rate": 3.492906121673898e-05,
+      "loss": 1.4349,
+      "step": 15928
+    },
+    {
+      "epoch": 0.7338185838669553,
+      "grad_norm": 0.886101871155736,
+      "learning_rate": 3.4917732111108795e-05,
+      "loss": 1.2448,
+      "step": 15929
+    },
+    {
+      "epoch": 0.7338646519555904,
+      "grad_norm": 2.369787515309592,
+      "learning_rate": 3.490640445442369e-05,
+      "loss": 1.3555,
+      "step": 15930
+    },
+    {
+      "epoch": 0.7339107200442254,
+      "grad_norm": 0.9597307200004649,
+      "learning_rate": 3.48950782469359e-05,
+      "loss": 1.0661,
+      "step": 15931
+    },
+    {
+      "epoch": 0.7339567881328604,
+      "grad_norm": 0.9095699114461766,
+      "learning_rate": 3.488375348889758e-05,
+      "loss": 1.2953,
+      "step": 15932
+    },
+    {
+      "epoch": 0.7340028562214954,
+      "grad_norm": 0.6908476122788942,
+      "learning_rate": 3.487243018056081e-05,
+      "loss": 1.4286,
+      "step": 15933
+    },
+    {
+      "epoch": 0.7340489243101304,
+      "grad_norm": 0.7507514838186122,
+      "learning_rate": 3.4861108322177715e-05,
+      "loss": 1.3855,
+      "step": 15934
+    },
+    {
+      "epoch": 0.7340949923987654,
+      "grad_norm": 0.7957161845004027,
+      "learning_rate": 3.4849787914000386e-05,
+      "loss": 1.1854,
+      "step": 15935
+    },
+    {
+      "epoch": 0.7341410604874004,
+      "grad_norm": 1.1073404468871322,
+      "learning_rate": 3.483846895628078e-05,
+      "loss": 1.0894,
+      "step": 15936
+    },
+    {
+      "epoch": 0.7341871285760354,
+      "grad_norm": 0.7722889370094723,
+      "learning_rate": 3.482715144927095e-05,
+      "loss": 1.1109,
+      "step": 15937
+    },
+    {
+      "epoch": 0.7342331966646704,
+      "grad_norm": 0.7364662337325464,
+      "learning_rate": 3.481583539322283e-05,
+      "loss": 1.3919,
+      "step": 15938
+    },
+    {
+      "epoch": 0.7342792647533054,
+      "grad_norm": 0.9087876282030698,
+      "learning_rate": 3.480452078838841e-05,
+      "loss": 1.209,
+      "step": 15939
+    },
+    {
+      "epoch": 0.7343253328419403,
+      "grad_norm": 1.568954668583523,
+      "learning_rate": 3.4793207635019506e-05,
+      "loss": 1.4228,
+      "step": 15940
+    },
+    {
+      "epoch": 0.7343714009305754,
+      "grad_norm": 0.9637747999354801,
+      "learning_rate": 3.4781895933368033e-05,
+      "loss": 0.9954,
+      "step": 15941
+    },
+    {
+      "epoch": 0.7344174690192103,
+      "grad_norm": 1.0899050545104363,
+      "learning_rate": 3.477058568368583e-05,
+      "loss": 1.2971,
+      "step": 15942
+    },
+    {
+      "epoch": 0.7344635371078454,
+      "grad_norm": 1.039863603107641,
+      "learning_rate": 3.475927688622472e-05,
+      "loss": 1.3356,
+      "step": 15943
+    },
+    {
+      "epoch": 0.7345096051964805,
+      "grad_norm": 0.7190223886882662,
+      "learning_rate": 3.4747969541236417e-05,
+      "loss": 1.4633,
+      "step": 15944
+    },
+    {
+      "epoch": 0.7345556732851154,
+      "grad_norm": 0.9545320034039124,
+      "learning_rate": 3.473666364897268e-05,
+      "loss": 1.1542,
+      "step": 15945
+    },
+    {
+      "epoch": 0.7346017413737504,
+      "grad_norm": 1.1098012462917686,
+      "learning_rate": 3.472535920968526e-05,
+      "loss": 1.2219,
+      "step": 15946
+    },
+    {
+      "epoch": 0.7346478094623854,
+      "grad_norm": 1.0214367698826365,
+      "learning_rate": 3.471405622362577e-05,
+      "loss": 1.4264,
+      "step": 15947
+    },
+    {
+      "epoch": 0.7346938775510204,
+      "grad_norm": 0.7621729347628813,
+      "learning_rate": 3.470275469104586e-05,
+      "loss": 1.4134,
+      "step": 15948
+    },
+    {
+      "epoch": 0.7347399456396554,
+      "grad_norm": 1.3857905941009176,
+      "learning_rate": 3.4691454612197175e-05,
+      "loss": 1.4782,
+      "step": 15949
+    },
+    {
+      "epoch": 0.7347860137282904,
+      "grad_norm": 0.8526083941744784,
+      "learning_rate": 3.4680155987331294e-05,
+      "loss": 1.6294,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7348320818169254,
+      "grad_norm": 1.2595264060646083,
+      "learning_rate": 3.466885881669971e-05,
+      "loss": 1.1468,
+      "step": 15951
+    },
+    {
+      "epoch": 0.7348781499055604,
+      "grad_norm": 2.0166573901793012,
+      "learning_rate": 3.465756310055397e-05,
+      "loss": 1.0467,
+      "step": 15952
+    },
+    {
+      "epoch": 0.7349242179941954,
+      "grad_norm": 1.1042827000701636,
+      "learning_rate": 3.464626883914558e-05,
+      "loss": 1.2388,
+      "step": 15953
+    },
+    {
+      "epoch": 0.7349702860828304,
+      "grad_norm": 1.0511401814057961,
+      "learning_rate": 3.463497603272592e-05,
+      "loss": 1.3699,
+      "step": 15954
+    },
+    {
+      "epoch": 0.7350163541714654,
+      "grad_norm": 0.7660118450618731,
+      "learning_rate": 3.462368468154644e-05,
+      "loss": 1.0987,
+      "step": 15955
+    },
+    {
+      "epoch": 0.7350624222601004,
+      "grad_norm": 2.0943842318138586,
+      "learning_rate": 3.4612394785858525e-05,
+      "loss": 1.0571,
+      "step": 15956
+    },
+    {
+      "epoch": 0.7351084903487354,
+      "grad_norm": 1.4460542108885537,
+      "learning_rate": 3.460110634591357e-05,
+      "loss": 1.1204,
+      "step": 15957
+    },
+    {
+      "epoch": 0.7351545584373704,
+      "grad_norm": 0.7198345844186027,
+      "learning_rate": 3.4589819361962804e-05,
+      "loss": 1.2942,
+      "step": 15958
+    },
+    {
+      "epoch": 0.7352006265260055,
+      "grad_norm": 1.623560075491387,
+      "learning_rate": 3.4578533834257554e-05,
+      "loss": 1.2693,
+      "step": 15959
+    },
+    {
+      "epoch": 0.7352466946146404,
+      "grad_norm": 1.7988039132598732,
+      "learning_rate": 3.456724976304907e-05,
+      "loss": 1.4999,
+      "step": 15960
+    },
+    {
+      "epoch": 0.7352927627032755,
+      "grad_norm": 0.7635524851132649,
+      "learning_rate": 3.455596714858862e-05,
+      "loss": 1.2092,
+      "step": 15961
+    },
+    {
+      "epoch": 0.7353388307919104,
+      "grad_norm": 0.792764681989691,
+      "learning_rate": 3.454468599112735e-05,
+      "loss": 1.1035,
+      "step": 15962
+    },
+    {
+      "epoch": 0.7353848988805455,
+      "grad_norm": 0.8033322114292891,
+      "learning_rate": 3.453340629091634e-05,
+      "loss": 1.2228,
+      "step": 15963
+    },
+    {
+      "epoch": 0.7354309669691804,
+      "grad_norm": 1.1226721618187687,
+      "learning_rate": 3.452212804820686e-05,
+      "loss": 1.1126,
+      "step": 15964
+    },
+    {
+      "epoch": 0.7354770350578155,
+      "grad_norm": 0.8620844784246102,
+      "learning_rate": 3.4510851263249886e-05,
+      "loss": 1.4337,
+      "step": 15965
+    },
+    {
+      "epoch": 0.7355231031464504,
+      "grad_norm": 0.9426117993009363,
+      "learning_rate": 3.449957593629657e-05,
+      "loss": 1.0655,
+      "step": 15966
+    },
+    {
+      "epoch": 0.7355691712350855,
+      "grad_norm": 0.8654740404222184,
+      "learning_rate": 3.448830206759781e-05,
+      "loss": 1.6478,
+      "step": 15967
+    },
+    {
+      "epoch": 0.7356152393237204,
+      "grad_norm": 0.7543680003274029,
+      "learning_rate": 3.4477029657404735e-05,
+      "loss": 1.1698,
+      "step": 15968
+    },
+    {
+      "epoch": 0.7356613074123555,
+      "grad_norm": 1.1135717407553316,
+      "learning_rate": 3.446575870596822e-05,
+      "loss": 1.3266,
+      "step": 15969
+    },
+    {
+      "epoch": 0.7357073755009904,
+      "grad_norm": 1.1776586704274723,
+      "learning_rate": 3.445448921353922e-05,
+      "loss": 1.2934,
+      "step": 15970
+    },
+    {
+      "epoch": 0.7357534435896255,
+      "grad_norm": 0.8128171936856262,
+      "learning_rate": 3.4443221180368666e-05,
+      "loss": 1.2355,
+      "step": 15971
+    },
+    {
+      "epoch": 0.7357995116782605,
+      "grad_norm": 1.0119639476040132,
+      "learning_rate": 3.4431954606707365e-05,
+      "loss": 1.2218,
+      "step": 15972
+    },
+    {
+      "epoch": 0.7358455797668955,
+      "grad_norm": 1.0714927544305253,
+      "learning_rate": 3.442068949280618e-05,
+      "loss": 1.0023,
+      "step": 15973
+    },
+    {
+      "epoch": 0.7358916478555305,
+      "grad_norm": 0.8173143021177439,
+      "learning_rate": 3.440942583891584e-05,
+      "loss": 1.2389,
+      "step": 15974
+    },
+    {
+      "epoch": 0.7359377159441655,
+      "grad_norm": 1.5863066307271176,
+      "learning_rate": 3.4398163645287254e-05,
+      "loss": 1.3286,
+      "step": 15975
+    },
+    {
+      "epoch": 0.7359837840328005,
+      "grad_norm": 0.8037690231970339,
+      "learning_rate": 3.4386902912171014e-05,
+      "loss": 1.1955,
+      "step": 15976
+    },
+    {
+      "epoch": 0.7360298521214355,
+      "grad_norm": 1.2659163878862763,
+      "learning_rate": 3.4375643639817925e-05,
+      "loss": 1.7311,
+      "step": 15977
+    },
+    {
+      "epoch": 0.7360759202100705,
+      "grad_norm": 0.926785134092498,
+      "learning_rate": 3.436438582847854e-05,
+      "loss": 1.2286,
+      "step": 15978
+    },
+    {
+      "epoch": 0.7361219882987055,
+      "grad_norm": 0.9155651407719033,
+      "learning_rate": 3.435312947840362e-05,
+      "loss": 1.3758,
+      "step": 15979
+    },
+    {
+      "epoch": 0.7361680563873405,
+      "grad_norm": 0.6583305150524521,
+      "learning_rate": 3.4341874589843725e-05,
+      "loss": 1.2372,
+      "step": 15980
+    },
+    {
+      "epoch": 0.7362141244759755,
+      "grad_norm": 0.9457458130631576,
+      "learning_rate": 3.433062116304935e-05,
+      "loss": 1.3323,
+      "step": 15981
+    },
+    {
+      "epoch": 0.7362601925646105,
+      "grad_norm": 1.0638605304207092,
+      "learning_rate": 3.4319369198271145e-05,
+      "loss": 1.4007,
+      "step": 15982
+    },
+    {
+      "epoch": 0.7363062606532454,
+      "grad_norm": 0.6321154156560452,
+      "learning_rate": 3.430811869575955e-05,
+      "loss": 1.1771,
+      "step": 15983
+    },
+    {
+      "epoch": 0.7363523287418805,
+      "grad_norm": 1.401019985546732,
+      "learning_rate": 3.4296869655765077e-05,
+      "loss": 1.4843,
+      "step": 15984
+    },
+    {
+      "epoch": 0.7363983968305156,
+      "grad_norm": 2.869531857979553,
+      "learning_rate": 3.4285622078538083e-05,
+      "loss": 1.2194,
+      "step": 15985
+    },
+    {
+      "epoch": 0.7364444649191505,
+      "grad_norm": 0.8954217779933237,
+      "learning_rate": 3.42743759643291e-05,
+      "loss": 0.9173,
+      "step": 15986
+    },
+    {
+      "epoch": 0.7364905330077856,
+      "grad_norm": 0.9996212492699712,
+      "learning_rate": 3.4263131313388406e-05,
+      "loss": 1.1581,
+      "step": 15987
+    },
+    {
+      "epoch": 0.7365366010964205,
+      "grad_norm": 0.9860049219701349,
+      "learning_rate": 3.425188812596643e-05,
+      "loss": 1.4327,
+      "step": 15988
+    },
+    {
+      "epoch": 0.7365826691850555,
+      "grad_norm": 1.4091708864491144,
+      "learning_rate": 3.424064640231338e-05,
+      "loss": 1.1769,
+      "step": 15989
+    },
+    {
+      "epoch": 0.7366287372736905,
+      "grad_norm": 0.7391981158481035,
+      "learning_rate": 3.422940614267959e-05,
+      "loss": 0.874,
+      "step": 15990
+    },
+    {
+      "epoch": 0.7366748053623255,
+      "grad_norm": 0.9470843535205293,
+      "learning_rate": 3.421816734731533e-05,
+      "loss": 0.941,
+      "step": 15991
+    },
+    {
+      "epoch": 0.7367208734509605,
+      "grad_norm": 1.2082565575659916,
+      "learning_rate": 3.4206930016470754e-05,
+      "loss": 1.3577,
+      "step": 15992
+    },
+    {
+      "epoch": 0.7367669415395955,
+      "grad_norm": 0.7116571401675799,
+      "learning_rate": 3.4195694150396065e-05,
+      "loss": 1.3984,
+      "step": 15993
+    },
+    {
+      "epoch": 0.7368130096282305,
+      "grad_norm": 0.9119806594069003,
+      "learning_rate": 3.418445974934141e-05,
+      "loss": 1.2139,
+      "step": 15994
+    },
+    {
+      "epoch": 0.7368590777168655,
+      "grad_norm": 1.2420846594723425,
+      "learning_rate": 3.4173226813556956e-05,
+      "loss": 0.8934,
+      "step": 15995
+    },
+    {
+      "epoch": 0.7369051458055005,
+      "grad_norm": 0.8849042815776116,
+      "learning_rate": 3.4161995343292664e-05,
+      "loss": 1.1336,
+      "step": 15996
+    },
+    {
+      "epoch": 0.7369512138941355,
+      "grad_norm": 1.006173457823642,
+      "learning_rate": 3.415076533879873e-05,
+      "loss": 1.1844,
+      "step": 15997
+    },
+    {
+      "epoch": 0.7369972819827705,
+      "grad_norm": 3.4538664771115926,
+      "learning_rate": 3.413953680032509e-05,
+      "loss": 1.0948,
+      "step": 15998
+    },
+    {
+      "epoch": 0.7370433500714055,
+      "grad_norm": 0.7753786531672708,
+      "learning_rate": 3.412830972812171e-05,
+      "loss": 1.2968,
+      "step": 15999
+    },
+    {
+      "epoch": 0.7370894181600406,
+      "grad_norm": 1.3122835626182476,
+      "learning_rate": 3.411708412243857e-05,
+      "loss": 0.9148,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7371354862486755,
+      "grad_norm": 1.3161655381821002,
+      "learning_rate": 3.410585998352558e-05,
+      "loss": 1.1422,
+      "step": 16001
+    },
+    {
+      "epoch": 0.7371815543373106,
+      "grad_norm": 0.7941101032126712,
+      "learning_rate": 3.409463731163267e-05,
+      "loss": 1.4167,
+      "step": 16002
+    },
+    {
+      "epoch": 0.7372276224259455,
+      "grad_norm": 1.0281366471614846,
+      "learning_rate": 3.4083416107009626e-05,
+      "loss": 1.4386,
+      "step": 16003
+    },
+    {
+      "epoch": 0.7372736905145806,
+      "grad_norm": 1.1261557790327248,
+      "learning_rate": 3.407219636990629e-05,
+      "loss": 1.0528,
+      "step": 16004
+    },
+    {
+      "epoch": 0.7373197586032155,
+      "grad_norm": 1.552638995738654,
+      "learning_rate": 3.406097810057247e-05,
+      "loss": 1.2185,
+      "step": 16005
+    },
+    {
+      "epoch": 0.7373658266918506,
+      "grad_norm": 0.7576470153017832,
+      "learning_rate": 3.404976129925794e-05,
+      "loss": 1.2449,
+      "step": 16006
+    },
+    {
+      "epoch": 0.7374118947804855,
+      "grad_norm": 1.4522297944520557,
+      "learning_rate": 3.4038545966212366e-05,
+      "loss": 1.0171,
+      "step": 16007
+    },
+    {
+      "epoch": 0.7374579628691206,
+      "grad_norm": 0.8178399554080364,
+      "learning_rate": 3.402733210168546e-05,
+      "loss": 1.1407,
+      "step": 16008
+    },
+    {
+      "epoch": 0.7375040309577555,
+      "grad_norm": 1.265071860750325,
+      "learning_rate": 3.401611970592693e-05,
+      "loss": 1.323,
+      "step": 16009
+    },
+    {
+      "epoch": 0.7375500990463906,
+      "grad_norm": 0.8794948194169588,
+      "learning_rate": 3.400490877918632e-05,
+      "loss": 1.5024,
+      "step": 16010
+    },
+    {
+      "epoch": 0.7375961671350255,
+      "grad_norm": 0.7368020363272452,
+      "learning_rate": 3.399369932171325e-05,
+      "loss": 1.3865,
+      "step": 16011
+    },
+    {
+      "epoch": 0.7376422352236606,
+      "grad_norm": 1.6498594804086024,
+      "learning_rate": 3.398249133375729e-05,
+      "loss": 1.6092,
+      "step": 16012
+    },
+    {
+      "epoch": 0.7376883033122956,
+      "grad_norm": 1.5747359879684226,
+      "learning_rate": 3.3971284815568005e-05,
+      "loss": 1.0294,
+      "step": 16013
+    },
+    {
+      "epoch": 0.7377343714009306,
+      "grad_norm": 0.8499624672417213,
+      "learning_rate": 3.3960079767394804e-05,
+      "loss": 1.5528,
+      "step": 16014
+    },
+    {
+      "epoch": 0.7377804394895656,
+      "grad_norm": 0.8182767801129439,
+      "learning_rate": 3.39488761894872e-05,
+      "loss": 1.4498,
+      "step": 16015
+    },
+    {
+      "epoch": 0.7378265075782006,
+      "grad_norm": 0.878281634604436,
+      "learning_rate": 3.393767408209465e-05,
+      "loss": 1.315,
+      "step": 16016
+    },
+    {
+      "epoch": 0.7378725756668356,
+      "grad_norm": 0.8672202207212656,
+      "learning_rate": 3.392647344546648e-05,
+      "loss": 1.2294,
+      "step": 16017
+    },
+    {
+      "epoch": 0.7379186437554706,
+      "grad_norm": 0.8529854721105802,
+      "learning_rate": 3.391527427985208e-05,
+      "loss": 1.2428,
+      "step": 16018
+    },
+    {
+      "epoch": 0.7379647118441056,
+      "grad_norm": 0.81496231231504,
+      "learning_rate": 3.39040765855008e-05,
+      "loss": 1.3097,
+      "step": 16019
+    },
+    {
+      "epoch": 0.7380107799327406,
+      "grad_norm": 2.0630945174426443,
+      "learning_rate": 3.389288036266194e-05,
+      "loss": 1.2029,
+      "step": 16020
+    },
+    {
+      "epoch": 0.7380568480213756,
+      "grad_norm": 1.29822939861873,
+      "learning_rate": 3.3881685611584734e-05,
+      "loss": 1.0224,
+      "step": 16021
+    },
+    {
+      "epoch": 0.7381029161100106,
+      "grad_norm": 0.72851562206342,
+      "learning_rate": 3.387049233251842e-05,
+      "loss": 1.221,
+      "step": 16022
+    },
+    {
+      "epoch": 0.7381489841986456,
+      "grad_norm": 1.232631495468304,
+      "learning_rate": 3.38593005257122e-05,
+      "loss": 1.3995,
+      "step": 16023
+    },
+    {
+      "epoch": 0.7381950522872806,
+      "grad_norm": 0.8317967892943382,
+      "learning_rate": 3.38481101914153e-05,
+      "loss": 1.3648,
+      "step": 16024
+    },
+    {
+      "epoch": 0.7382411203759156,
+      "grad_norm": 0.9799979058341537,
+      "learning_rate": 3.383692132987675e-05,
+      "loss": 1.0956,
+      "step": 16025
+    },
+    {
+      "epoch": 0.7382871884645507,
+      "grad_norm": 1.0440482467045593,
+      "learning_rate": 3.38257339413457e-05,
+      "loss": 0.779,
+      "step": 16026
+    },
+    {
+      "epoch": 0.7383332565531856,
+      "grad_norm": 0.8220104333890127,
+      "learning_rate": 3.381454802607126e-05,
+      "loss": 1.3123,
+      "step": 16027
+    },
+    {
+      "epoch": 0.7383793246418207,
+      "grad_norm": 0.8425309403897646,
+      "learning_rate": 3.3803363584302397e-05,
+      "loss": 1.1726,
+      "step": 16028
+    },
+    {
+      "epoch": 0.7384253927304556,
+      "grad_norm": 0.7964542438149204,
+      "learning_rate": 3.3792180616288125e-05,
+      "loss": 1.2301,
+      "step": 16029
+    },
+    {
+      "epoch": 0.7384714608190907,
+      "grad_norm": 1.0449792013303565,
+      "learning_rate": 3.3780999122277444e-05,
+      "loss": 1.3833,
+      "step": 16030
+    },
+    {
+      "epoch": 0.7385175289077256,
+      "grad_norm": 0.7691689843751562,
+      "learning_rate": 3.376981910251929e-05,
+      "loss": 1.6224,
+      "step": 16031
+    },
+    {
+      "epoch": 0.7385635969963606,
+      "grad_norm": 2.7303102280837304,
+      "learning_rate": 3.3758640557262544e-05,
+      "loss": 1.2923,
+      "step": 16032
+    },
+    {
+      "epoch": 0.7386096650849956,
+      "grad_norm": 0.6924331454299774,
+      "learning_rate": 3.374746348675608e-05,
+      "loss": 1.3603,
+      "step": 16033
+    },
+    {
+      "epoch": 0.7386557331736306,
+      "grad_norm": 1.1820336359544201,
+      "learning_rate": 3.373628789124876e-05,
+      "loss": 1.0996,
+      "step": 16034
+    },
+    {
+      "epoch": 0.7387018012622656,
+      "grad_norm": 0.6020073164539754,
+      "learning_rate": 3.372511377098936e-05,
+      "loss": 1.3625,
+      "step": 16035
+    },
+    {
+      "epoch": 0.7387478693509006,
+      "grad_norm": 1.1593565516714206,
+      "learning_rate": 3.3713941126226654e-05,
+      "loss": 1.0569,
+      "step": 16036
+    },
+    {
+      "epoch": 0.7387939374395356,
+      "grad_norm": 3.863361207903253,
+      "learning_rate": 3.370276995720939e-05,
+      "loss": 1.251,
+      "step": 16037
+    },
+    {
+      "epoch": 0.7388400055281706,
+      "grad_norm": 1.9374218930604674,
+      "learning_rate": 3.369160026418632e-05,
+      "loss": 1.1655,
+      "step": 16038
+    },
+    {
+      "epoch": 0.7388860736168057,
+      "grad_norm": 0.711231203695703,
+      "learning_rate": 3.368043204740604e-05,
+      "loss": 1.3913,
+      "step": 16039
+    },
+    {
+      "epoch": 0.7389321417054406,
+      "grad_norm": 1.2848546463905623,
+      "learning_rate": 3.3669265307117224e-05,
+      "loss": 1.064,
+      "step": 16040
+    },
+    {
+      "epoch": 0.7389782097940757,
+      "grad_norm": 1.0198477690517735,
+      "learning_rate": 3.3658100043568484e-05,
+      "loss": 1.1521,
+      "step": 16041
+    },
+    {
+      "epoch": 0.7390242778827106,
+      "grad_norm": 0.8417344121241073,
+      "learning_rate": 3.364693625700842e-05,
+      "loss": 1.4949,
+      "step": 16042
+    },
+    {
+      "epoch": 0.7390703459713457,
+      "grad_norm": 0.7199052993636199,
+      "learning_rate": 3.3635773947685534e-05,
+      "loss": 1.4259,
+      "step": 16043
+    },
+    {
+      "epoch": 0.7391164140599806,
+      "grad_norm": 1.0384154896672217,
+      "learning_rate": 3.3624613115848336e-05,
+      "loss": 1.4376,
+      "step": 16044
+    },
+    {
+      "epoch": 0.7391624821486157,
+      "grad_norm": 0.8256620938583593,
+      "learning_rate": 3.361345376174535e-05,
+      "loss": 1.5697,
+      "step": 16045
+    },
+    {
+      "epoch": 0.7392085502372506,
+      "grad_norm": 0.7390245535903902,
+      "learning_rate": 3.360229588562497e-05,
+      "loss": 0.9642,
+      "step": 16046
+    },
+    {
+      "epoch": 0.7392546183258857,
+      "grad_norm": 0.9551168558419135,
+      "learning_rate": 3.35911394877356e-05,
+      "loss": 1.2205,
+      "step": 16047
+    },
+    {
+      "epoch": 0.7393006864145206,
+      "grad_norm": 0.9081407457541794,
+      "learning_rate": 3.357998456832566e-05,
+      "loss": 1.3969,
+      "step": 16048
+    },
+    {
+      "epoch": 0.7393467545031557,
+      "grad_norm": 0.8199011810709391,
+      "learning_rate": 3.356883112764351e-05,
+      "loss": 1.4718,
+      "step": 16049
+    },
+    {
+      "epoch": 0.7393928225917906,
+      "grad_norm": 1.9367952533855533,
+      "learning_rate": 3.3557679165937395e-05,
+      "loss": 1.0858,
+      "step": 16050
+    },
+    {
+      "epoch": 0.7394388906804257,
+      "grad_norm": 0.9011228173629972,
+      "learning_rate": 3.354652868345562e-05,
+      "loss": 1.1617,
+      "step": 16051
+    },
+    {
+      "epoch": 0.7394849587690606,
+      "grad_norm": 0.830665066564,
+      "learning_rate": 3.3535379680446456e-05,
+      "loss": 1.3031,
+      "step": 16052
+    },
+    {
+      "epoch": 0.7395310268576957,
+      "grad_norm": 1.07667137450128,
+      "learning_rate": 3.352423215715812e-05,
+      "loss": 1.0818,
+      "step": 16053
+    },
+    {
+      "epoch": 0.7395770949463307,
+      "grad_norm": 2.784551621713365,
+      "learning_rate": 3.351308611383875e-05,
+      "loss": 1.0578,
+      "step": 16054
+    },
+    {
+      "epoch": 0.7396231630349657,
+      "grad_norm": 0.856491195384805,
+      "learning_rate": 3.350194155073652e-05,
+      "loss": 1.2347,
+      "step": 16055
+    },
+    {
+      "epoch": 0.7396692311236007,
+      "grad_norm": 0.6611715473133494,
+      "learning_rate": 3.349079846809956e-05,
+      "loss": 1.2944,
+      "step": 16056
+    },
+    {
+      "epoch": 0.7397152992122357,
+      "grad_norm": 0.8138408212778689,
+      "learning_rate": 3.347965686617591e-05,
+      "loss": 1.0639,
+      "step": 16057
+    },
+    {
+      "epoch": 0.7397613673008707,
+      "grad_norm": 0.59121767991129,
+      "learning_rate": 3.346851674521364e-05,
+      "loss": 1.415,
+      "step": 16058
+    },
+    {
+      "epoch": 0.7398074353895057,
+      "grad_norm": 1.1485178431649572,
+      "learning_rate": 3.345737810546078e-05,
+      "loss": 1.4052,
+      "step": 16059
+    },
+    {
+      "epoch": 0.7398535034781407,
+      "grad_norm": 0.7045896371826573,
+      "learning_rate": 3.344624094716532e-05,
+      "loss": 1.1286,
+      "step": 16060
+    },
+    {
+      "epoch": 0.7398995715667757,
+      "grad_norm": 0.7225562958558941,
+      "learning_rate": 3.343510527057515e-05,
+      "loss": 1.2024,
+      "step": 16061
+    },
+    {
+      "epoch": 0.7399456396554107,
+      "grad_norm": 1.0827371832458612,
+      "learning_rate": 3.342397107593824e-05,
+      "loss": 1.1546,
+      "step": 16062
+    },
+    {
+      "epoch": 0.7399917077440457,
+      "grad_norm": 1.1038017049028692,
+      "learning_rate": 3.3412838363502485e-05,
+      "loss": 0.8528,
+      "step": 16063
+    },
+    {
+      "epoch": 0.7400377758326807,
+      "grad_norm": 0.7098053778185541,
+      "learning_rate": 3.340170713351569e-05,
+      "loss": 1.2218,
+      "step": 16064
+    },
+    {
+      "epoch": 0.7400838439213157,
+      "grad_norm": 0.7271667290645488,
+      "learning_rate": 3.339057738622572e-05,
+      "loss": 1.5087,
+      "step": 16065
+    },
+    {
+      "epoch": 0.7401299120099507,
+      "grad_norm": 0.9682281825165199,
+      "learning_rate": 3.337944912188027e-05,
+      "loss": 1.0751,
+      "step": 16066
+    },
+    {
+      "epoch": 0.7401759800985858,
+      "grad_norm": 2.160809577697057,
+      "learning_rate": 3.336832234072722e-05,
+      "loss": 0.9761,
+      "step": 16067
+    },
+    {
+      "epoch": 0.7402220481872207,
+      "grad_norm": 1.28625363577128,
+      "learning_rate": 3.335719704301419e-05,
+      "loss": 1.2685,
+      "step": 16068
+    },
+    {
+      "epoch": 0.7402681162758558,
+      "grad_norm": 0.6571518510737945,
+      "learning_rate": 3.3346073228988895e-05,
+      "loss": 1.5022,
+      "step": 16069
+    },
+    {
+      "epoch": 0.7403141843644907,
+      "grad_norm": 1.0606808642168994,
+      "learning_rate": 3.3334950898899e-05,
+      "loss": 1.0418,
+      "step": 16070
+    },
+    {
+      "epoch": 0.7403602524531258,
+      "grad_norm": 0.771181856492836,
+      "learning_rate": 3.3323830052992146e-05,
+      "loss": 1.23,
+      "step": 16071
+    },
+    {
+      "epoch": 0.7404063205417607,
+      "grad_norm": 7.194504725652362,
+      "learning_rate": 3.331271069151589e-05,
+      "loss": 1.1199,
+      "step": 16072
+    },
+    {
+      "epoch": 0.7404523886303958,
+      "grad_norm": 0.9398950127289953,
+      "learning_rate": 3.330159281471773e-05,
+      "loss": 1.1058,
+      "step": 16073
+    },
+    {
+      "epoch": 0.7404984567190307,
+      "grad_norm": 1.1633198577357837,
+      "learning_rate": 3.3290476422845295e-05,
+      "loss": 1.2387,
+      "step": 16074
+    },
+    {
+      "epoch": 0.7405445248076657,
+      "grad_norm": 0.7695559054094806,
+      "learning_rate": 3.3279361516146e-05,
+      "loss": 1.1924,
+      "step": 16075
+    },
+    {
+      "epoch": 0.7405905928963007,
+      "grad_norm": 1.8026208645933204,
+      "learning_rate": 3.326824809486735e-05,
+      "loss": 1.3672,
+      "step": 16076
+    },
+    {
+      "epoch": 0.7406366609849357,
+      "grad_norm": 1.0140876989714567,
+      "learning_rate": 3.3257136159256664e-05,
+      "loss": 1.0158,
+      "step": 16077
+    },
+    {
+      "epoch": 0.7406827290735707,
+      "grad_norm": 0.6991141137146042,
+      "learning_rate": 3.324602570956147e-05,
+      "loss": 1.2847,
+      "step": 16078
+    },
+    {
+      "epoch": 0.7407287971622057,
+      "grad_norm": 1.0211099110594672,
+      "learning_rate": 3.323491674602902e-05,
+      "loss": 1.0782,
+      "step": 16079
+    },
+    {
+      "epoch": 0.7407748652508408,
+      "grad_norm": 0.6587619550824283,
+      "learning_rate": 3.3223809268906706e-05,
+      "loss": 1.2762,
+      "step": 16080
+    },
+    {
+      "epoch": 0.7408209333394757,
+      "grad_norm": 0.7094860004155221,
+      "learning_rate": 3.321270327844175e-05,
+      "loss": 1.3073,
+      "step": 16081
+    },
+    {
+      "epoch": 0.7408670014281108,
+      "grad_norm": 1.093245506346001,
+      "learning_rate": 3.320159877488144e-05,
+      "loss": 1.255,
+      "step": 16082
+    },
+    {
+      "epoch": 0.7409130695167457,
+      "grad_norm": 1.9653705030606348,
+      "learning_rate": 3.319049575847303e-05,
+      "loss": 1.4338,
+      "step": 16083
+    },
+    {
+      "epoch": 0.7409591376053808,
+      "grad_norm": 0.7005160837205492,
+      "learning_rate": 3.317939422946361e-05,
+      "loss": 1.1584,
+      "step": 16084
+    },
+    {
+      "epoch": 0.7410052056940157,
+      "grad_norm": 0.7271584785372361,
+      "learning_rate": 3.316829418810048e-05,
+      "loss": 1.232,
+      "step": 16085
+    },
+    {
+      "epoch": 0.7410512737826508,
+      "grad_norm": 0.9363023694236485,
+      "learning_rate": 3.315719563463064e-05,
+      "loss": 1.2081,
+      "step": 16086
+    },
+    {
+      "epoch": 0.7410973418712857,
+      "grad_norm": 1.0719205845177555,
+      "learning_rate": 3.314609856930128e-05,
+      "loss": 0.9798,
+      "step": 16087
+    },
+    {
+      "epoch": 0.7411434099599208,
+      "grad_norm": 1.0177212344116997,
+      "learning_rate": 3.313500299235933e-05,
+      "loss": 1.2121,
+      "step": 16088
+    },
+    {
+      "epoch": 0.7411894780485557,
+      "grad_norm": 1.7514401862934261,
+      "learning_rate": 3.312390890405195e-05,
+      "loss": 1.4052,
+      "step": 16089
+    },
+    {
+      "epoch": 0.7412355461371908,
+      "grad_norm": 0.8373535482297985,
+      "learning_rate": 3.3112816304626085e-05,
+      "loss": 1.3754,
+      "step": 16090
+    },
+    {
+      "epoch": 0.7412816142258257,
+      "grad_norm": 1.3152063998215793,
+      "learning_rate": 3.310172519432863e-05,
+      "loss": 1.4553,
+      "step": 16091
+    },
+    {
+      "epoch": 0.7413276823144608,
+      "grad_norm": 0.9433811334469302,
+      "learning_rate": 3.309063557340656e-05,
+      "loss": 1.3218,
+      "step": 16092
+    },
+    {
+      "epoch": 0.7413737504030957,
+      "grad_norm": 1.2967913384976026,
+      "learning_rate": 3.3079547442106776e-05,
+      "loss": 0.9781,
+      "step": 16093
+    },
+    {
+      "epoch": 0.7414198184917308,
+      "grad_norm": 0.6929236680494988,
+      "learning_rate": 3.3068460800676136e-05,
+      "loss": 1.3621,
+      "step": 16094
+    },
+    {
+      "epoch": 0.7414658865803658,
+      "grad_norm": 0.9608275328599352,
+      "learning_rate": 3.305737564936141e-05,
+      "loss": 1.3433,
+      "step": 16095
+    },
+    {
+      "epoch": 0.7415119546690008,
+      "grad_norm": 1.1013366181143356,
+      "learning_rate": 3.3046291988409486e-05,
+      "loss": 1.1648,
+      "step": 16096
+    },
+    {
+      "epoch": 0.7415580227576358,
+      "grad_norm": 0.8435287777330485,
+      "learning_rate": 3.303520981806704e-05,
+      "loss": 1.1511,
+      "step": 16097
+    },
+    {
+      "epoch": 0.7416040908462708,
+      "grad_norm": 1.268561094200287,
+      "learning_rate": 3.302412913858086e-05,
+      "loss": 1.3753,
+      "step": 16098
+    },
+    {
+      "epoch": 0.7416501589349058,
+      "grad_norm": 4.644520356038372,
+      "learning_rate": 3.3013049950197574e-05,
+      "loss": 1.0691,
+      "step": 16099
+    },
+    {
+      "epoch": 0.7416962270235408,
+      "grad_norm": 0.9626609320064619,
+      "learning_rate": 3.3001972253163874e-05,
+      "loss": 1.1372,
+      "step": 16100
+    },
+    {
+      "epoch": 0.7417422951121758,
+      "grad_norm": 0.8190497691466472,
+      "learning_rate": 3.29908960477264e-05,
+      "loss": 1.4361,
+      "step": 16101
+    },
+    {
+      "epoch": 0.7417883632008108,
+      "grad_norm": 1.140674205009228,
+      "learning_rate": 3.297982133413172e-05,
+      "loss": 1.2702,
+      "step": 16102
+    },
+    {
+      "epoch": 0.7418344312894458,
+      "grad_norm": 0.9686391281179907,
+      "learning_rate": 3.29687481126264e-05,
+      "loss": 1.1413,
+      "step": 16103
+    },
+    {
+      "epoch": 0.7418804993780808,
+      "grad_norm": 1.009927623466213,
+      "learning_rate": 3.295767638345696e-05,
+      "loss": 1.354,
+      "step": 16104
+    },
+    {
+      "epoch": 0.7419265674667158,
+      "grad_norm": 0.9505806597782877,
+      "learning_rate": 3.294660614686994e-05,
+      "loss": 1.0979,
+      "step": 16105
+    },
+    {
+      "epoch": 0.7419726355553508,
+      "grad_norm": 3.4477248748652376,
+      "learning_rate": 3.2935537403111726e-05,
+      "loss": 0.9721,
+      "step": 16106
+    },
+    {
+      "epoch": 0.7420187036439858,
+      "grad_norm": 0.8057244979384801,
+      "learning_rate": 3.292447015242879e-05,
+      "loss": 1.209,
+      "step": 16107
+    },
+    {
+      "epoch": 0.7420647717326209,
+      "grad_norm": 1.0666848450760764,
+      "learning_rate": 3.291340439506754e-05,
+      "loss": 1.2031,
+      "step": 16108
+    },
+    {
+      "epoch": 0.7421108398212558,
+      "grad_norm": 0.9598463467834841,
+      "learning_rate": 3.290234013127429e-05,
+      "loss": 1.1455,
+      "step": 16109
+    },
+    {
+      "epoch": 0.7421569079098909,
+      "grad_norm": 2.2249040653759042,
+      "learning_rate": 3.289127736129538e-05,
+      "loss": 1.3979,
+      "step": 16110
+    },
+    {
+      "epoch": 0.7422029759985258,
+      "grad_norm": 0.9069478246172312,
+      "learning_rate": 3.2880216085377114e-05,
+      "loss": 1.1831,
+      "step": 16111
+    },
+    {
+      "epoch": 0.7422490440871609,
+      "grad_norm": 1.1117944252974568,
+      "learning_rate": 3.286915630376578e-05,
+      "loss": 0.9616,
+      "step": 16112
+    },
+    {
+      "epoch": 0.7422951121757958,
+      "grad_norm": 0.9468055649414902,
+      "learning_rate": 3.285809801670755e-05,
+      "loss": 1.4024,
+      "step": 16113
+    },
+    {
+      "epoch": 0.7423411802644309,
+      "grad_norm": 1.2227873503307,
+      "learning_rate": 3.284704122444864e-05,
+      "loss": 1.1697,
+      "step": 16114
+    },
+    {
+      "epoch": 0.7423872483530658,
+      "grad_norm": 1.0143468969247846,
+      "learning_rate": 3.283598592723521e-05,
+      "loss": 1.2911,
+      "step": 16115
+    },
+    {
+      "epoch": 0.7424333164417009,
+      "grad_norm": 0.6727509329633817,
+      "learning_rate": 3.2824932125313435e-05,
+      "loss": 1.3263,
+      "step": 16116
+    },
+    {
+      "epoch": 0.7424793845303358,
+      "grad_norm": 0.8386265304296141,
+      "learning_rate": 3.2813879818929326e-05,
+      "loss": 1.4163,
+      "step": 16117
+    },
+    {
+      "epoch": 0.7425254526189708,
+      "grad_norm": 1.075430173324688,
+      "learning_rate": 3.280282900832899e-05,
+      "loss": 1.0009,
+      "step": 16118
+    },
+    {
+      "epoch": 0.7425715207076058,
+      "grad_norm": 0.9298155634550581,
+      "learning_rate": 3.2791779693758476e-05,
+      "loss": 1.3176,
+      "step": 16119
+    },
+    {
+      "epoch": 0.7426175887962408,
+      "grad_norm": 0.8991135079522976,
+      "learning_rate": 3.2780731875463724e-05,
+      "loss": 1.5007,
+      "step": 16120
+    },
+    {
+      "epoch": 0.7426636568848759,
+      "grad_norm": 0.8793540441401579,
+      "learning_rate": 3.276968555369071e-05,
+      "loss": 1.0564,
+      "step": 16121
+    },
+    {
+      "epoch": 0.7427097249735108,
+      "grad_norm": 0.8937108230555706,
+      "learning_rate": 3.275864072868537e-05,
+      "loss": 1.2636,
+      "step": 16122
+    },
+    {
+      "epoch": 0.7427557930621459,
+      "grad_norm": 0.8006177326095971,
+      "learning_rate": 3.274759740069364e-05,
+      "loss": 1.1768,
+      "step": 16123
+    },
+    {
+      "epoch": 0.7428018611507808,
+      "grad_norm": 2.2037865810141204,
+      "learning_rate": 3.273655556996131e-05,
+      "loss": 1.3745,
+      "step": 16124
+    },
+    {
+      "epoch": 0.7428479292394159,
+      "grad_norm": 1.2246972148912811,
+      "learning_rate": 3.272551523673424e-05,
+      "loss": 1.1169,
+      "step": 16125
+    },
+    {
+      "epoch": 0.7428939973280508,
+      "grad_norm": 0.9308323812419171,
+      "learning_rate": 3.2714476401258244e-05,
+      "loss": 0.9968,
+      "step": 16126
+    },
+    {
+      "epoch": 0.7429400654166859,
+      "grad_norm": 0.8937415182382905,
+      "learning_rate": 3.270343906377903e-05,
+      "loss": 1.2205,
+      "step": 16127
+    },
+    {
+      "epoch": 0.7429861335053208,
+      "grad_norm": 0.9526123462669084,
+      "learning_rate": 3.269240322454237e-05,
+      "loss": 1.3324,
+      "step": 16128
+    },
+    {
+      "epoch": 0.7430322015939559,
+      "grad_norm": 0.8003629130409684,
+      "learning_rate": 3.268136888379393e-05,
+      "loss": 1.3623,
+      "step": 16129
+    },
+    {
+      "epoch": 0.7430782696825908,
+      "grad_norm": 1.033672782760318,
+      "learning_rate": 3.267033604177943e-05,
+      "loss": 1.3343,
+      "step": 16130
+    },
+    {
+      "epoch": 0.7431243377712259,
+      "grad_norm": 0.9988718004835336,
+      "learning_rate": 3.26593046987444e-05,
+      "loss": 1.1544,
+      "step": 16131
+    },
+    {
+      "epoch": 0.7431704058598608,
+      "grad_norm": 0.8187202420798226,
+      "learning_rate": 3.2648274854934504e-05,
+      "loss": 1.3525,
+      "step": 16132
+    },
+    {
+      "epoch": 0.7432164739484959,
+      "grad_norm": 1.2189498159341265,
+      "learning_rate": 3.2637246510595275e-05,
+      "loss": 1.5552,
+      "step": 16133
+    },
+    {
+      "epoch": 0.7432625420371308,
+      "grad_norm": 1.1592093951479456,
+      "learning_rate": 3.262621966597229e-05,
+      "loss": 1.2665,
+      "step": 16134
+    },
+    {
+      "epoch": 0.7433086101257659,
+      "grad_norm": 0.6534850736068286,
+      "learning_rate": 3.2615194321310974e-05,
+      "loss": 1.3134,
+      "step": 16135
+    },
+    {
+      "epoch": 0.7433546782144009,
+      "grad_norm": 2.3332577552619886,
+      "learning_rate": 3.26041704768568e-05,
+      "loss": 1.4099,
+      "step": 16136
+    },
+    {
+      "epoch": 0.7434007463030359,
+      "grad_norm": 1.6033700437240004,
+      "learning_rate": 3.259314813285526e-05,
+      "loss": 1.3543,
+      "step": 16137
+    },
+    {
+      "epoch": 0.7434468143916709,
+      "grad_norm": 0.9759983748099617,
+      "learning_rate": 3.2582127289551655e-05,
+      "loss": 1.4472,
+      "step": 16138
+    },
+    {
+      "epoch": 0.7434928824803059,
+      "grad_norm": 1.2821810930887139,
+      "learning_rate": 3.257110794719139e-05,
+      "loss": 1.2637,
+      "step": 16139
+    },
+    {
+      "epoch": 0.7435389505689409,
+      "grad_norm": 0.9067455500766576,
+      "learning_rate": 3.25600901060198e-05,
+      "loss": 0.992,
+      "step": 16140
+    },
+    {
+      "epoch": 0.7435850186575759,
+      "grad_norm": 0.9009364765949813,
+      "learning_rate": 3.254907376628219e-05,
+      "loss": 0.9284,
+      "step": 16141
+    },
+    {
+      "epoch": 0.7436310867462109,
+      "grad_norm": 0.7304263671820417,
+      "learning_rate": 3.253805892822378e-05,
+      "loss": 1.3057,
+      "step": 16142
+    },
+    {
+      "epoch": 0.7436771548348459,
+      "grad_norm": 0.7393033202421977,
+      "learning_rate": 3.252704559208981e-05,
+      "loss": 1.3541,
+      "step": 16143
+    },
+    {
+      "epoch": 0.7437232229234809,
+      "grad_norm": 0.885739257303583,
+      "learning_rate": 3.2516033758125505e-05,
+      "loss": 1.2358,
+      "step": 16144
+    },
+    {
+      "epoch": 0.7437692910121159,
+      "grad_norm": 1.0348576284260367,
+      "learning_rate": 3.2505023426575974e-05,
+      "loss": 1.2363,
+      "step": 16145
+    },
+    {
+      "epoch": 0.7438153591007509,
+      "grad_norm": 0.782309358420985,
+      "learning_rate": 3.249401459768636e-05,
+      "loss": 1.1719,
+      "step": 16146
+    },
+    {
+      "epoch": 0.7438614271893859,
+      "grad_norm": 0.9225278618214684,
+      "learning_rate": 3.248300727170177e-05,
+      "loss": 1.4024,
+      "step": 16147
+    },
+    {
+      "epoch": 0.7439074952780209,
+      "grad_norm": 0.8499690329716656,
+      "learning_rate": 3.247200144886728e-05,
+      "loss": 1.02,
+      "step": 16148
+    },
+    {
+      "epoch": 0.743953563366656,
+      "grad_norm": 0.694407898888324,
+      "learning_rate": 3.246099712942786e-05,
+      "loss": 1.0593,
+      "step": 16149
+    },
+    {
+      "epoch": 0.7439996314552909,
+      "grad_norm": 0.6007269959303407,
+      "learning_rate": 3.244999431362855e-05,
+      "loss": 1.5557,
+      "step": 16150
+    },
+    {
+      "epoch": 0.744045699543926,
+      "grad_norm": 1.0033962018354123,
+      "learning_rate": 3.243899300171428e-05,
+      "loss": 1.3327,
+      "step": 16151
+    },
+    {
+      "epoch": 0.7440917676325609,
+      "grad_norm": 0.7610280525781387,
+      "learning_rate": 3.242799319393002e-05,
+      "loss": 1.2157,
+      "step": 16152
+    },
+    {
+      "epoch": 0.744137835721196,
+      "grad_norm": 0.8905869491247026,
+      "learning_rate": 3.24169948905206e-05,
+      "loss": 1.3825,
+      "step": 16153
+    },
+    {
+      "epoch": 0.7441839038098309,
+      "grad_norm": 1.2109325836898897,
+      "learning_rate": 3.240599809173091e-05,
+      "loss": 1.3745,
+      "step": 16154
+    },
+    {
+      "epoch": 0.744229971898466,
+      "grad_norm": 0.7112836759398486,
+      "learning_rate": 3.239500279780581e-05,
+      "loss": 1.4202,
+      "step": 16155
+    },
+    {
+      "epoch": 0.7442760399871009,
+      "grad_norm": 1.393090804314161,
+      "learning_rate": 3.238400900899001e-05,
+      "loss": 1.1343,
+      "step": 16156
+    },
+    {
+      "epoch": 0.744322108075736,
+      "grad_norm": 0.9646569851705209,
+      "learning_rate": 3.237301672552833e-05,
+      "loss": 1.0766,
+      "step": 16157
+    },
+    {
+      "epoch": 0.7443681761643709,
+      "grad_norm": 1.5055771222142387,
+      "learning_rate": 3.2362025947665464e-05,
+      "loss": 1.0071,
+      "step": 16158
+    },
+    {
+      "epoch": 0.744414244253006,
+      "grad_norm": 0.8037178630461866,
+      "learning_rate": 3.2351036675646165e-05,
+      "loss": 1.362,
+      "step": 16159
+    },
+    {
+      "epoch": 0.7444603123416409,
+      "grad_norm": 0.7317848129625678,
+      "learning_rate": 3.2340048909714995e-05,
+      "loss": 1.5031,
+      "step": 16160
+    },
+    {
+      "epoch": 0.744506380430276,
+      "grad_norm": 1.2201771470517784,
+      "learning_rate": 3.232906265011662e-05,
+      "loss": 1.4205,
+      "step": 16161
+    },
+    {
+      "epoch": 0.744552448518911,
+      "grad_norm": 0.8935386758203472,
+      "learning_rate": 3.231807789709568e-05,
+      "loss": 1.1705,
+      "step": 16162
+    },
+    {
+      "epoch": 0.744598516607546,
+      "grad_norm": 1.0089515854895814,
+      "learning_rate": 3.230709465089664e-05,
+      "loss": 1.1891,
+      "step": 16163
+    },
+    {
+      "epoch": 0.744644584696181,
+      "grad_norm": 0.6725644860502175,
+      "learning_rate": 3.229611291176411e-05,
+      "loss": 1.2965,
+      "step": 16164
+    },
+    {
+      "epoch": 0.7446906527848159,
+      "grad_norm": 0.6523499800078566,
+      "learning_rate": 3.2285132679942466e-05,
+      "loss": 1.3678,
+      "step": 16165
+    },
+    {
+      "epoch": 0.744736720873451,
+      "grad_norm": 0.9429399324087334,
+      "learning_rate": 3.227415395567631e-05,
+      "loss": 1.109,
+      "step": 16166
+    },
+    {
+      "epoch": 0.7447827889620859,
+      "grad_norm": 0.6432807205175812,
+      "learning_rate": 3.226317673920996e-05,
+      "loss": 1.3548,
+      "step": 16167
+    },
+    {
+      "epoch": 0.744828857050721,
+      "grad_norm": 1.1086147250725151,
+      "learning_rate": 3.225220103078783e-05,
+      "loss": 0.9083,
+      "step": 16168
+    },
+    {
+      "epoch": 0.7448749251393559,
+      "grad_norm": 0.8870144969847341,
+      "learning_rate": 3.224122683065428e-05,
+      "loss": 1.5268,
+      "step": 16169
+    },
+    {
+      "epoch": 0.744920993227991,
+      "grad_norm": 0.9396927671099695,
+      "learning_rate": 3.2230254139053675e-05,
+      "loss": 1.267,
+      "step": 16170
+    },
+    {
+      "epoch": 0.7449670613166259,
+      "grad_norm": 0.7779696535936718,
+      "learning_rate": 3.221928295623022e-05,
+      "loss": 1.3043,
+      "step": 16171
+    },
+    {
+      "epoch": 0.745013129405261,
+      "grad_norm": 0.655193353343856,
+      "learning_rate": 3.220831328242822e-05,
+      "loss": 1.1484,
+      "step": 16172
+    },
+    {
+      "epoch": 0.7450591974938959,
+      "grad_norm": 0.852194952702046,
+      "learning_rate": 3.219734511789191e-05,
+      "loss": 1.1193,
+      "step": 16173
+    },
+    {
+      "epoch": 0.745105265582531,
+      "grad_norm": 1.4503091503238887,
+      "learning_rate": 3.218637846286543e-05,
+      "loss": 1.0076,
+      "step": 16174
+    },
+    {
+      "epoch": 0.745151333671166,
+      "grad_norm": 0.682085705797853,
+      "learning_rate": 3.2175413317592983e-05,
+      "loss": 1.2431,
+      "step": 16175
+    },
+    {
+      "epoch": 0.745197401759801,
+      "grad_norm": 0.6758415769097321,
+      "learning_rate": 3.216444968231861e-05,
+      "loss": 1.0432,
+      "step": 16176
+    },
+    {
+      "epoch": 0.745243469848436,
+      "grad_norm": 0.9767085153933238,
+      "learning_rate": 3.21534875572865e-05,
+      "loss": 1.0915,
+      "step": 16177
+    },
+    {
+      "epoch": 0.745289537937071,
+      "grad_norm": 0.8966206064799547,
+      "learning_rate": 3.2142526942740634e-05,
+      "loss": 1.2151,
+      "step": 16178
+    },
+    {
+      "epoch": 0.745335606025706,
+      "grad_norm": 0.9778150719002221,
+      "learning_rate": 3.213156783892509e-05,
+      "loss": 1.3673,
+      "step": 16179
+    },
+    {
+      "epoch": 0.745381674114341,
+      "grad_norm": 0.7930580767613491,
+      "learning_rate": 3.212061024608376e-05,
+      "loss": 1.1878,
+      "step": 16180
+    },
+    {
+      "epoch": 0.745427742202976,
+      "grad_norm": 1.9174702972207835,
+      "learning_rate": 3.21096541644607e-05,
+      "loss": 1.0517,
+      "step": 16181
+    },
+    {
+      "epoch": 0.745473810291611,
+      "grad_norm": 0.8609972690790891,
+      "learning_rate": 3.20986995942998e-05,
+      "loss": 1.4832,
+      "step": 16182
+    },
+    {
+      "epoch": 0.745519878380246,
+      "grad_norm": 0.6859234509790646,
+      "learning_rate": 3.2087746535844854e-05,
+      "loss": 1.5431,
+      "step": 16183
+    },
+    {
+      "epoch": 0.745565946468881,
+      "grad_norm": 1.376918704284689,
+      "learning_rate": 3.2076794989339864e-05,
+      "loss": 1.4908,
+      "step": 16184
+    },
+    {
+      "epoch": 0.745612014557516,
+      "grad_norm": 1.2265578335002891,
+      "learning_rate": 3.2065844955028524e-05,
+      "loss": 1.1643,
+      "step": 16185
+    },
+    {
+      "epoch": 0.745658082646151,
+      "grad_norm": 0.8399736340424321,
+      "learning_rate": 3.2054896433154705e-05,
+      "loss": 1.0716,
+      "step": 16186
+    },
+    {
+      "epoch": 0.745704150734786,
+      "grad_norm": 2.8914162557344096,
+      "learning_rate": 3.204394942396205e-05,
+      "loss": 1.2245,
+      "step": 16187
+    },
+    {
+      "epoch": 0.745750218823421,
+      "grad_norm": 1.0470059108698606,
+      "learning_rate": 3.2033003927694395e-05,
+      "loss": 1.0288,
+      "step": 16188
+    },
+    {
+      "epoch": 0.745796286912056,
+      "grad_norm": 0.874666016388645,
+      "learning_rate": 3.202205994459535e-05,
+      "loss": 1.2653,
+      "step": 16189
+    },
+    {
+      "epoch": 0.7458423550006911,
+      "grad_norm": 0.9535379969290507,
+      "learning_rate": 3.2011117474908614e-05,
+      "loss": 1.2773,
+      "step": 16190
+    },
+    {
+      "epoch": 0.745888423089326,
+      "grad_norm": 1.5481795096120088,
+      "learning_rate": 3.200017651887773e-05,
+      "loss": 1.283,
+      "step": 16191
+    },
+    {
+      "epoch": 0.7459344911779611,
+      "grad_norm": 2.1996563175719697,
+      "learning_rate": 3.1989237076746325e-05,
+      "loss": 1.2884,
+      "step": 16192
+    },
+    {
+      "epoch": 0.745980559266596,
+      "grad_norm": 1.6935134394754476,
+      "learning_rate": 3.197829914875798e-05,
+      "loss": 1.4063,
+      "step": 16193
+    },
+    {
+      "epoch": 0.7460266273552311,
+      "grad_norm": 3.262770834609742,
+      "learning_rate": 3.19673627351561e-05,
+      "loss": 1.145,
+      "step": 16194
+    },
+    {
+      "epoch": 0.746072695443866,
+      "grad_norm": 0.9222217327993899,
+      "learning_rate": 3.195642783618431e-05,
+      "loss": 1.1471,
+      "step": 16195
+    },
+    {
+      "epoch": 0.7461187635325011,
+      "grad_norm": 0.7208046774219282,
+      "learning_rate": 3.194549445208596e-05,
+      "loss": 1.1567,
+      "step": 16196
+    },
+    {
+      "epoch": 0.746164831621136,
+      "grad_norm": 0.8619403513939679,
+      "learning_rate": 3.193456258310451e-05,
+      "loss": 1.1346,
+      "step": 16197
+    },
+    {
+      "epoch": 0.7462108997097711,
+      "grad_norm": 2.4584647367778345,
+      "learning_rate": 3.192363222948325e-05,
+      "loss": 1.1993,
+      "step": 16198
+    },
+    {
+      "epoch": 0.746256967798406,
+      "grad_norm": 1.3719185506880789,
+      "learning_rate": 3.191270339146566e-05,
+      "loss": 1.0468,
+      "step": 16199
+    },
+    {
+      "epoch": 0.746303035887041,
+      "grad_norm": 1.4376470827059442,
+      "learning_rate": 3.1901776069294974e-05,
+      "loss": 1.1529,
+      "step": 16200
+    },
+    {
+      "epoch": 0.746349103975676,
+      "grad_norm": 0.8160424313362938,
+      "learning_rate": 3.1890850263214465e-05,
+      "loss": 1.0918,
+      "step": 16201
+    },
+    {
+      "epoch": 0.746395172064311,
+      "grad_norm": 0.7082608827538894,
+      "learning_rate": 3.187992597346738e-05,
+      "loss": 1.2133,
+      "step": 16202
+    },
+    {
+      "epoch": 0.7464412401529461,
+      "grad_norm": 2.0941798671711878,
+      "learning_rate": 3.186900320029696e-05,
+      "loss": 1.1351,
+      "step": 16203
+    },
+    {
+      "epoch": 0.746487308241581,
+      "grad_norm": 1.5973586511411326,
+      "learning_rate": 3.185808194394637e-05,
+      "loss": 1.1418,
+      "step": 16204
+    },
+    {
+      "epoch": 0.7465333763302161,
+      "grad_norm": 1.734664195209028,
+      "learning_rate": 3.184716220465873e-05,
+      "loss": 1.1255,
+      "step": 16205
+    },
+    {
+      "epoch": 0.746579444418851,
+      "grad_norm": 1.95633109252842,
+      "learning_rate": 3.183624398267717e-05,
+      "loss": 1.3471,
+      "step": 16206
+    },
+    {
+      "epoch": 0.7466255125074861,
+      "grad_norm": 3.056011858072079,
+      "learning_rate": 3.1825327278244755e-05,
+      "loss": 1.3147,
+      "step": 16207
+    },
+    {
+      "epoch": 0.746671580596121,
+      "grad_norm": 2.9876314416720313,
+      "learning_rate": 3.181441209160457e-05,
+      "loss": 1.2474,
+      "step": 16208
+    },
+    {
+      "epoch": 0.7467176486847561,
+      "grad_norm": 0.9546180376685673,
+      "learning_rate": 3.1803498422999556e-05,
+      "loss": 1.1903,
+      "step": 16209
+    },
+    {
+      "epoch": 0.746763716773391,
+      "grad_norm": 1.0740360531565643,
+      "learning_rate": 3.1792586272672706e-05,
+      "loss": 1.0551,
+      "step": 16210
+    },
+    {
+      "epoch": 0.7468097848620261,
+      "grad_norm": 1.1190424903306597,
+      "learning_rate": 3.178167564086702e-05,
+      "loss": 1.2609,
+      "step": 16211
+    },
+    {
+      "epoch": 0.746855852950661,
+      "grad_norm": 1.9214013123155451,
+      "learning_rate": 3.177076652782531e-05,
+      "loss": 1.2474,
+      "step": 16212
+    },
+    {
+      "epoch": 0.7469019210392961,
+      "grad_norm": 0.8811388672147898,
+      "learning_rate": 3.175985893379051e-05,
+      "loss": 1.3437,
+      "step": 16213
+    },
+    {
+      "epoch": 0.746947989127931,
+      "grad_norm": 0.9535817752745066,
+      "learning_rate": 3.174895285900543e-05,
+      "loss": 1.1662,
+      "step": 16214
+    },
+    {
+      "epoch": 0.7469940572165661,
+      "grad_norm": 0.9828856987289475,
+      "learning_rate": 3.173804830371293e-05,
+      "loss": 1.3358,
+      "step": 16215
+    },
+    {
+      "epoch": 0.7470401253052011,
+      "grad_norm": 0.6820551459807872,
+      "learning_rate": 3.172714526815571e-05,
+      "loss": 1.3728,
+      "step": 16216
+    },
+    {
+      "epoch": 0.7470861933938361,
+      "grad_norm": 0.9107672884991507,
+      "learning_rate": 3.1716243752576526e-05,
+      "loss": 1.3633,
+      "step": 16217
+    },
+    {
+      "epoch": 0.7471322614824711,
+      "grad_norm": 0.9199561361311437,
+      "learning_rate": 3.1705343757218134e-05,
+      "loss": 1.0509,
+      "step": 16218
+    },
+    {
+      "epoch": 0.7471783295711061,
+      "grad_norm": 0.8603639602437183,
+      "learning_rate": 3.169444528232314e-05,
+      "loss": 1.194,
+      "step": 16219
+    },
+    {
+      "epoch": 0.7472243976597411,
+      "grad_norm": 1.9374752727964122,
+      "learning_rate": 3.168354832813418e-05,
+      "loss": 1.4235,
+      "step": 16220
+    },
+    {
+      "epoch": 0.7472704657483761,
+      "grad_norm": 0.7438759693944555,
+      "learning_rate": 3.1672652894893895e-05,
+      "loss": 1.1898,
+      "step": 16221
+    },
+    {
+      "epoch": 0.7473165338370111,
+      "grad_norm": 0.630084972555974,
+      "learning_rate": 3.166175898284487e-05,
+      "loss": 1.4304,
+      "step": 16222
+    },
+    {
+      "epoch": 0.7473626019256461,
+      "grad_norm": 0.8747704694891021,
+      "learning_rate": 3.1650866592229566e-05,
+      "loss": 0.9974,
+      "step": 16223
+    },
+    {
+      "epoch": 0.7474086700142811,
+      "grad_norm": 0.8082609359796478,
+      "learning_rate": 3.163997572329053e-05,
+      "loss": 1.2794,
+      "step": 16224
+    },
+    {
+      "epoch": 0.7474547381029161,
+      "grad_norm": 1.00760258791475,
+      "learning_rate": 3.162908637627022e-05,
+      "loss": 1.4023,
+      "step": 16225
+    },
+    {
+      "epoch": 0.7475008061915511,
+      "grad_norm": 1.4358963660731285,
+      "learning_rate": 3.1618198551411106e-05,
+      "loss": 1.1534,
+      "step": 16226
+    },
+    {
+      "epoch": 0.7475468742801861,
+      "grad_norm": 1.197114036649519,
+      "learning_rate": 3.160731224895551e-05,
+      "loss": 1.2216,
+      "step": 16227
+    },
+    {
+      "epoch": 0.7475929423688211,
+      "grad_norm": 1.1076169601964103,
+      "learning_rate": 3.159642746914584e-05,
+      "loss": 1.1958,
+      "step": 16228
+    },
+    {
+      "epoch": 0.7476390104574561,
+      "grad_norm": 0.9901304633556968,
+      "learning_rate": 3.158554421222447e-05,
+      "loss": 1.2354,
+      "step": 16229
+    },
+    {
+      "epoch": 0.7476850785460911,
+      "grad_norm": 0.7607702766335069,
+      "learning_rate": 3.1574662478433615e-05,
+      "loss": 1.1575,
+      "step": 16230
+    },
+    {
+      "epoch": 0.7477311466347262,
+      "grad_norm": 0.8486453042371227,
+      "learning_rate": 3.1563782268015565e-05,
+      "loss": 1.531,
+      "step": 16231
+    },
+    {
+      "epoch": 0.7477772147233611,
+      "grad_norm": 0.7098806943176829,
+      "learning_rate": 3.155290358121257e-05,
+      "loss": 1.4546,
+      "step": 16232
+    },
+    {
+      "epoch": 0.7478232828119962,
+      "grad_norm": 0.8807502616205831,
+      "learning_rate": 3.154202641826684e-05,
+      "loss": 1.1415,
+      "step": 16233
+    },
+    {
+      "epoch": 0.7478693509006311,
+      "grad_norm": 1.1644806551894282,
+      "learning_rate": 3.153115077942049e-05,
+      "loss": 1.386,
+      "step": 16234
+    },
+    {
+      "epoch": 0.7479154189892662,
+      "grad_norm": 1.2151968358031273,
+      "learning_rate": 3.152027666491566e-05,
+      "loss": 0.9877,
+      "step": 16235
+    },
+    {
+      "epoch": 0.7479614870779011,
+      "grad_norm": 0.7835224869558106,
+      "learning_rate": 3.15094040749945e-05,
+      "loss": 1.3854,
+      "step": 16236
+    },
+    {
+      "epoch": 0.7480075551665362,
+      "grad_norm": 0.7565485980741368,
+      "learning_rate": 3.149853300989897e-05,
+      "loss": 1.3334,
+      "step": 16237
+    },
+    {
+      "epoch": 0.7480536232551711,
+      "grad_norm": 0.8680284351760855,
+      "learning_rate": 3.148766346987116e-05,
+      "loss": 1.0324,
+      "step": 16238
+    },
+    {
+      "epoch": 0.7480996913438062,
+      "grad_norm": 1.043690399482744,
+      "learning_rate": 3.147679545515304e-05,
+      "loss": 1.0448,
+      "step": 16239
+    },
+    {
+      "epoch": 0.7481457594324411,
+      "grad_norm": 0.8440239357480591,
+      "learning_rate": 3.146592896598661e-05,
+      "loss": 1.1688,
+      "step": 16240
+    },
+    {
+      "epoch": 0.7481918275210762,
+      "grad_norm": 1.4088007781539456,
+      "learning_rate": 3.1455064002613745e-05,
+      "loss": 1.3965,
+      "step": 16241
+    },
+    {
+      "epoch": 0.7482378956097111,
+      "grad_norm": 0.8058779064943329,
+      "learning_rate": 3.144420056527634e-05,
+      "loss": 1.2769,
+      "step": 16242
+    },
+    {
+      "epoch": 0.7482839636983462,
+      "grad_norm": 1.5041766103241314,
+      "learning_rate": 3.143333865421625e-05,
+      "loss": 1.3316,
+      "step": 16243
+    },
+    {
+      "epoch": 0.7483300317869812,
+      "grad_norm": 0.9729172514266614,
+      "learning_rate": 3.142247826967536e-05,
+      "loss": 1.0293,
+      "step": 16244
+    },
+    {
+      "epoch": 0.7483760998756162,
+      "grad_norm": 0.9822723838901964,
+      "learning_rate": 3.141161941189535e-05,
+      "loss": 1.413,
+      "step": 16245
+    },
+    {
+      "epoch": 0.7484221679642512,
+      "grad_norm": 0.9197862704608024,
+      "learning_rate": 3.140076208111805e-05,
+      "loss": 1.2582,
+      "step": 16246
+    },
+    {
+      "epoch": 0.7484682360528861,
+      "grad_norm": 1.4897789538049648,
+      "learning_rate": 3.138990627758519e-05,
+      "loss": 1.2405,
+      "step": 16247
+    },
+    {
+      "epoch": 0.7485143041415212,
+      "grad_norm": 0.9717845080556736,
+      "learning_rate": 3.137905200153839e-05,
+      "loss": 1.1219,
+      "step": 16248
+    },
+    {
+      "epoch": 0.7485603722301561,
+      "grad_norm": 0.995619012168504,
+      "learning_rate": 3.136819925321934e-05,
+      "loss": 1.4569,
+      "step": 16249
+    },
+    {
+      "epoch": 0.7486064403187912,
+      "grad_norm": 5.799214713546559,
+      "learning_rate": 3.135734803286965e-05,
+      "loss": 1.2988,
+      "step": 16250
+    },
+    {
+      "epoch": 0.7486525084074261,
+      "grad_norm": 0.9275430052354113,
+      "learning_rate": 3.1346498340730946e-05,
+      "loss": 1.133,
+      "step": 16251
+    },
+    {
+      "epoch": 0.7486985764960612,
+      "grad_norm": 0.9525109474241468,
+      "learning_rate": 3.133565017704471e-05,
+      "loss": 1.3562,
+      "step": 16252
+    },
+    {
+      "epoch": 0.7487446445846961,
+      "grad_norm": 1.5232058812102045,
+      "learning_rate": 3.1324803542052484e-05,
+      "loss": 1.0596,
+      "step": 16253
+    },
+    {
+      "epoch": 0.7487907126733312,
+      "grad_norm": 0.7916031983206657,
+      "learning_rate": 3.131395843599578e-05,
+      "loss": 0.9097,
+      "step": 16254
+    },
+    {
+      "epoch": 0.7488367807619661,
+      "grad_norm": 1.185020979971286,
+      "learning_rate": 3.1303114859116e-05,
+      "loss": 1.2135,
+      "step": 16255
+    },
+    {
+      "epoch": 0.7488828488506012,
+      "grad_norm": 1.0376138815815017,
+      "learning_rate": 3.129227281165458e-05,
+      "loss": 1.2231,
+      "step": 16256
+    },
+    {
+      "epoch": 0.7489289169392362,
+      "grad_norm": 0.7533286411170882,
+      "learning_rate": 3.128143229385288e-05,
+      "loss": 0.8972,
+      "step": 16257
+    },
+    {
+      "epoch": 0.7489749850278712,
+      "grad_norm": 1.02526370982314,
+      "learning_rate": 3.127059330595231e-05,
+      "loss": 1.0147,
+      "step": 16258
+    },
+    {
+      "epoch": 0.7490210531165062,
+      "grad_norm": 0.8527660764550059,
+      "learning_rate": 3.1259755848194106e-05,
+      "loss": 1.3292,
+      "step": 16259
+    },
+    {
+      "epoch": 0.7490671212051412,
+      "grad_norm": 0.7971007212404324,
+      "learning_rate": 3.124891992081956e-05,
+      "loss": 1.0811,
+      "step": 16260
+    },
+    {
+      "epoch": 0.7491131892937762,
+      "grad_norm": 1.5386231492296745,
+      "learning_rate": 3.123808552406994e-05,
+      "loss": 1.2695,
+      "step": 16261
+    },
+    {
+      "epoch": 0.7491592573824112,
+      "grad_norm": 2.985490073491664,
+      "learning_rate": 3.122725265818647e-05,
+      "loss": 1.0937,
+      "step": 16262
+    },
+    {
+      "epoch": 0.7492053254710462,
+      "grad_norm": 0.6639084209652352,
+      "learning_rate": 3.1216421323410304e-05,
+      "loss": 1.2365,
+      "step": 16263
+    },
+    {
+      "epoch": 0.7492513935596812,
+      "grad_norm": 1.0828445196906813,
+      "learning_rate": 3.1205591519982515e-05,
+      "loss": 1.0912,
+      "step": 16264
+    },
+    {
+      "epoch": 0.7492974616483162,
+      "grad_norm": 0.8320049715632,
+      "learning_rate": 3.119476324814433e-05,
+      "loss": 1.1156,
+      "step": 16265
+    },
+    {
+      "epoch": 0.7493435297369512,
+      "grad_norm": 0.8352405015280618,
+      "learning_rate": 3.1183936508136735e-05,
+      "loss": 1.2229,
+      "step": 16266
+    },
+    {
+      "epoch": 0.7493895978255862,
+      "grad_norm": 0.6893510184754377,
+      "learning_rate": 3.117311130020081e-05,
+      "loss": 1.2663,
+      "step": 16267
+    },
+    {
+      "epoch": 0.7494356659142212,
+      "grad_norm": 1.3980080664698515,
+      "learning_rate": 3.1162287624577534e-05,
+      "loss": 0.8788,
+      "step": 16268
+    },
+    {
+      "epoch": 0.7494817340028562,
+      "grad_norm": 1.1371563979196104,
+      "learning_rate": 3.115146548150793e-05,
+      "loss": 1.2813,
+      "step": 16269
+    },
+    {
+      "epoch": 0.7495278020914912,
+      "grad_norm": 1.0279916460509264,
+      "learning_rate": 3.114064487123287e-05,
+      "loss": 1.3383,
+      "step": 16270
+    },
+    {
+      "epoch": 0.7495738701801262,
+      "grad_norm": 1.093188608854988,
+      "learning_rate": 3.1129825793993285e-05,
+      "loss": 1.131,
+      "step": 16271
+    },
+    {
+      "epoch": 0.7496199382687613,
+      "grad_norm": 0.8342795807172093,
+      "learning_rate": 3.1119008250030065e-05,
+      "loss": 0.929,
+      "step": 16272
+    },
+    {
+      "epoch": 0.7496660063573962,
+      "grad_norm": 22.799347920060296,
+      "learning_rate": 3.1108192239584e-05,
+      "loss": 1.3837,
+      "step": 16273
+    },
+    {
+      "epoch": 0.7497120744460313,
+      "grad_norm": 0.7497157507671371,
+      "learning_rate": 3.1097377762895926e-05,
+      "loss": 1.3617,
+      "step": 16274
+    },
+    {
+      "epoch": 0.7497581425346662,
+      "grad_norm": 1.8764675340143215,
+      "learning_rate": 3.108656482020654e-05,
+      "loss": 1.2074,
+      "step": 16275
+    },
+    {
+      "epoch": 0.7498042106233013,
+      "grad_norm": 1.0069094597755062,
+      "learning_rate": 3.1075753411756705e-05,
+      "loss": 1.4646,
+      "step": 16276
+    },
+    {
+      "epoch": 0.7498502787119362,
+      "grad_norm": 2.5466841199705414,
+      "learning_rate": 3.1064943537786984e-05,
+      "loss": 1.5761,
+      "step": 16277
+    },
+    {
+      "epoch": 0.7498963468005713,
+      "grad_norm": 1.053387464698536,
+      "learning_rate": 3.105413519853815e-05,
+      "loss": 1.3083,
+      "step": 16278
+    },
+    {
+      "epoch": 0.7499424148892062,
+      "grad_norm": 1.165902439652623,
+      "learning_rate": 3.104332839425069e-05,
+      "loss": 1.3932,
+      "step": 16279
+    },
+    {
+      "epoch": 0.7499884829778413,
+      "grad_norm": 10.927557247846618,
+      "learning_rate": 3.103252312516537e-05,
+      "loss": 1.2853,
+      "step": 16280
+    },
+    {
+      "epoch": 0.7500345510664762,
+      "grad_norm": 5.69334481864087,
+      "learning_rate": 3.102171939152268e-05,
+      "loss": 1.2389,
+      "step": 16281
+    },
+    {
+      "epoch": 0.7500806191551113,
+      "grad_norm": 2.149433840041296,
+      "learning_rate": 3.101091719356305e-05,
+      "loss": 1.271,
+      "step": 16282
+    },
+    {
+      "epoch": 0.7501266872437462,
+      "grad_norm": 0.7057586046308417,
+      "learning_rate": 3.1000116531527125e-05,
+      "loss": 1.5139,
+      "step": 16283
+    },
+    {
+      "epoch": 0.7501727553323813,
+      "grad_norm": 2.6252430427128584,
+      "learning_rate": 3.098931740565526e-05,
+      "loss": 1.1371,
+      "step": 16284
+    },
+    {
+      "epoch": 0.7502188234210163,
+      "grad_norm": 1.3743607356660466,
+      "learning_rate": 3.097851981618796e-05,
+      "loss": 1.0693,
+      "step": 16285
+    },
+    {
+      "epoch": 0.7502648915096513,
+      "grad_norm": 0.9482389676332315,
+      "learning_rate": 3.0967723763365494e-05,
+      "loss": 1.3495,
+      "step": 16286
+    },
+    {
+      "epoch": 0.7503109595982863,
+      "grad_norm": 0.7593537403505078,
+      "learning_rate": 3.095692924742837e-05,
+      "loss": 1.2576,
+      "step": 16287
+    },
+    {
+      "epoch": 0.7503570276869213,
+      "grad_norm": 1.6941744147575202,
+      "learning_rate": 3.094613626861679e-05,
+      "loss": 1.2832,
+      "step": 16288
+    },
+    {
+      "epoch": 0.7504030957755563,
+      "grad_norm": 2.779202479829126,
+      "learning_rate": 3.093534482717112e-05,
+      "loss": 1.1819,
+      "step": 16289
+    },
+    {
+      "epoch": 0.7504491638641912,
+      "grad_norm": 0.8414239778236682,
+      "learning_rate": 3.092455492333155e-05,
+      "loss": 1.31,
+      "step": 16290
+    },
+    {
+      "epoch": 0.7504952319528263,
+      "grad_norm": 1.4871943377697001,
+      "learning_rate": 3.091376655733832e-05,
+      "loss": 1.1832,
+      "step": 16291
+    },
+    {
+      "epoch": 0.7505413000414612,
+      "grad_norm": 1.2781664088194935,
+      "learning_rate": 3.0902979729431646e-05,
+      "loss": 1.0983,
+      "step": 16292
+    },
+    {
+      "epoch": 0.7505873681300963,
+      "grad_norm": 2.026433409957018,
+      "learning_rate": 3.08921944398516e-05,
+      "loss": 1.1344,
+      "step": 16293
+    },
+    {
+      "epoch": 0.7506334362187312,
+      "grad_norm": 1.7260290110688536,
+      "learning_rate": 3.088141068883841e-05,
+      "loss": 1.1639,
+      "step": 16294
+    },
+    {
+      "epoch": 0.7506795043073663,
+      "grad_norm": 2.497890285772993,
+      "learning_rate": 3.0870628476632065e-05,
+      "loss": 0.9863,
+      "step": 16295
+    },
+    {
+      "epoch": 0.7507255723960012,
+      "grad_norm": 0.86561924846534,
+      "learning_rate": 3.085984780347267e-05,
+      "loss": 1.1219,
+      "step": 16296
+    },
+    {
+      "epoch": 0.7507716404846363,
+      "grad_norm": 0.9044402099825047,
+      "learning_rate": 3.0849068669600167e-05,
+      "loss": 1.0735,
+      "step": 16297
+    },
+    {
+      "epoch": 0.7508177085732713,
+      "grad_norm": 1.2838097204241175,
+      "learning_rate": 3.083829107525463e-05,
+      "loss": 0.9921,
+      "step": 16298
+    },
+    {
+      "epoch": 0.7508637766619063,
+      "grad_norm": 1.3240185820741037,
+      "learning_rate": 3.0827515020675936e-05,
+      "loss": 1.2126,
+      "step": 16299
+    },
+    {
+      "epoch": 0.7509098447505413,
+      "grad_norm": 0.873870827384251,
+      "learning_rate": 3.081674050610404e-05,
+      "loss": 1.3058,
+      "step": 16300
+    },
+    {
+      "epoch": 0.7509559128391763,
+      "grad_norm": 0.8079657829320557,
+      "learning_rate": 3.0805967531778756e-05,
+      "loss": 1.2845,
+      "step": 16301
+    },
+    {
+      "epoch": 0.7510019809278113,
+      "grad_norm": 0.8473298643203743,
+      "learning_rate": 3.079519609793997e-05,
+      "loss": 1.027,
+      "step": 16302
+    },
+    {
+      "epoch": 0.7510480490164463,
+      "grad_norm": 4.306429888266106,
+      "learning_rate": 3.078442620482751e-05,
+      "loss": 1.058,
+      "step": 16303
+    },
+    {
+      "epoch": 0.7510941171050813,
+      "grad_norm": 0.8237910496642167,
+      "learning_rate": 3.0773657852681094e-05,
+      "loss": 1.1134,
+      "step": 16304
+    },
+    {
+      "epoch": 0.7511401851937163,
+      "grad_norm": 0.7222971064519721,
+      "learning_rate": 3.076289104174048e-05,
+      "loss": 1.372,
+      "step": 16305
+    },
+    {
+      "epoch": 0.7511862532823513,
+      "grad_norm": 0.9113662320243575,
+      "learning_rate": 3.075212577224539e-05,
+      "loss": 1.3505,
+      "step": 16306
+    },
+    {
+      "epoch": 0.7512323213709863,
+      "grad_norm": 1.5544550150283205,
+      "learning_rate": 3.074136204443552e-05,
+      "loss": 1.4793,
+      "step": 16307
+    },
+    {
+      "epoch": 0.7512783894596213,
+      "grad_norm": 1.0616360096369164,
+      "learning_rate": 3.0730599858550404e-05,
+      "loss": 1.3218,
+      "step": 16308
+    },
+    {
+      "epoch": 0.7513244575482563,
+      "grad_norm": 0.7411085192761792,
+      "learning_rate": 3.071983921482976e-05,
+      "loss": 1.463,
+      "step": 16309
+    },
+    {
+      "epoch": 0.7513705256368913,
+      "grad_norm": 1.1425973125915048,
+      "learning_rate": 3.070908011351312e-05,
+      "loss": 1.0613,
+      "step": 16310
+    },
+    {
+      "epoch": 0.7514165937255264,
+      "grad_norm": 0.865094739169675,
+      "learning_rate": 3.0698322554839964e-05,
+      "loss": 1.2682,
+      "step": 16311
+    },
+    {
+      "epoch": 0.7514626618141613,
+      "grad_norm": 1.3361001149981875,
+      "learning_rate": 3.068756653904984e-05,
+      "loss": 1.0847,
+      "step": 16312
+    },
+    {
+      "epoch": 0.7515087299027964,
+      "grad_norm": 0.8815531847602465,
+      "learning_rate": 3.0676812066382186e-05,
+      "loss": 1.5144,
+      "step": 16313
+    },
+    {
+      "epoch": 0.7515547979914313,
+      "grad_norm": 4.44196404439163,
+      "learning_rate": 3.0666059137076496e-05,
+      "loss": 1.2186,
+      "step": 16314
+    },
+    {
+      "epoch": 0.7516008660800664,
+      "grad_norm": 1.120719844068771,
+      "learning_rate": 3.065530775137208e-05,
+      "loss": 1.2165,
+      "step": 16315
+    },
+    {
+      "epoch": 0.7516469341687013,
+      "grad_norm": 1.3005594771773175,
+      "learning_rate": 3.064455790950833e-05,
+      "loss": 1.2496,
+      "step": 16316
+    },
+    {
+      "epoch": 0.7516930022573364,
+      "grad_norm": 1.5457023153454714,
+      "learning_rate": 3.063380961172458e-05,
+      "loss": 1.5281,
+      "step": 16317
+    },
+    {
+      "epoch": 0.7517390703459713,
+      "grad_norm": 1.2914485448887876,
+      "learning_rate": 3.0623062858260157e-05,
+      "loss": 1.4388,
+      "step": 16318
+    },
+    {
+      "epoch": 0.7517851384346064,
+      "grad_norm": 1.9165484820871426,
+      "learning_rate": 3.061231764935425e-05,
+      "loss": 1.0813,
+      "step": 16319
+    },
+    {
+      "epoch": 0.7518312065232413,
+      "grad_norm": 0.6711019154443247,
+      "learning_rate": 3.060157398524611e-05,
+      "loss": 1.3331,
+      "step": 16320
+    },
+    {
+      "epoch": 0.7518772746118764,
+      "grad_norm": 0.8045256557255297,
+      "learning_rate": 3.0590831866174974e-05,
+      "loss": 1.3152,
+      "step": 16321
+    },
+    {
+      "epoch": 0.7519233427005113,
+      "grad_norm": 1.4202088112391105,
+      "learning_rate": 3.058009129237992e-05,
+      "loss": 1.2152,
+      "step": 16322
+    },
+    {
+      "epoch": 0.7519694107891464,
+      "grad_norm": 1.619793562854742,
+      "learning_rate": 3.0569352264100104e-05,
+      "loss": 1.1852,
+      "step": 16323
+    },
+    {
+      "epoch": 0.7520154788777813,
+      "grad_norm": 1.4159163678753122,
+      "learning_rate": 3.0558614781574615e-05,
+      "loss": 1.4297,
+      "step": 16324
+    },
+    {
+      "epoch": 0.7520615469664164,
+      "grad_norm": 2.0337021243109263,
+      "learning_rate": 3.0547878845042534e-05,
+      "loss": 1.0147,
+      "step": 16325
+    },
+    {
+      "epoch": 0.7521076150550514,
+      "grad_norm": 1.9606985946689464,
+      "learning_rate": 3.053714445474281e-05,
+      "loss": 1.1105,
+      "step": 16326
+    },
+    {
+      "epoch": 0.7521536831436864,
+      "grad_norm": 1.6930497997053866,
+      "learning_rate": 3.052641161091446e-05,
+      "loss": 1.1516,
+      "step": 16327
+    },
+    {
+      "epoch": 0.7521997512323214,
+      "grad_norm": 1.1330583334027937,
+      "learning_rate": 3.051568031379647e-05,
+      "loss": 1.2469,
+      "step": 16328
+    },
+    {
+      "epoch": 0.7522458193209564,
+      "grad_norm": 1.9015205263167523,
+      "learning_rate": 3.0504950563627676e-05,
+      "loss": 1.0924,
+      "step": 16329
+    },
+    {
+      "epoch": 0.7522918874095914,
+      "grad_norm": 0.8856540949381017,
+      "learning_rate": 3.0494222360647005e-05,
+      "loss": 1.107,
+      "step": 16330
+    },
+    {
+      "epoch": 0.7523379554982264,
+      "grad_norm": 0.8078461933075519,
+      "learning_rate": 3.0483495705093302e-05,
+      "loss": 1.2885,
+      "step": 16331
+    },
+    {
+      "epoch": 0.7523840235868614,
+      "grad_norm": 1.1188994958394682,
+      "learning_rate": 3.0472770597205403e-05,
+      "loss": 1.1819,
+      "step": 16332
+    },
+    {
+      "epoch": 0.7524300916754963,
+      "grad_norm": 1.1675753504407034,
+      "learning_rate": 3.046204703722202e-05,
+      "loss": 1.3291,
+      "step": 16333
+    },
+    {
+      "epoch": 0.7524761597641314,
+      "grad_norm": 0.7579624533742746,
+      "learning_rate": 3.0451325025381918e-05,
+      "loss": 1.1051,
+      "step": 16334
+    },
+    {
+      "epoch": 0.7525222278527663,
+      "grad_norm": 0.8974869495908605,
+      "learning_rate": 3.0440604561923824e-05,
+      "loss": 1.1976,
+      "step": 16335
+    },
+    {
+      "epoch": 0.7525682959414014,
+      "grad_norm": 2.610341508271391,
+      "learning_rate": 3.0429885647086433e-05,
+      "loss": 1.3118,
+      "step": 16336
+    },
+    {
+      "epoch": 0.7526143640300363,
+      "grad_norm": 0.6449188607576242,
+      "learning_rate": 3.041916828110831e-05,
+      "loss": 1.2385,
+      "step": 16337
+    },
+    {
+      "epoch": 0.7526604321186714,
+      "grad_norm": 0.7805007867401762,
+      "learning_rate": 3.040845246422811e-05,
+      "loss": 1.2334,
+      "step": 16338
+    },
+    {
+      "epoch": 0.7527065002073065,
+      "grad_norm": 1.0034635782222694,
+      "learning_rate": 3.039773819668441e-05,
+      "loss": 0.8756,
+      "step": 16339
+    },
+    {
+      "epoch": 0.7527525682959414,
+      "grad_norm": 1.1607907155167758,
+      "learning_rate": 3.0387025478715714e-05,
+      "loss": 1.3868,
+      "step": 16340
+    },
+    {
+      "epoch": 0.7527986363845764,
+      "grad_norm": 1.3146666000997884,
+      "learning_rate": 3.0376314310560518e-05,
+      "loss": 1.0491,
+      "step": 16341
+    },
+    {
+      "epoch": 0.7528447044732114,
+      "grad_norm": 2.4187067694184075,
+      "learning_rate": 3.0365604692457316e-05,
+      "loss": 1.0778,
+      "step": 16342
+    },
+    {
+      "epoch": 0.7528907725618464,
+      "grad_norm": 0.9078196835437271,
+      "learning_rate": 3.0354896624644557e-05,
+      "loss": 1.2451,
+      "step": 16343
+    },
+    {
+      "epoch": 0.7529368406504814,
+      "grad_norm": 0.68679867249156,
+      "learning_rate": 3.034419010736057e-05,
+      "loss": 1.662,
+      "step": 16344
+    },
+    {
+      "epoch": 0.7529829087391164,
+      "grad_norm": 0.6415230003464408,
+      "learning_rate": 3.0333485140843763e-05,
+      "loss": 1.3059,
+      "step": 16345
+    },
+    {
+      "epoch": 0.7530289768277514,
+      "grad_norm": 1.2105500824408049,
+      "learning_rate": 3.0322781725332495e-05,
+      "loss": 1.133,
+      "step": 16346
+    },
+    {
+      "epoch": 0.7530750449163864,
+      "grad_norm": 0.8545060382158876,
+      "learning_rate": 3.031207986106498e-05,
+      "loss": 1.4647,
+      "step": 16347
+    },
+    {
+      "epoch": 0.7531211130050214,
+      "grad_norm": 1.1119670816165839,
+      "learning_rate": 3.0301379548279518e-05,
+      "loss": 1.2588,
+      "step": 16348
+    },
+    {
+      "epoch": 0.7531671810936564,
+      "grad_norm": 0.8671365585519167,
+      "learning_rate": 3.0290680787214332e-05,
+      "loss": 1.2758,
+      "step": 16349
+    },
+    {
+      "epoch": 0.7532132491822914,
+      "grad_norm": 1.4121923273634545,
+      "learning_rate": 3.0279983578107652e-05,
+      "loss": 1.3504,
+      "step": 16350
+    },
+    {
+      "epoch": 0.7532593172709264,
+      "grad_norm": 2.0935007301479507,
+      "learning_rate": 3.0269287921197552e-05,
+      "loss": 1.2795,
+      "step": 16351
+    },
+    {
+      "epoch": 0.7533053853595615,
+      "grad_norm": 0.911992598608148,
+      "learning_rate": 3.02585938167222e-05,
+      "loss": 1.3996,
+      "step": 16352
+    },
+    {
+      "epoch": 0.7533514534481964,
+      "grad_norm": 0.9222020886033151,
+      "learning_rate": 3.0247901264919676e-05,
+      "loss": 1.3521,
+      "step": 16353
+    },
+    {
+      "epoch": 0.7533975215368315,
+      "grad_norm": 0.8597478646352071,
+      "learning_rate": 3.023721026602806e-05,
+      "loss": 1.3742,
+      "step": 16354
+    },
+    {
+      "epoch": 0.7534435896254664,
+      "grad_norm": 0.787462950252117,
+      "learning_rate": 3.0226520820285308e-05,
+      "loss": 1.3305,
+      "step": 16355
+    },
+    {
+      "epoch": 0.7534896577141015,
+      "grad_norm": 2.2359134741435787,
+      "learning_rate": 3.0215832927929432e-05,
+      "loss": 1.1011,
+      "step": 16356
+    },
+    {
+      "epoch": 0.7535357258027364,
+      "grad_norm": 1.610514515448171,
+      "learning_rate": 3.0205146589198418e-05,
+      "loss": 1.0738,
+      "step": 16357
+    },
+    {
+      "epoch": 0.7535817938913715,
+      "grad_norm": 0.8407815329528102,
+      "learning_rate": 3.019446180433011e-05,
+      "loss": 1.1729,
+      "step": 16358
+    },
+    {
+      "epoch": 0.7536278619800064,
+      "grad_norm": 0.9255995980752906,
+      "learning_rate": 3.0183778573562428e-05,
+      "loss": 1.4507,
+      "step": 16359
+    },
+    {
+      "epoch": 0.7536739300686415,
+      "grad_norm": 1.0463533804881093,
+      "learning_rate": 3.017309689713319e-05,
+      "loss": 1.4268,
+      "step": 16360
+    },
+    {
+      "epoch": 0.7537199981572764,
+      "grad_norm": 1.0125080553750998,
+      "learning_rate": 3.0162416775280266e-05,
+      "loss": 1.2623,
+      "step": 16361
+    },
+    {
+      "epoch": 0.7537660662459115,
+      "grad_norm": 0.8344749611282567,
+      "learning_rate": 3.0151738208241355e-05,
+      "loss": 1.25,
+      "step": 16362
+    },
+    {
+      "epoch": 0.7538121343345464,
+      "grad_norm": 0.7414556757247699,
+      "learning_rate": 3.0141061196254215e-05,
+      "loss": 1.3463,
+      "step": 16363
+    },
+    {
+      "epoch": 0.7538582024231815,
+      "grad_norm": 3.560612270742039,
+      "learning_rate": 3.013038573955661e-05,
+      "loss": 1.312,
+      "step": 16364
+    },
+    {
+      "epoch": 0.7539042705118164,
+      "grad_norm": 0.7864834277447826,
+      "learning_rate": 3.0119711838386132e-05,
+      "loss": 1.0913,
+      "step": 16365
+    },
+    {
+      "epoch": 0.7539503386004515,
+      "grad_norm": 0.9105902767304523,
+      "learning_rate": 3.0109039492980452e-05,
+      "loss": 1.1094,
+      "step": 16366
+    },
+    {
+      "epoch": 0.7539964066890865,
+      "grad_norm": 1.6717669496422576,
+      "learning_rate": 3.0098368703577163e-05,
+      "loss": 1.2394,
+      "step": 16367
+    },
+    {
+      "epoch": 0.7540424747777215,
+      "grad_norm": 0.7227403870334983,
+      "learning_rate": 3.0087699470413876e-05,
+      "loss": 1.2568,
+      "step": 16368
+    },
+    {
+      "epoch": 0.7540885428663565,
+      "grad_norm": 1.311572811820021,
+      "learning_rate": 3.007703179372805e-05,
+      "loss": 1.2913,
+      "step": 16369
+    },
+    {
+      "epoch": 0.7541346109549915,
+      "grad_norm": 0.7344662749261953,
+      "learning_rate": 3.0066365673757225e-05,
+      "loss": 1.383,
+      "step": 16370
+    },
+    {
+      "epoch": 0.7541806790436265,
+      "grad_norm": 0.8064473105969919,
+      "learning_rate": 3.0055701110738855e-05,
+      "loss": 1.0857,
+      "step": 16371
+    },
+    {
+      "epoch": 0.7542267471322615,
+      "grad_norm": 2.3418085475714476,
+      "learning_rate": 3.0045038104910406e-05,
+      "loss": 1.3996,
+      "step": 16372
+    },
+    {
+      "epoch": 0.7542728152208965,
+      "grad_norm": 1.2993807496240064,
+      "learning_rate": 3.0034376656509233e-05,
+      "loss": 1.2093,
+      "step": 16373
+    },
+    {
+      "epoch": 0.7543188833095315,
+      "grad_norm": 0.9808728267684977,
+      "learning_rate": 3.0023716765772637e-05,
+      "loss": 1.2314,
+      "step": 16374
+    },
+    {
+      "epoch": 0.7543649513981665,
+      "grad_norm": 0.9931718361944557,
+      "learning_rate": 3.001305843293807e-05,
+      "loss": 1.4601,
+      "step": 16375
+    },
+    {
+      "epoch": 0.7544110194868014,
+      "grad_norm": 1.423950670386735,
+      "learning_rate": 3.000240165824272e-05,
+      "loss": 0.9751,
+      "step": 16376
+    },
+    {
+      "epoch": 0.7544570875754365,
+      "grad_norm": 0.7911919131985102,
+      "learning_rate": 2.9991746441923908e-05,
+      "loss": 1.3789,
+      "step": 16377
+    },
+    {
+      "epoch": 0.7545031556640714,
+      "grad_norm": 2.5494460342755705,
+      "learning_rate": 2.9981092784218757e-05,
+      "loss": 0.9302,
+      "step": 16378
+    },
+    {
+      "epoch": 0.7545492237527065,
+      "grad_norm": 3.6561615247726205,
+      "learning_rate": 2.9970440685364586e-05,
+      "loss": 1.1257,
+      "step": 16379
+    },
+    {
+      "epoch": 0.7545952918413416,
+      "grad_norm": 3.0791990225379107,
+      "learning_rate": 2.995979014559844e-05,
+      "loss": 1.2228,
+      "step": 16380
+    },
+    {
+      "epoch": 0.7546413599299765,
+      "grad_norm": 2.2576518211410512,
+      "learning_rate": 2.994914116515748e-05,
+      "loss": 1.2914,
+      "step": 16381
+    },
+    {
+      "epoch": 0.7546874280186116,
+      "grad_norm": 0.633224151178189,
+      "learning_rate": 2.9938493744278807e-05,
+      "loss": 1.3729,
+      "step": 16382
+    },
+    {
+      "epoch": 0.7547334961072465,
+      "grad_norm": 1.2349811501326509,
+      "learning_rate": 2.992784788319941e-05,
+      "loss": 1.3437,
+      "step": 16383
+    },
+    {
+      "epoch": 0.7547795641958815,
+      "grad_norm": 1.399103179081473,
+      "learning_rate": 2.9917203582156373e-05,
+      "loss": 1.1948,
+      "step": 16384
+    },
+    {
+      "epoch": 0.7548256322845165,
+      "grad_norm": 1.5394957691795803,
+      "learning_rate": 2.9906560841386564e-05,
+      "loss": 0.9871,
+      "step": 16385
+    },
+    {
+      "epoch": 0.7548717003731515,
+      "grad_norm": 1.0219419410804342,
+      "learning_rate": 2.989591966112706e-05,
+      "loss": 1.1773,
+      "step": 16386
+    },
+    {
+      "epoch": 0.7549177684617865,
+      "grad_norm": 1.1702796757491962,
+      "learning_rate": 2.988528004161466e-05,
+      "loss": 1.511,
+      "step": 16387
+    },
+    {
+      "epoch": 0.7549638365504215,
+      "grad_norm": 1.0580396352955066,
+      "learning_rate": 2.9874641983086316e-05,
+      "loss": 1.5056,
+      "step": 16388
+    },
+    {
+      "epoch": 0.7550099046390565,
+      "grad_norm": 1.6427488764718479,
+      "learning_rate": 2.9864005485778766e-05,
+      "loss": 1.4283,
+      "step": 16389
+    },
+    {
+      "epoch": 0.7550559727276915,
+      "grad_norm": 1.059992410476865,
+      "learning_rate": 2.9853370549928937e-05,
+      "loss": 1.3265,
+      "step": 16390
+    },
+    {
+      "epoch": 0.7551020408163265,
+      "grad_norm": 2.019429141042841,
+      "learning_rate": 2.984273717577354e-05,
+      "loss": 1.3741,
+      "step": 16391
+    },
+    {
+      "epoch": 0.7551481089049615,
+      "grad_norm": 1.0043845246797178,
+      "learning_rate": 2.9832105363549247e-05,
+      "loss": 1.0259,
+      "step": 16392
+    },
+    {
+      "epoch": 0.7551941769935966,
+      "grad_norm": 0.8949786233781774,
+      "learning_rate": 2.9821475113492868e-05,
+      "loss": 1.257,
+      "step": 16393
+    },
+    {
+      "epoch": 0.7552402450822315,
+      "grad_norm": 1.2592357644783008,
+      "learning_rate": 2.9810846425840988e-05,
+      "loss": 1.0684,
+      "step": 16394
+    },
+    {
+      "epoch": 0.7552863131708666,
+      "grad_norm": 0.985550593048758,
+      "learning_rate": 2.9800219300830302e-05,
+      "loss": 1.3542,
+      "step": 16395
+    },
+    {
+      "epoch": 0.7553323812595015,
+      "grad_norm": 0.807556748409824,
+      "learning_rate": 2.9789593738697295e-05,
+      "loss": 1.1331,
+      "step": 16396
+    },
+    {
+      "epoch": 0.7553784493481366,
+      "grad_norm": 1.0730307830841328,
+      "learning_rate": 2.9778969739678663e-05,
+      "loss": 1.534,
+      "step": 16397
+    },
+    {
+      "epoch": 0.7554245174367715,
+      "grad_norm": 1.297296243361992,
+      "learning_rate": 2.9768347304010834e-05,
+      "loss": 1.3581,
+      "step": 16398
+    },
+    {
+      "epoch": 0.7554705855254066,
+      "grad_norm": 1.1686321957029444,
+      "learning_rate": 2.9757726431930376e-05,
+      "loss": 0.8299,
+      "step": 16399
+    },
+    {
+      "epoch": 0.7555166536140415,
+      "grad_norm": 1.5963077018986318,
+      "learning_rate": 2.9747107123673655e-05,
+      "loss": 1.2309,
+      "step": 16400
+    },
+    {
+      "epoch": 0.7555627217026766,
+      "grad_norm": 1.1762534081039644,
+      "learning_rate": 2.973648937947714e-05,
+      "loss": 0.9209,
+      "step": 16401
+    },
+    {
+      "epoch": 0.7556087897913115,
+      "grad_norm": 5.221882841526327,
+      "learning_rate": 2.9725873199577238e-05,
+      "loss": 1.4795,
+      "step": 16402
+    },
+    {
+      "epoch": 0.7556548578799466,
+      "grad_norm": 0.8441658622404482,
+      "learning_rate": 2.9715258584210247e-05,
+      "loss": 1.1841,
+      "step": 16403
+    },
+    {
+      "epoch": 0.7557009259685815,
+      "grad_norm": 0.7894935814589717,
+      "learning_rate": 2.970464553361252e-05,
+      "loss": 1.0407,
+      "step": 16404
+    },
+    {
+      "epoch": 0.7557469940572166,
+      "grad_norm": 0.8012648728309789,
+      "learning_rate": 2.9694034048020314e-05,
+      "loss": 1.4391,
+      "step": 16405
+    },
+    {
+      "epoch": 0.7557930621458516,
+      "grad_norm": 1.1250640590411867,
+      "learning_rate": 2.9683424127669936e-05,
+      "loss": 1.2477,
+      "step": 16406
+    },
+    {
+      "epoch": 0.7558391302344866,
+      "grad_norm": 1.052687543230252,
+      "learning_rate": 2.967281577279748e-05,
+      "loss": 1.5568,
+      "step": 16407
+    },
+    {
+      "epoch": 0.7558851983231216,
+      "grad_norm": 0.7306240810619705,
+      "learning_rate": 2.966220898363926e-05,
+      "loss": 1.076,
+      "step": 16408
+    },
+    {
+      "epoch": 0.7559312664117566,
+      "grad_norm": 2.9758328443739295,
+      "learning_rate": 2.9651603760431358e-05,
+      "loss": 1.2395,
+      "step": 16409
+    },
+    {
+      "epoch": 0.7559773345003916,
+      "grad_norm": 0.9757562840546892,
+      "learning_rate": 2.9641000103409844e-05,
+      "loss": 1.1796,
+      "step": 16410
+    },
+    {
+      "epoch": 0.7560234025890266,
+      "grad_norm": 0.8055763505230786,
+      "learning_rate": 2.9630398012810822e-05,
+      "loss": 1.295,
+      "step": 16411
+    },
+    {
+      "epoch": 0.7560694706776616,
+      "grad_norm": 1.156287121793365,
+      "learning_rate": 2.9619797488870326e-05,
+      "loss": 1.2064,
+      "step": 16412
+    },
+    {
+      "epoch": 0.7561155387662966,
+      "grad_norm": 0.8482780291675854,
+      "learning_rate": 2.96091985318244e-05,
+      "loss": 1.2015,
+      "step": 16413
+    },
+    {
+      "epoch": 0.7561616068549316,
+      "grad_norm": 1.344933694410211,
+      "learning_rate": 2.9598601141908943e-05,
+      "loss": 1.511,
+      "step": 16414
+    },
+    {
+      "epoch": 0.7562076749435666,
+      "grad_norm": 16.331819784038494,
+      "learning_rate": 2.9588005319359914e-05,
+      "loss": 1.2358,
+      "step": 16415
+    },
+    {
+      "epoch": 0.7562537430322016,
+      "grad_norm": 1.0102460884068776,
+      "learning_rate": 2.957741106441322e-05,
+      "loss": 1.1281,
+      "step": 16416
+    },
+    {
+      "epoch": 0.7562998111208366,
+      "grad_norm": 0.7884181111353213,
+      "learning_rate": 2.956681837730475e-05,
+      "loss": 1.1137,
+      "step": 16417
+    },
+    {
+      "epoch": 0.7563458792094716,
+      "grad_norm": 0.8573704475016601,
+      "learning_rate": 2.9556227258270243e-05,
+      "loss": 1.1863,
+      "step": 16418
+    },
+    {
+      "epoch": 0.7563919472981065,
+      "grad_norm": 0.837489108295114,
+      "learning_rate": 2.9545637707545603e-05,
+      "loss": 1.0807,
+      "step": 16419
+    },
+    {
+      "epoch": 0.7564380153867416,
+      "grad_norm": 0.7964002912779465,
+      "learning_rate": 2.953504972536655e-05,
+      "loss": 1.1978,
+      "step": 16420
+    },
+    {
+      "epoch": 0.7564840834753767,
+      "grad_norm": 0.8294886090322292,
+      "learning_rate": 2.9524463311968754e-05,
+      "loss": 1.1896,
+      "step": 16421
+    },
+    {
+      "epoch": 0.7565301515640116,
+      "grad_norm": 1.0807931376199067,
+      "learning_rate": 2.9513878467587942e-05,
+      "loss": 1.2014,
+      "step": 16422
+    },
+    {
+      "epoch": 0.7565762196526467,
+      "grad_norm": 0.8116363035531927,
+      "learning_rate": 2.950329519245977e-05,
+      "loss": 1.1869,
+      "step": 16423
+    },
+    {
+      "epoch": 0.7566222877412816,
+      "grad_norm": 0.9483129826834731,
+      "learning_rate": 2.949271348681989e-05,
+      "loss": 1.4188,
+      "step": 16424
+    },
+    {
+      "epoch": 0.7566683558299167,
+      "grad_norm": 1.0869778524871763,
+      "learning_rate": 2.9482133350903806e-05,
+      "loss": 1.1742,
+      "step": 16425
+    },
+    {
+      "epoch": 0.7567144239185516,
+      "grad_norm": 1.2606686502599296,
+      "learning_rate": 2.947155478494712e-05,
+      "loss": 1.0326,
+      "step": 16426
+    },
+    {
+      "epoch": 0.7567604920071866,
+      "grad_norm": 0.8431604597284509,
+      "learning_rate": 2.946097778918534e-05,
+      "loss": 1.3128,
+      "step": 16427
+    },
+    {
+      "epoch": 0.7568065600958216,
+      "grad_norm": 1.5432763430680168,
+      "learning_rate": 2.9450402363853967e-05,
+      "loss": 1.6122,
+      "step": 16428
+    },
+    {
+      "epoch": 0.7568526281844566,
+      "grad_norm": 0.8719397270874867,
+      "learning_rate": 2.9439828509188393e-05,
+      "loss": 1.2045,
+      "step": 16429
+    },
+    {
+      "epoch": 0.7568986962730916,
+      "grad_norm": 0.6632627938881508,
+      "learning_rate": 2.942925622542404e-05,
+      "loss": 1.2255,
+      "step": 16430
+    },
+    {
+      "epoch": 0.7569447643617266,
+      "grad_norm": 0.7245542093481874,
+      "learning_rate": 2.941868551279634e-05,
+      "loss": 1.484,
+      "step": 16431
+    },
+    {
+      "epoch": 0.7569908324503616,
+      "grad_norm": 0.9569968574393872,
+      "learning_rate": 2.940811637154055e-05,
+      "loss": 1.3902,
+      "step": 16432
+    },
+    {
+      "epoch": 0.7570369005389966,
+      "grad_norm": 0.908203672444329,
+      "learning_rate": 2.9397548801892006e-05,
+      "loss": 1.0172,
+      "step": 16433
+    },
+    {
+      "epoch": 0.7570829686276317,
+      "grad_norm": 4.581050814401609,
+      "learning_rate": 2.9386982804085984e-05,
+      "loss": 1.2401,
+      "step": 16434
+    },
+    {
+      "epoch": 0.7571290367162666,
+      "grad_norm": 0.8421704963834576,
+      "learning_rate": 2.937641837835774e-05,
+      "loss": 1.1087,
+      "step": 16435
+    },
+    {
+      "epoch": 0.7571751048049017,
+      "grad_norm": 0.9709733617595463,
+      "learning_rate": 2.936585552494242e-05,
+      "loss": 1.1956,
+      "step": 16436
+    },
+    {
+      "epoch": 0.7572211728935366,
+      "grad_norm": 0.720653054695481,
+      "learning_rate": 2.9355294244075214e-05,
+      "loss": 1.0921,
+      "step": 16437
+    },
+    {
+      "epoch": 0.7572672409821717,
+      "grad_norm": 0.7093899177763657,
+      "learning_rate": 2.9344734535991292e-05,
+      "loss": 1.2395,
+      "step": 16438
+    },
+    {
+      "epoch": 0.7573133090708066,
+      "grad_norm": 1.167183767286826,
+      "learning_rate": 2.9334176400925673e-05,
+      "loss": 1.1815,
+      "step": 16439
+    },
+    {
+      "epoch": 0.7573593771594417,
+      "grad_norm": 0.9008107198105064,
+      "learning_rate": 2.9323619839113447e-05,
+      "loss": 1.408,
+      "step": 16440
+    },
+    {
+      "epoch": 0.7574054452480766,
+      "grad_norm": 0.7441794707298708,
+      "learning_rate": 2.931306485078964e-05,
+      "loss": 1.2702,
+      "step": 16441
+    },
+    {
+      "epoch": 0.7574515133367117,
+      "grad_norm": 1.1331112306020128,
+      "learning_rate": 2.9302511436189272e-05,
+      "loss": 1.2088,
+      "step": 16442
+    },
+    {
+      "epoch": 0.7574975814253466,
+      "grad_norm": 1.412396340689116,
+      "learning_rate": 2.9291959595547237e-05,
+      "loss": 1.341,
+      "step": 16443
+    },
+    {
+      "epoch": 0.7575436495139817,
+      "grad_norm": 0.729683857065645,
+      "learning_rate": 2.9281409329098485e-05,
+      "loss": 1.4773,
+      "step": 16444
+    },
+    {
+      "epoch": 0.7575897176026166,
+      "grad_norm": 0.940894833096794,
+      "learning_rate": 2.9270860637077903e-05,
+      "loss": 1.3177,
+      "step": 16445
+    },
+    {
+      "epoch": 0.7576357856912517,
+      "grad_norm": 0.835553132356743,
+      "learning_rate": 2.9260313519720362e-05,
+      "loss": 1.3017,
+      "step": 16446
+    },
+    {
+      "epoch": 0.7576818537798867,
+      "grad_norm": 1.0785755321186445,
+      "learning_rate": 2.9249767977260613e-05,
+      "loss": 1.1955,
+      "step": 16447
+    },
+    {
+      "epoch": 0.7577279218685217,
+      "grad_norm": 0.807832283820701,
+      "learning_rate": 2.9239224009933462e-05,
+      "loss": 1.1156,
+      "step": 16448
+    },
+    {
+      "epoch": 0.7577739899571567,
+      "grad_norm": 0.7926271841550426,
+      "learning_rate": 2.922868161797371e-05,
+      "loss": 1.4172,
+      "step": 16449
+    },
+    {
+      "epoch": 0.7578200580457917,
+      "grad_norm": 1.2716273044868445,
+      "learning_rate": 2.921814080161597e-05,
+      "loss": 1.3777,
+      "step": 16450
+    },
+    {
+      "epoch": 0.7578661261344267,
+      "grad_norm": 0.942337715216288,
+      "learning_rate": 2.9207601561094967e-05,
+      "loss": 1.2184,
+      "step": 16451
+    },
+    {
+      "epoch": 0.7579121942230617,
+      "grad_norm": 0.795916443360289,
+      "learning_rate": 2.9197063896645328e-05,
+      "loss": 1.2563,
+      "step": 16452
+    },
+    {
+      "epoch": 0.7579582623116967,
+      "grad_norm": 1.1397744807303292,
+      "learning_rate": 2.9186527808501696e-05,
+      "loss": 1.3619,
+      "step": 16453
+    },
+    {
+      "epoch": 0.7580043304003317,
+      "grad_norm": 0.8668240594031005,
+      "learning_rate": 2.9175993296898564e-05,
+      "loss": 1.2806,
+      "step": 16454
+    },
+    {
+      "epoch": 0.7580503984889667,
+      "grad_norm": 0.8289724678050258,
+      "learning_rate": 2.9165460362070506e-05,
+      "loss": 1.2875,
+      "step": 16455
+    },
+    {
+      "epoch": 0.7580964665776017,
+      "grad_norm": 0.7531337427579594,
+      "learning_rate": 2.915492900425205e-05,
+      "loss": 1.339,
+      "step": 16456
+    },
+    {
+      "epoch": 0.7581425346662367,
+      "grad_norm": 1.0188530859148703,
+      "learning_rate": 2.91443992236776e-05,
+      "loss": 1.215,
+      "step": 16457
+    },
+    {
+      "epoch": 0.7581886027548717,
+      "grad_norm": 0.810048257676218,
+      "learning_rate": 2.9133871020581606e-05,
+      "loss": 1.4048,
+      "step": 16458
+    },
+    {
+      "epoch": 0.7582346708435067,
+      "grad_norm": 0.7938673004856487,
+      "learning_rate": 2.912334439519847e-05,
+      "loss": 1.3482,
+      "step": 16459
+    },
+    {
+      "epoch": 0.7582807389321417,
+      "grad_norm": 0.8048886306878784,
+      "learning_rate": 2.9112819347762565e-05,
+      "loss": 1.2575,
+      "step": 16460
+    },
+    {
+      "epoch": 0.7583268070207767,
+      "grad_norm": 4.761779579627597,
+      "learning_rate": 2.9102295878508167e-05,
+      "loss": 1.3711,
+      "step": 16461
+    },
+    {
+      "epoch": 0.7583728751094118,
+      "grad_norm": 0.7944428063808693,
+      "learning_rate": 2.909177398766959e-05,
+      "loss": 1.3,
+      "step": 16462
+    },
+    {
+      "epoch": 0.7584189431980467,
+      "grad_norm": 0.9667831350251491,
+      "learning_rate": 2.9081253675481078e-05,
+      "loss": 1.2509,
+      "step": 16463
+    },
+    {
+      "epoch": 0.7584650112866818,
+      "grad_norm": 0.9943493323093153,
+      "learning_rate": 2.907073494217689e-05,
+      "loss": 1.1336,
+      "step": 16464
+    },
+    {
+      "epoch": 0.7585110793753167,
+      "grad_norm": 1.0413994229678805,
+      "learning_rate": 2.906021778799114e-05,
+      "loss": 0.9893,
+      "step": 16465
+    },
+    {
+      "epoch": 0.7585571474639518,
+      "grad_norm": 0.771115861985214,
+      "learning_rate": 2.9049702213158002e-05,
+      "loss": 1.1394,
+      "step": 16466
+    },
+    {
+      "epoch": 0.7586032155525867,
+      "grad_norm": 0.7553341440267578,
+      "learning_rate": 2.9039188217911618e-05,
+      "loss": 1.0698,
+      "step": 16467
+    },
+    {
+      "epoch": 0.7586492836412217,
+      "grad_norm": 0.9831162602173327,
+      "learning_rate": 2.9028675802486014e-05,
+      "loss": 1.262,
+      "step": 16468
+    },
+    {
+      "epoch": 0.7586953517298567,
+      "grad_norm": 1.0945244490948904,
+      "learning_rate": 2.9018164967115247e-05,
+      "loss": 1.2144,
+      "step": 16469
+    },
+    {
+      "epoch": 0.7587414198184917,
+      "grad_norm": 1.3522801285456576,
+      "learning_rate": 2.900765571203333e-05,
+      "loss": 1.3291,
+      "step": 16470
+    },
+    {
+      "epoch": 0.7587874879071267,
+      "grad_norm": 1.1726646691684057,
+      "learning_rate": 2.8997148037474264e-05,
+      "loss": 1.036,
+      "step": 16471
+    },
+    {
+      "epoch": 0.7588335559957617,
+      "grad_norm": 1.3197270059432138,
+      "learning_rate": 2.8986641943671922e-05,
+      "loss": 1.3629,
+      "step": 16472
+    },
+    {
+      "epoch": 0.7588796240843967,
+      "grad_norm": 1.1215426669278856,
+      "learning_rate": 2.897613743086024e-05,
+      "loss": 1.3485,
+      "step": 16473
+    },
+    {
+      "epoch": 0.7589256921730317,
+      "grad_norm": 0.693678264147324,
+      "learning_rate": 2.8965634499273107e-05,
+      "loss": 1.2993,
+      "step": 16474
+    },
+    {
+      "epoch": 0.7589717602616668,
+      "grad_norm": 5.007782863549114,
+      "learning_rate": 2.8955133149144288e-05,
+      "loss": 1.3508,
+      "step": 16475
+    },
+    {
+      "epoch": 0.7590178283503017,
+      "grad_norm": 0.8415545846698208,
+      "learning_rate": 2.8944633380707653e-05,
+      "loss": 1.1511,
+      "step": 16476
+    },
+    {
+      "epoch": 0.7590638964389368,
+      "grad_norm": 1.2886983530753027,
+      "learning_rate": 2.8934135194196856e-05,
+      "loss": 1.0884,
+      "step": 16477
+    },
+    {
+      "epoch": 0.7591099645275717,
+      "grad_norm": 0.5976188741484013,
+      "learning_rate": 2.8923638589845748e-05,
+      "loss": 1.3593,
+      "step": 16478
+    },
+    {
+      "epoch": 0.7591560326162068,
+      "grad_norm": 0.9770477213475919,
+      "learning_rate": 2.8913143567887924e-05,
+      "loss": 1.1499,
+      "step": 16479
+    },
+    {
+      "epoch": 0.7592021007048417,
+      "grad_norm": 1.000671493916144,
+      "learning_rate": 2.8902650128557074e-05,
+      "loss": 1.4797,
+      "step": 16480
+    },
+    {
+      "epoch": 0.7592481687934768,
+      "grad_norm": 1.456248599139167,
+      "learning_rate": 2.8892158272086813e-05,
+      "loss": 1.2149,
+      "step": 16481
+    },
+    {
+      "epoch": 0.7592942368821117,
+      "grad_norm": 1.0014249576517746,
+      "learning_rate": 2.888166799871075e-05,
+      "loss": 1.3071,
+      "step": 16482
+    },
+    {
+      "epoch": 0.7593403049707468,
+      "grad_norm": 0.8780939492295506,
+      "learning_rate": 2.887117930866241e-05,
+      "loss": 1.3452,
+      "step": 16483
+    },
+    {
+      "epoch": 0.7593863730593817,
+      "grad_norm": 0.8436663510130199,
+      "learning_rate": 2.8860692202175243e-05,
+      "loss": 1.3828,
+      "step": 16484
+    },
+    {
+      "epoch": 0.7594324411480168,
+      "grad_norm": 1.108664447765566,
+      "learning_rate": 2.885020667948285e-05,
+      "loss": 1.1974,
+      "step": 16485
+    },
+    {
+      "epoch": 0.7594785092366517,
+      "grad_norm": 1.2004878841499513,
+      "learning_rate": 2.8839722740818574e-05,
+      "loss": 1.2652,
+      "step": 16486
+    },
+    {
+      "epoch": 0.7595245773252868,
+      "grad_norm": 1.2261745080834696,
+      "learning_rate": 2.882924038641589e-05,
+      "loss": 1.1184,
+      "step": 16487
+    },
+    {
+      "epoch": 0.7595706454139218,
+      "grad_norm": 1.2656699295692142,
+      "learning_rate": 2.8818759616508073e-05,
+      "loss": 1.284,
+      "step": 16488
+    },
+    {
+      "epoch": 0.7596167135025568,
+      "grad_norm": 1.8218338078305298,
+      "learning_rate": 2.8808280431328593e-05,
+      "loss": 1.0003,
+      "step": 16489
+    },
+    {
+      "epoch": 0.7596627815911918,
+      "grad_norm": 1.7075356501876546,
+      "learning_rate": 2.879780283111064e-05,
+      "loss": 1.0887,
+      "step": 16490
+    },
+    {
+      "epoch": 0.7597088496798268,
+      "grad_norm": 2.743427083098995,
+      "learning_rate": 2.8787326816087533e-05,
+      "loss": 1.244,
+      "step": 16491
+    },
+    {
+      "epoch": 0.7597549177684618,
+      "grad_norm": 0.7529129425100328,
+      "learning_rate": 2.8776852386492515e-05,
+      "loss": 1.0612,
+      "step": 16492
+    },
+    {
+      "epoch": 0.7598009858570968,
+      "grad_norm": 0.8175165477171283,
+      "learning_rate": 2.876637954255873e-05,
+      "loss": 1.5208,
+      "step": 16493
+    },
+    {
+      "epoch": 0.7598470539457318,
+      "grad_norm": 2.300966947905619,
+      "learning_rate": 2.8755908284519396e-05,
+      "loss": 1.4459,
+      "step": 16494
+    },
+    {
+      "epoch": 0.7598931220343668,
+      "grad_norm": 1.9867312704090616,
+      "learning_rate": 2.8745438612607534e-05,
+      "loss": 1.4025,
+      "step": 16495
+    },
+    {
+      "epoch": 0.7599391901230018,
+      "grad_norm": 2.701195317896777,
+      "learning_rate": 2.873497052705638e-05,
+      "loss": 1.3182,
+      "step": 16496
+    },
+    {
+      "epoch": 0.7599852582116368,
+      "grad_norm": 0.7732743674817616,
+      "learning_rate": 2.8724504028098875e-05,
+      "loss": 1.4509,
+      "step": 16497
+    },
+    {
+      "epoch": 0.7600313263002718,
+      "grad_norm": 1.673269579441216,
+      "learning_rate": 2.8714039115968118e-05,
+      "loss": 1.3454,
+      "step": 16498
+    },
+    {
+      "epoch": 0.7600773943889068,
+      "grad_norm": 3.127928472181679,
+      "learning_rate": 2.8703575790896976e-05,
+      "loss": 1.1515,
+      "step": 16499
+    },
+    {
+      "epoch": 0.7601234624775418,
+      "grad_norm": 0.8673857048906237,
+      "learning_rate": 2.869311405311854e-05,
+      "loss": 1.5192,
+      "step": 16500
+    },
+    {
+      "epoch": 0.7601695305661768,
+      "grad_norm": 0.8071786622676377,
+      "learning_rate": 2.8682653902865663e-05,
+      "loss": 1.2761,
+      "step": 16501
+    },
+    {
+      "epoch": 0.7602155986548118,
+      "grad_norm": 1.063336570323637,
+      "learning_rate": 2.8672195340371167e-05,
+      "loss": 1.4419,
+      "step": 16502
+    },
+    {
+      "epoch": 0.7602616667434469,
+      "grad_norm": 1.1150063943625907,
+      "learning_rate": 2.866173836586795e-05,
+      "loss": 1.2466,
+      "step": 16503
+    },
+    {
+      "epoch": 0.7603077348320818,
+      "grad_norm": 1.0178200016346008,
+      "learning_rate": 2.865128297958881e-05,
+      "loss": 1.2663,
+      "step": 16504
+    },
+    {
+      "epoch": 0.7603538029207169,
+      "grad_norm": 1.2751564472010015,
+      "learning_rate": 2.864082918176655e-05,
+      "loss": 1.3922,
+      "step": 16505
+    },
+    {
+      "epoch": 0.7603998710093518,
+      "grad_norm": 0.7603637537744482,
+      "learning_rate": 2.86303769726338e-05,
+      "loss": 1.3301,
+      "step": 16506
+    },
+    {
+      "epoch": 0.7604459390979869,
+      "grad_norm": 1.2416925934712653,
+      "learning_rate": 2.861992635242341e-05,
+      "loss": 1.4298,
+      "step": 16507
+    },
+    {
+      "epoch": 0.7604920071866218,
+      "grad_norm": 1.0333166112796026,
+      "learning_rate": 2.8609477321367927e-05,
+      "loss": 1.3364,
+      "step": 16508
+    },
+    {
+      "epoch": 0.7605380752752569,
+      "grad_norm": 0.6490105708979852,
+      "learning_rate": 2.8599029879700058e-05,
+      "loss": 1.3577,
+      "step": 16509
+    },
+    {
+      "epoch": 0.7605841433638918,
+      "grad_norm": 0.8293076024192286,
+      "learning_rate": 2.858858402765233e-05,
+      "loss": 1.4675,
+      "step": 16510
+    },
+    {
+      "epoch": 0.7606302114525268,
+      "grad_norm": 2.3713278585864974,
+      "learning_rate": 2.8578139765457345e-05,
+      "loss": 1.5154,
+      "step": 16511
+    },
+    {
+      "epoch": 0.7606762795411618,
+      "grad_norm": 1.5980303723888065,
+      "learning_rate": 2.8567697093347635e-05,
+      "loss": 0.8272,
+      "step": 16512
+    },
+    {
+      "epoch": 0.7607223476297968,
+      "grad_norm": 0.6758298282040829,
+      "learning_rate": 2.8557256011555645e-05,
+      "loss": 1.4464,
+      "step": 16513
+    },
+    {
+      "epoch": 0.7607684157184318,
+      "grad_norm": 0.7416893178169027,
+      "learning_rate": 2.8546816520313847e-05,
+      "loss": 1.0062,
+      "step": 16514
+    },
+    {
+      "epoch": 0.7608144838070668,
+      "grad_norm": 1.169059810036364,
+      "learning_rate": 2.8536378619854665e-05,
+      "loss": 1.3778,
+      "step": 16515
+    },
+    {
+      "epoch": 0.7608605518957019,
+      "grad_norm": 1.4178136078888843,
+      "learning_rate": 2.8525942310410503e-05,
+      "loss": 1.3436,
+      "step": 16516
+    },
+    {
+      "epoch": 0.7609066199843368,
+      "grad_norm": 1.4335856489221785,
+      "learning_rate": 2.851550759221363e-05,
+      "loss": 1.2013,
+      "step": 16517
+    },
+    {
+      "epoch": 0.7609526880729719,
+      "grad_norm": 16.009423779759455,
+      "learning_rate": 2.850507446549647e-05,
+      "loss": 1.2523,
+      "step": 16518
+    },
+    {
+      "epoch": 0.7609987561616068,
+      "grad_norm": 0.9004052051986036,
+      "learning_rate": 2.8494642930491244e-05,
+      "loss": 1.1946,
+      "step": 16519
+    },
+    {
+      "epoch": 0.7610448242502419,
+      "grad_norm": 0.9598160631494367,
+      "learning_rate": 2.848421298743016e-05,
+      "loss": 1.1329,
+      "step": 16520
+    },
+    {
+      "epoch": 0.7610908923388768,
+      "grad_norm": 1.1305443655458876,
+      "learning_rate": 2.847378463654544e-05,
+      "loss": 1.2651,
+      "step": 16521
+    },
+    {
+      "epoch": 0.7611369604275119,
+      "grad_norm": 0.8822130425818574,
+      "learning_rate": 2.8463357878069263e-05,
+      "loss": 1.1603,
+      "step": 16522
+    },
+    {
+      "epoch": 0.7611830285161468,
+      "grad_norm": 1.1951027177280544,
+      "learning_rate": 2.8452932712233804e-05,
+      "loss": 1.2362,
+      "step": 16523
+    },
+    {
+      "epoch": 0.7612290966047819,
+      "grad_norm": 0.9910891085449359,
+      "learning_rate": 2.844250913927109e-05,
+      "loss": 1.2603,
+      "step": 16524
+    },
+    {
+      "epoch": 0.7612751646934168,
+      "grad_norm": 1.4588793265937907,
+      "learning_rate": 2.8432087159413213e-05,
+      "loss": 1.0383,
+      "step": 16525
+    },
+    {
+      "epoch": 0.7613212327820519,
+      "grad_norm": 1.0701303153823183,
+      "learning_rate": 2.8421666772892197e-05,
+      "loss": 1.3338,
+      "step": 16526
+    },
+    {
+      "epoch": 0.7613673008706868,
+      "grad_norm": 0.7149398538086076,
+      "learning_rate": 2.841124797994007e-05,
+      "loss": 1.071,
+      "step": 16527
+    },
+    {
+      "epoch": 0.7614133689593219,
+      "grad_norm": 1.117478160991373,
+      "learning_rate": 2.8400830780788733e-05,
+      "loss": 1.0952,
+      "step": 16528
+    },
+    {
+      "epoch": 0.7614594370479569,
+      "grad_norm": 0.8512027232364152,
+      "learning_rate": 2.8390415175670137e-05,
+      "loss": 1.2799,
+      "step": 16529
+    },
+    {
+      "epoch": 0.7615055051365919,
+      "grad_norm": 2.9104439722625597,
+      "learning_rate": 2.8380001164816185e-05,
+      "loss": 1.1541,
+      "step": 16530
+    },
+    {
+      "epoch": 0.7615515732252269,
+      "grad_norm": 1.1468261712840944,
+      "learning_rate": 2.8369588748458676e-05,
+      "loss": 1.1485,
+      "step": 16531
+    },
+    {
+      "epoch": 0.7615976413138619,
+      "grad_norm": 0.7667036442868055,
+      "learning_rate": 2.835917792682946e-05,
+      "loss": 1.2986,
+      "step": 16532
+    },
+    {
+      "epoch": 0.7616437094024969,
+      "grad_norm": 2.4892023204545515,
+      "learning_rate": 2.83487687001603e-05,
+      "loss": 1.0974,
+      "step": 16533
+    },
+    {
+      "epoch": 0.7616897774911319,
+      "grad_norm": 1.5539507329110214,
+      "learning_rate": 2.8338361068682994e-05,
+      "loss": 1.1911,
+      "step": 16534
+    },
+    {
+      "epoch": 0.7617358455797669,
+      "grad_norm": 0.7040757626921256,
+      "learning_rate": 2.832795503262916e-05,
+      "loss": 1.222,
+      "step": 16535
+    },
+    {
+      "epoch": 0.7617819136684019,
+      "grad_norm": 1.0769384433686127,
+      "learning_rate": 2.8317550592230523e-05,
+      "loss": 1.1963,
+      "step": 16536
+    },
+    {
+      "epoch": 0.7618279817570369,
+      "grad_norm": 0.9946893949628086,
+      "learning_rate": 2.830714774771871e-05,
+      "loss": 1.3713,
+      "step": 16537
+    },
+    {
+      "epoch": 0.7618740498456719,
+      "grad_norm": 1.0662814287542142,
+      "learning_rate": 2.8296746499325366e-05,
+      "loss": 1.085,
+      "step": 16538
+    },
+    {
+      "epoch": 0.7619201179343069,
+      "grad_norm": 1.1249164435865826,
+      "learning_rate": 2.8286346847281975e-05,
+      "loss": 1.3557,
+      "step": 16539
+    },
+    {
+      "epoch": 0.7619661860229419,
+      "grad_norm": 1.0804451969505176,
+      "learning_rate": 2.8275948791820107e-05,
+      "loss": 1.1866,
+      "step": 16540
+    },
+    {
+      "epoch": 0.7620122541115769,
+      "grad_norm": 2.4984685125484107,
+      "learning_rate": 2.8265552333171297e-05,
+      "loss": 0.9285,
+      "step": 16541
+    },
+    {
+      "epoch": 0.762058322200212,
+      "grad_norm": 1.6213682547244264,
+      "learning_rate": 2.825515747156694e-05,
+      "loss": 1.1015,
+      "step": 16542
+    },
+    {
+      "epoch": 0.7621043902888469,
+      "grad_norm": 1.2467696671194561,
+      "learning_rate": 2.824476420723847e-05,
+      "loss": 1.0852,
+      "step": 16543
+    },
+    {
+      "epoch": 0.762150458377482,
+      "grad_norm": 0.8697577494835338,
+      "learning_rate": 2.82343725404173e-05,
+      "loss": 1.4721,
+      "step": 16544
+    },
+    {
+      "epoch": 0.7621965264661169,
+      "grad_norm": 3.1252506375401077,
+      "learning_rate": 2.8223982471334807e-05,
+      "loss": 1.2156,
+      "step": 16545
+    },
+    {
+      "epoch": 0.762242594554752,
+      "grad_norm": 1.0782968587038781,
+      "learning_rate": 2.8213594000222242e-05,
+      "loss": 0.8654,
+      "step": 16546
+    },
+    {
+      "epoch": 0.7622886626433869,
+      "grad_norm": 1.0024222147588453,
+      "learning_rate": 2.8203207127310916e-05,
+      "loss": 1.1395,
+      "step": 16547
+    },
+    {
+      "epoch": 0.762334730732022,
+      "grad_norm": 0.7912611373332956,
+      "learning_rate": 2.8192821852832117e-05,
+      "loss": 1.3597,
+      "step": 16548
+    },
+    {
+      "epoch": 0.7623807988206569,
+      "grad_norm": 1.0118930430978548,
+      "learning_rate": 2.8182438177016978e-05,
+      "loss": 1.4573,
+      "step": 16549
+    },
+    {
+      "epoch": 0.762426866909292,
+      "grad_norm": 0.8641097096568908,
+      "learning_rate": 2.817205610009671e-05,
+      "loss": 1.4892,
+      "step": 16550
+    },
+    {
+      "epoch": 0.7624729349979269,
+      "grad_norm": 1.5241438461474164,
+      "learning_rate": 2.8161675622302465e-05,
+      "loss": 1.0416,
+      "step": 16551
+    },
+    {
+      "epoch": 0.762519003086562,
+      "grad_norm": 0.9223769556422519,
+      "learning_rate": 2.8151296743865363e-05,
+      "loss": 0.9632,
+      "step": 16552
+    },
+    {
+      "epoch": 0.7625650711751969,
+      "grad_norm": 0.8998202475327988,
+      "learning_rate": 2.8140919465016415e-05,
+      "loss": 1.1708,
+      "step": 16553
+    },
+    {
+      "epoch": 0.762611139263832,
+      "grad_norm": 1.7969484612422368,
+      "learning_rate": 2.8130543785986675e-05,
+      "loss": 1.1091,
+      "step": 16554
+    },
+    {
+      "epoch": 0.7626572073524669,
+      "grad_norm": 1.2834885705404204,
+      "learning_rate": 2.812016970700715e-05,
+      "loss": 1.1632,
+      "step": 16555
+    },
+    {
+      "epoch": 0.762703275441102,
+      "grad_norm": 0.8670220932969063,
+      "learning_rate": 2.8109797228308832e-05,
+      "loss": 1.4635,
+      "step": 16556
+    },
+    {
+      "epoch": 0.762749343529737,
+      "grad_norm": 2.072092942421221,
+      "learning_rate": 2.809942635012258e-05,
+      "loss": 1.0037,
+      "step": 16557
+    },
+    {
+      "epoch": 0.7627954116183719,
+      "grad_norm": 0.7845156048767307,
+      "learning_rate": 2.8089057072679324e-05,
+      "loss": 1.2216,
+      "step": 16558
+    },
+    {
+      "epoch": 0.762841479707007,
+      "grad_norm": 1.1204071040194858,
+      "learning_rate": 2.8078689396209947e-05,
+      "loss": 1.3457,
+      "step": 16559
+    },
+    {
+      "epoch": 0.7628875477956419,
+      "grad_norm": 1.592523968828195,
+      "learning_rate": 2.8068323320945188e-05,
+      "loss": 1.1343,
+      "step": 16560
+    },
+    {
+      "epoch": 0.762933615884277,
+      "grad_norm": 8.449476508131577,
+      "learning_rate": 2.8057958847115885e-05,
+      "loss": 1.1522,
+      "step": 16561
+    },
+    {
+      "epoch": 0.7629796839729119,
+      "grad_norm": 0.981373671869725,
+      "learning_rate": 2.8047595974952756e-05,
+      "loss": 1.3268,
+      "step": 16562
+    },
+    {
+      "epoch": 0.763025752061547,
+      "grad_norm": 1.025886966857572,
+      "learning_rate": 2.8037234704686576e-05,
+      "loss": 1.2338,
+      "step": 16563
+    },
+    {
+      "epoch": 0.7630718201501819,
+      "grad_norm": 2.4424929033585534,
+      "learning_rate": 2.802687503654795e-05,
+      "loss": 1.0298,
+      "step": 16564
+    },
+    {
+      "epoch": 0.763117888238817,
+      "grad_norm": 1.668511579736359,
+      "learning_rate": 2.8016516970767536e-05,
+      "loss": 1.1342,
+      "step": 16565
+    },
+    {
+      "epoch": 0.7631639563274519,
+      "grad_norm": 1.7004557550347166,
+      "learning_rate": 2.8006160507575973e-05,
+      "loss": 1.1096,
+      "step": 16566
+    },
+    {
+      "epoch": 0.763210024416087,
+      "grad_norm": 1.0673187023313437,
+      "learning_rate": 2.7995805647203778e-05,
+      "loss": 1.1786,
+      "step": 16567
+    },
+    {
+      "epoch": 0.7632560925047219,
+      "grad_norm": 1.048432706606706,
+      "learning_rate": 2.7985452389881507e-05,
+      "loss": 0.8194,
+      "step": 16568
+    },
+    {
+      "epoch": 0.763302160593357,
+      "grad_norm": 1.319968691072294,
+      "learning_rate": 2.7975100735839665e-05,
+      "loss": 1.2682,
+      "step": 16569
+    },
+    {
+      "epoch": 0.763348228681992,
+      "grad_norm": 7.217114393049458,
+      "learning_rate": 2.7964750685308737e-05,
+      "loss": 1.427,
+      "step": 16570
+    },
+    {
+      "epoch": 0.763394296770627,
+      "grad_norm": 2.506579541441181,
+      "learning_rate": 2.7954402238519084e-05,
+      "loss": 1.3164,
+      "step": 16571
+    },
+    {
+      "epoch": 0.763440364859262,
+      "grad_norm": 2.411993408047094,
+      "learning_rate": 2.794405539570114e-05,
+      "loss": 1.3824,
+      "step": 16572
+    },
+    {
+      "epoch": 0.763486432947897,
+      "grad_norm": 0.8264139601530258,
+      "learning_rate": 2.793371015708526e-05,
+      "loss": 1.4729,
+      "step": 16573
+    },
+    {
+      "epoch": 0.763532501036532,
+      "grad_norm": 1.189861593813485,
+      "learning_rate": 2.792336652290177e-05,
+      "loss": 1.3003,
+      "step": 16574
+    },
+    {
+      "epoch": 0.763578569125167,
+      "grad_norm": 1.062906189887033,
+      "learning_rate": 2.7913024493380958e-05,
+      "loss": 1.5366,
+      "step": 16575
+    },
+    {
+      "epoch": 0.763624637213802,
+      "grad_norm": 0.8785520152616823,
+      "learning_rate": 2.7902684068752983e-05,
+      "loss": 1.2667,
+      "step": 16576
+    },
+    {
+      "epoch": 0.763670705302437,
+      "grad_norm": 1.2375837417514102,
+      "learning_rate": 2.7892345249248186e-05,
+      "loss": 1.399,
+      "step": 16577
+    },
+    {
+      "epoch": 0.763716773391072,
+      "grad_norm": 1.0060089292294245,
+      "learning_rate": 2.788200803509665e-05,
+      "loss": 1.1393,
+      "step": 16578
+    },
+    {
+      "epoch": 0.763762841479707,
+      "grad_norm": 1.2410081293455817,
+      "learning_rate": 2.787167242652855e-05,
+      "loss": 1.4428,
+      "step": 16579
+    },
+    {
+      "epoch": 0.763808909568342,
+      "grad_norm": 0.8963654445728878,
+      "learning_rate": 2.7861338423773996e-05,
+      "loss": 1.343,
+      "step": 16580
+    },
+    {
+      "epoch": 0.763854977656977,
+      "grad_norm": 0.8468953732539141,
+      "learning_rate": 2.7851006027063077e-05,
+      "loss": 1.3137,
+      "step": 16581
+    },
+    {
+      "epoch": 0.763901045745612,
+      "grad_norm": 1.8611177837692447,
+      "learning_rate": 2.784067523662578e-05,
+      "loss": 1.1649,
+      "step": 16582
+    },
+    {
+      "epoch": 0.7639471138342471,
+      "grad_norm": 0.866356476811017,
+      "learning_rate": 2.7830346052692113e-05,
+      "loss": 1.4016,
+      "step": 16583
+    },
+    {
+      "epoch": 0.763993181922882,
+      "grad_norm": 0.799092831189386,
+      "learning_rate": 2.7820018475492073e-05,
+      "loss": 1.2841,
+      "step": 16584
+    },
+    {
+      "epoch": 0.7640392500115171,
+      "grad_norm": 1.1886510285454919,
+      "learning_rate": 2.7809692505255536e-05,
+      "loss": 1.2051,
+      "step": 16585
+    },
+    {
+      "epoch": 0.764085318100152,
+      "grad_norm": 1.8210160973873672,
+      "learning_rate": 2.7799368142212447e-05,
+      "loss": 1.6888,
+      "step": 16586
+    },
+    {
+      "epoch": 0.7641313861887871,
+      "grad_norm": 2.1870627881470264,
+      "learning_rate": 2.778904538659256e-05,
+      "loss": 1.1612,
+      "step": 16587
+    },
+    {
+      "epoch": 0.764177454277422,
+      "grad_norm": 1.6133265640692822,
+      "learning_rate": 2.7778724238625832e-05,
+      "loss": 1.4549,
+      "step": 16588
+    },
+    {
+      "epoch": 0.7642235223660571,
+      "grad_norm": 1.3758958100796026,
+      "learning_rate": 2.776840469854194e-05,
+      "loss": 1.1432,
+      "step": 16589
+    },
+    {
+      "epoch": 0.764269590454692,
+      "grad_norm": 1.3366802854433466,
+      "learning_rate": 2.7758086766570668e-05,
+      "loss": 0.9842,
+      "step": 16590
+    },
+    {
+      "epoch": 0.7643156585433271,
+      "grad_norm": 1.330090764695585,
+      "learning_rate": 2.7747770442941724e-05,
+      "loss": 1.3333,
+      "step": 16591
+    },
+    {
+      "epoch": 0.764361726631962,
+      "grad_norm": 0.9596573608108885,
+      "learning_rate": 2.773745572788482e-05,
+      "loss": 1.5019,
+      "step": 16592
+    },
+    {
+      "epoch": 0.7644077947205971,
+      "grad_norm": 0.7536477401634368,
+      "learning_rate": 2.7727142621629554e-05,
+      "loss": 1.2491,
+      "step": 16593
+    },
+    {
+      "epoch": 0.764453862809232,
+      "grad_norm": 1.2199659453656788,
+      "learning_rate": 2.7716831124405474e-05,
+      "loss": 1.2201,
+      "step": 16594
+    },
+    {
+      "epoch": 0.764499930897867,
+      "grad_norm": 1.3895919445610925,
+      "learning_rate": 2.7706521236442285e-05,
+      "loss": 1.1941,
+      "step": 16595
+    },
+    {
+      "epoch": 0.764545998986502,
+      "grad_norm": 2.1734103072224693,
+      "learning_rate": 2.76962129579694e-05,
+      "loss": 1.0776,
+      "step": 16596
+    },
+    {
+      "epoch": 0.764592067075137,
+      "grad_norm": 1.1431031475817248,
+      "learning_rate": 2.768590628921639e-05,
+      "loss": 1.2458,
+      "step": 16597
+    },
+    {
+      "epoch": 0.7646381351637721,
+      "grad_norm": 2.040561776783177,
+      "learning_rate": 2.767560123041263e-05,
+      "loss": 0.962,
+      "step": 16598
+    },
+    {
+      "epoch": 0.764684203252407,
+      "grad_norm": 1.4926127959117437,
+      "learning_rate": 2.766529778178766e-05,
+      "loss": 1.2426,
+      "step": 16599
+    },
+    {
+      "epoch": 0.7647302713410421,
+      "grad_norm": 2.3736875553677366,
+      "learning_rate": 2.765499594357077e-05,
+      "loss": 1.074,
+      "step": 16600
+    },
+    {
+      "epoch": 0.764776339429677,
+      "grad_norm": 1.2291792063248506,
+      "learning_rate": 2.7644695715991386e-05,
+      "loss": 1.3002,
+      "step": 16601
+    },
+    {
+      "epoch": 0.7648224075183121,
+      "grad_norm": 1.2166157442603955,
+      "learning_rate": 2.7634397099278763e-05,
+      "loss": 1.625,
+      "step": 16602
+    },
+    {
+      "epoch": 0.764868475606947,
+      "grad_norm": 0.9735347967843205,
+      "learning_rate": 2.7624100093662198e-05,
+      "loss": 1.2326,
+      "step": 16603
+    },
+    {
+      "epoch": 0.7649145436955821,
+      "grad_norm": 5.138106266781634,
+      "learning_rate": 2.7613804699370982e-05,
+      "loss": 1.2897,
+      "step": 16604
+    },
+    {
+      "epoch": 0.764960611784217,
+      "grad_norm": 0.9402313705234158,
+      "learning_rate": 2.7603510916634222e-05,
+      "loss": 1.3365,
+      "step": 16605
+    },
+    {
+      "epoch": 0.7650066798728521,
+      "grad_norm": 0.8763232103738414,
+      "learning_rate": 2.7593218745681228e-05,
+      "loss": 1.0519,
+      "step": 16606
+    },
+    {
+      "epoch": 0.765052747961487,
+      "grad_norm": 0.8408665676167738,
+      "learning_rate": 2.758292818674103e-05,
+      "loss": 1.0805,
+      "step": 16607
+    },
+    {
+      "epoch": 0.7650988160501221,
+      "grad_norm": 1.7757510275768367,
+      "learning_rate": 2.757263924004281e-05,
+      "loss": 1.4388,
+      "step": 16608
+    },
+    {
+      "epoch": 0.765144884138757,
+      "grad_norm": 5.879339230279858,
+      "learning_rate": 2.756235190581552e-05,
+      "loss": 1.3616,
+      "step": 16609
+    },
+    {
+      "epoch": 0.7651909522273921,
+      "grad_norm": 3.143954372853699,
+      "learning_rate": 2.7552066184288315e-05,
+      "loss": 1.2328,
+      "step": 16610
+    },
+    {
+      "epoch": 0.7652370203160271,
+      "grad_norm": 0.9284982097049462,
+      "learning_rate": 2.7541782075690148e-05,
+      "loss": 1.4161,
+      "step": 16611
+    },
+    {
+      "epoch": 0.7652830884046621,
+      "grad_norm": 1.79670406284219,
+      "learning_rate": 2.7531499580249932e-05,
+      "loss": 1.1461,
+      "step": 16612
+    },
+    {
+      "epoch": 0.7653291564932971,
+      "grad_norm": 1.0191870614622482,
+      "learning_rate": 2.7521218698196617e-05,
+      "loss": 1.1445,
+      "step": 16613
+    },
+    {
+      "epoch": 0.7653752245819321,
+      "grad_norm": 1.1233458806573966,
+      "learning_rate": 2.7510939429759087e-05,
+      "loss": 1.1134,
+      "step": 16614
+    },
+    {
+      "epoch": 0.7654212926705671,
+      "grad_norm": 1.2757688294373104,
+      "learning_rate": 2.7500661775166236e-05,
+      "loss": 1.2657,
+      "step": 16615
+    },
+    {
+      "epoch": 0.7654673607592021,
+      "grad_norm": 1.1372805536436446,
+      "learning_rate": 2.7490385734646785e-05,
+      "loss": 1.3421,
+      "step": 16616
+    },
+    {
+      "epoch": 0.7655134288478371,
+      "grad_norm": 1.3148912720938128,
+      "learning_rate": 2.7480111308429624e-05,
+      "loss": 1.2604,
+      "step": 16617
+    },
+    {
+      "epoch": 0.7655594969364721,
+      "grad_norm": 1.6255695266753087,
+      "learning_rate": 2.7469838496743407e-05,
+      "loss": 1.3361,
+      "step": 16618
+    },
+    {
+      "epoch": 0.7656055650251071,
+      "grad_norm": 1.0626380966954758,
+      "learning_rate": 2.74595672998169e-05,
+      "loss": 1.1891,
+      "step": 16619
+    },
+    {
+      "epoch": 0.7656516331137421,
+      "grad_norm": 0.8881240597553671,
+      "learning_rate": 2.744929771787873e-05,
+      "loss": 1.4549,
+      "step": 16620
+    },
+    {
+      "epoch": 0.7656977012023771,
+      "grad_norm": 1.48247240455576,
+      "learning_rate": 2.743902975115754e-05,
+      "loss": 1.0387,
+      "step": 16621
+    },
+    {
+      "epoch": 0.7657437692910121,
+      "grad_norm": 0.7193323923565097,
+      "learning_rate": 2.7428763399881964e-05,
+      "loss": 1.43,
+      "step": 16622
+    },
+    {
+      "epoch": 0.7657898373796471,
+      "grad_norm": 1.0177730717106215,
+      "learning_rate": 2.7418498664280513e-05,
+      "loss": 1.4275,
+      "step": 16623
+    },
+    {
+      "epoch": 0.7658359054682822,
+      "grad_norm": 1.3034324086823743,
+      "learning_rate": 2.740823554458174e-05,
+      "loss": 1.3679,
+      "step": 16624
+    },
+    {
+      "epoch": 0.7658819735569171,
+      "grad_norm": 1.4510753464678992,
+      "learning_rate": 2.7397974041014128e-05,
+      "loss": 1.319,
+      "step": 16625
+    },
+    {
+      "epoch": 0.7659280416455522,
+      "grad_norm": 2.2347077732879903,
+      "learning_rate": 2.7387714153806178e-05,
+      "loss": 1.3472,
+      "step": 16626
+    },
+    {
+      "epoch": 0.7659741097341871,
+      "grad_norm": 0.8722247355649437,
+      "learning_rate": 2.737745588318623e-05,
+      "loss": 1.1289,
+      "step": 16627
+    },
+    {
+      "epoch": 0.7660201778228222,
+      "grad_norm": 1.6777769457564935,
+      "learning_rate": 2.736719922938271e-05,
+      "loss": 1.4405,
+      "step": 16628
+    },
+    {
+      "epoch": 0.7660662459114571,
+      "grad_norm": 1.620338208470096,
+      "learning_rate": 2.735694419262399e-05,
+      "loss": 1.5769,
+      "step": 16629
+    },
+    {
+      "epoch": 0.7661123140000922,
+      "grad_norm": 0.8949511617481605,
+      "learning_rate": 2.7346690773138318e-05,
+      "loss": 1.4237,
+      "step": 16630
+    },
+    {
+      "epoch": 0.7661583820887271,
+      "grad_norm": 1.5591042673902662,
+      "learning_rate": 2.733643897115401e-05,
+      "loss": 1.3344,
+      "step": 16631
+    },
+    {
+      "epoch": 0.7662044501773622,
+      "grad_norm": 1.3290899020315765,
+      "learning_rate": 2.7326188786899286e-05,
+      "loss": 1.0703,
+      "step": 16632
+    },
+    {
+      "epoch": 0.7662505182659971,
+      "grad_norm": 1.6436828216827173,
+      "learning_rate": 2.73159402206024e-05,
+      "loss": 1.5118,
+      "step": 16633
+    },
+    {
+      "epoch": 0.7662965863546322,
+      "grad_norm": 1.1581681604390615,
+      "learning_rate": 2.730569327249144e-05,
+      "loss": 1.3429,
+      "step": 16634
+    },
+    {
+      "epoch": 0.7663426544432671,
+      "grad_norm": 1.6234501173991622,
+      "learning_rate": 2.7295447942794584e-05,
+      "loss": 1.3288,
+      "step": 16635
+    },
+    {
+      "epoch": 0.7663887225319022,
+      "grad_norm": 1.5986208690849786,
+      "learning_rate": 2.7285204231739915e-05,
+      "loss": 0.9484,
+      "step": 16636
+    },
+    {
+      "epoch": 0.7664347906205371,
+      "grad_norm": 1.1942372923925395,
+      "learning_rate": 2.7274962139555527e-05,
+      "loss": 1.4879,
+      "step": 16637
+    },
+    {
+      "epoch": 0.7664808587091722,
+      "grad_norm": 2.3245955931815185,
+      "learning_rate": 2.7264721666469373e-05,
+      "loss": 1.1007,
+      "step": 16638
+    },
+    {
+      "epoch": 0.7665269267978072,
+      "grad_norm": 2.3231193366600857,
+      "learning_rate": 2.725448281270948e-05,
+      "loss": 1.4159,
+      "step": 16639
+    },
+    {
+      "epoch": 0.7665729948864421,
+      "grad_norm": 1.0363097090087485,
+      "learning_rate": 2.7244245578503824e-05,
+      "loss": 1.3192,
+      "step": 16640
+    },
+    {
+      "epoch": 0.7666190629750772,
+      "grad_norm": 1.5150471641028576,
+      "learning_rate": 2.7234009964080265e-05,
+      "loss": 1.1958,
+      "step": 16641
+    },
+    {
+      "epoch": 0.7666651310637121,
+      "grad_norm": 0.8996274851084772,
+      "learning_rate": 2.722377596966671e-05,
+      "loss": 1.2462,
+      "step": 16642
+    },
+    {
+      "epoch": 0.7667111991523472,
+      "grad_norm": 1.3624258663729436,
+      "learning_rate": 2.7213543595490986e-05,
+      "loss": 1.2605,
+      "step": 16643
+    },
+    {
+      "epoch": 0.7667572672409821,
+      "grad_norm": 1.6659101865059727,
+      "learning_rate": 2.7203312841780947e-05,
+      "loss": 1.1848,
+      "step": 16644
+    },
+    {
+      "epoch": 0.7668033353296172,
+      "grad_norm": 3.5140001434206884,
+      "learning_rate": 2.719308370876431e-05,
+      "loss": 1.0756,
+      "step": 16645
+    },
+    {
+      "epoch": 0.7668494034182521,
+      "grad_norm": 0.9277641821867996,
+      "learning_rate": 2.7182856196668816e-05,
+      "loss": 1.1225,
+      "step": 16646
+    },
+    {
+      "epoch": 0.7668954715068872,
+      "grad_norm": 2.4954448077293123,
+      "learning_rate": 2.71726303057222e-05,
+      "loss": 1.4532,
+      "step": 16647
+    },
+    {
+      "epoch": 0.7669415395955221,
+      "grad_norm": 2.631119792097606,
+      "learning_rate": 2.7162406036152077e-05,
+      "loss": 1.4742,
+      "step": 16648
+    },
+    {
+      "epoch": 0.7669876076841572,
+      "grad_norm": 1.2483218111441274,
+      "learning_rate": 2.7152183388186093e-05,
+      "loss": 1.2674,
+      "step": 16649
+    },
+    {
+      "epoch": 0.7670336757727921,
+      "grad_norm": 3.531727920194822,
+      "learning_rate": 2.714196236205183e-05,
+      "loss": 1.2042,
+      "step": 16650
+    },
+    {
+      "epoch": 0.7670797438614272,
+      "grad_norm": 0.9823264622086358,
+      "learning_rate": 2.713174295797689e-05,
+      "loss": 1.113,
+      "step": 16651
+    },
+    {
+      "epoch": 0.7671258119500622,
+      "grad_norm": 0.7652498276510155,
+      "learning_rate": 2.712152517618872e-05,
+      "loss": 1.3033,
+      "step": 16652
+    },
+    {
+      "epoch": 0.7671718800386972,
+      "grad_norm": 0.9838216846800893,
+      "learning_rate": 2.711130901691483e-05,
+      "loss": 1.3258,
+      "step": 16653
+    },
+    {
+      "epoch": 0.7672179481273322,
+      "grad_norm": 1.0334406644522827,
+      "learning_rate": 2.710109448038266e-05,
+      "loss": 1.325,
+      "step": 16654
+    },
+    {
+      "epoch": 0.7672640162159672,
+      "grad_norm": 4.153384296400725,
+      "learning_rate": 2.709088156681967e-05,
+      "loss": 1.6385,
+      "step": 16655
+    },
+    {
+      "epoch": 0.7673100843046022,
+      "grad_norm": 0.7899072772679001,
+      "learning_rate": 2.7080670276453157e-05,
+      "loss": 1.1532,
+      "step": 16656
+    },
+    {
+      "epoch": 0.7673561523932372,
+      "grad_norm": 1.5361630527105243,
+      "learning_rate": 2.7070460609510484e-05,
+      "loss": 1.1232,
+      "step": 16657
+    },
+    {
+      "epoch": 0.7674022204818722,
+      "grad_norm": 1.4953690844581211,
+      "learning_rate": 2.7060252566218992e-05,
+      "loss": 1.3067,
+      "step": 16658
+    },
+    {
+      "epoch": 0.7674482885705072,
+      "grad_norm": 2.016279198946354,
+      "learning_rate": 2.7050046146805873e-05,
+      "loss": 1.1414,
+      "step": 16659
+    },
+    {
+      "epoch": 0.7674943566591422,
+      "grad_norm": 0.8304613420508642,
+      "learning_rate": 2.7039841351498406e-05,
+      "loss": 1.1697,
+      "step": 16660
+    },
+    {
+      "epoch": 0.7675404247477772,
+      "grad_norm": 1.026690636888629,
+      "learning_rate": 2.7029638180523765e-05,
+      "loss": 1.3804,
+      "step": 16661
+    },
+    {
+      "epoch": 0.7675864928364122,
+      "grad_norm": 2.034687715184302,
+      "learning_rate": 2.7019436634109152e-05,
+      "loss": 1.3586,
+      "step": 16662
+    },
+    {
+      "epoch": 0.7676325609250472,
+      "grad_norm": 1.9581353471756475,
+      "learning_rate": 2.70092367124816e-05,
+      "loss": 1.2484,
+      "step": 16663
+    },
+    {
+      "epoch": 0.7676786290136822,
+      "grad_norm": 1.0898691563836986,
+      "learning_rate": 2.699903841586826e-05,
+      "loss": 1.3432,
+      "step": 16664
+    },
+    {
+      "epoch": 0.7677246971023173,
+      "grad_norm": 1.4227536559079779,
+      "learning_rate": 2.698884174449614e-05,
+      "loss": 1.3307,
+      "step": 16665
+    },
+    {
+      "epoch": 0.7677707651909522,
+      "grad_norm": 1.1426729612404607,
+      "learning_rate": 2.6978646698592312e-05,
+      "loss": 1.3402,
+      "step": 16666
+    },
+    {
+      "epoch": 0.7678168332795873,
+      "grad_norm": 3.5644781924293003,
+      "learning_rate": 2.6968453278383686e-05,
+      "loss": 1.4711,
+      "step": 16667
+    },
+    {
+      "epoch": 0.7678629013682222,
+      "grad_norm": 1.7108971803855035,
+      "learning_rate": 2.6958261484097224e-05,
+      "loss": 1.0391,
+      "step": 16668
+    },
+    {
+      "epoch": 0.7679089694568573,
+      "grad_norm": 0.9045742804919537,
+      "learning_rate": 2.6948071315959855e-05,
+      "loss": 1.2385,
+      "step": 16669
+    },
+    {
+      "epoch": 0.7679550375454922,
+      "grad_norm": 1.280305445137589,
+      "learning_rate": 2.6937882774198386e-05,
+      "loss": 1.6466,
+      "step": 16670
+    },
+    {
+      "epoch": 0.7680011056341273,
+      "grad_norm": 1.0948466961690233,
+      "learning_rate": 2.6927695859039693e-05,
+      "loss": 1.2515,
+      "step": 16671
+    },
+    {
+      "epoch": 0.7680471737227622,
+      "grad_norm": 2.058707502965718,
+      "learning_rate": 2.6917510570710558e-05,
+      "loss": 1.2818,
+      "step": 16672
+    },
+    {
+      "epoch": 0.7680932418113973,
+      "grad_norm": 2.776935307099664,
+      "learning_rate": 2.6907326909437758e-05,
+      "loss": 1.2085,
+      "step": 16673
+    },
+    {
+      "epoch": 0.7681393099000322,
+      "grad_norm": 1.16987232941411,
+      "learning_rate": 2.6897144875447976e-05,
+      "loss": 1.1849,
+      "step": 16674
+    },
+    {
+      "epoch": 0.7681853779886673,
+      "grad_norm": 5.334414550255211,
+      "learning_rate": 2.6886964468967923e-05,
+      "loss": 1.3486,
+      "step": 16675
+    },
+    {
+      "epoch": 0.7682314460773022,
+      "grad_norm": 1.475107828848532,
+      "learning_rate": 2.687678569022427e-05,
+      "loss": 1.1901,
+      "step": 16676
+    },
+    {
+      "epoch": 0.7682775141659373,
+      "grad_norm": 1.0902505198874526,
+      "learning_rate": 2.6866608539443585e-05,
+      "loss": 1.4127,
+      "step": 16677
+    },
+    {
+      "epoch": 0.7683235822545723,
+      "grad_norm": 1.6078985146925544,
+      "learning_rate": 2.6856433016852445e-05,
+      "loss": 1.223,
+      "step": 16678
+    },
+    {
+      "epoch": 0.7683696503432073,
+      "grad_norm": 0.869027519697439,
+      "learning_rate": 2.684625912267742e-05,
+      "loss": 1.2489,
+      "step": 16679
+    },
+    {
+      "epoch": 0.7684157184318423,
+      "grad_norm": 1.368292344924616,
+      "learning_rate": 2.6836086857145037e-05,
+      "loss": 1.4562,
+      "step": 16680
+    },
+    {
+      "epoch": 0.7684617865204773,
+      "grad_norm": 0.9297010201839122,
+      "learning_rate": 2.6825916220481694e-05,
+      "loss": 1.2329,
+      "step": 16681
+    },
+    {
+      "epoch": 0.7685078546091123,
+      "grad_norm": 1.5058792334174953,
+      "learning_rate": 2.6815747212913854e-05,
+      "loss": 1.5332,
+      "step": 16682
+    },
+    {
+      "epoch": 0.7685539226977472,
+      "grad_norm": 1.1977364385049796,
+      "learning_rate": 2.6805579834667925e-05,
+      "loss": 1.365,
+      "step": 16683
+    },
+    {
+      "epoch": 0.7685999907863823,
+      "grad_norm": 1.5531010026485024,
+      "learning_rate": 2.6795414085970293e-05,
+      "loss": 1.4165,
+      "step": 16684
+    },
+    {
+      "epoch": 0.7686460588750172,
+      "grad_norm": 1.1118469593529354,
+      "learning_rate": 2.6785249967047253e-05,
+      "loss": 1.1882,
+      "step": 16685
+    },
+    {
+      "epoch": 0.7686921269636523,
+      "grad_norm": 1.3587066657784683,
+      "learning_rate": 2.6775087478125015e-05,
+      "loss": 0.9959,
+      "step": 16686
+    },
+    {
+      "epoch": 0.7687381950522872,
+      "grad_norm": 0.9683439449074152,
+      "learning_rate": 2.6764926619429963e-05,
+      "loss": 1.1939,
+      "step": 16687
+    },
+    {
+      "epoch": 0.7687842631409223,
+      "grad_norm": 0.931984865959757,
+      "learning_rate": 2.6754767391188218e-05,
+      "loss": 1.15,
+      "step": 16688
+    },
+    {
+      "epoch": 0.7688303312295572,
+      "grad_norm": 1.8942191030245024,
+      "learning_rate": 2.6744609793625984e-05,
+      "loss": 0.9865,
+      "step": 16689
+    },
+    {
+      "epoch": 0.7688763993181923,
+      "grad_norm": 1.9464460832264787,
+      "learning_rate": 2.6734453826969418e-05,
+      "loss": 1.391,
+      "step": 16690
+    },
+    {
+      "epoch": 0.7689224674068272,
+      "grad_norm": 1.8960044967359406,
+      "learning_rate": 2.6724299491444638e-05,
+      "loss": 1.37,
+      "step": 16691
+    },
+    {
+      "epoch": 0.7689685354954623,
+      "grad_norm": 0.9109210544359576,
+      "learning_rate": 2.671414678727766e-05,
+      "loss": 1.3942,
+      "step": 16692
+    },
+    {
+      "epoch": 0.7690146035840973,
+      "grad_norm": 1.2150057473375173,
+      "learning_rate": 2.6703995714694553e-05,
+      "loss": 1.2815,
+      "step": 16693
+    },
+    {
+      "epoch": 0.7690606716727323,
+      "grad_norm": 2.1344699193545322,
+      "learning_rate": 2.669384627392132e-05,
+      "loss": 1.6066,
+      "step": 16694
+    },
+    {
+      "epoch": 0.7691067397613673,
+      "grad_norm": 1.109702691845269,
+      "learning_rate": 2.6683698465183882e-05,
+      "loss": 1.3151,
+      "step": 16695
+    },
+    {
+      "epoch": 0.7691528078500023,
+      "grad_norm": 0.8798045024555846,
+      "learning_rate": 2.6673552288708226e-05,
+      "loss": 1.3029,
+      "step": 16696
+    },
+    {
+      "epoch": 0.7691988759386373,
+      "grad_norm": 1.409762057470914,
+      "learning_rate": 2.6663407744720124e-05,
+      "loss": 1.2753,
+      "step": 16697
+    },
+    {
+      "epoch": 0.7692449440272723,
+      "grad_norm": 4.657539523249302,
+      "learning_rate": 2.665326483344558e-05,
+      "loss": 1.1801,
+      "step": 16698
+    },
+    {
+      "epoch": 0.7692910121159073,
+      "grad_norm": 1.375026852668946,
+      "learning_rate": 2.6643123555110295e-05,
+      "loss": 1.2286,
+      "step": 16699
+    },
+    {
+      "epoch": 0.7693370802045423,
+      "grad_norm": 0.9130545940876295,
+      "learning_rate": 2.6632983909940113e-05,
+      "loss": 1.2548,
+      "step": 16700
+    },
+    {
+      "epoch": 0.7693831482931773,
+      "grad_norm": 0.9000720710100135,
+      "learning_rate": 2.6622845898160688e-05,
+      "loss": 1.5118,
+      "step": 16701
+    },
+    {
+      "epoch": 0.7694292163818123,
+      "grad_norm": 1.3230865294746281,
+      "learning_rate": 2.6612709519997848e-05,
+      "loss": 1.2171,
+      "step": 16702
+    },
+    {
+      "epoch": 0.7694752844704473,
+      "grad_norm": 1.4251457331829156,
+      "learning_rate": 2.660257477567719e-05,
+      "loss": 1.1975,
+      "step": 16703
+    },
+    {
+      "epoch": 0.7695213525590823,
+      "grad_norm": 1.0283506445302455,
+      "learning_rate": 2.6592441665424294e-05,
+      "loss": 1.2011,
+      "step": 16704
+    },
+    {
+      "epoch": 0.7695674206477173,
+      "grad_norm": 1.0324577898030884,
+      "learning_rate": 2.6582310189464887e-05,
+      "loss": 1.0394,
+      "step": 16705
+    },
+    {
+      "epoch": 0.7696134887363524,
+      "grad_norm": 1.4489270037069075,
+      "learning_rate": 2.6572180348024423e-05,
+      "loss": 0.9395,
+      "step": 16706
+    },
+    {
+      "epoch": 0.7696595568249873,
+      "grad_norm": 1.4901593227563426,
+      "learning_rate": 2.6562052141328497e-05,
+      "loss": 1.2784,
+      "step": 16707
+    },
+    {
+      "epoch": 0.7697056249136224,
+      "grad_norm": 1.7193967149618496,
+      "learning_rate": 2.65519255696025e-05,
+      "loss": 1.1276,
+      "step": 16708
+    },
+    {
+      "epoch": 0.7697516930022573,
+      "grad_norm": 1.2223188776873062,
+      "learning_rate": 2.6541800633072002e-05,
+      "loss": 1.4759,
+      "step": 16709
+    },
+    {
+      "epoch": 0.7697977610908924,
+      "grad_norm": 1.6527374636902519,
+      "learning_rate": 2.6531677331962325e-05,
+      "loss": 1.5289,
+      "step": 16710
+    },
+    {
+      "epoch": 0.7698438291795273,
+      "grad_norm": 0.9403302478136025,
+      "learning_rate": 2.652155566649892e-05,
+      "loss": 1.1895,
+      "step": 16711
+    },
+    {
+      "epoch": 0.7698898972681624,
+      "grad_norm": 2.7174516934367743,
+      "learning_rate": 2.6511435636907045e-05,
+      "loss": 1.1544,
+      "step": 16712
+    },
+    {
+      "epoch": 0.7699359653567973,
+      "grad_norm": 2.765416660253985,
+      "learning_rate": 2.650131724341206e-05,
+      "loss": 1.5635,
+      "step": 16713
+    },
+    {
+      "epoch": 0.7699820334454324,
+      "grad_norm": 1.5852859008987945,
+      "learning_rate": 2.6491200486239255e-05,
+      "loss": 1.426,
+      "step": 16714
+    },
+    {
+      "epoch": 0.7700281015340673,
+      "grad_norm": 1.0866182371247335,
+      "learning_rate": 2.6481085365613767e-05,
+      "loss": 1.4043,
+      "step": 16715
+    },
+    {
+      "epoch": 0.7700741696227024,
+      "grad_norm": 0.9227974611192712,
+      "learning_rate": 2.647097188176092e-05,
+      "loss": 1.4416,
+      "step": 16716
+    },
+    {
+      "epoch": 0.7701202377113373,
+      "grad_norm": 0.9599541239795361,
+      "learning_rate": 2.6460860034905777e-05,
+      "loss": 1.4481,
+      "step": 16717
+    },
+    {
+      "epoch": 0.7701663057999724,
+      "grad_norm": 11.123699396823229,
+      "learning_rate": 2.6450749825273513e-05,
+      "loss": 1.1336,
+      "step": 16718
+    },
+    {
+      "epoch": 0.7702123738886074,
+      "grad_norm": 2.126376204941333,
+      "learning_rate": 2.644064125308915e-05,
+      "loss": 0.9523,
+      "step": 16719
+    },
+    {
+      "epoch": 0.7702584419772424,
+      "grad_norm": 0.8926426027964774,
+      "learning_rate": 2.6430534318577826e-05,
+      "loss": 1.4808,
+      "step": 16720
+    },
+    {
+      "epoch": 0.7703045100658774,
+      "grad_norm": 0.959040157377365,
+      "learning_rate": 2.642042902196452e-05,
+      "loss": 1.2769,
+      "step": 16721
+    },
+    {
+      "epoch": 0.7703505781545124,
+      "grad_norm": 1.6495931061752944,
+      "learning_rate": 2.6410325363474176e-05,
+      "loss": 1.267,
+      "step": 16722
+    },
+    {
+      "epoch": 0.7703966462431474,
+      "grad_norm": 3.6993124187846695,
+      "learning_rate": 2.6400223343331752e-05,
+      "loss": 1.1037,
+      "step": 16723
+    },
+    {
+      "epoch": 0.7704427143317824,
+      "grad_norm": 0.984656327268573,
+      "learning_rate": 2.6390122961762154e-05,
+      "loss": 1.4133,
+      "step": 16724
+    },
+    {
+      "epoch": 0.7704887824204174,
+      "grad_norm": 2.16448263738957,
+      "learning_rate": 2.6380024218990286e-05,
+      "loss": 0.88,
+      "step": 16725
+    },
+    {
+      "epoch": 0.7705348505090523,
+      "grad_norm": 1.5390670476026256,
+      "learning_rate": 2.6369927115240923e-05,
+      "loss": 1.2296,
+      "step": 16726
+    },
+    {
+      "epoch": 0.7705809185976874,
+      "grad_norm": 1.706063669043385,
+      "learning_rate": 2.6359831650738885e-05,
+      "loss": 1.0852,
+      "step": 16727
+    },
+    {
+      "epoch": 0.7706269866863223,
+      "grad_norm": 7.716765160577941,
+      "learning_rate": 2.634973782570892e-05,
+      "loss": 1.2749,
+      "step": 16728
+    },
+    {
+      "epoch": 0.7706730547749574,
+      "grad_norm": 1.6854991926609761,
+      "learning_rate": 2.6339645640375798e-05,
+      "loss": 1.2774,
+      "step": 16729
+    },
+    {
+      "epoch": 0.7707191228635923,
+      "grad_norm": 0.8452682747225775,
+      "learning_rate": 2.632955509496413e-05,
+      "loss": 1.4428,
+      "step": 16730
+    },
+    {
+      "epoch": 0.7707651909522274,
+      "grad_norm": 1.2273336837409798,
+      "learning_rate": 2.6319466189698603e-05,
+      "loss": 1.0392,
+      "step": 16731
+    },
+    {
+      "epoch": 0.7708112590408623,
+      "grad_norm": 2.404401758961928,
+      "learning_rate": 2.6309378924803862e-05,
+      "loss": 0.8243,
+      "step": 16732
+    },
+    {
+      "epoch": 0.7708573271294974,
+      "grad_norm": 1.520097481742834,
+      "learning_rate": 2.62992933005044e-05,
+      "loss": 0.891,
+      "step": 16733
+    },
+    {
+      "epoch": 0.7709033952181324,
+      "grad_norm": 1.3793212003341555,
+      "learning_rate": 2.6289209317024822e-05,
+      "loss": 1.102,
+      "step": 16734
+    },
+    {
+      "epoch": 0.7709494633067674,
+      "grad_norm": 1.2390998401414655,
+      "learning_rate": 2.627912697458961e-05,
+      "loss": 1.3106,
+      "step": 16735
+    },
+    {
+      "epoch": 0.7709955313954024,
+      "grad_norm": 1.117677775518331,
+      "learning_rate": 2.6269046273423246e-05,
+      "loss": 1.3348,
+      "step": 16736
+    },
+    {
+      "epoch": 0.7710415994840374,
+      "grad_norm": 1.3710257495672233,
+      "learning_rate": 2.625896721375013e-05,
+      "loss": 1.1552,
+      "step": 16737
+    },
+    {
+      "epoch": 0.7710876675726724,
+      "grad_norm": 2.1361187147066127,
+      "learning_rate": 2.6248889795794662e-05,
+      "loss": 1.2029,
+      "step": 16738
+    },
+    {
+      "epoch": 0.7711337356613074,
+      "grad_norm": 1.0393035187339528,
+      "learning_rate": 2.6238814019781243e-05,
+      "loss": 1.2784,
+      "step": 16739
+    },
+    {
+      "epoch": 0.7711798037499424,
+      "grad_norm": 1.2932633207288635,
+      "learning_rate": 2.6228739885934116e-05,
+      "loss": 1.3339,
+      "step": 16740
+    },
+    {
+      "epoch": 0.7712258718385774,
+      "grad_norm": 1.3596335914983642,
+      "learning_rate": 2.6218667394477613e-05,
+      "loss": 1.2191,
+      "step": 16741
+    },
+    {
+      "epoch": 0.7712719399272124,
+      "grad_norm": 1.6920817824595116,
+      "learning_rate": 2.6208596545635976e-05,
+      "loss": 1.0613,
+      "step": 16742
+    },
+    {
+      "epoch": 0.7713180080158474,
+      "grad_norm": 2.499712929568243,
+      "learning_rate": 2.619852733963344e-05,
+      "loss": 1.0581,
+      "step": 16743
+    },
+    {
+      "epoch": 0.7713640761044824,
+      "grad_norm": 1.1660321855706384,
+      "learning_rate": 2.6188459776694108e-05,
+      "loss": 1.1393,
+      "step": 16744
+    },
+    {
+      "epoch": 0.7714101441931174,
+      "grad_norm": 3.307245231442703,
+      "learning_rate": 2.617839385704217e-05,
+      "loss": 1.291,
+      "step": 16745
+    },
+    {
+      "epoch": 0.7714562122817524,
+      "grad_norm": 1.0789691355166457,
+      "learning_rate": 2.6168329580901718e-05,
+      "loss": 1.2853,
+      "step": 16746
+    },
+    {
+      "epoch": 0.7715022803703875,
+      "grad_norm": 0.9995657074567651,
+      "learning_rate": 2.6158266948496836e-05,
+      "loss": 1.2188,
+      "step": 16747
+    },
+    {
+      "epoch": 0.7715483484590224,
+      "grad_norm": 0.8929719465255975,
+      "learning_rate": 2.614820596005151e-05,
+      "loss": 1.595,
+      "step": 16748
+    },
+    {
+      "epoch": 0.7715944165476575,
+      "grad_norm": 1.1886251135698085,
+      "learning_rate": 2.613814661578974e-05,
+      "loss": 1.4948,
+      "step": 16749
+    },
+    {
+      "epoch": 0.7716404846362924,
+      "grad_norm": 1.0546874138027549,
+      "learning_rate": 2.6128088915935522e-05,
+      "loss": 1.1156,
+      "step": 16750
+    },
+    {
+      "epoch": 0.7716865527249275,
+      "grad_norm": 1.6632683072165533,
+      "learning_rate": 2.6118032860712716e-05,
+      "loss": 1.2569,
+      "step": 16751
+    },
+    {
+      "epoch": 0.7717326208135624,
+      "grad_norm": 1.8676752148582076,
+      "learning_rate": 2.6107978450345217e-05,
+      "loss": 1.2682,
+      "step": 16752
+    },
+    {
+      "epoch": 0.7717786889021975,
+      "grad_norm": 1.3701202414694997,
+      "learning_rate": 2.609792568505689e-05,
+      "loss": 1.4233,
+      "step": 16753
+    },
+    {
+      "epoch": 0.7718247569908324,
+      "grad_norm": 1.7566872334666526,
+      "learning_rate": 2.6087874565071556e-05,
+      "loss": 1.1096,
+      "step": 16754
+    },
+    {
+      "epoch": 0.7718708250794675,
+      "grad_norm": 5.032790442345844,
+      "learning_rate": 2.607782509061294e-05,
+      "loss": 1.2593,
+      "step": 16755
+    },
+    {
+      "epoch": 0.7719168931681024,
+      "grad_norm": 0.8932373431418812,
+      "learning_rate": 2.606777726190479e-05,
+      "loss": 1.2401,
+      "step": 16756
+    },
+    {
+      "epoch": 0.7719629612567375,
+      "grad_norm": 1.383935506260883,
+      "learning_rate": 2.6057731079170856e-05,
+      "loss": 1.4255,
+      "step": 16757
+    },
+    {
+      "epoch": 0.7720090293453724,
+      "grad_norm": 0.8401177332694791,
+      "learning_rate": 2.604768654263472e-05,
+      "loss": 1.1192,
+      "step": 16758
+    },
+    {
+      "epoch": 0.7720550974340075,
+      "grad_norm": 1.141250131264426,
+      "learning_rate": 2.6037643652520037e-05,
+      "loss": 1.1435,
+      "step": 16759
+    },
+    {
+      "epoch": 0.7721011655226425,
+      "grad_norm": 0.9733742329999903,
+      "learning_rate": 2.60276024090504e-05,
+      "loss": 1.3215,
+      "step": 16760
+    },
+    {
+      "epoch": 0.7721472336112775,
+      "grad_norm": 0.7717810049712955,
+      "learning_rate": 2.601756281244939e-05,
+      "loss": 1.2831,
+      "step": 16761
+    },
+    {
+      "epoch": 0.7721933016999125,
+      "grad_norm": 1.3838262731214859,
+      "learning_rate": 2.600752486294047e-05,
+      "loss": 1.1786,
+      "step": 16762
+    },
+    {
+      "epoch": 0.7722393697885475,
+      "grad_norm": 1.641102680627331,
+      "learning_rate": 2.599748856074713e-05,
+      "loss": 1.0388,
+      "step": 16763
+    },
+    {
+      "epoch": 0.7722854378771825,
+      "grad_norm": 1.7443801895159745,
+      "learning_rate": 2.598745390609283e-05,
+      "loss": 1.3749,
+      "step": 16764
+    },
+    {
+      "epoch": 0.7723315059658175,
+      "grad_norm": 1.7371171967633645,
+      "learning_rate": 2.5977420899200976e-05,
+      "loss": 1.3679,
+      "step": 16765
+    },
+    {
+      "epoch": 0.7723775740544525,
+      "grad_norm": 1.6024566776891391,
+      "learning_rate": 2.596738954029492e-05,
+      "loss": 1.4425,
+      "step": 16766
+    },
+    {
+      "epoch": 0.7724236421430875,
+      "grad_norm": 0.9867698489061957,
+      "learning_rate": 2.595735982959797e-05,
+      "loss": 1.3802,
+      "step": 16767
+    },
+    {
+      "epoch": 0.7724697102317225,
+      "grad_norm": 1.703601850234639,
+      "learning_rate": 2.5947331767333492e-05,
+      "loss": 1.5107,
+      "step": 16768
+    },
+    {
+      "epoch": 0.7725157783203574,
+      "grad_norm": 1.0354462114745364,
+      "learning_rate": 2.5937305353724672e-05,
+      "loss": 1.1186,
+      "step": 16769
+    },
+    {
+      "epoch": 0.7725618464089925,
+      "grad_norm": 1.399563032152242,
+      "learning_rate": 2.592728058899475e-05,
+      "loss": 1.1195,
+      "step": 16770
+    },
+    {
+      "epoch": 0.7726079144976274,
+      "grad_norm": 1.4892848311244484,
+      "learning_rate": 2.591725747336693e-05,
+      "loss": 1.4613,
+      "step": 16771
+    },
+    {
+      "epoch": 0.7726539825862625,
+      "grad_norm": 0.883959272539984,
+      "learning_rate": 2.5907236007064363e-05,
+      "loss": 1.3003,
+      "step": 16772
+    },
+    {
+      "epoch": 0.7727000506748976,
+      "grad_norm": 0.7393238367061029,
+      "learning_rate": 2.5897216190310126e-05,
+      "loss": 1.5488,
+      "step": 16773
+    },
+    {
+      "epoch": 0.7727461187635325,
+      "grad_norm": 1.0349005292143743,
+      "learning_rate": 2.5887198023327297e-05,
+      "loss": 1.4255,
+      "step": 16774
+    },
+    {
+      "epoch": 0.7727921868521676,
+      "grad_norm": 1.3123390066276766,
+      "learning_rate": 2.5877181506338977e-05,
+      "loss": 1.0881,
+      "step": 16775
+    },
+    {
+      "epoch": 0.7728382549408025,
+      "grad_norm": 1.4452188760187048,
+      "learning_rate": 2.5867166639568075e-05,
+      "loss": 1.5779,
+      "step": 16776
+    },
+    {
+      "epoch": 0.7728843230294375,
+      "grad_norm": 1.4257712401325426,
+      "learning_rate": 2.5857153423237602e-05,
+      "loss": 1.2844,
+      "step": 16777
+    },
+    {
+      "epoch": 0.7729303911180725,
+      "grad_norm": 0.864275520915378,
+      "learning_rate": 2.584714185757048e-05,
+      "loss": 1.2176,
+      "step": 16778
+    },
+    {
+      "epoch": 0.7729764592067075,
+      "grad_norm": 13.428552198152895,
+      "learning_rate": 2.5837131942789617e-05,
+      "loss": 1.1935,
+      "step": 16779
+    },
+    {
+      "epoch": 0.7730225272953425,
+      "grad_norm": 2.333185009806196,
+      "learning_rate": 2.582712367911784e-05,
+      "loss": 1.0954,
+      "step": 16780
+    },
+    {
+      "epoch": 0.7730685953839775,
+      "grad_norm": 1.247105766577171,
+      "learning_rate": 2.5817117066777964e-05,
+      "loss": 1.2163,
+      "step": 16781
+    },
+    {
+      "epoch": 0.7731146634726125,
+      "grad_norm": 1.0061080253111563,
+      "learning_rate": 2.5807112105992782e-05,
+      "loss": 1.3311,
+      "step": 16782
+    },
+    {
+      "epoch": 0.7731607315612475,
+      "grad_norm": 2.5524879713575106,
+      "learning_rate": 2.579710879698506e-05,
+      "loss": 1.2759,
+      "step": 16783
+    },
+    {
+      "epoch": 0.7732067996498825,
+      "grad_norm": 1.5957335248155209,
+      "learning_rate": 2.578710713997745e-05,
+      "loss": 1.0942,
+      "step": 16784
+    },
+    {
+      "epoch": 0.7732528677385175,
+      "grad_norm": 1.8206365810567626,
+      "learning_rate": 2.577710713519266e-05,
+      "loss": 1.106,
+      "step": 16785
+    },
+    {
+      "epoch": 0.7732989358271525,
+      "grad_norm": 1.4619439637595544,
+      "learning_rate": 2.5767108782853343e-05,
+      "loss": 1.371,
+      "step": 16786
+    },
+    {
+      "epoch": 0.7733450039157875,
+      "grad_norm": 1.6004102928868498,
+      "learning_rate": 2.5757112083182034e-05,
+      "loss": 1.1534,
+      "step": 16787
+    },
+    {
+      "epoch": 0.7733910720044226,
+      "grad_norm": 0.9916980677597436,
+      "learning_rate": 2.5747117036401325e-05,
+      "loss": 1.6668,
+      "step": 16788
+    },
+    {
+      "epoch": 0.7734371400930575,
+      "grad_norm": 0.8557250579965319,
+      "learning_rate": 2.573712364273374e-05,
+      "loss": 1.2165,
+      "step": 16789
+    },
+    {
+      "epoch": 0.7734832081816926,
+      "grad_norm": 0.8316696777705214,
+      "learning_rate": 2.572713190240179e-05,
+      "loss": 1.609,
+      "step": 16790
+    },
+    {
+      "epoch": 0.7735292762703275,
+      "grad_norm": 1.574464834103629,
+      "learning_rate": 2.5717141815627876e-05,
+      "loss": 0.9409,
+      "step": 16791
+    },
+    {
+      "epoch": 0.7735753443589626,
+      "grad_norm": 2.645893118199831,
+      "learning_rate": 2.570715338263443e-05,
+      "loss": 1.3316,
+      "step": 16792
+    },
+    {
+      "epoch": 0.7736214124475975,
+      "grad_norm": 0.9874895439231829,
+      "learning_rate": 2.569716660364382e-05,
+      "loss": 1.5214,
+      "step": 16793
+    },
+    {
+      "epoch": 0.7736674805362326,
+      "grad_norm": 1.301320592952614,
+      "learning_rate": 2.568718147887843e-05,
+      "loss": 1.5415,
+      "step": 16794
+    },
+    {
+      "epoch": 0.7737135486248675,
+      "grad_norm": 0.8378964296224908,
+      "learning_rate": 2.567719800856052e-05,
+      "loss": 1.2032,
+      "step": 16795
+    },
+    {
+      "epoch": 0.7737596167135026,
+      "grad_norm": 1.0249142233658406,
+      "learning_rate": 2.5667216192912314e-05,
+      "loss": 1.1432,
+      "step": 16796
+    },
+    {
+      "epoch": 0.7738056848021375,
+      "grad_norm": 1.0834177899480835,
+      "learning_rate": 2.565723603215613e-05,
+      "loss": 1.0032,
+      "step": 16797
+    },
+    {
+      "epoch": 0.7738517528907726,
+      "grad_norm": 1.3530225946873664,
+      "learning_rate": 2.564725752651409e-05,
+      "loss": 1.563,
+      "step": 16798
+    },
+    {
+      "epoch": 0.7738978209794075,
+      "grad_norm": 6.2955724837146905,
+      "learning_rate": 2.563728067620841e-05,
+      "loss": 1.176,
+      "step": 16799
+    },
+    {
+      "epoch": 0.7739438890680426,
+      "grad_norm": 2.067653579539905,
+      "learning_rate": 2.5627305481461105e-05,
+      "loss": 1.1623,
+      "step": 16800
+    },
+    {
+      "epoch": 0.7739899571566776,
+      "grad_norm": 2.0448625139919683,
+      "learning_rate": 2.561733194249438e-05,
+      "loss": 1.196,
+      "step": 16801
+    },
+    {
+      "epoch": 0.7740360252453126,
+      "grad_norm": 1.1396211708238713,
+      "learning_rate": 2.5607360059530196e-05,
+      "loss": 1.344,
+      "step": 16802
+    },
+    {
+      "epoch": 0.7740820933339476,
+      "grad_norm": 1.0261371390539873,
+      "learning_rate": 2.5597389832790586e-05,
+      "loss": 1.1036,
+      "step": 16803
+    },
+    {
+      "epoch": 0.7741281614225826,
+      "grad_norm": 2.9344531038812476,
+      "learning_rate": 2.5587421262497536e-05,
+      "loss": 1.2305,
+      "step": 16804
+    },
+    {
+      "epoch": 0.7741742295112176,
+      "grad_norm": 1.4960416382702721,
+      "learning_rate": 2.5577454348872943e-05,
+      "loss": 1.0588,
+      "step": 16805
+    },
+    {
+      "epoch": 0.7742202975998526,
+      "grad_norm": 1.4884481275165475,
+      "learning_rate": 2.5567489092138742e-05,
+      "loss": 1.3887,
+      "step": 16806
+    },
+    {
+      "epoch": 0.7742663656884876,
+      "grad_norm": 1.3622149289976118,
+      "learning_rate": 2.555752549251672e-05,
+      "loss": 1.4544,
+      "step": 16807
+    },
+    {
+      "epoch": 0.7743124337771226,
+      "grad_norm": 0.8439821169608825,
+      "learning_rate": 2.5547563550228816e-05,
+      "loss": 1.5833,
+      "step": 16808
+    },
+    {
+      "epoch": 0.7743585018657576,
+      "grad_norm": 1.9345500369267141,
+      "learning_rate": 2.5537603265496713e-05,
+      "loss": 1.3961,
+      "step": 16809
+    },
+    {
+      "epoch": 0.7744045699543926,
+      "grad_norm": 1.034810101876919,
+      "learning_rate": 2.5527644638542224e-05,
+      "loss": 1.5144,
+      "step": 16810
+    },
+    {
+      "epoch": 0.7744506380430276,
+      "grad_norm": 1.1254894881622517,
+      "learning_rate": 2.5517687669586977e-05,
+      "loss": 1.4564,
+      "step": 16811
+    },
+    {
+      "epoch": 0.7744967061316625,
+      "grad_norm": 0.9290727533956445,
+      "learning_rate": 2.5507732358852766e-05,
+      "loss": 1.5576,
+      "step": 16812
+    },
+    {
+      "epoch": 0.7745427742202976,
+      "grad_norm": 2.153931773240603,
+      "learning_rate": 2.549777870656117e-05,
+      "loss": 0.9553,
+      "step": 16813
+    },
+    {
+      "epoch": 0.7745888423089327,
+      "grad_norm": 1.3491452918524653,
+      "learning_rate": 2.5487826712933728e-05,
+      "loss": 1.518,
+      "step": 16814
+    },
+    {
+      "epoch": 0.7746349103975676,
+      "grad_norm": 2.782042573690993,
+      "learning_rate": 2.5477876378192123e-05,
+      "loss": 1.5575,
+      "step": 16815
+    },
+    {
+      "epoch": 0.7746809784862027,
+      "grad_norm": 0.8488700045243254,
+      "learning_rate": 2.546792770255778e-05,
+      "loss": 1.377,
+      "step": 16816
+    },
+    {
+      "epoch": 0.7747270465748376,
+      "grad_norm": 0.8519004426938855,
+      "learning_rate": 2.545798068625228e-05,
+      "loss": 1.3608,
+      "step": 16817
+    },
+    {
+      "epoch": 0.7747731146634727,
+      "grad_norm": 1.7735696930999045,
+      "learning_rate": 2.5448035329496955e-05,
+      "loss": 1.1336,
+      "step": 16818
+    },
+    {
+      "epoch": 0.7748191827521076,
+      "grad_norm": 1.6225321593001558,
+      "learning_rate": 2.543809163251336e-05,
+      "loss": 1.6053,
+      "step": 16819
+    },
+    {
+      "epoch": 0.7748652508407426,
+      "grad_norm": 1.4538404725064191,
+      "learning_rate": 2.5428149595522775e-05,
+      "loss": 1.4172,
+      "step": 16820
+    },
+    {
+      "epoch": 0.7749113189293776,
+      "grad_norm": 0.8308125701549699,
+      "learning_rate": 2.5418209218746602e-05,
+      "loss": 1.279,
+      "step": 16821
+    },
+    {
+      "epoch": 0.7749573870180126,
+      "grad_norm": 1.4065988402192215,
+      "learning_rate": 2.5408270502406107e-05,
+      "loss": 1.1844,
+      "step": 16822
+    },
+    {
+      "epoch": 0.7750034551066476,
+      "grad_norm": 1.9143036696225961,
+      "learning_rate": 2.5398333446722555e-05,
+      "loss": 1.0787,
+      "step": 16823
+    },
+    {
+      "epoch": 0.7750495231952826,
+      "grad_norm": 1.0335454418188932,
+      "learning_rate": 2.5388398051917228e-05,
+      "loss": 1.2139,
+      "step": 16824
+    },
+    {
+      "epoch": 0.7750955912839176,
+      "grad_norm": 1.1006507035613995,
+      "learning_rate": 2.537846431821126e-05,
+      "loss": 1.0984,
+      "step": 16825
+    },
+    {
+      "epoch": 0.7751416593725526,
+      "grad_norm": 1.7764870202369751,
+      "learning_rate": 2.5368532245825838e-05,
+      "loss": 1.1507,
+      "step": 16826
+    },
+    {
+      "epoch": 0.7751877274611876,
+      "grad_norm": 1.9946452284029366,
+      "learning_rate": 2.5358601834982075e-05,
+      "loss": 1.0752,
+      "step": 16827
+    },
+    {
+      "epoch": 0.7752337955498226,
+      "grad_norm": 1.3523726462701067,
+      "learning_rate": 2.5348673085901087e-05,
+      "loss": 1.4223,
+      "step": 16828
+    },
+    {
+      "epoch": 0.7752798636384577,
+      "grad_norm": 0.9653489887253629,
+      "learning_rate": 2.5338745998803848e-05,
+      "loss": 1.3501,
+      "step": 16829
+    },
+    {
+      "epoch": 0.7753259317270926,
+      "grad_norm": 0.738497417753209,
+      "learning_rate": 2.532882057391146e-05,
+      "loss": 1.3612,
+      "step": 16830
+    },
+    {
+      "epoch": 0.7753719998157277,
+      "grad_norm": 1.0524166705452425,
+      "learning_rate": 2.531889681144485e-05,
+      "loss": 1.2824,
+      "step": 16831
+    },
+    {
+      "epoch": 0.7754180679043626,
+      "grad_norm": 1.565627301768609,
+      "learning_rate": 2.5308974711624922e-05,
+      "loss": 1.1662,
+      "step": 16832
+    },
+    {
+      "epoch": 0.7754641359929977,
+      "grad_norm": 0.7943623520394139,
+      "learning_rate": 2.5299054274672606e-05,
+      "loss": 1.304,
+      "step": 16833
+    },
+    {
+      "epoch": 0.7755102040816326,
+      "grad_norm": 0.8645994597628163,
+      "learning_rate": 2.528913550080876e-05,
+      "loss": 1.1782,
+      "step": 16834
+    },
+    {
+      "epoch": 0.7755562721702677,
+      "grad_norm": 1.7153847472775903,
+      "learning_rate": 2.5279218390254235e-05,
+      "loss": 0.9975,
+      "step": 16835
+    },
+    {
+      "epoch": 0.7756023402589026,
+      "grad_norm": 3.066220662935636,
+      "learning_rate": 2.526930294322978e-05,
+      "loss": 1.5102,
+      "step": 16836
+    },
+    {
+      "epoch": 0.7756484083475377,
+      "grad_norm": 1.1509852269341905,
+      "learning_rate": 2.525938915995614e-05,
+      "loss": 1.3731,
+      "step": 16837
+    },
+    {
+      "epoch": 0.7756944764361726,
+      "grad_norm": 1.44681165572135,
+      "learning_rate": 2.5249477040654058e-05,
+      "loss": 1.1146,
+      "step": 16838
+    },
+    {
+      "epoch": 0.7757405445248077,
+      "grad_norm": 2.7514011510998255,
+      "learning_rate": 2.523956658554423e-05,
+      "loss": 1.2775,
+      "step": 16839
+    },
+    {
+      "epoch": 0.7757866126134426,
+      "grad_norm": 1.088965733142395,
+      "learning_rate": 2.522965779484724e-05,
+      "loss": 1.1705,
+      "step": 16840
+    },
+    {
+      "epoch": 0.7758326807020777,
+      "grad_norm": 2.163020393346584,
+      "learning_rate": 2.5219750668783705e-05,
+      "loss": 1.5731,
+      "step": 16841
+    },
+    {
+      "epoch": 0.7758787487907127,
+      "grad_norm": 1.2609512062905852,
+      "learning_rate": 2.5209845207574233e-05,
+      "loss": 0.9247,
+      "step": 16842
+    },
+    {
+      "epoch": 0.7759248168793477,
+      "grad_norm": 1.3990225740735618,
+      "learning_rate": 2.51999414114393e-05,
+      "loss": 0.9831,
+      "step": 16843
+    },
+    {
+      "epoch": 0.7759708849679827,
+      "grad_norm": 2.155632523120073,
+      "learning_rate": 2.5190039280599408e-05,
+      "loss": 0.9427,
+      "step": 16844
+    },
+    {
+      "epoch": 0.7760169530566177,
+      "grad_norm": 0.9607135564216891,
+      "learning_rate": 2.5180138815275013e-05,
+      "loss": 1.383,
+      "step": 16845
+    },
+    {
+      "epoch": 0.7760630211452527,
+      "grad_norm": 2.647059169872616,
+      "learning_rate": 2.5170240015686587e-05,
+      "loss": 1.2481,
+      "step": 16846
+    },
+    {
+      "epoch": 0.7761090892338877,
+      "grad_norm": 1.1425728794697072,
+      "learning_rate": 2.516034288205441e-05,
+      "loss": 1.1385,
+      "step": 16847
+    },
+    {
+      "epoch": 0.7761551573225227,
+      "grad_norm": 0.8423896989785091,
+      "learning_rate": 2.5150447414598888e-05,
+      "loss": 1.2705,
+      "step": 16848
+    },
+    {
+      "epoch": 0.7762012254111577,
+      "grad_norm": 1.6070589886180453,
+      "learning_rate": 2.5140553613540352e-05,
+      "loss": 0.9122,
+      "step": 16849
+    },
+    {
+      "epoch": 0.7762472934997927,
+      "grad_norm": 1.2580243737087733,
+      "learning_rate": 2.5130661479098992e-05,
+      "loss": 1.3671,
+      "step": 16850
+    },
+    {
+      "epoch": 0.7762933615884277,
+      "grad_norm": 1.0036760118620216,
+      "learning_rate": 2.512077101149508e-05,
+      "loss": 1.2671,
+      "step": 16851
+    },
+    {
+      "epoch": 0.7763394296770627,
+      "grad_norm": 1.397374616592324,
+      "learning_rate": 2.5110882210948817e-05,
+      "loss": 1.0599,
+      "step": 16852
+    },
+    {
+      "epoch": 0.7763854977656977,
+      "grad_norm": 1.2857921932889258,
+      "learning_rate": 2.510099507768039e-05,
+      "loss": 1.2913,
+      "step": 16853
+    },
+    {
+      "epoch": 0.7764315658543327,
+      "grad_norm": 1.3793954573165317,
+      "learning_rate": 2.5091109611909848e-05,
+      "loss": 1.2054,
+      "step": 16854
+    },
+    {
+      "epoch": 0.7764776339429678,
+      "grad_norm": 0.8867468929292188,
+      "learning_rate": 2.50812258138573e-05,
+      "loss": 1.4346,
+      "step": 16855
+    },
+    {
+      "epoch": 0.7765237020316027,
+      "grad_norm": 1.0181211302572126,
+      "learning_rate": 2.507134368374281e-05,
+      "loss": 1.0517,
+      "step": 16856
+    },
+    {
+      "epoch": 0.7765697701202378,
+      "grad_norm": 1.0066474102788654,
+      "learning_rate": 2.5061463221786408e-05,
+      "loss": 1.2215,
+      "step": 16857
+    },
+    {
+      "epoch": 0.7766158382088727,
+      "grad_norm": 1.469188136879067,
+      "learning_rate": 2.5051584428208008e-05,
+      "loss": 1.4568,
+      "step": 16858
+    },
+    {
+      "epoch": 0.7766619062975078,
+      "grad_norm": 1.4250773661445797,
+      "learning_rate": 2.504170730322757e-05,
+      "loss": 1.1404,
+      "step": 16859
+    },
+    {
+      "epoch": 0.7767079743861427,
+      "grad_norm": 1.1320309112110603,
+      "learning_rate": 2.5031831847065014e-05,
+      "loss": 1.0782,
+      "step": 16860
+    },
+    {
+      "epoch": 0.7767540424747778,
+      "grad_norm": 1.4874726472901543,
+      "learning_rate": 2.5021958059940155e-05,
+      "loss": 1.245,
+      "step": 16861
+    },
+    {
+      "epoch": 0.7768001105634127,
+      "grad_norm": 1.9540940258782298,
+      "learning_rate": 2.501208594207284e-05,
+      "loss": 1.272,
+      "step": 16862
+    },
+    {
+      "epoch": 0.7768461786520477,
+      "grad_norm": 0.9239480942565793,
+      "learning_rate": 2.5002215493682857e-05,
+      "loss": 1.1487,
+      "step": 16863
+    },
+    {
+      "epoch": 0.7768922467406827,
+      "grad_norm": 1.0326614027370717,
+      "learning_rate": 2.4992346714989978e-05,
+      "loss": 1.378,
+      "step": 16864
+    },
+    {
+      "epoch": 0.7769383148293177,
+      "grad_norm": 0.7573764084532963,
+      "learning_rate": 2.498247960621386e-05,
+      "loss": 1.3489,
+      "step": 16865
+    },
+    {
+      "epoch": 0.7769843829179527,
+      "grad_norm": 1.0024670902330066,
+      "learning_rate": 2.4972614167574217e-05,
+      "loss": 1.3447,
+      "step": 16866
+    },
+    {
+      "epoch": 0.7770304510065877,
+      "grad_norm": 0.9009125164499113,
+      "learning_rate": 2.4962750399290692e-05,
+      "loss": 1.0958,
+      "step": 16867
+    },
+    {
+      "epoch": 0.7770765190952227,
+      "grad_norm": 1.5264047330603931,
+      "learning_rate": 2.4952888301582845e-05,
+      "loss": 1.3405,
+      "step": 16868
+    },
+    {
+      "epoch": 0.7771225871838577,
+      "grad_norm": 3.797302662784083,
+      "learning_rate": 2.4943027874670255e-05,
+      "loss": 1.1565,
+      "step": 16869
+    },
+    {
+      "epoch": 0.7771686552724928,
+      "grad_norm": 0.9578686496026613,
+      "learning_rate": 2.493316911877246e-05,
+      "loss": 1.2714,
+      "step": 16870
+    },
+    {
+      "epoch": 0.7772147233611277,
+      "grad_norm": 1.4151801536878779,
+      "learning_rate": 2.492331203410897e-05,
+      "loss": 1.2413,
+      "step": 16871
+    },
+    {
+      "epoch": 0.7772607914497628,
+      "grad_norm": 0.7651011782388585,
+      "learning_rate": 2.491345662089918e-05,
+      "loss": 1.2127,
+      "step": 16872
+    },
+    {
+      "epoch": 0.7773068595383977,
+      "grad_norm": 0.9522767912460497,
+      "learning_rate": 2.490360287936253e-05,
+      "loss": 1.2064,
+      "step": 16873
+    },
+    {
+      "epoch": 0.7773529276270328,
+      "grad_norm": 1.056896622674337,
+      "learning_rate": 2.4893750809718397e-05,
+      "loss": 1.1241,
+      "step": 16874
+    },
+    {
+      "epoch": 0.7773989957156677,
+      "grad_norm": 1.3514850741574505,
+      "learning_rate": 2.4883900412186155e-05,
+      "loss": 1.0716,
+      "step": 16875
+    },
+    {
+      "epoch": 0.7774450638043028,
+      "grad_norm": 1.0356835865770353,
+      "learning_rate": 2.4874051686985044e-05,
+      "loss": 1.3859,
+      "step": 16876
+    },
+    {
+      "epoch": 0.7774911318929377,
+      "grad_norm": 0.9280322300381596,
+      "learning_rate": 2.4864204634334364e-05,
+      "loss": 1.5019,
+      "step": 16877
+    },
+    {
+      "epoch": 0.7775371999815728,
+      "grad_norm": 1.0895264347614773,
+      "learning_rate": 2.4854359254453364e-05,
+      "loss": 1.0525,
+      "step": 16878
+    },
+    {
+      "epoch": 0.7775832680702077,
+      "grad_norm": 1.8404819911294505,
+      "learning_rate": 2.4844515547561176e-05,
+      "loss": 0.9115,
+      "step": 16879
+    },
+    {
+      "epoch": 0.7776293361588428,
+      "grad_norm": 1.1871632644347683,
+      "learning_rate": 2.4834673513876994e-05,
+      "loss": 1.2362,
+      "step": 16880
+    },
+    {
+      "epoch": 0.7776754042474777,
+      "grad_norm": 2.020531541155339,
+      "learning_rate": 2.4824833153619918e-05,
+      "loss": 1.105,
+      "step": 16881
+    },
+    {
+      "epoch": 0.7777214723361128,
+      "grad_norm": 0.9118225977559388,
+      "learning_rate": 2.4814994467009056e-05,
+      "loss": 1.3948,
+      "step": 16882
+    },
+    {
+      "epoch": 0.7777675404247478,
+      "grad_norm": 2.1818335807299687,
+      "learning_rate": 2.4805157454263407e-05,
+      "loss": 1.2549,
+      "step": 16883
+    },
+    {
+      "epoch": 0.7778136085133828,
+      "grad_norm": 0.9798987616324998,
+      "learning_rate": 2.4795322115602003e-05,
+      "loss": 0.9503,
+      "step": 16884
+    },
+    {
+      "epoch": 0.7778596766020178,
+      "grad_norm": 1.5266420774661962,
+      "learning_rate": 2.478548845124382e-05,
+      "loss": 1.2335,
+      "step": 16885
+    },
+    {
+      "epoch": 0.7779057446906528,
+      "grad_norm": 1.2515061219106864,
+      "learning_rate": 2.4775656461407747e-05,
+      "loss": 0.9883,
+      "step": 16886
+    },
+    {
+      "epoch": 0.7779518127792878,
+      "grad_norm": 1.2326028019377442,
+      "learning_rate": 2.4765826146312697e-05,
+      "loss": 1.3156,
+      "step": 16887
+    },
+    {
+      "epoch": 0.7779978808679228,
+      "grad_norm": 2.647368660100771,
+      "learning_rate": 2.475599750617754e-05,
+      "loss": 1.0059,
+      "step": 16888
+    },
+    {
+      "epoch": 0.7780439489565578,
+      "grad_norm": 0.7331074498244741,
+      "learning_rate": 2.4746170541221102e-05,
+      "loss": 1.43,
+      "step": 16889
+    },
+    {
+      "epoch": 0.7780900170451928,
+      "grad_norm": 0.6766059503230123,
+      "learning_rate": 2.4736345251662117e-05,
+      "loss": 1.2853,
+      "step": 16890
+    },
+    {
+      "epoch": 0.7781360851338278,
+      "grad_norm": 1.12495667887443,
+      "learning_rate": 2.4726521637719358e-05,
+      "loss": 1.4313,
+      "step": 16891
+    },
+    {
+      "epoch": 0.7781821532224628,
+      "grad_norm": 1.0995420115901067,
+      "learning_rate": 2.4716699699611522e-05,
+      "loss": 1.2464,
+      "step": 16892
+    },
+    {
+      "epoch": 0.7782282213110978,
+      "grad_norm": 0.9365877678687462,
+      "learning_rate": 2.4706879437557317e-05,
+      "loss": 0.6931,
+      "step": 16893
+    },
+    {
+      "epoch": 0.7782742893997328,
+      "grad_norm": 0.8516538649105861,
+      "learning_rate": 2.4697060851775344e-05,
+      "loss": 1.1007,
+      "step": 16894
+    },
+    {
+      "epoch": 0.7783203574883678,
+      "grad_norm": 1.8393219653191102,
+      "learning_rate": 2.468724394248414e-05,
+      "loss": 1.4036,
+      "step": 16895
+    },
+    {
+      "epoch": 0.7783664255770029,
+      "grad_norm": 0.9163948650505904,
+      "learning_rate": 2.467742870990236e-05,
+      "loss": 1.2464,
+      "step": 16896
+    },
+    {
+      "epoch": 0.7784124936656378,
+      "grad_norm": 1.2535113552891397,
+      "learning_rate": 2.466761515424847e-05,
+      "loss": 1.1948,
+      "step": 16897
+    },
+    {
+      "epoch": 0.7784585617542729,
+      "grad_norm": 1.0300327503032694,
+      "learning_rate": 2.4657803275740976e-05,
+      "loss": 1.464,
+      "step": 16898
+    },
+    {
+      "epoch": 0.7785046298429078,
+      "grad_norm": 0.8047430862530958,
+      "learning_rate": 2.4647993074598252e-05,
+      "loss": 1.4903,
+      "step": 16899
+    },
+    {
+      "epoch": 0.7785506979315429,
+      "grad_norm": 2.410941936641759,
+      "learning_rate": 2.4638184551038822e-05,
+      "loss": 1.4528,
+      "step": 16900
+    },
+    {
+      "epoch": 0.7785967660201778,
+      "grad_norm": 1.2520154316857002,
+      "learning_rate": 2.4628377705280957e-05,
+      "loss": 1.3097,
+      "step": 16901
+    },
+    {
+      "epoch": 0.7786428341088129,
+      "grad_norm": 0.8366855353325378,
+      "learning_rate": 2.4618572537543038e-05,
+      "loss": 1.155,
+      "step": 16902
+    },
+    {
+      "epoch": 0.7786889021974478,
+      "grad_norm": 0.8405630893755474,
+      "learning_rate": 2.4608769048043345e-05,
+      "loss": 1.0988,
+      "step": 16903
+    },
+    {
+      "epoch": 0.7787349702860829,
+      "grad_norm": 1.0627434457938647,
+      "learning_rate": 2.459896723700017e-05,
+      "loss": 1.3589,
+      "step": 16904
+    },
+    {
+      "epoch": 0.7787810383747178,
+      "grad_norm": 1.285138801289161,
+      "learning_rate": 2.4589167104631705e-05,
+      "loss": 1.3109,
+      "step": 16905
+    },
+    {
+      "epoch": 0.7788271064633528,
+      "grad_norm": 1.3341243749956337,
+      "learning_rate": 2.4579368651156076e-05,
+      "loss": 1.2796,
+      "step": 16906
+    },
+    {
+      "epoch": 0.7788731745519878,
+      "grad_norm": 1.0210231753093437,
+      "learning_rate": 2.4569571876791542e-05,
+      "loss": 1.0889,
+      "step": 16907
+    },
+    {
+      "epoch": 0.7789192426406228,
+      "grad_norm": 1.0191039540831446,
+      "learning_rate": 2.455977678175613e-05,
+      "loss": 0.8342,
+      "step": 16908
+    },
+    {
+      "epoch": 0.7789653107292579,
+      "grad_norm": 1.2561175575488301,
+      "learning_rate": 2.454998336626796e-05,
+      "loss": 1.1775,
+      "step": 16909
+    },
+    {
+      "epoch": 0.7790113788178928,
+      "grad_norm": 0.719016772372158,
+      "learning_rate": 2.4540191630544985e-05,
+      "loss": 1.0382,
+      "step": 16910
+    },
+    {
+      "epoch": 0.7790574469065279,
+      "grad_norm": 0.7714717000621726,
+      "learning_rate": 2.4530401574805317e-05,
+      "loss": 1.1758,
+      "step": 16911
+    },
+    {
+      "epoch": 0.7791035149951628,
+      "grad_norm": 3.3998403788492633,
+      "learning_rate": 2.4520613199266828e-05,
+      "loss": 1.1554,
+      "step": 16912
+    },
+    {
+      "epoch": 0.7791495830837979,
+      "grad_norm": 1.395536822722743,
+      "learning_rate": 2.4510826504147465e-05,
+      "loss": 1.4415,
+      "step": 16913
+    },
+    {
+      "epoch": 0.7791956511724328,
+      "grad_norm": 1.1392861527852707,
+      "learning_rate": 2.4501041489665132e-05,
+      "loss": 1.4179,
+      "step": 16914
+    },
+    {
+      "epoch": 0.7792417192610679,
+      "grad_norm": 1.1080236370740348,
+      "learning_rate": 2.4491258156037644e-05,
+      "loss": 1.1693,
+      "step": 16915
+    },
+    {
+      "epoch": 0.7792877873497028,
+      "grad_norm": 1.7019949933713647,
+      "learning_rate": 2.448147650348285e-05,
+      "loss": 1.0162,
+      "step": 16916
+    },
+    {
+      "epoch": 0.7793338554383379,
+      "grad_norm": 1.2721172024361522,
+      "learning_rate": 2.4471696532218425e-05,
+      "loss": 0.8554,
+      "step": 16917
+    },
+    {
+      "epoch": 0.7793799235269728,
+      "grad_norm": 0.7291150109140364,
+      "learning_rate": 2.4461918242462246e-05,
+      "loss": 1.3601,
+      "step": 16918
+    },
+    {
+      "epoch": 0.7794259916156079,
+      "grad_norm": 1.0338022751794198,
+      "learning_rate": 2.44521416344319e-05,
+      "loss": 1.1698,
+      "step": 16919
+    },
+    {
+      "epoch": 0.7794720597042428,
+      "grad_norm": 0.9392007118081489,
+      "learning_rate": 2.444236670834511e-05,
+      "loss": 1.3843,
+      "step": 16920
+    },
+    {
+      "epoch": 0.7795181277928779,
+      "grad_norm": 1.1301980575871606,
+      "learning_rate": 2.4432593464419418e-05,
+      "loss": 1.0104,
+      "step": 16921
+    },
+    {
+      "epoch": 0.7795641958815128,
+      "grad_norm": 1.070902132955548,
+      "learning_rate": 2.4422821902872517e-05,
+      "loss": 0.8418,
+      "step": 16922
+    },
+    {
+      "epoch": 0.7796102639701479,
+      "grad_norm": 0.9141235846238441,
+      "learning_rate": 2.4413052023921912e-05,
+      "loss": 1.0615,
+      "step": 16923
+    },
+    {
+      "epoch": 0.7796563320587829,
+      "grad_norm": 0.8634887111234929,
+      "learning_rate": 2.4403283827785063e-05,
+      "loss": 1.2055,
+      "step": 16924
+    },
+    {
+      "epoch": 0.7797024001474179,
+      "grad_norm": 1.3336891524496572,
+      "learning_rate": 2.439351731467947e-05,
+      "loss": 1.2681,
+      "step": 16925
+    },
+    {
+      "epoch": 0.7797484682360529,
+      "grad_norm": 0.8187511678115167,
+      "learning_rate": 2.4383752484822588e-05,
+      "loss": 1.2695,
+      "step": 16926
+    },
+    {
+      "epoch": 0.7797945363246879,
+      "grad_norm": 1.199669011209016,
+      "learning_rate": 2.4373989338431825e-05,
+      "loss": 0.956,
+      "step": 16927
+    },
+    {
+      "epoch": 0.7798406044133229,
+      "grad_norm": 0.9410369100472157,
+      "learning_rate": 2.4364227875724467e-05,
+      "loss": 1.1528,
+      "step": 16928
+    },
+    {
+      "epoch": 0.7798866725019579,
+      "grad_norm": 0.7167467901718935,
+      "learning_rate": 2.4354468096917938e-05,
+      "loss": 1.3802,
+      "step": 16929
+    },
+    {
+      "epoch": 0.7799327405905929,
+      "grad_norm": 0.7451521798947487,
+      "learning_rate": 2.4344710002229453e-05,
+      "loss": 1.2697,
+      "step": 16930
+    },
+    {
+      "epoch": 0.7799788086792279,
+      "grad_norm": 1.2978727049605663,
+      "learning_rate": 2.4334953591876298e-05,
+      "loss": 1.1808,
+      "step": 16931
+    },
+    {
+      "epoch": 0.7800248767678629,
+      "grad_norm": 1.00851175966765,
+      "learning_rate": 2.4325198866075637e-05,
+      "loss": 1.4185,
+      "step": 16932
+    },
+    {
+      "epoch": 0.7800709448564979,
+      "grad_norm": 1.0076645985583972,
+      "learning_rate": 2.431544582504467e-05,
+      "loss": 1.0798,
+      "step": 16933
+    },
+    {
+      "epoch": 0.7801170129451329,
+      "grad_norm": 1.1698400635370245,
+      "learning_rate": 2.430569446900056e-05,
+      "loss": 1.2522,
+      "step": 16934
+    },
+    {
+      "epoch": 0.7801630810337679,
+      "grad_norm": 1.024038778570186,
+      "learning_rate": 2.429594479816035e-05,
+      "loss": 1.4165,
+      "step": 16935
+    },
+    {
+      "epoch": 0.7802091491224029,
+      "grad_norm": 1.0588304038767504,
+      "learning_rate": 2.4286196812741124e-05,
+      "loss": 1.1831,
+      "step": 16936
+    },
+    {
+      "epoch": 0.780255217211038,
+      "grad_norm": 1.4439910240759335,
+      "learning_rate": 2.4276450512959904e-05,
+      "loss": 1.106,
+      "step": 16937
+    },
+    {
+      "epoch": 0.7803012852996729,
+      "grad_norm": 1.159602903754902,
+      "learning_rate": 2.4266705899033692e-05,
+      "loss": 1.4262,
+      "step": 16938
+    },
+    {
+      "epoch": 0.780347353388308,
+      "grad_norm": 0.8912903721895998,
+      "learning_rate": 2.4256962971179376e-05,
+      "loss": 1.2065,
+      "step": 16939
+    },
+    {
+      "epoch": 0.7803934214769429,
+      "grad_norm": 1.0016710901686292,
+      "learning_rate": 2.4247221729613955e-05,
+      "loss": 1.3665,
+      "step": 16940
+    },
+    {
+      "epoch": 0.780439489565578,
+      "grad_norm": 1.9385677525549283,
+      "learning_rate": 2.423748217455426e-05,
+      "loss": 1.287,
+      "step": 16941
+    },
+    {
+      "epoch": 0.7804855576542129,
+      "grad_norm": 0.9592920270468205,
+      "learning_rate": 2.4227744306217093e-05,
+      "loss": 1.1386,
+      "step": 16942
+    },
+    {
+      "epoch": 0.780531625742848,
+      "grad_norm": 1.1144891584527612,
+      "learning_rate": 2.421800812481927e-05,
+      "loss": 1.1914,
+      "step": 16943
+    },
+    {
+      "epoch": 0.7805776938314829,
+      "grad_norm": 1.3925389337683958,
+      "learning_rate": 2.420827363057756e-05,
+      "loss": 1.1504,
+      "step": 16944
+    },
+    {
+      "epoch": 0.780623761920118,
+      "grad_norm": 1.3243591710714218,
+      "learning_rate": 2.4198540823708714e-05,
+      "loss": 1.2911,
+      "step": 16945
+    },
+    {
+      "epoch": 0.7806698300087529,
+      "grad_norm": 1.4964459834920272,
+      "learning_rate": 2.418880970442936e-05,
+      "loss": 1.3271,
+      "step": 16946
+    },
+    {
+      "epoch": 0.780715898097388,
+      "grad_norm": 1.071468647050874,
+      "learning_rate": 2.4179080272956166e-05,
+      "loss": 1.2215,
+      "step": 16947
+    },
+    {
+      "epoch": 0.7807619661860229,
+      "grad_norm": 1.7981470369733563,
+      "learning_rate": 2.4169352529505738e-05,
+      "loss": 1.1083,
+      "step": 16948
+    },
+    {
+      "epoch": 0.780808034274658,
+      "grad_norm": 0.8420466266806527,
+      "learning_rate": 2.4159626474294695e-05,
+      "loss": 0.9464,
+      "step": 16949
+    },
+    {
+      "epoch": 0.780854102363293,
+      "grad_norm": 2.25047697908172,
+      "learning_rate": 2.414990210753949e-05,
+      "loss": 1.1896,
+      "step": 16950
+    },
+    {
+      "epoch": 0.7809001704519279,
+      "grad_norm": 0.813196323910374,
+      "learning_rate": 2.4140179429456667e-05,
+      "loss": 1.3269,
+      "step": 16951
+    },
+    {
+      "epoch": 0.780946238540563,
+      "grad_norm": 1.3903373296600616,
+      "learning_rate": 2.4130458440262705e-05,
+      "loss": 1.0814,
+      "step": 16952
+    },
+    {
+      "epoch": 0.7809923066291979,
+      "grad_norm": 1.1185874652362928,
+      "learning_rate": 2.4120739140173966e-05,
+      "loss": 1.2368,
+      "step": 16953
+    },
+    {
+      "epoch": 0.781038374717833,
+      "grad_norm": 0.8425141834078983,
+      "learning_rate": 2.4111021529406862e-05,
+      "loss": 1.2816,
+      "step": 16954
+    },
+    {
+      "epoch": 0.7810844428064679,
+      "grad_norm": 1.1981470877170368,
+      "learning_rate": 2.410130560817775e-05,
+      "loss": 1.5927,
+      "step": 16955
+    },
+    {
+      "epoch": 0.781130510895103,
+      "grad_norm": 1.1059180846250936,
+      "learning_rate": 2.409159137670296e-05,
+      "loss": 1.3238,
+      "step": 16956
+    },
+    {
+      "epoch": 0.7811765789837379,
+      "grad_norm": 1.08846854155142,
+      "learning_rate": 2.40818788351987e-05,
+      "loss": 0.8785,
+      "step": 16957
+    },
+    {
+      "epoch": 0.781222647072373,
+      "grad_norm": 1.1280720408382037,
+      "learning_rate": 2.4072167983881233e-05,
+      "loss": 1.2249,
+      "step": 16958
+    },
+    {
+      "epoch": 0.7812687151610079,
+      "grad_norm": 0.8118684429309099,
+      "learning_rate": 2.4062458822966794e-05,
+      "loss": 1.4353,
+      "step": 16959
+    },
+    {
+      "epoch": 0.781314783249643,
+      "grad_norm": 1.8500191974589484,
+      "learning_rate": 2.4052751352671465e-05,
+      "loss": 1.2495,
+      "step": 16960
+    },
+    {
+      "epoch": 0.7813608513382779,
+      "grad_norm": 1.1777560425868323,
+      "learning_rate": 2.404304557321141e-05,
+      "loss": 1.3738,
+      "step": 16961
+    },
+    {
+      "epoch": 0.781406919426913,
+      "grad_norm": 1.2554784132087915,
+      "learning_rate": 2.4033341484802707e-05,
+      "loss": 1.3001,
+      "step": 16962
+    },
+    {
+      "epoch": 0.7814529875155479,
+      "grad_norm": 0.7898532988865297,
+      "learning_rate": 2.402363908766143e-05,
+      "loss": 1.2006,
+      "step": 16963
+    },
+    {
+      "epoch": 0.781499055604183,
+      "grad_norm": 1.680154542526309,
+      "learning_rate": 2.4013938382003532e-05,
+      "loss": 1.1295,
+      "step": 16964
+    },
+    {
+      "epoch": 0.781545123692818,
+      "grad_norm": 0.7708195215869712,
+      "learning_rate": 2.400423936804501e-05,
+      "loss": 1.247,
+      "step": 16965
+    },
+    {
+      "epoch": 0.781591191781453,
+      "grad_norm": 1.7516476436944761,
+      "learning_rate": 2.399454204600179e-05,
+      "loss": 1.3437,
+      "step": 16966
+    },
+    {
+      "epoch": 0.781637259870088,
+      "grad_norm": 1.5277801681077103,
+      "learning_rate": 2.398484641608979e-05,
+      "loss": 1.1943,
+      "step": 16967
+    },
+    {
+      "epoch": 0.781683327958723,
+      "grad_norm": 1.3305116696237467,
+      "learning_rate": 2.3975152478524832e-05,
+      "loss": 1.1606,
+      "step": 16968
+    },
+    {
+      "epoch": 0.781729396047358,
+      "grad_norm": 1.2356927332510075,
+      "learning_rate": 2.3965460233522742e-05,
+      "loss": 0.8591,
+      "step": 16969
+    },
+    {
+      "epoch": 0.781775464135993,
+      "grad_norm": 0.7080270230707538,
+      "learning_rate": 2.3955769681299334e-05,
+      "loss": 1.2729,
+      "step": 16970
+    },
+    {
+      "epoch": 0.781821532224628,
+      "grad_norm": 0.9076611477306286,
+      "learning_rate": 2.3946080822070303e-05,
+      "loss": 1.3832,
+      "step": 16971
+    },
+    {
+      "epoch": 0.781867600313263,
+      "grad_norm": 1.4559468210302906,
+      "learning_rate": 2.3936393656051383e-05,
+      "loss": 1.1282,
+      "step": 16972
+    },
+    {
+      "epoch": 0.781913668401898,
+      "grad_norm": 1.6601683771656044,
+      "learning_rate": 2.3926708183458223e-05,
+      "loss": 1.3591,
+      "step": 16973
+    },
+    {
+      "epoch": 0.781959736490533,
+      "grad_norm": 1.5823682985230836,
+      "learning_rate": 2.3917024404506504e-05,
+      "loss": 1.4797,
+      "step": 16974
+    },
+    {
+      "epoch": 0.782005804579168,
+      "grad_norm": 0.9486232499495023,
+      "learning_rate": 2.3907342319411753e-05,
+      "loss": 1.4312,
+      "step": 16975
+    },
+    {
+      "epoch": 0.782051872667803,
+      "grad_norm": 0.8853310577331329,
+      "learning_rate": 2.3897661928389557e-05,
+      "loss": 1.0284,
+      "step": 16976
+    },
+    {
+      "epoch": 0.782097940756438,
+      "grad_norm": 1.0582298267967158,
+      "learning_rate": 2.3887983231655453e-05,
+      "loss": 1.1208,
+      "step": 16977
+    },
+    {
+      "epoch": 0.7821440088450731,
+      "grad_norm": 0.9882288449984377,
+      "learning_rate": 2.3878306229424875e-05,
+      "loss": 1.0691,
+      "step": 16978
+    },
+    {
+      "epoch": 0.782190076933708,
+      "grad_norm": 0.9348403039117158,
+      "learning_rate": 2.386863092191329e-05,
+      "loss": 1.2779,
+      "step": 16979
+    },
+    {
+      "epoch": 0.7822361450223431,
+      "grad_norm": 1.1798361736670655,
+      "learning_rate": 2.3858957309336096e-05,
+      "loss": 1.105,
+      "step": 16980
+    },
+    {
+      "epoch": 0.782282213110978,
+      "grad_norm": 1.1971357943048435,
+      "learning_rate": 2.3849285391908695e-05,
+      "loss": 1.2083,
+      "step": 16981
+    },
+    {
+      "epoch": 0.7823282811996131,
+      "grad_norm": 0.7009828993009595,
+      "learning_rate": 2.383961516984635e-05,
+      "loss": 1.2107,
+      "step": 16982
+    },
+    {
+      "epoch": 0.782374349288248,
+      "grad_norm": 0.8601072661661389,
+      "learning_rate": 2.3829946643364386e-05,
+      "loss": 1.5482,
+      "step": 16983
+    },
+    {
+      "epoch": 0.7824204173768831,
+      "grad_norm": 2.2690364848439324,
+      "learning_rate": 2.382027981267807e-05,
+      "loss": 1.2232,
+      "step": 16984
+    },
+    {
+      "epoch": 0.782466485465518,
+      "grad_norm": 0.9373871783104123,
+      "learning_rate": 2.3810614678002617e-05,
+      "loss": 1.1721,
+      "step": 16985
+    },
+    {
+      "epoch": 0.7825125535541531,
+      "grad_norm": 1.0685593868371215,
+      "learning_rate": 2.380095123955317e-05,
+      "loss": 1.3405,
+      "step": 16986
+    },
+    {
+      "epoch": 0.782558621642788,
+      "grad_norm": 1.126037023805163,
+      "learning_rate": 2.379128949754489e-05,
+      "loss": 1.1891,
+      "step": 16987
+    },
+    {
+      "epoch": 0.782604689731423,
+      "grad_norm": 1.3080951917249275,
+      "learning_rate": 2.3781629452192912e-05,
+      "loss": 1.071,
+      "step": 16988
+    },
+    {
+      "epoch": 0.782650757820058,
+      "grad_norm": 0.9702927200698477,
+      "learning_rate": 2.377197110371223e-05,
+      "loss": 0.9271,
+      "step": 16989
+    },
+    {
+      "epoch": 0.782696825908693,
+      "grad_norm": 0.7053703335397116,
+      "learning_rate": 2.3762314452317914e-05,
+      "loss": 1.2589,
+      "step": 16990
+    },
+    {
+      "epoch": 0.7827428939973281,
+      "grad_norm": 1.1306563937570329,
+      "learning_rate": 2.3752659498224937e-05,
+      "loss": 1.4144,
+      "step": 16991
+    },
+    {
+      "epoch": 0.782788962085963,
+      "grad_norm": 0.9143399082808705,
+      "learning_rate": 2.37430062416483e-05,
+      "loss": 1.3456,
+      "step": 16992
+    },
+    {
+      "epoch": 0.7828350301745981,
+      "grad_norm": 0.9124794463694947,
+      "learning_rate": 2.3733354682802843e-05,
+      "loss": 1.2988,
+      "step": 16993
+    },
+    {
+      "epoch": 0.782881098263233,
+      "grad_norm": 1.2663816225297848,
+      "learning_rate": 2.372370482190348e-05,
+      "loss": 1.2435,
+      "step": 16994
+    },
+    {
+      "epoch": 0.7829271663518681,
+      "grad_norm": 0.9922531646221849,
+      "learning_rate": 2.3714056659165075e-05,
+      "loss": 1.2977,
+      "step": 16995
+    },
+    {
+      "epoch": 0.782973234440503,
+      "grad_norm": 1.014880567850513,
+      "learning_rate": 2.3704410194802352e-05,
+      "loss": 1.0061,
+      "step": 16996
+    },
+    {
+      "epoch": 0.7830193025291381,
+      "grad_norm": 0.7594855988218981,
+      "learning_rate": 2.3694765429030163e-05,
+      "loss": 1.2416,
+      "step": 16997
+    },
+    {
+      "epoch": 0.783065370617773,
+      "grad_norm": 0.8412585937597254,
+      "learning_rate": 2.3685122362063117e-05,
+      "loss": 1.1142,
+      "step": 16998
+    },
+    {
+      "epoch": 0.7831114387064081,
+      "grad_norm": 0.9256585922078685,
+      "learning_rate": 2.3675480994116027e-05,
+      "loss": 1.2595,
+      "step": 16999
+    },
+    {
+      "epoch": 0.783157506795043,
+      "grad_norm": 1.052072423154504,
+      "learning_rate": 2.366584132540346e-05,
+      "loss": 1.3488,
+      "step": 17000
+    },
+    {
+      "epoch": 0.7832035748836781,
+      "grad_norm": 1.4725334238933514,
+      "learning_rate": 2.365620335614005e-05,
+      "loss": 1.1871,
+      "step": 17001
+    },
+    {
+      "epoch": 0.783249642972313,
+      "grad_norm": 1.1071978375605065,
+      "learning_rate": 2.364656708654037e-05,
+      "loss": 1.2638,
+      "step": 17002
+    },
+    {
+      "epoch": 0.7832957110609481,
+      "grad_norm": 1.72019254077313,
+      "learning_rate": 2.3636932516818987e-05,
+      "loss": 1.2827,
+      "step": 17003
+    },
+    {
+      "epoch": 0.783341779149583,
+      "grad_norm": 0.7327902956525348,
+      "learning_rate": 2.362729964719037e-05,
+      "loss": 1.3887,
+      "step": 17004
+    },
+    {
+      "epoch": 0.7833878472382181,
+      "grad_norm": 0.8151909542916269,
+      "learning_rate": 2.3617668477868916e-05,
+      "loss": 1.0755,
+      "step": 17005
+    },
+    {
+      "epoch": 0.7834339153268531,
+      "grad_norm": 1.1425923480174256,
+      "learning_rate": 2.360803900906916e-05,
+      "loss": 1.2214,
+      "step": 17006
+    },
+    {
+      "epoch": 0.7834799834154881,
+      "grad_norm": 1.1225383099167205,
+      "learning_rate": 2.35984112410054e-05,
+      "loss": 1.426,
+      "step": 17007
+    },
+    {
+      "epoch": 0.7835260515041231,
+      "grad_norm": 1.072918243664254,
+      "learning_rate": 2.3588785173892048e-05,
+      "loss": 1.1725,
+      "step": 17008
+    },
+    {
+      "epoch": 0.7835721195927581,
+      "grad_norm": 1.249956107149958,
+      "learning_rate": 2.3579160807943312e-05,
+      "loss": 1.1989,
+      "step": 17009
+    },
+    {
+      "epoch": 0.7836181876813931,
+      "grad_norm": 1.9355545816647841,
+      "learning_rate": 2.3569538143373593e-05,
+      "loss": 1.039,
+      "step": 17010
+    },
+    {
+      "epoch": 0.7836642557700281,
+      "grad_norm": 1.0956345845714206,
+      "learning_rate": 2.355991718039702e-05,
+      "loss": 1.3114,
+      "step": 17011
+    },
+    {
+      "epoch": 0.7837103238586631,
+      "grad_norm": 0.8408652427955626,
+      "learning_rate": 2.3550297919227828e-05,
+      "loss": 1.2886,
+      "step": 17012
+    },
+    {
+      "epoch": 0.7837563919472981,
+      "grad_norm": 0.9761975303088875,
+      "learning_rate": 2.3540680360080203e-05,
+      "loss": 1.2336,
+      "step": 17013
+    },
+    {
+      "epoch": 0.7838024600359331,
+      "grad_norm": 1.1553631865431817,
+      "learning_rate": 2.35310645031682e-05,
+      "loss": 1.0614,
+      "step": 17014
+    },
+    {
+      "epoch": 0.7838485281245681,
+      "grad_norm": 0.8575470446635475,
+      "learning_rate": 2.3521450348705953e-05,
+      "loss": 1.0623,
+      "step": 17015
+    },
+    {
+      "epoch": 0.7838945962132031,
+      "grad_norm": 1.200943856362289,
+      "learning_rate": 2.3511837896907418e-05,
+      "loss": 1.0981,
+      "step": 17016
+    },
+    {
+      "epoch": 0.7839406643018381,
+      "grad_norm": 0.9851389850013413,
+      "learning_rate": 2.3502227147986733e-05,
+      "loss": 1.3378,
+      "step": 17017
+    },
+    {
+      "epoch": 0.7839867323904731,
+      "grad_norm": 1.4741449761007281,
+      "learning_rate": 2.349261810215776e-05,
+      "loss": 1.2513,
+      "step": 17018
+    },
+    {
+      "epoch": 0.7840328004791082,
+      "grad_norm": 1.1191966592320748,
+      "learning_rate": 2.3483010759634495e-05,
+      "loss": 1.3621,
+      "step": 17019
+    },
+    {
+      "epoch": 0.7840788685677431,
+      "grad_norm": 2.045501467023028,
+      "learning_rate": 2.347340512063074e-05,
+      "loss": 1.0934,
+      "step": 17020
+    },
+    {
+      "epoch": 0.7841249366563782,
+      "grad_norm": 1.0618223610902229,
+      "learning_rate": 2.346380118536047e-05,
+      "loss": 1.3001,
+      "step": 17021
+    },
+    {
+      "epoch": 0.7841710047450131,
+      "grad_norm": 1.341709001790242,
+      "learning_rate": 2.3454198954037387e-05,
+      "loss": 1.3343,
+      "step": 17022
+    },
+    {
+      "epoch": 0.7842170728336482,
+      "grad_norm": 1.5586323508369282,
+      "learning_rate": 2.3444598426875363e-05,
+      "loss": 1.169,
+      "step": 17023
+    },
+    {
+      "epoch": 0.7842631409222831,
+      "grad_norm": 1.212762097973612,
+      "learning_rate": 2.3434999604088047e-05,
+      "loss": 1.0538,
+      "step": 17024
+    },
+    {
+      "epoch": 0.7843092090109182,
+      "grad_norm": 1.0916259769789214,
+      "learning_rate": 2.342540248588919e-05,
+      "loss": 0.8984,
+      "step": 17025
+    },
+    {
+      "epoch": 0.7843552770995531,
+      "grad_norm": 1.5471489532303906,
+      "learning_rate": 2.341580707249249e-05,
+      "loss": 1.1724,
+      "step": 17026
+    },
+    {
+      "epoch": 0.7844013451881882,
+      "grad_norm": 0.964370540451354,
+      "learning_rate": 2.3406213364111463e-05,
+      "loss": 1.0679,
+      "step": 17027
+    },
+    {
+      "epoch": 0.7844474132768231,
+      "grad_norm": 1.9231712242032097,
+      "learning_rate": 2.3396621360959824e-05,
+      "loss": 1.1862,
+      "step": 17028
+    },
+    {
+      "epoch": 0.7844934813654582,
+      "grad_norm": 1.3738385332239584,
+      "learning_rate": 2.338703106325103e-05,
+      "loss": 1.062,
+      "step": 17029
+    },
+    {
+      "epoch": 0.7845395494540931,
+      "grad_norm": 1.171739993692107,
+      "learning_rate": 2.3377442471198664e-05,
+      "loss": 1.3982,
+      "step": 17030
+    },
+    {
+      "epoch": 0.7845856175427282,
+      "grad_norm": 1.3816361199331808,
+      "learning_rate": 2.3367855585016096e-05,
+      "loss": 1.2884,
+      "step": 17031
+    },
+    {
+      "epoch": 0.7846316856313632,
+      "grad_norm": 2.6463298662499333,
+      "learning_rate": 2.3358270404916895e-05,
+      "loss": 1.245,
+      "step": 17032
+    },
+    {
+      "epoch": 0.7846777537199982,
+      "grad_norm": 1.0749386637945315,
+      "learning_rate": 2.334868693111438e-05,
+      "loss": 1.0921,
+      "step": 17033
+    },
+    {
+      "epoch": 0.7847238218086332,
+      "grad_norm": 0.8603946966275188,
+      "learning_rate": 2.3339105163821896e-05,
+      "loss": 1.1485,
+      "step": 17034
+    },
+    {
+      "epoch": 0.7847698898972681,
+      "grad_norm": 0.8541907292216103,
+      "learning_rate": 2.332952510325279e-05,
+      "loss": 1.4149,
+      "step": 17035
+    },
+    {
+      "epoch": 0.7848159579859032,
+      "grad_norm": 0.7301008851339751,
+      "learning_rate": 2.331994674962036e-05,
+      "loss": 1.3104,
+      "step": 17036
+    },
+    {
+      "epoch": 0.7848620260745381,
+      "grad_norm": 1.9871751251043202,
+      "learning_rate": 2.3310370103137846e-05,
+      "loss": 1.0874,
+      "step": 17037
+    },
+    {
+      "epoch": 0.7849080941631732,
+      "grad_norm": 1.0234833520838598,
+      "learning_rate": 2.3300795164018398e-05,
+      "loss": 1.1495,
+      "step": 17038
+    },
+    {
+      "epoch": 0.7849541622518081,
+      "grad_norm": 2.369351081016177,
+      "learning_rate": 2.3291221932475294e-05,
+      "loss": 1.1566,
+      "step": 17039
+    },
+    {
+      "epoch": 0.7850002303404432,
+      "grad_norm": 0.8017891404240411,
+      "learning_rate": 2.3281650408721577e-05,
+      "loss": 1.1451,
+      "step": 17040
+    },
+    {
+      "epoch": 0.7850462984290781,
+      "grad_norm": 1.5553664430534333,
+      "learning_rate": 2.32720805929704e-05,
+      "loss": 1.4734,
+      "step": 17041
+    },
+    {
+      "epoch": 0.7850923665177132,
+      "grad_norm": 3.6492889811545783,
+      "learning_rate": 2.326251248543475e-05,
+      "loss": 1.1707,
+      "step": 17042
+    },
+    {
+      "epoch": 0.7851384346063481,
+      "grad_norm": 1.019809578111078,
+      "learning_rate": 2.32529460863277e-05,
+      "loss": 1.4041,
+      "step": 17043
+    },
+    {
+      "epoch": 0.7851845026949832,
+      "grad_norm": 1.1081681731226811,
+      "learning_rate": 2.324338139586223e-05,
+      "loss": 1.2874,
+      "step": 17044
+    },
+    {
+      "epoch": 0.7852305707836182,
+      "grad_norm": 1.1710921138103323,
+      "learning_rate": 2.3233818414251228e-05,
+      "loss": 1.3163,
+      "step": 17045
+    },
+    {
+      "epoch": 0.7852766388722532,
+      "grad_norm": 1.1403821208461031,
+      "learning_rate": 2.322425714170765e-05,
+      "loss": 1.22,
+      "step": 17046
+    },
+    {
+      "epoch": 0.7853227069608882,
+      "grad_norm": 2.8687091184788476,
+      "learning_rate": 2.3214697578444334e-05,
+      "loss": 1.1496,
+      "step": 17047
+    },
+    {
+      "epoch": 0.7853687750495232,
+      "grad_norm": 1.3094854928218287,
+      "learning_rate": 2.320513972467415e-05,
+      "loss": 0.9982,
+      "step": 17048
+    },
+    {
+      "epoch": 0.7854148431381582,
+      "grad_norm": 1.012381395953667,
+      "learning_rate": 2.3195583580609825e-05,
+      "loss": 1.3266,
+      "step": 17049
+    },
+    {
+      "epoch": 0.7854609112267932,
+      "grad_norm": 1.3655501646910773,
+      "learning_rate": 2.318602914646414e-05,
+      "loss": 1.0176,
+      "step": 17050
+    },
+    {
+      "epoch": 0.7855069793154282,
+      "grad_norm": 1.2129559186450696,
+      "learning_rate": 2.3176476422449834e-05,
+      "loss": 0.9705,
+      "step": 17051
+    },
+    {
+      "epoch": 0.7855530474040632,
+      "grad_norm": 1.2103501516389832,
+      "learning_rate": 2.316692540877954e-05,
+      "loss": 1.0771,
+      "step": 17052
+    },
+    {
+      "epoch": 0.7855991154926982,
+      "grad_norm": 1.4208253971902827,
+      "learning_rate": 2.3157376105665897e-05,
+      "loss": 1.1003,
+      "step": 17053
+    },
+    {
+      "epoch": 0.7856451835813332,
+      "grad_norm": 0.8038305193119345,
+      "learning_rate": 2.3147828513321524e-05,
+      "loss": 1.2175,
+      "step": 17054
+    },
+    {
+      "epoch": 0.7856912516699682,
+      "grad_norm": 1.2102093666463647,
+      "learning_rate": 2.3138282631959006e-05,
+      "loss": 1.4639,
+      "step": 17055
+    },
+    {
+      "epoch": 0.7857373197586032,
+      "grad_norm": 0.9154389737921058,
+      "learning_rate": 2.31287384617908e-05,
+      "loss": 1.2872,
+      "step": 17056
+    },
+    {
+      "epoch": 0.7857833878472382,
+      "grad_norm": 0.9264210555826157,
+      "learning_rate": 2.311919600302943e-05,
+      "loss": 1.4655,
+      "step": 17057
+    },
+    {
+      "epoch": 0.7858294559358732,
+      "grad_norm": 0.9438501270545303,
+      "learning_rate": 2.3109655255887342e-05,
+      "loss": 1.031,
+      "step": 17058
+    },
+    {
+      "epoch": 0.7858755240245082,
+      "grad_norm": 1.7044954322265093,
+      "learning_rate": 2.3100116220576974e-05,
+      "loss": 1.2891,
+      "step": 17059
+    },
+    {
+      "epoch": 0.7859215921131433,
+      "grad_norm": 1.369590051324067,
+      "learning_rate": 2.3090578897310623e-05,
+      "loss": 0.8917,
+      "step": 17060
+    },
+    {
+      "epoch": 0.7859676602017782,
+      "grad_norm": 1.239584357485959,
+      "learning_rate": 2.3081043286300676e-05,
+      "loss": 1.2127,
+      "step": 17061
+    },
+    {
+      "epoch": 0.7860137282904133,
+      "grad_norm": 1.2613708576012845,
+      "learning_rate": 2.3071509387759427e-05,
+      "loss": 1.1552,
+      "step": 17062
+    },
+    {
+      "epoch": 0.7860597963790482,
+      "grad_norm": 1.0246688689197487,
+      "learning_rate": 2.306197720189909e-05,
+      "loss": 0.9699,
+      "step": 17063
+    },
+    {
+      "epoch": 0.7861058644676833,
+      "grad_norm": 0.6788007844805417,
+      "learning_rate": 2.305244672893192e-05,
+      "loss": 1.2553,
+      "step": 17064
+    },
+    {
+      "epoch": 0.7861519325563182,
+      "grad_norm": 1.1742086915920633,
+      "learning_rate": 2.3042917969070075e-05,
+      "loss": 1.0934,
+      "step": 17065
+    },
+    {
+      "epoch": 0.7861980006449533,
+      "grad_norm": 1.5248753739380039,
+      "learning_rate": 2.3033390922525743e-05,
+      "loss": 1.1837,
+      "step": 17066
+    },
+    {
+      "epoch": 0.7862440687335882,
+      "grad_norm": 1.2888108012109216,
+      "learning_rate": 2.3023865589510963e-05,
+      "loss": 0.8731,
+      "step": 17067
+    },
+    {
+      "epoch": 0.7862901368222233,
+      "grad_norm": 1.202954967400806,
+      "learning_rate": 2.301434197023783e-05,
+      "loss": 1.3745,
+      "step": 17068
+    },
+    {
+      "epoch": 0.7863362049108582,
+      "grad_norm": 0.8942392386751269,
+      "learning_rate": 2.300482006491841e-05,
+      "loss": 1.1384,
+      "step": 17069
+    },
+    {
+      "epoch": 0.7863822729994933,
+      "grad_norm": 1.993559571162771,
+      "learning_rate": 2.299529987376461e-05,
+      "loss": 1.137,
+      "step": 17070
+    },
+    {
+      "epoch": 0.7864283410881282,
+      "grad_norm": 0.8924291039037686,
+      "learning_rate": 2.2985781396988426e-05,
+      "loss": 1.6037,
+      "step": 17071
+    },
+    {
+      "epoch": 0.7864744091767633,
+      "grad_norm": 0.9118595121844038,
+      "learning_rate": 2.297626463480177e-05,
+      "loss": 1.0445,
+      "step": 17072
+    },
+    {
+      "epoch": 0.7865204772653983,
+      "grad_norm": 1.0162623286799737,
+      "learning_rate": 2.2966749587416548e-05,
+      "loss": 1.3109,
+      "step": 17073
+    },
+    {
+      "epoch": 0.7865665453540333,
+      "grad_norm": 0.7415259620332182,
+      "learning_rate": 2.2957236255044524e-05,
+      "loss": 1.2333,
+      "step": 17074
+    },
+    {
+      "epoch": 0.7866126134426683,
+      "grad_norm": 1.40852511229601,
+      "learning_rate": 2.2947724637897537e-05,
+      "loss": 1.1327,
+      "step": 17075
+    },
+    {
+      "epoch": 0.7866586815313032,
+      "grad_norm": 4.0507243339806,
+      "learning_rate": 2.2938214736187358e-05,
+      "loss": 1.0066,
+      "step": 17076
+    },
+    {
+      "epoch": 0.7867047496199383,
+      "grad_norm": 0.8644724484391874,
+      "learning_rate": 2.2928706550125713e-05,
+      "loss": 1.2939,
+      "step": 17077
+    },
+    {
+      "epoch": 0.7867508177085732,
+      "grad_norm": 2.6235327105407706,
+      "learning_rate": 2.291920007992423e-05,
+      "loss": 1.2989,
+      "step": 17078
+    },
+    {
+      "epoch": 0.7867968857972083,
+      "grad_norm": 0.9472607650131492,
+      "learning_rate": 2.2909695325794613e-05,
+      "loss": 1.0103,
+      "step": 17079
+    },
+    {
+      "epoch": 0.7868429538858432,
+      "grad_norm": 4.816299172825617,
+      "learning_rate": 2.2900192287948464e-05,
+      "loss": 1.1252,
+      "step": 17080
+    },
+    {
+      "epoch": 0.7868890219744783,
+      "grad_norm": 1.44916722310091,
+      "learning_rate": 2.2890690966597306e-05,
+      "loss": 0.8749,
+      "step": 17081
+    },
+    {
+      "epoch": 0.7869350900631132,
+      "grad_norm": 1.0932711739257652,
+      "learning_rate": 2.2881191361952702e-05,
+      "loss": 1.1539,
+      "step": 17082
+    },
+    {
+      "epoch": 0.7869811581517483,
+      "grad_norm": 5.234362580741101,
+      "learning_rate": 2.287169347422614e-05,
+      "loss": 1.1729,
+      "step": 17083
+    },
+    {
+      "epoch": 0.7870272262403832,
+      "grad_norm": 0.9978209164307199,
+      "learning_rate": 2.286219730362911e-05,
+      "loss": 1.4427,
+      "step": 17084
+    },
+    {
+      "epoch": 0.7870732943290183,
+      "grad_norm": 0.8796238962257389,
+      "learning_rate": 2.2852702850372965e-05,
+      "loss": 1.3269,
+      "step": 17085
+    },
+    {
+      "epoch": 0.7871193624176533,
+      "grad_norm": 1.2067139198291348,
+      "learning_rate": 2.28432101146691e-05,
+      "loss": 1.1957,
+      "step": 17086
+    },
+    {
+      "epoch": 0.7871654305062883,
+      "grad_norm": 1.1720163523077851,
+      "learning_rate": 2.2833719096728913e-05,
+      "loss": 1.2149,
+      "step": 17087
+    },
+    {
+      "epoch": 0.7872114985949233,
+      "grad_norm": 1.08342844470881,
+      "learning_rate": 2.282422979676362e-05,
+      "loss": 0.9693,
+      "step": 17088
+    },
+    {
+      "epoch": 0.7872575666835583,
+      "grad_norm": 1.1423135126514243,
+      "learning_rate": 2.281474221498452e-05,
+      "loss": 1.2484,
+      "step": 17089
+    },
+    {
+      "epoch": 0.7873036347721933,
+      "grad_norm": 1.2011314745466046,
+      "learning_rate": 2.2805256351602844e-05,
+      "loss": 1.2782,
+      "step": 17090
+    },
+    {
+      "epoch": 0.7873497028608283,
+      "grad_norm": 1.0094063961460062,
+      "learning_rate": 2.2795772206829802e-05,
+      "loss": 1.1885,
+      "step": 17091
+    },
+    {
+      "epoch": 0.7873957709494633,
+      "grad_norm": 0.7160602973082728,
+      "learning_rate": 2.278628978087649e-05,
+      "loss": 1.3324,
+      "step": 17092
+    },
+    {
+      "epoch": 0.7874418390380983,
+      "grad_norm": 0.9292535928899148,
+      "learning_rate": 2.2776809073954043e-05,
+      "loss": 1.3521,
+      "step": 17093
+    },
+    {
+      "epoch": 0.7874879071267333,
+      "grad_norm": 4.204516964601903,
+      "learning_rate": 2.276733008627353e-05,
+      "loss": 1.2475,
+      "step": 17094
+    },
+    {
+      "epoch": 0.7875339752153683,
+      "grad_norm": 1.4678560710908564,
+      "learning_rate": 2.275785281804601e-05,
+      "loss": 1.0167,
+      "step": 17095
+    },
+    {
+      "epoch": 0.7875800433040033,
+      "grad_norm": 0.8032723236782048,
+      "learning_rate": 2.2748377269482456e-05,
+      "loss": 1.1687,
+      "step": 17096
+    },
+    {
+      "epoch": 0.7876261113926383,
+      "grad_norm": 0.9635926827744833,
+      "learning_rate": 2.273890344079378e-05,
+      "loss": 1.3635,
+      "step": 17097
+    },
+    {
+      "epoch": 0.7876721794812733,
+      "grad_norm": 0.850907877477039,
+      "learning_rate": 2.272943133219099e-05,
+      "loss": 1.1712,
+      "step": 17098
+    },
+    {
+      "epoch": 0.7877182475699083,
+      "grad_norm": 1.7294828731821912,
+      "learning_rate": 2.27199609438849e-05,
+      "loss": 1.134,
+      "step": 17099
+    },
+    {
+      "epoch": 0.7877643156585433,
+      "grad_norm": 1.1891947366314664,
+      "learning_rate": 2.2710492276086358e-05,
+      "loss": 1.2516,
+      "step": 17100
+    },
+    {
+      "epoch": 0.7878103837471784,
+      "grad_norm": 1.05366964852302,
+      "learning_rate": 2.2701025329006198e-05,
+      "loss": 1.174,
+      "step": 17101
+    },
+    {
+      "epoch": 0.7878564518358133,
+      "grad_norm": 12.422961790101853,
+      "learning_rate": 2.2691560102855192e-05,
+      "loss": 1.3307,
+      "step": 17102
+    },
+    {
+      "epoch": 0.7879025199244484,
+      "grad_norm": 1.1191609112121625,
+      "learning_rate": 2.2682096597844016e-05,
+      "loss": 1.1288,
+      "step": 17103
+    },
+    {
+      "epoch": 0.7879485880130833,
+      "grad_norm": 0.9541070920202003,
+      "learning_rate": 2.267263481418339e-05,
+      "loss": 1.4106,
+      "step": 17104
+    },
+    {
+      "epoch": 0.7879946561017184,
+      "grad_norm": 1.049275673410881,
+      "learning_rate": 2.2663174752083992e-05,
+      "loss": 1.1782,
+      "step": 17105
+    },
+    {
+      "epoch": 0.7880407241903533,
+      "grad_norm": 0.8671875378146837,
+      "learning_rate": 2.2653716411756367e-05,
+      "loss": 1.3967,
+      "step": 17106
+    },
+    {
+      "epoch": 0.7880867922789884,
+      "grad_norm": 0.9878093813662389,
+      "learning_rate": 2.264425979341116e-05,
+      "loss": 1.0988,
+      "step": 17107
+    },
+    {
+      "epoch": 0.7881328603676233,
+      "grad_norm": 0.9422597741551402,
+      "learning_rate": 2.263480489725881e-05,
+      "loss": 1.4119,
+      "step": 17108
+    },
+    {
+      "epoch": 0.7881789284562584,
+      "grad_norm": 1.6412843384020541,
+      "learning_rate": 2.2625351723509947e-05,
+      "loss": 1.1273,
+      "step": 17109
+    },
+    {
+      "epoch": 0.7882249965448933,
+      "grad_norm": 0.9724363129316193,
+      "learning_rate": 2.261590027237491e-05,
+      "loss": 1.2679,
+      "step": 17110
+    },
+    {
+      "epoch": 0.7882710646335284,
+      "grad_norm": 2.1980807387834673,
+      "learning_rate": 2.2606450544064184e-05,
+      "loss": 1.4529,
+      "step": 17111
+    },
+    {
+      "epoch": 0.7883171327221633,
+      "grad_norm": 1.3869345930031918,
+      "learning_rate": 2.2597002538788126e-05,
+      "loss": 1.2693,
+      "step": 17112
+    },
+    {
+      "epoch": 0.7883632008107984,
+      "grad_norm": 0.7207623620656789,
+      "learning_rate": 2.2587556256757126e-05,
+      "loss": 1.5114,
+      "step": 17113
+    },
+    {
+      "epoch": 0.7884092688994334,
+      "grad_norm": 1.0290830350390843,
+      "learning_rate": 2.257811169818145e-05,
+      "loss": 1.2168,
+      "step": 17114
+    },
+    {
+      "epoch": 0.7884553369880684,
+      "grad_norm": 1.1471949408417084,
+      "learning_rate": 2.2568668863271303e-05,
+      "loss": 1.3747,
+      "step": 17115
+    },
+    {
+      "epoch": 0.7885014050767034,
+      "grad_norm": 0.8709521995315169,
+      "learning_rate": 2.2559227752237045e-05,
+      "loss": 1.3742,
+      "step": 17116
+    },
+    {
+      "epoch": 0.7885474731653384,
+      "grad_norm": 1.6168139451586048,
+      "learning_rate": 2.254978836528876e-05,
+      "loss": 1.1634,
+      "step": 17117
+    },
+    {
+      "epoch": 0.7885935412539734,
+      "grad_norm": 2.6056615009795716,
+      "learning_rate": 2.254035070263667e-05,
+      "loss": 1.2589,
+      "step": 17118
+    },
+    {
+      "epoch": 0.7886396093426083,
+      "grad_norm": 0.8396428817622761,
+      "learning_rate": 2.2530914764490806e-05,
+      "loss": 1.3658,
+      "step": 17119
+    },
+    {
+      "epoch": 0.7886856774312434,
+      "grad_norm": 0.8342069536585105,
+      "learning_rate": 2.2521480551061336e-05,
+      "loss": 1.0904,
+      "step": 17120
+    },
+    {
+      "epoch": 0.7887317455198783,
+      "grad_norm": 0.8515139144461982,
+      "learning_rate": 2.2512048062558244e-05,
+      "loss": 1.3681,
+      "step": 17121
+    },
+    {
+      "epoch": 0.7887778136085134,
+      "grad_norm": 1.5671645559605674,
+      "learning_rate": 2.2502617299191553e-05,
+      "loss": 1.0694,
+      "step": 17122
+    },
+    {
+      "epoch": 0.7888238816971483,
+      "grad_norm": 1.2306871842499778,
+      "learning_rate": 2.249318826117118e-05,
+      "loss": 1.2174,
+      "step": 17123
+    },
+    {
+      "epoch": 0.7888699497857834,
+      "grad_norm": 0.8074488479689883,
+      "learning_rate": 2.248376094870709e-05,
+      "loss": 1.4862,
+      "step": 17124
+    },
+    {
+      "epoch": 0.7889160178744183,
+      "grad_norm": 5.885252937067111,
+      "learning_rate": 2.2474335362009168e-05,
+      "loss": 1.1126,
+      "step": 17125
+    },
+    {
+      "epoch": 0.7889620859630534,
+      "grad_norm": 2.32658431635386,
+      "learning_rate": 2.246491150128718e-05,
+      "loss": 0.9218,
+      "step": 17126
+    },
+    {
+      "epoch": 0.7890081540516884,
+      "grad_norm": 1.0780662161547585,
+      "learning_rate": 2.245548936675106e-05,
+      "loss": 1.3051,
+      "step": 17127
+    },
+    {
+      "epoch": 0.7890542221403234,
+      "grad_norm": 1.2829353516221833,
+      "learning_rate": 2.2446068958610477e-05,
+      "loss": 1.0314,
+      "step": 17128
+    },
+    {
+      "epoch": 0.7891002902289584,
+      "grad_norm": 0.7708044262337893,
+      "learning_rate": 2.2436650277075222e-05,
+      "loss": 1.4171,
+      "step": 17129
+    },
+    {
+      "epoch": 0.7891463583175934,
+      "grad_norm": 0.9712202284686973,
+      "learning_rate": 2.242723332235489e-05,
+      "loss": 1.148,
+      "step": 17130
+    },
+    {
+      "epoch": 0.7891924264062284,
+      "grad_norm": 2.0950450390006865,
+      "learning_rate": 2.241781809465926e-05,
+      "loss": 1.3506,
+      "step": 17131
+    },
+    {
+      "epoch": 0.7892384944948634,
+      "grad_norm": 0.9252565805427345,
+      "learning_rate": 2.2408404594197897e-05,
+      "loss": 1.097,
+      "step": 17132
+    },
+    {
+      "epoch": 0.7892845625834984,
+      "grad_norm": 0.9455252181894633,
+      "learning_rate": 2.239899282118032e-05,
+      "loss": 1.2392,
+      "step": 17133
+    },
+    {
+      "epoch": 0.7893306306721334,
+      "grad_norm": 0.9483209743067467,
+      "learning_rate": 2.2389582775816122e-05,
+      "loss": 1.1845,
+      "step": 17134
+    },
+    {
+      "epoch": 0.7893766987607684,
+      "grad_norm": 0.6430866312558673,
+      "learning_rate": 2.2380174458314784e-05,
+      "loss": 1.3718,
+      "step": 17135
+    },
+    {
+      "epoch": 0.7894227668494034,
+      "grad_norm": 0.7319294066367701,
+      "learning_rate": 2.23707678688858e-05,
+      "loss": 1.1813,
+      "step": 17136
+    },
+    {
+      "epoch": 0.7894688349380384,
+      "grad_norm": 1.4718954842405867,
+      "learning_rate": 2.236136300773851e-05,
+      "loss": 0.9879,
+      "step": 17137
+    },
+    {
+      "epoch": 0.7895149030266734,
+      "grad_norm": 0.8015995011035069,
+      "learning_rate": 2.2351959875082406e-05,
+      "loss": 1.0749,
+      "step": 17138
+    },
+    {
+      "epoch": 0.7895609711153084,
+      "grad_norm": 1.007261572265246,
+      "learning_rate": 2.2342558471126753e-05,
+      "loss": 1.2422,
+      "step": 17139
+    },
+    {
+      "epoch": 0.7896070392039435,
+      "grad_norm": 0.8537148539072924,
+      "learning_rate": 2.233315879608091e-05,
+      "loss": 1.3569,
+      "step": 17140
+    },
+    {
+      "epoch": 0.7896531072925784,
+      "grad_norm": 1.2150023649290822,
+      "learning_rate": 2.2323760850154086e-05,
+      "loss": 1.4874,
+      "step": 17141
+    },
+    {
+      "epoch": 0.7896991753812135,
+      "grad_norm": 1.2913799687206784,
+      "learning_rate": 2.231436463355554e-05,
+      "loss": 1.4284,
+      "step": 17142
+    },
+    {
+      "epoch": 0.7897452434698484,
+      "grad_norm": 1.1079307638674576,
+      "learning_rate": 2.2304970146494486e-05,
+      "loss": 1.5983,
+      "step": 17143
+    },
+    {
+      "epoch": 0.7897913115584835,
+      "grad_norm": 1.0195733004514982,
+      "learning_rate": 2.2295577389180032e-05,
+      "loss": 1.4456,
+      "step": 17144
+    },
+    {
+      "epoch": 0.7898373796471184,
+      "grad_norm": 1.178600787115619,
+      "learning_rate": 2.2286186361821316e-05,
+      "loss": 1.3994,
+      "step": 17145
+    },
+    {
+      "epoch": 0.7898834477357535,
+      "grad_norm": 0.7815496449191401,
+      "learning_rate": 2.2276797064627396e-05,
+      "loss": 0.9484,
+      "step": 17146
+    },
+    {
+      "epoch": 0.7899295158243884,
+      "grad_norm": 1.0600337285532306,
+      "learning_rate": 2.2267409497807366e-05,
+      "loss": 1.2512,
+      "step": 17147
+    },
+    {
+      "epoch": 0.7899755839130235,
+      "grad_norm": 0.7502779556918682,
+      "learning_rate": 2.225802366157015e-05,
+      "loss": 1.2014,
+      "step": 17148
+    },
+    {
+      "epoch": 0.7900216520016584,
+      "grad_norm": 0.9295919788889725,
+      "learning_rate": 2.224863955612473e-05,
+      "loss": 0.896,
+      "step": 17149
+    },
+    {
+      "epoch": 0.7900677200902935,
+      "grad_norm": 1.3326706099014314,
+      "learning_rate": 2.223925718168004e-05,
+      "loss": 1.1338,
+      "step": 17150
+    },
+    {
+      "epoch": 0.7901137881789284,
+      "grad_norm": 0.9266783653916184,
+      "learning_rate": 2.2229876538444983e-05,
+      "loss": 1.2771,
+      "step": 17151
+    },
+    {
+      "epoch": 0.7901598562675635,
+      "grad_norm": 1.068459821969249,
+      "learning_rate": 2.2220497626628358e-05,
+      "loss": 1.2553,
+      "step": 17152
+    },
+    {
+      "epoch": 0.7902059243561984,
+      "grad_norm": 0.9767044744614296,
+      "learning_rate": 2.2211120446438984e-05,
+      "loss": 1.4675,
+      "step": 17153
+    },
+    {
+      "epoch": 0.7902519924448335,
+      "grad_norm": 1.2596221055078909,
+      "learning_rate": 2.220174499808566e-05,
+      "loss": 1.2797,
+      "step": 17154
+    },
+    {
+      "epoch": 0.7902980605334685,
+      "grad_norm": 0.8570387139925577,
+      "learning_rate": 2.219237128177707e-05,
+      "loss": 1.5141,
+      "step": 17155
+    },
+    {
+      "epoch": 0.7903441286221035,
+      "grad_norm": 1.1996734013985524,
+      "learning_rate": 2.2182999297721908e-05,
+      "loss": 0.9583,
+      "step": 17156
+    },
+    {
+      "epoch": 0.7903901967107385,
+      "grad_norm": 0.8324388866005576,
+      "learning_rate": 2.2173629046128852e-05,
+      "loss": 1.52,
+      "step": 17157
+    },
+    {
+      "epoch": 0.7904362647993735,
+      "grad_norm": 1.2452996874583113,
+      "learning_rate": 2.2164260527206514e-05,
+      "loss": 1.3944,
+      "step": 17158
+    },
+    {
+      "epoch": 0.7904823328880085,
+      "grad_norm": 2.7359504799986065,
+      "learning_rate": 2.2154893741163442e-05,
+      "loss": 1.2618,
+      "step": 17159
+    },
+    {
+      "epoch": 0.7905284009766435,
+      "grad_norm": 1.782514277166812,
+      "learning_rate": 2.2145528688208184e-05,
+      "loss": 1.4372,
+      "step": 17160
+    },
+    {
+      "epoch": 0.7905744690652785,
+      "grad_norm": 0.8280902581933589,
+      "learning_rate": 2.2136165368549267e-05,
+      "loss": 0.9938,
+      "step": 17161
+    },
+    {
+      "epoch": 0.7906205371539134,
+      "grad_norm": 2.494281202745011,
+      "learning_rate": 2.2126803782395078e-05,
+      "loss": 1.1125,
+      "step": 17162
+    },
+    {
+      "epoch": 0.7906666052425485,
+      "grad_norm": 0.8960402912018902,
+      "learning_rate": 2.2117443929954095e-05,
+      "loss": 1.1844,
+      "step": 17163
+    },
+    {
+      "epoch": 0.7907126733311834,
+      "grad_norm": 3.098447847579405,
+      "learning_rate": 2.2108085811434675e-05,
+      "loss": 1.096,
+      "step": 17164
+    },
+    {
+      "epoch": 0.7907587414198185,
+      "grad_norm": 0.7033295734161332,
+      "learning_rate": 2.2098729427045194e-05,
+      "loss": 1.3474,
+      "step": 17165
+    },
+    {
+      "epoch": 0.7908048095084534,
+      "grad_norm": 1.036853563316973,
+      "learning_rate": 2.2089374776993908e-05,
+      "loss": 1.3889,
+      "step": 17166
+    },
+    {
+      "epoch": 0.7908508775970885,
+      "grad_norm": 0.9914106716792922,
+      "learning_rate": 2.208002186148911e-05,
+      "loss": 1.0685,
+      "step": 17167
+    },
+    {
+      "epoch": 0.7908969456857236,
+      "grad_norm": 1.560870307810293,
+      "learning_rate": 2.2070670680739004e-05,
+      "loss": 0.9643,
+      "step": 17168
+    },
+    {
+      "epoch": 0.7909430137743585,
+      "grad_norm": 1.186799200012558,
+      "learning_rate": 2.2061321234951838e-05,
+      "loss": 1.3824,
+      "step": 17169
+    },
+    {
+      "epoch": 0.7909890818629935,
+      "grad_norm": 0.8907651321618839,
+      "learning_rate": 2.205197352433569e-05,
+      "loss": 1.3676,
+      "step": 17170
+    },
+    {
+      "epoch": 0.7910351499516285,
+      "grad_norm": 0.9346893703841378,
+      "learning_rate": 2.2042627549098695e-05,
+      "loss": 1.2252,
+      "step": 17171
+    },
+    {
+      "epoch": 0.7910812180402635,
+      "grad_norm": 0.9660894961376797,
+      "learning_rate": 2.203328330944895e-05,
+      "loss": 1.2329,
+      "step": 17172
+    },
+    {
+      "epoch": 0.7911272861288985,
+      "grad_norm": 1.3655040534858212,
+      "learning_rate": 2.2023940805594446e-05,
+      "loss": 1.1918,
+      "step": 17173
+    },
+    {
+      "epoch": 0.7911733542175335,
+      "grad_norm": 0.8814791338859875,
+      "learning_rate": 2.2014600037743206e-05,
+      "loss": 1.3478,
+      "step": 17174
+    },
+    {
+      "epoch": 0.7912194223061685,
+      "grad_norm": 1.3991634443558172,
+      "learning_rate": 2.200526100610317e-05,
+      "loss": 0.9364,
+      "step": 17175
+    },
+    {
+      "epoch": 0.7912654903948035,
+      "grad_norm": 0.9267160158957174,
+      "learning_rate": 2.1995923710882305e-05,
+      "loss": 1.2422,
+      "step": 17176
+    },
+    {
+      "epoch": 0.7913115584834385,
+      "grad_norm": 0.914196324906719,
+      "learning_rate": 2.198658815228841e-05,
+      "loss": 1.5204,
+      "step": 17177
+    },
+    {
+      "epoch": 0.7913576265720735,
+      "grad_norm": 0.9559816397506923,
+      "learning_rate": 2.1977254330529374e-05,
+      "loss": 1.4169,
+      "step": 17178
+    },
+    {
+      "epoch": 0.7914036946607085,
+      "grad_norm": 1.1048292410217948,
+      "learning_rate": 2.196792224581301e-05,
+      "loss": 1.4209,
+      "step": 17179
+    },
+    {
+      "epoch": 0.7914497627493435,
+      "grad_norm": 1.0806564498753752,
+      "learning_rate": 2.1958591898347036e-05,
+      "loss": 1.1751,
+      "step": 17180
+    },
+    {
+      "epoch": 0.7914958308379786,
+      "grad_norm": 0.8112955542764572,
+      "learning_rate": 2.1949263288339206e-05,
+      "loss": 1.2898,
+      "step": 17181
+    },
+    {
+      "epoch": 0.7915418989266135,
+      "grad_norm": 1.1080046264710597,
+      "learning_rate": 2.1939936415997198e-05,
+      "loss": 1.3313,
+      "step": 17182
+    },
+    {
+      "epoch": 0.7915879670152486,
+      "grad_norm": 1.8696118096469843,
+      "learning_rate": 2.193061128152869e-05,
+      "loss": 1.2949,
+      "step": 17183
+    },
+    {
+      "epoch": 0.7916340351038835,
+      "grad_norm": 1.1184727916994979,
+      "learning_rate": 2.1921287885141228e-05,
+      "loss": 1.0217,
+      "step": 17184
+    },
+    {
+      "epoch": 0.7916801031925186,
+      "grad_norm": 0.9816053673331793,
+      "learning_rate": 2.1911966227042424e-05,
+      "loss": 1.3363,
+      "step": 17185
+    },
+    {
+      "epoch": 0.7917261712811535,
+      "grad_norm": 1.1089592803700368,
+      "learning_rate": 2.190264630743979e-05,
+      "loss": 1.3507,
+      "step": 17186
+    },
+    {
+      "epoch": 0.7917722393697886,
+      "grad_norm": 0.9841795803941078,
+      "learning_rate": 2.1893328126540857e-05,
+      "loss": 1.1146,
+      "step": 17187
+    },
+    {
+      "epoch": 0.7918183074584235,
+      "grad_norm": 0.8444842231353764,
+      "learning_rate": 2.188401168455303e-05,
+      "loss": 1.1828,
+      "step": 17188
+    },
+    {
+      "epoch": 0.7918643755470586,
+      "grad_norm": 0.7370758594987041,
+      "learning_rate": 2.187469698168374e-05,
+      "loss": 1.2233,
+      "step": 17189
+    },
+    {
+      "epoch": 0.7919104436356935,
+      "grad_norm": 0.7311689058685746,
+      "learning_rate": 2.1865384018140388e-05,
+      "loss": 1.2665,
+      "step": 17190
+    },
+    {
+      "epoch": 0.7919565117243286,
+      "grad_norm": 0.9359488374786649,
+      "learning_rate": 2.1856072794130267e-05,
+      "loss": 1.2655,
+      "step": 17191
+    },
+    {
+      "epoch": 0.7920025798129635,
+      "grad_norm": 1.0870838042036695,
+      "learning_rate": 2.1846763309860686e-05,
+      "loss": 1.2846,
+      "step": 17192
+    },
+    {
+      "epoch": 0.7920486479015986,
+      "grad_norm": 2.3621569972927676,
+      "learning_rate": 2.183745556553892e-05,
+      "loss": 1.2728,
+      "step": 17193
+    },
+    {
+      "epoch": 0.7920947159902335,
+      "grad_norm": 2.985674307378414,
+      "learning_rate": 2.1828149561372224e-05,
+      "loss": 1.0738,
+      "step": 17194
+    },
+    {
+      "epoch": 0.7921407840788686,
+      "grad_norm": 1.0369589816283566,
+      "learning_rate": 2.1818845297567704e-05,
+      "loss": 1.0556,
+      "step": 17195
+    },
+    {
+      "epoch": 0.7921868521675036,
+      "grad_norm": 1.1138000882586052,
+      "learning_rate": 2.1809542774332547e-05,
+      "loss": 1.3347,
+      "step": 17196
+    },
+    {
+      "epoch": 0.7922329202561386,
+      "grad_norm": 1.6364897141176282,
+      "learning_rate": 2.180024199187388e-05,
+      "loss": 1.0984,
+      "step": 17197
+    },
+    {
+      "epoch": 0.7922789883447736,
+      "grad_norm": 1.5038644579354674,
+      "learning_rate": 2.179094295039871e-05,
+      "loss": 1.0007,
+      "step": 17198
+    },
+    {
+      "epoch": 0.7923250564334086,
+      "grad_norm": 1.0944125912400893,
+      "learning_rate": 2.1781645650114102e-05,
+      "loss": 1.4931,
+      "step": 17199
+    },
+    {
+      "epoch": 0.7923711245220436,
+      "grad_norm": 0.7267628452568206,
+      "learning_rate": 2.1772350091227022e-05,
+      "loss": 1.4596,
+      "step": 17200
+    },
+    {
+      "epoch": 0.7924171926106786,
+      "grad_norm": 0.88729624913093,
+      "learning_rate": 2.1763056273944483e-05,
+      "loss": 1.1276,
+      "step": 17201
+    },
+    {
+      "epoch": 0.7924632606993136,
+      "grad_norm": 1.3739948506343689,
+      "learning_rate": 2.1753764198473304e-05,
+      "loss": 0.9123,
+      "step": 17202
+    },
+    {
+      "epoch": 0.7925093287879486,
+      "grad_norm": 1.4392245524319947,
+      "learning_rate": 2.1744473865020408e-05,
+      "loss": 1.1473,
+      "step": 17203
+    },
+    {
+      "epoch": 0.7925553968765836,
+      "grad_norm": 2.131128859471124,
+      "learning_rate": 2.1735185273792623e-05,
+      "loss": 1.1362,
+      "step": 17204
+    },
+    {
+      "epoch": 0.7926014649652185,
+      "grad_norm": 1.0303771084654414,
+      "learning_rate": 2.172589842499677e-05,
+      "loss": 0.9527,
+      "step": 17205
+    },
+    {
+      "epoch": 0.7926475330538536,
+      "grad_norm": 2.303828865482185,
+      "learning_rate": 2.1716613318839583e-05,
+      "loss": 1.5882,
+      "step": 17206
+    },
+    {
+      "epoch": 0.7926936011424885,
+      "grad_norm": 1.2215491871134456,
+      "learning_rate": 2.170732995552771e-05,
+      "loss": 1.0615,
+      "step": 17207
+    },
+    {
+      "epoch": 0.7927396692311236,
+      "grad_norm": 1.0715648465879672,
+      "learning_rate": 2.1698048335267952e-05,
+      "loss": 1.2425,
+      "step": 17208
+    },
+    {
+      "epoch": 0.7927857373197587,
+      "grad_norm": 1.8496911472063018,
+      "learning_rate": 2.168876845826686e-05,
+      "loss": 1.5984,
+      "step": 17209
+    },
+    {
+      "epoch": 0.7928318054083936,
+      "grad_norm": 0.753238850600364,
+      "learning_rate": 2.1679490324731067e-05,
+      "loss": 1.4282,
+      "step": 17210
+    },
+    {
+      "epoch": 0.7928778734970287,
+      "grad_norm": 1.238931905106109,
+      "learning_rate": 2.1670213934867123e-05,
+      "loss": 1.3154,
+      "step": 17211
+    },
+    {
+      "epoch": 0.7929239415856636,
+      "grad_norm": 0.8079844747980663,
+      "learning_rate": 2.1660939288881587e-05,
+      "loss": 1.3187,
+      "step": 17212
+    },
+    {
+      "epoch": 0.7929700096742986,
+      "grad_norm": 0.7153252407380748,
+      "learning_rate": 2.1651666386980895e-05,
+      "loss": 1.2709,
+      "step": 17213
+    },
+    {
+      "epoch": 0.7930160777629336,
+      "grad_norm": 0.8054622563169479,
+      "learning_rate": 2.1642395229371505e-05,
+      "loss": 1.3759,
+      "step": 17214
+    },
+    {
+      "epoch": 0.7930621458515686,
+      "grad_norm": 0.8948775909107944,
+      "learning_rate": 2.163312581625987e-05,
+      "loss": 1.38,
+      "step": 17215
+    },
+    {
+      "epoch": 0.7931082139402036,
+      "grad_norm": 2.1166855245318508,
+      "learning_rate": 2.162385814785227e-05,
+      "loss": 1.1996,
+      "step": 17216
+    },
+    {
+      "epoch": 0.7931542820288386,
+      "grad_norm": 1.9894880046296237,
+      "learning_rate": 2.1614592224355124e-05,
+      "loss": 1.1831,
+      "step": 17217
+    },
+    {
+      "epoch": 0.7932003501174736,
+      "grad_norm": 1.144622050156351,
+      "learning_rate": 2.1605328045974625e-05,
+      "loss": 1.1495,
+      "step": 17218
+    },
+    {
+      "epoch": 0.7932464182061086,
+      "grad_norm": 0.6924325301785154,
+      "learning_rate": 2.159606561291714e-05,
+      "loss": 1.259,
+      "step": 17219
+    },
+    {
+      "epoch": 0.7932924862947436,
+      "grad_norm": 1.0480181655680998,
+      "learning_rate": 2.1586804925388783e-05,
+      "loss": 1.3845,
+      "step": 17220
+    },
+    {
+      "epoch": 0.7933385543833786,
+      "grad_norm": 1.0188788234440753,
+      "learning_rate": 2.157754598359579e-05,
+      "loss": 1.4311,
+      "step": 17221
+    },
+    {
+      "epoch": 0.7933846224720137,
+      "grad_norm": 0.8860026175783965,
+      "learning_rate": 2.1568288787744217e-05,
+      "loss": 1.2262,
+      "step": 17222
+    },
+    {
+      "epoch": 0.7934306905606486,
+      "grad_norm": 2.4465076803419277,
+      "learning_rate": 2.155903333804027e-05,
+      "loss": 1.4068,
+      "step": 17223
+    },
+    {
+      "epoch": 0.7934767586492837,
+      "grad_norm": 1.0259840404571565,
+      "learning_rate": 2.1549779634689947e-05,
+      "loss": 1.0562,
+      "step": 17224
+    },
+    {
+      "epoch": 0.7935228267379186,
+      "grad_norm": 1.4700945603296607,
+      "learning_rate": 2.1540527677899213e-05,
+      "loss": 1.2655,
+      "step": 17225
+    },
+    {
+      "epoch": 0.7935688948265537,
+      "grad_norm": 1.8254153107730384,
+      "learning_rate": 2.1531277467874155e-05,
+      "loss": 1.2447,
+      "step": 17226
+    },
+    {
+      "epoch": 0.7936149629151886,
+      "grad_norm": 2.6397421300277237,
+      "learning_rate": 2.152202900482063e-05,
+      "loss": 1.2396,
+      "step": 17227
+    },
+    {
+      "epoch": 0.7936610310038237,
+      "grad_norm": 1.2460784027913545,
+      "learning_rate": 2.15127822889446e-05,
+      "loss": 1.2207,
+      "step": 17228
+    },
+    {
+      "epoch": 0.7937070990924586,
+      "grad_norm": 1.0418356160717508,
+      "learning_rate": 2.150353732045184e-05,
+      "loss": 1.1185,
+      "step": 17229
+    },
+    {
+      "epoch": 0.7937531671810937,
+      "grad_norm": 0.9347549567785363,
+      "learning_rate": 2.1494294099548284e-05,
+      "loss": 1.0778,
+      "step": 17230
+    },
+    {
+      "epoch": 0.7937992352697286,
+      "grad_norm": 1.2128276068994461,
+      "learning_rate": 2.148505262643963e-05,
+      "loss": 1.4177,
+      "step": 17231
+    },
+    {
+      "epoch": 0.7938453033583637,
+      "grad_norm": 1.2000248648604914,
+      "learning_rate": 2.147581290133168e-05,
+      "loss": 1.1701,
+      "step": 17232
+    },
+    {
+      "epoch": 0.7938913714469986,
+      "grad_norm": 1.3515176830995794,
+      "learning_rate": 2.1466574924430093e-05,
+      "loss": 1.2589,
+      "step": 17233
+    },
+    {
+      "epoch": 0.7939374395356337,
+      "grad_norm": 0.8861962342853613,
+      "learning_rate": 2.1457338695940565e-05,
+      "loss": 1.4279,
+      "step": 17234
+    },
+    {
+      "epoch": 0.7939835076242686,
+      "grad_norm": 1.6245233541915094,
+      "learning_rate": 2.1448104216068743e-05,
+      "loss": 1.2436,
+      "step": 17235
+    },
+    {
+      "epoch": 0.7940295757129037,
+      "grad_norm": 1.0114665123739146,
+      "learning_rate": 2.143887148502013e-05,
+      "loss": 1.285,
+      "step": 17236
+    },
+    {
+      "epoch": 0.7940756438015387,
+      "grad_norm": 0.9252924931973724,
+      "learning_rate": 2.1429640503000404e-05,
+      "loss": 1.337,
+      "step": 17237
+    },
+    {
+      "epoch": 0.7941217118901737,
+      "grad_norm": 0.9792329628108565,
+      "learning_rate": 2.1420411270214978e-05,
+      "loss": 1.5423,
+      "step": 17238
+    },
+    {
+      "epoch": 0.7941677799788087,
+      "grad_norm": 1.5981608068491178,
+      "learning_rate": 2.1411183786869392e-05,
+      "loss": 1.0401,
+      "step": 17239
+    },
+    {
+      "epoch": 0.7942138480674437,
+      "grad_norm": 1.2956734641018375,
+      "learning_rate": 2.1401958053169e-05,
+      "loss": 1.3533,
+      "step": 17240
+    },
+    {
+      "epoch": 0.7942599161560787,
+      "grad_norm": 1.226300194002348,
+      "learning_rate": 2.1392734069319288e-05,
+      "loss": 1.2414,
+      "step": 17241
+    },
+    {
+      "epoch": 0.7943059842447137,
+      "grad_norm": 0.9871799739053049,
+      "learning_rate": 2.1383511835525584e-05,
+      "loss": 1.2769,
+      "step": 17242
+    },
+    {
+      "epoch": 0.7943520523333487,
+      "grad_norm": 0.9308569714184504,
+      "learning_rate": 2.137429135199315e-05,
+      "loss": 1.1057,
+      "step": 17243
+    },
+    {
+      "epoch": 0.7943981204219837,
+      "grad_norm": 0.9331601783316456,
+      "learning_rate": 2.1365072618927306e-05,
+      "loss": 1.2873,
+      "step": 17244
+    },
+    {
+      "epoch": 0.7944441885106187,
+      "grad_norm": 1.6596039469092565,
+      "learning_rate": 2.1355855636533297e-05,
+      "loss": 1.2036,
+      "step": 17245
+    },
+    {
+      "epoch": 0.7944902565992537,
+      "grad_norm": 1.2767778822669815,
+      "learning_rate": 2.1346640405016338e-05,
+      "loss": 1.4678,
+      "step": 17246
+    },
+    {
+      "epoch": 0.7945363246878887,
+      "grad_norm": 1.3611402463266657,
+      "learning_rate": 2.133742692458154e-05,
+      "loss": 1.3084,
+      "step": 17247
+    },
+    {
+      "epoch": 0.7945823927765236,
+      "grad_norm": 1.3414093827333746,
+      "learning_rate": 2.1328215195434063e-05,
+      "loss": 1.2489,
+      "step": 17248
+    },
+    {
+      "epoch": 0.7946284608651587,
+      "grad_norm": 0.9116238908121357,
+      "learning_rate": 2.1319005217778975e-05,
+      "loss": 1.3141,
+      "step": 17249
+    },
+    {
+      "epoch": 0.7946745289537938,
+      "grad_norm": 1.0607523643363406,
+      "learning_rate": 2.130979699182135e-05,
+      "loss": 1.0612,
+      "step": 17250
+    },
+    {
+      "epoch": 0.7947205970424287,
+      "grad_norm": 1.7623227594810138,
+      "learning_rate": 2.130059051776615e-05,
+      "loss": 1.111,
+      "step": 17251
+    },
+    {
+      "epoch": 0.7947666651310638,
+      "grad_norm": 0.9837705580579491,
+      "learning_rate": 2.1291385795818364e-05,
+      "loss": 1.2356,
+      "step": 17252
+    },
+    {
+      "epoch": 0.7948127332196987,
+      "grad_norm": 2.800561286820635,
+      "learning_rate": 2.1282182826182938e-05,
+      "loss": 1.1714,
+      "step": 17253
+    },
+    {
+      "epoch": 0.7948588013083338,
+      "grad_norm": 2.5065146743528053,
+      "learning_rate": 2.127298160906471e-05,
+      "loss": 1.0024,
+      "step": 17254
+    },
+    {
+      "epoch": 0.7949048693969687,
+      "grad_norm": 1.430807795628969,
+      "learning_rate": 2.1263782144668564e-05,
+      "loss": 0.9926,
+      "step": 17255
+    },
+    {
+      "epoch": 0.7949509374856037,
+      "grad_norm": 0.8642709522262116,
+      "learning_rate": 2.1254584433199298e-05,
+      "loss": 1.4085,
+      "step": 17256
+    },
+    {
+      "epoch": 0.7949970055742387,
+      "grad_norm": 0.7130350832939675,
+      "learning_rate": 2.1245388474861726e-05,
+      "loss": 1.393,
+      "step": 17257
+    },
+    {
+      "epoch": 0.7950430736628737,
+      "grad_norm": 1.2028665843227346,
+      "learning_rate": 2.123619426986051e-05,
+      "loss": 1.475,
+      "step": 17258
+    },
+    {
+      "epoch": 0.7950891417515087,
+      "grad_norm": 0.9686712250528604,
+      "learning_rate": 2.1227001818400382e-05,
+      "loss": 1.2779,
+      "step": 17259
+    },
+    {
+      "epoch": 0.7951352098401437,
+      "grad_norm": 1.6900684431366066,
+      "learning_rate": 2.1217811120686026e-05,
+      "loss": 0.9559,
+      "step": 17260
+    },
+    {
+      "epoch": 0.7951812779287787,
+      "grad_norm": 1.1238894481149841,
+      "learning_rate": 2.120862217692199e-05,
+      "loss": 1.3993,
+      "step": 17261
+    },
+    {
+      "epoch": 0.7952273460174137,
+      "grad_norm": 0.6911809934023814,
+      "learning_rate": 2.119943498731287e-05,
+      "loss": 1.2921,
+      "step": 17262
+    },
+    {
+      "epoch": 0.7952734141060488,
+      "grad_norm": 2.910233948723144,
+      "learning_rate": 2.1190249552063234e-05,
+      "loss": 1.2993,
+      "step": 17263
+    },
+    {
+      "epoch": 0.7953194821946837,
+      "grad_norm": 1.5264445153606003,
+      "learning_rate": 2.1181065871377583e-05,
+      "loss": 1.4565,
+      "step": 17264
+    },
+    {
+      "epoch": 0.7953655502833188,
+      "grad_norm": 1.1365528146998651,
+      "learning_rate": 2.1171883945460324e-05,
+      "loss": 1.2692,
+      "step": 17265
+    },
+    {
+      "epoch": 0.7954116183719537,
+      "grad_norm": 0.9565941111350176,
+      "learning_rate": 2.1162703774515914e-05,
+      "loss": 1.2691,
+      "step": 17266
+    },
+    {
+      "epoch": 0.7954576864605888,
+      "grad_norm": 1.0679663398236314,
+      "learning_rate": 2.1153525358748714e-05,
+      "loss": 1.429,
+      "step": 17267
+    },
+    {
+      "epoch": 0.7955037545492237,
+      "grad_norm": 1.079728127842092,
+      "learning_rate": 2.1144348698363125e-05,
+      "loss": 1.1441,
+      "step": 17268
+    },
+    {
+      "epoch": 0.7955498226378588,
+      "grad_norm": 0.8868845757155501,
+      "learning_rate": 2.113517379356337e-05,
+      "loss": 1.0697,
+      "step": 17269
+    },
+    {
+      "epoch": 0.7955958907264937,
+      "grad_norm": 0.934304103651799,
+      "learning_rate": 2.1126000644553744e-05,
+      "loss": 1.0411,
+      "step": 17270
+    },
+    {
+      "epoch": 0.7956419588151288,
+      "grad_norm": 1.3599499607661556,
+      "learning_rate": 2.1116829251538496e-05,
+      "loss": 1.4912,
+      "step": 17271
+    },
+    {
+      "epoch": 0.7956880269037637,
+      "grad_norm": 1.0035305531298369,
+      "learning_rate": 2.1107659614721763e-05,
+      "loss": 1.1308,
+      "step": 17272
+    },
+    {
+      "epoch": 0.7957340949923988,
+      "grad_norm": 1.7048699003199141,
+      "learning_rate": 2.1098491734307722e-05,
+      "loss": 1.393,
+      "step": 17273
+    },
+    {
+      "epoch": 0.7957801630810337,
+      "grad_norm": 0.902328238152007,
+      "learning_rate": 2.1089325610500477e-05,
+      "loss": 1.0302,
+      "step": 17274
+    },
+    {
+      "epoch": 0.7958262311696688,
+      "grad_norm": 0.8474488966282266,
+      "learning_rate": 2.1080161243504126e-05,
+      "loss": 1.3822,
+      "step": 17275
+    },
+    {
+      "epoch": 0.7958722992583038,
+      "grad_norm": 0.8601173723039917,
+      "learning_rate": 2.1070998633522633e-05,
+      "loss": 1.0358,
+      "step": 17276
+    },
+    {
+      "epoch": 0.7959183673469388,
+      "grad_norm": 0.903411630154089,
+      "learning_rate": 2.106183778076002e-05,
+      "loss": 1.1775,
+      "step": 17277
+    },
+    {
+      "epoch": 0.7959644354355738,
+      "grad_norm": 1.1289453679094685,
+      "learning_rate": 2.1052678685420236e-05,
+      "loss": 1.1386,
+      "step": 17278
+    },
+    {
+      "epoch": 0.7960105035242088,
+      "grad_norm": 2.9514334252866643,
+      "learning_rate": 2.1043521347707228e-05,
+      "loss": 1.0572,
+      "step": 17279
+    },
+    {
+      "epoch": 0.7960565716128438,
+      "grad_norm": 0.9141063995579057,
+      "learning_rate": 2.1034365767824803e-05,
+      "loss": 1.1533,
+      "step": 17280
+    },
+    {
+      "epoch": 0.7961026397014788,
+      "grad_norm": 0.9021051050918527,
+      "learning_rate": 2.102521194597683e-05,
+      "loss": 1.2556,
+      "step": 17281
+    },
+    {
+      "epoch": 0.7961487077901138,
+      "grad_norm": 4.706231657681648,
+      "learning_rate": 2.1016059882367135e-05,
+      "loss": 1.2773,
+      "step": 17282
+    },
+    {
+      "epoch": 0.7961947758787488,
+      "grad_norm": 0.9030320820846123,
+      "learning_rate": 2.1006909577199406e-05,
+      "loss": 1.3488,
+      "step": 17283
+    },
+    {
+      "epoch": 0.7962408439673838,
+      "grad_norm": 0.8374133397350819,
+      "learning_rate": 2.0997761030677386e-05,
+      "loss": 1.1583,
+      "step": 17284
+    },
+    {
+      "epoch": 0.7962869120560188,
+      "grad_norm": 1.5960703744722327,
+      "learning_rate": 2.0988614243004766e-05,
+      "loss": 1.3346,
+      "step": 17285
+    },
+    {
+      "epoch": 0.7963329801446538,
+      "grad_norm": 1.1901846822705138,
+      "learning_rate": 2.0979469214385194e-05,
+      "loss": 1.0788,
+      "step": 17286
+    },
+    {
+      "epoch": 0.7963790482332888,
+      "grad_norm": 1.773891330475572,
+      "learning_rate": 2.0970325945022228e-05,
+      "loss": 0.9424,
+      "step": 17287
+    },
+    {
+      "epoch": 0.7964251163219238,
+      "grad_norm": 1.2149885889233543,
+      "learning_rate": 2.096118443511944e-05,
+      "loss": 1.0455,
+      "step": 17288
+    },
+    {
+      "epoch": 0.7964711844105588,
+      "grad_norm": 0.9276132978835872,
+      "learning_rate": 2.095204468488039e-05,
+      "loss": 1.0391,
+      "step": 17289
+    },
+    {
+      "epoch": 0.7965172524991938,
+      "grad_norm": 1.475219319048815,
+      "learning_rate": 2.0942906694508502e-05,
+      "loss": 1.3591,
+      "step": 17290
+    },
+    {
+      "epoch": 0.7965633205878289,
+      "grad_norm": 0.8993237024061839,
+      "learning_rate": 2.0933770464207236e-05,
+      "loss": 1.6556,
+      "step": 17291
+    },
+    {
+      "epoch": 0.7966093886764638,
+      "grad_norm": 0.6250734178580977,
+      "learning_rate": 2.0924635994180007e-05,
+      "loss": 1.5239,
+      "step": 17292
+    },
+    {
+      "epoch": 0.7966554567650989,
+      "grad_norm": 0.6969431692314032,
+      "learning_rate": 2.09155032846302e-05,
+      "loss": 1.1551,
+      "step": 17293
+    },
+    {
+      "epoch": 0.7967015248537338,
+      "grad_norm": 1.1397993252284686,
+      "learning_rate": 2.0906372335761082e-05,
+      "loss": 1.0791,
+      "step": 17294
+    },
+    {
+      "epoch": 0.7967475929423689,
+      "grad_norm": 0.7395012299391809,
+      "learning_rate": 2.0897243147775958e-05,
+      "loss": 1.3853,
+      "step": 17295
+    },
+    {
+      "epoch": 0.7967936610310038,
+      "grad_norm": 0.8246894274529509,
+      "learning_rate": 2.0888115720878098e-05,
+      "loss": 1.1823,
+      "step": 17296
+    },
+    {
+      "epoch": 0.7968397291196389,
+      "grad_norm": 2.2456775209316238,
+      "learning_rate": 2.087899005527072e-05,
+      "loss": 1.0476,
+      "step": 17297
+    },
+    {
+      "epoch": 0.7968857972082738,
+      "grad_norm": 1.0075776023758694,
+      "learning_rate": 2.0869866151156924e-05,
+      "loss": 1.0042,
+      "step": 17298
+    },
+    {
+      "epoch": 0.7969318652969088,
+      "grad_norm": 1.3174129986282754,
+      "learning_rate": 2.0860744008739898e-05,
+      "loss": 1.0854,
+      "step": 17299
+    },
+    {
+      "epoch": 0.7969779333855438,
+      "grad_norm": 18.363425573463566,
+      "learning_rate": 2.0851623628222726e-05,
+      "loss": 0.9951,
+      "step": 17300
+    },
+    {
+      "epoch": 0.7970240014741788,
+      "grad_norm": 0.665668035586526,
+      "learning_rate": 2.084250500980841e-05,
+      "loss": 1.204,
+      "step": 17301
+    },
+    {
+      "epoch": 0.7970700695628138,
+      "grad_norm": 1.3174660833474057,
+      "learning_rate": 2.0833388153700007e-05,
+      "loss": 1.3676,
+      "step": 17302
+    },
+    {
+      "epoch": 0.7971161376514488,
+      "grad_norm": 0.917877084777953,
+      "learning_rate": 2.0824273060100473e-05,
+      "loss": 1.5238,
+      "step": 17303
+    },
+    {
+      "epoch": 0.7971622057400839,
+      "grad_norm": 1.407758676706409,
+      "learning_rate": 2.0815159729212773e-05,
+      "loss": 1.5504,
+      "step": 17304
+    },
+    {
+      "epoch": 0.7972082738287188,
+      "grad_norm": 1.5493588934044602,
+      "learning_rate": 2.0806048161239733e-05,
+      "loss": 1.4136,
+      "step": 17305
+    },
+    {
+      "epoch": 0.7972543419173539,
+      "grad_norm": 1.1604481274091198,
+      "learning_rate": 2.0796938356384245e-05,
+      "loss": 1.4219,
+      "step": 17306
+    },
+    {
+      "epoch": 0.7973004100059888,
+      "grad_norm": 1.0489279981713202,
+      "learning_rate": 2.0787830314849143e-05,
+      "loss": 1.2382,
+      "step": 17307
+    },
+    {
+      "epoch": 0.7973464780946239,
+      "grad_norm": 0.8045995422352762,
+      "learning_rate": 2.0778724036837162e-05,
+      "loss": 1.3554,
+      "step": 17308
+    },
+    {
+      "epoch": 0.7973925461832588,
+      "grad_norm": 1.0784899273853856,
+      "learning_rate": 2.0769619522551055e-05,
+      "loss": 1.1205,
+      "step": 17309
+    },
+    {
+      "epoch": 0.7974386142718939,
+      "grad_norm": 0.9642155955847522,
+      "learning_rate": 2.0760516772193506e-05,
+      "loss": 1.1622,
+      "step": 17310
+    },
+    {
+      "epoch": 0.7974846823605288,
+      "grad_norm": 0.8491726270161121,
+      "learning_rate": 2.0751415785967222e-05,
+      "loss": 1.143,
+      "step": 17311
+    },
+    {
+      "epoch": 0.7975307504491639,
+      "grad_norm": 2.443316650440481,
+      "learning_rate": 2.074231656407476e-05,
+      "loss": 1.0621,
+      "step": 17312
+    },
+    {
+      "epoch": 0.7975768185377988,
+      "grad_norm": 0.7121317507157379,
+      "learning_rate": 2.0733219106718703e-05,
+      "loss": 1.2738,
+      "step": 17313
+    },
+    {
+      "epoch": 0.7976228866264339,
+      "grad_norm": 1.1856029586327708,
+      "learning_rate": 2.0724123414101624e-05,
+      "loss": 0.8158,
+      "step": 17314
+    },
+    {
+      "epoch": 0.7976689547150688,
+      "grad_norm": 1.0959549344853616,
+      "learning_rate": 2.071502948642604e-05,
+      "loss": 1.4752,
+      "step": 17315
+    },
+    {
+      "epoch": 0.7977150228037039,
+      "grad_norm": 0.9953236253392147,
+      "learning_rate": 2.0705937323894363e-05,
+      "loss": 1.2282,
+      "step": 17316
+    },
+    {
+      "epoch": 0.7977610908923389,
+      "grad_norm": 0.8158802864829869,
+      "learning_rate": 2.069684692670899e-05,
+      "loss": 1.4804,
+      "step": 17317
+    },
+    {
+      "epoch": 0.7978071589809739,
+      "grad_norm": 1.745608474409768,
+      "learning_rate": 2.06877582950724e-05,
+      "loss": 1.1696,
+      "step": 17318
+    },
+    {
+      "epoch": 0.7978532270696089,
+      "grad_norm": 0.8329503601879779,
+      "learning_rate": 2.0678671429186847e-05,
+      "loss": 1.2288,
+      "step": 17319
+    },
+    {
+      "epoch": 0.7978992951582439,
+      "grad_norm": 1.5648487410609453,
+      "learning_rate": 2.0669586329254698e-05,
+      "loss": 1.2876,
+      "step": 17320
+    },
+    {
+      "epoch": 0.7979453632468789,
+      "grad_norm": 0.9018288162291044,
+      "learning_rate": 2.0660502995478126e-05,
+      "loss": 1.189,
+      "step": 17321
+    },
+    {
+      "epoch": 0.7979914313355139,
+      "grad_norm": 1.5264147498974847,
+      "learning_rate": 2.065142142805948e-05,
+      "loss": 0.9719,
+      "step": 17322
+    },
+    {
+      "epoch": 0.7980374994241489,
+      "grad_norm": 1.2722243905313113,
+      "learning_rate": 2.0642341627200856e-05,
+      "loss": 1.3291,
+      "step": 17323
+    },
+    {
+      "epoch": 0.7980835675127839,
+      "grad_norm": 1.1910614368849224,
+      "learning_rate": 2.0633263593104413e-05,
+      "loss": 1.1642,
+      "step": 17324
+    },
+    {
+      "epoch": 0.7981296356014189,
+      "grad_norm": 0.872031421642431,
+      "learning_rate": 2.062418732597231e-05,
+      "loss": 1.2714,
+      "step": 17325
+    },
+    {
+      "epoch": 0.7981757036900539,
+      "grad_norm": 2.9752535841549963,
+      "learning_rate": 2.0615112826006545e-05,
+      "loss": 0.8352,
+      "step": 17326
+    },
+    {
+      "epoch": 0.7982217717786889,
+      "grad_norm": 0.9941491261459784,
+      "learning_rate": 2.0606040093409206e-05,
+      "loss": 1.1613,
+      "step": 17327
+    },
+    {
+      "epoch": 0.7982678398673239,
+      "grad_norm": 1.349635469892607,
+      "learning_rate": 2.0596969128382203e-05,
+      "loss": 1.1909,
+      "step": 17328
+    },
+    {
+      "epoch": 0.7983139079559589,
+      "grad_norm": 1.0903242255034058,
+      "learning_rate": 2.0587899931127585e-05,
+      "loss": 1.3752,
+      "step": 17329
+    },
+    {
+      "epoch": 0.7983599760445939,
+      "grad_norm": 0.8179837314533975,
+      "learning_rate": 2.0578832501847178e-05,
+      "loss": 1.4703,
+      "step": 17330
+    },
+    {
+      "epoch": 0.7984060441332289,
+      "grad_norm": 0.9368165901943409,
+      "learning_rate": 2.0569766840742922e-05,
+      "loss": 1.2664,
+      "step": 17331
+    },
+    {
+      "epoch": 0.798452112221864,
+      "grad_norm": 1.8524481340889014,
+      "learning_rate": 2.0560702948016564e-05,
+      "loss": 1.2211,
+      "step": 17332
+    },
+    {
+      "epoch": 0.7984981803104989,
+      "grad_norm": 1.1448036222747657,
+      "learning_rate": 2.055164082386999e-05,
+      "loss": 1.6164,
+      "step": 17333
+    },
+    {
+      "epoch": 0.798544248399134,
+      "grad_norm": 1.1495952571195265,
+      "learning_rate": 2.054258046850491e-05,
+      "loss": 1.2187,
+      "step": 17334
+    },
+    {
+      "epoch": 0.7985903164877689,
+      "grad_norm": 0.8995398669250797,
+      "learning_rate": 2.0533521882122984e-05,
+      "loss": 1.25,
+      "step": 17335
+    },
+    {
+      "epoch": 0.798636384576404,
+      "grad_norm": 1.1009297830729539,
+      "learning_rate": 2.052446506492599e-05,
+      "loss": 0.9145,
+      "step": 17336
+    },
+    {
+      "epoch": 0.7986824526650389,
+      "grad_norm": 1.1547768281332866,
+      "learning_rate": 2.0515410017115487e-05,
+      "loss": 1.1091,
+      "step": 17337
+    },
+    {
+      "epoch": 0.798728520753674,
+      "grad_norm": 0.9576973870271627,
+      "learning_rate": 2.050635673889312e-05,
+      "loss": 1.5613,
+      "step": 17338
+    },
+    {
+      "epoch": 0.7987745888423089,
+      "grad_norm": 0.7596058932549566,
+      "learning_rate": 2.0497305230460363e-05,
+      "loss": 0.8875,
+      "step": 17339
+    },
+    {
+      "epoch": 0.798820656930944,
+      "grad_norm": 1.0362145550260473,
+      "learning_rate": 2.048825549201884e-05,
+      "loss": 1.3145,
+      "step": 17340
+    },
+    {
+      "epoch": 0.7988667250195789,
+      "grad_norm": 2.2052085732014897,
+      "learning_rate": 2.047920752376995e-05,
+      "loss": 1.483,
+      "step": 17341
+    },
+    {
+      "epoch": 0.798912793108214,
+      "grad_norm": 0.8685422569052207,
+      "learning_rate": 2.047016132591517e-05,
+      "loss": 1.1551,
+      "step": 17342
+    },
+    {
+      "epoch": 0.7989588611968489,
+      "grad_norm": 1.3701910062090457,
+      "learning_rate": 2.0461116898655875e-05,
+      "loss": 1.0951,
+      "step": 17343
+    },
+    {
+      "epoch": 0.799004929285484,
+      "grad_norm": 1.0888980934249601,
+      "learning_rate": 2.045207424219342e-05,
+      "loss": 1.2014,
+      "step": 17344
+    },
+    {
+      "epoch": 0.799050997374119,
+      "grad_norm": 6.115509008691398,
+      "learning_rate": 2.044303335672917e-05,
+      "loss": 1.4042,
+      "step": 17345
+    },
+    {
+      "epoch": 0.7990970654627539,
+      "grad_norm": 1.244699534339347,
+      "learning_rate": 2.0433994242464338e-05,
+      "loss": 1.0001,
+      "step": 17346
+    },
+    {
+      "epoch": 0.799143133551389,
+      "grad_norm": 0.9540037935419112,
+      "learning_rate": 2.0424956899600213e-05,
+      "loss": 1.1955,
+      "step": 17347
+    },
+    {
+      "epoch": 0.7991892016400239,
+      "grad_norm": 2.047019482586443,
+      "learning_rate": 2.0415921328337963e-05,
+      "loss": 1.3345,
+      "step": 17348
+    },
+    {
+      "epoch": 0.799235269728659,
+      "grad_norm": 1.0518393747924613,
+      "learning_rate": 2.0406887528878804e-05,
+      "loss": 1.2355,
+      "step": 17349
+    },
+    {
+      "epoch": 0.7992813378172939,
+      "grad_norm": 1.087482130207275,
+      "learning_rate": 2.0397855501423756e-05,
+      "loss": 1.0098,
+      "step": 17350
+    },
+    {
+      "epoch": 0.799327405905929,
+      "grad_norm": 0.9418357836196456,
+      "learning_rate": 2.0388825246174038e-05,
+      "loss": 1.2754,
+      "step": 17351
+    },
+    {
+      "epoch": 0.7993734739945639,
+      "grad_norm": 1.1915518716739333,
+      "learning_rate": 2.0379796763330605e-05,
+      "loss": 1.2465,
+      "step": 17352
+    },
+    {
+      "epoch": 0.799419542083199,
+      "grad_norm": 0.7208559961484704,
+      "learning_rate": 2.0370770053094467e-05,
+      "loss": 1.3132,
+      "step": 17353
+    },
+    {
+      "epoch": 0.7994656101718339,
+      "grad_norm": 1.0929622824603604,
+      "learning_rate": 2.0361745115666586e-05,
+      "loss": 0.7224,
+      "step": 17354
+    },
+    {
+      "epoch": 0.799511678260469,
+      "grad_norm": 1.2329403237713359,
+      "learning_rate": 2.035272195124791e-05,
+      "loss": 1.3489,
+      "step": 17355
+    },
+    {
+      "epoch": 0.7995577463491039,
+      "grad_norm": 2.7273257337224894,
+      "learning_rate": 2.0343700560039336e-05,
+      "loss": 1.1691,
+      "step": 17356
+    },
+    {
+      "epoch": 0.799603814437739,
+      "grad_norm": 1.089495353067806,
+      "learning_rate": 2.033468094224166e-05,
+      "loss": 1.0199,
+      "step": 17357
+    },
+    {
+      "epoch": 0.799649882526374,
+      "grad_norm": 0.8017496514819727,
+      "learning_rate": 2.0325663098055715e-05,
+      "loss": 1.5281,
+      "step": 17358
+    },
+    {
+      "epoch": 0.799695950615009,
+      "grad_norm": 0.876485332340774,
+      "learning_rate": 2.0316647027682266e-05,
+      "loss": 0.8468,
+      "step": 17359
+    },
+    {
+      "epoch": 0.799742018703644,
+      "grad_norm": 0.7781165005632348,
+      "learning_rate": 2.030763273132208e-05,
+      "loss": 1.5931,
+      "step": 17360
+    },
+    {
+      "epoch": 0.799788086792279,
+      "grad_norm": 0.8920443322189382,
+      "learning_rate": 2.0298620209175766e-05,
+      "loss": 1.1368,
+      "step": 17361
+    },
+    {
+      "epoch": 0.799834154880914,
+      "grad_norm": 0.926795490052203,
+      "learning_rate": 2.0289609461444025e-05,
+      "loss": 1.0693,
+      "step": 17362
+    },
+    {
+      "epoch": 0.799880222969549,
+      "grad_norm": 1.6395449810001392,
+      "learning_rate": 2.0280600488327484e-05,
+      "loss": 1.3213,
+      "step": 17363
+    },
+    {
+      "epoch": 0.799926291058184,
+      "grad_norm": 0.8610592284086939,
+      "learning_rate": 2.0271593290026658e-05,
+      "loss": 1.0212,
+      "step": 17364
+    },
+    {
+      "epoch": 0.799972359146819,
+      "grad_norm": 1.3525675850061742,
+      "learning_rate": 2.026258786674209e-05,
+      "loss": 1.101,
+      "step": 17365
+    },
+    {
+      "epoch": 0.800018427235454,
+      "grad_norm": 0.9317942874658188,
+      "learning_rate": 2.0253584218674293e-05,
+      "loss": 1.0725,
+      "step": 17366
+    },
+    {
+      "epoch": 0.800064495324089,
+      "grad_norm": 1.1854741604327608,
+      "learning_rate": 2.0244582346023733e-05,
+      "loss": 1.2479,
+      "step": 17367
+    },
+    {
+      "epoch": 0.800110563412724,
+      "grad_norm": 0.89758879935207,
+      "learning_rate": 2.023558224899077e-05,
+      "loss": 1.2096,
+      "step": 17368
+    },
+    {
+      "epoch": 0.800156631501359,
+      "grad_norm": 0.8588719429799812,
+      "learning_rate": 2.02265839277758e-05,
+      "loss": 1.2752,
+      "step": 17369
+    },
+    {
+      "epoch": 0.800202699589994,
+      "grad_norm": 1.5261600407983642,
+      "learning_rate": 2.0217587382579184e-05,
+      "loss": 1.548,
+      "step": 17370
+    },
+    {
+      "epoch": 0.800248767678629,
+      "grad_norm": 3.108156745243168,
+      "learning_rate": 2.0208592613601163e-05,
+      "loss": 1.1063,
+      "step": 17371
+    },
+    {
+      "epoch": 0.800294835767264,
+      "grad_norm": 0.90951907609718,
+      "learning_rate": 2.019959962104201e-05,
+      "loss": 1.3127,
+      "step": 17372
+    },
+    {
+      "epoch": 0.8003409038558991,
+      "grad_norm": 0.792256557529202,
+      "learning_rate": 2.019060840510194e-05,
+      "loss": 1.0635,
+      "step": 17373
+    },
+    {
+      "epoch": 0.800386971944534,
+      "grad_norm": 0.9679654029127931,
+      "learning_rate": 2.0181618965981163e-05,
+      "loss": 1.2972,
+      "step": 17374
+    },
+    {
+      "epoch": 0.8004330400331691,
+      "grad_norm": 1.1892635689619757,
+      "learning_rate": 2.0172631303879753e-05,
+      "loss": 1.2067,
+      "step": 17375
+    },
+    {
+      "epoch": 0.800479108121804,
+      "grad_norm": 0.8524354928932981,
+      "learning_rate": 2.0163645418997835e-05,
+      "loss": 1.2768,
+      "step": 17376
+    },
+    {
+      "epoch": 0.8005251762104391,
+      "grad_norm": 0.865635225177987,
+      "learning_rate": 2.015466131153546e-05,
+      "loss": 1.0044,
+      "step": 17377
+    },
+    {
+      "epoch": 0.800571244299074,
+      "grad_norm": 1.1609104830625308,
+      "learning_rate": 2.0145678981692672e-05,
+      "loss": 1.1028,
+      "step": 17378
+    },
+    {
+      "epoch": 0.8006173123877091,
+      "grad_norm": 0.7284116681365729,
+      "learning_rate": 2.0136698429669386e-05,
+      "loss": 1.4555,
+      "step": 17379
+    },
+    {
+      "epoch": 0.800663380476344,
+      "grad_norm": 1.0343195154106057,
+      "learning_rate": 2.0127719655665588e-05,
+      "loss": 1.0218,
+      "step": 17380
+    },
+    {
+      "epoch": 0.800709448564979,
+      "grad_norm": 1.2315025715401895,
+      "learning_rate": 2.0118742659881173e-05,
+      "loss": 1.4486,
+      "step": 17381
+    },
+    {
+      "epoch": 0.800755516653614,
+      "grad_norm": 1.0160107371238454,
+      "learning_rate": 2.010976744251597e-05,
+      "loss": 1.3534,
+      "step": 17382
+    },
+    {
+      "epoch": 0.800801584742249,
+      "grad_norm": 0.7904131780288113,
+      "learning_rate": 2.0100794003769808e-05,
+      "loss": 1.3703,
+      "step": 17383
+    },
+    {
+      "epoch": 0.800847652830884,
+      "grad_norm": 0.9009951151011636,
+      "learning_rate": 2.009182234384247e-05,
+      "loss": 1.1574,
+      "step": 17384
+    },
+    {
+      "epoch": 0.800893720919519,
+      "grad_norm": 1.3411541486787597,
+      "learning_rate": 2.008285246293372e-05,
+      "loss": 1.325,
+      "step": 17385
+    },
+    {
+      "epoch": 0.8009397890081541,
+      "grad_norm": 1.318636679704874,
+      "learning_rate": 2.00738843612432e-05,
+      "loss": 1.0669,
+      "step": 17386
+    },
+    {
+      "epoch": 0.800985857096789,
+      "grad_norm": 0.9436133607251501,
+      "learning_rate": 2.0064918038970615e-05,
+      "loss": 1.159,
+      "step": 17387
+    },
+    {
+      "epoch": 0.8010319251854241,
+      "grad_norm": 1.019385597647713,
+      "learning_rate": 2.005595349631556e-05,
+      "loss": 1.3184,
+      "step": 17388
+    },
+    {
+      "epoch": 0.801077993274059,
+      "grad_norm": 0.8134237125188677,
+      "learning_rate": 2.0046990733477656e-05,
+      "loss": 1.4901,
+      "step": 17389
+    },
+    {
+      "epoch": 0.8011240613626941,
+      "grad_norm": 1.0444585538818572,
+      "learning_rate": 2.0038029750656383e-05,
+      "loss": 1.1226,
+      "step": 17390
+    },
+    {
+      "epoch": 0.801170129451329,
+      "grad_norm": 0.965369995512765,
+      "learning_rate": 2.0029070548051275e-05,
+      "loss": 1.5413,
+      "step": 17391
+    },
+    {
+      "epoch": 0.8012161975399641,
+      "grad_norm": 1.3726270076670857,
+      "learning_rate": 2.0020113125861815e-05,
+      "loss": 1.0556,
+      "step": 17392
+    },
+    {
+      "epoch": 0.801262265628599,
+      "grad_norm": 1.914681634354617,
+      "learning_rate": 2.001115748428737e-05,
+      "loss": 1.3635,
+      "step": 17393
+    },
+    {
+      "epoch": 0.8013083337172341,
+      "grad_norm": 0.8002272527085562,
+      "learning_rate": 2.000220362352736e-05,
+      "loss": 1.3429,
+      "step": 17394
+    },
+    {
+      "epoch": 0.801354401805869,
+      "grad_norm": 1.1629528470513806,
+      "learning_rate": 1.999325154378112e-05,
+      "loss": 0.8795,
+      "step": 17395
+    },
+    {
+      "epoch": 0.8014004698945041,
+      "grad_norm": 0.9646791371875753,
+      "learning_rate": 1.9984301245247972e-05,
+      "loss": 1.3266,
+      "step": 17396
+    },
+    {
+      "epoch": 0.801446537983139,
+      "grad_norm": 0.8542017961285144,
+      "learning_rate": 1.997535272812714e-05,
+      "loss": 1.5088,
+      "step": 17397
+    },
+    {
+      "epoch": 0.8014926060717741,
+      "grad_norm": 0.8936364774550274,
+      "learning_rate": 1.9966405992617853e-05,
+      "loss": 1.2321,
+      "step": 17398
+    },
+    {
+      "epoch": 0.8015386741604091,
+      "grad_norm": 0.7561188797337326,
+      "learning_rate": 1.9957461038919344e-05,
+      "loss": 1.4639,
+      "step": 17399
+    },
+    {
+      "epoch": 0.8015847422490441,
+      "grad_norm": 1.7830552884099318,
+      "learning_rate": 1.99485178672307e-05,
+      "loss": 1.1195,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8016308103376791,
+      "grad_norm": 1.2445159105814194,
+      "learning_rate": 1.9939576477751044e-05,
+      "loss": 1.3422,
+      "step": 17401
+    },
+    {
+      "epoch": 0.8016768784263141,
+      "grad_norm": 1.1617833649058822,
+      "learning_rate": 1.993063687067944e-05,
+      "loss": 1.0366,
+      "step": 17402
+    },
+    {
+      "epoch": 0.8017229465149491,
+      "grad_norm": 1.0049188046402742,
+      "learning_rate": 1.9921699046214948e-05,
+      "loss": 1.2011,
+      "step": 17403
+    },
+    {
+      "epoch": 0.8017690146035841,
+      "grad_norm": 0.859283514774612,
+      "learning_rate": 1.991276300455649e-05,
+      "loss": 1.4618,
+      "step": 17404
+    },
+    {
+      "epoch": 0.8018150826922191,
+      "grad_norm": 0.7360771823579618,
+      "learning_rate": 1.990382874590304e-05,
+      "loss": 1.1063,
+      "step": 17405
+    },
+    {
+      "epoch": 0.8018611507808541,
+      "grad_norm": 1.390410813839459,
+      "learning_rate": 1.989489627045352e-05,
+      "loss": 1.1683,
+      "step": 17406
+    },
+    {
+      "epoch": 0.8019072188694891,
+      "grad_norm": 1.0461660575806861,
+      "learning_rate": 1.9885965578406805e-05,
+      "loss": 1.1385,
+      "step": 17407
+    },
+    {
+      "epoch": 0.8019532869581241,
+      "grad_norm": 2.0451882236269636,
+      "learning_rate": 1.987703666996167e-05,
+      "loss": 1.381,
+      "step": 17408
+    },
+    {
+      "epoch": 0.8019993550467591,
+      "grad_norm": 1.1952223423525363,
+      "learning_rate": 1.9868109545316937e-05,
+      "loss": 1.0084,
+      "step": 17409
+    },
+    {
+      "epoch": 0.8020454231353941,
+      "grad_norm": 1.2007859223879864,
+      "learning_rate": 1.9859184204671367e-05,
+      "loss": 1.2972,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8020914912240291,
+      "grad_norm": 0.7412878549845198,
+      "learning_rate": 1.9850260648223628e-05,
+      "loss": 0.9,
+      "step": 17411
+    },
+    {
+      "epoch": 0.8021375593126642,
+      "grad_norm": 1.9194490720557096,
+      "learning_rate": 1.9841338876172412e-05,
+      "loss": 1.0633,
+      "step": 17412
+    },
+    {
+      "epoch": 0.8021836274012991,
+      "grad_norm": 1.008761921892779,
+      "learning_rate": 1.9832418888716332e-05,
+      "loss": 1.0019,
+      "step": 17413
+    },
+    {
+      "epoch": 0.8022296954899342,
+      "grad_norm": 1.4031549965264034,
+      "learning_rate": 1.9823500686054018e-05,
+      "loss": 1.1879,
+      "step": 17414
+    },
+    {
+      "epoch": 0.8022757635785691,
+      "grad_norm": 1.1114397837321706,
+      "learning_rate": 1.9814584268383973e-05,
+      "loss": 1.3968,
+      "step": 17415
+    },
+    {
+      "epoch": 0.8023218316672042,
+      "grad_norm": 0.8473649798368622,
+      "learning_rate": 1.980566963590471e-05,
+      "loss": 1.4173,
+      "step": 17416
+    },
+    {
+      "epoch": 0.8023678997558391,
+      "grad_norm": 1.0332703894710051,
+      "learning_rate": 1.979675678881474e-05,
+      "loss": 1.2208,
+      "step": 17417
+    },
+    {
+      "epoch": 0.8024139678444742,
+      "grad_norm": 0.6651830858818336,
+      "learning_rate": 1.9787845727312426e-05,
+      "loss": 1.2105,
+      "step": 17418
+    },
+    {
+      "epoch": 0.8024600359331091,
+      "grad_norm": 1.0925525128388587,
+      "learning_rate": 1.977893645159623e-05,
+      "loss": 1.3317,
+      "step": 17419
+    },
+    {
+      "epoch": 0.8025061040217442,
+      "grad_norm": 1.0567561794770355,
+      "learning_rate": 1.9770028961864406e-05,
+      "loss": 1.5695,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8025521721103791,
+      "grad_norm": 1.4814612130146154,
+      "learning_rate": 1.976112325831537e-05,
+      "loss": 1.3441,
+      "step": 17421
+    },
+    {
+      "epoch": 0.8025982401990142,
+      "grad_norm": 1.331339830956233,
+      "learning_rate": 1.9752219341147326e-05,
+      "loss": 1.2032,
+      "step": 17422
+    },
+    {
+      "epoch": 0.8026443082876491,
+      "grad_norm": 0.8676130141182686,
+      "learning_rate": 1.974331721055852e-05,
+      "loss": 1.1402,
+      "step": 17423
+    },
+    {
+      "epoch": 0.8026903763762842,
+      "grad_norm": 0.9462969200715159,
+      "learning_rate": 1.973441686674715e-05,
+      "loss": 1.385,
+      "step": 17424
+    },
+    {
+      "epoch": 0.8027364444649191,
+      "grad_norm": 0.8386331056637815,
+      "learning_rate": 1.9725518309911383e-05,
+      "loss": 1.0765,
+      "step": 17425
+    },
+    {
+      "epoch": 0.8027825125535542,
+      "grad_norm": 1.3603947046578886,
+      "learning_rate": 1.9716621540249314e-05,
+      "loss": 1.2754,
+      "step": 17426
+    },
+    {
+      "epoch": 0.8028285806421892,
+      "grad_norm": 0.9725898216673132,
+      "learning_rate": 1.970772655795895e-05,
+      "loss": 1.4345,
+      "step": 17427
+    },
+    {
+      "epoch": 0.8028746487308241,
+      "grad_norm": 1.480659910542381,
+      "learning_rate": 1.9698833363238443e-05,
+      "loss": 1.143,
+      "step": 17428
+    },
+    {
+      "epoch": 0.8029207168194592,
+      "grad_norm": 1.8219863055941259,
+      "learning_rate": 1.9689941956285694e-05,
+      "loss": 1.037,
+      "step": 17429
+    },
+    {
+      "epoch": 0.8029667849080941,
+      "grad_norm": 0.7135332978431036,
+      "learning_rate": 1.968105233729871e-05,
+      "loss": 1.2409,
+      "step": 17430
+    },
+    {
+      "epoch": 0.8030128529967292,
+      "grad_norm": 0.9002974384857307,
+      "learning_rate": 1.9672164506475334e-05,
+      "loss": 1.1023,
+      "step": 17431
+    },
+    {
+      "epoch": 0.8030589210853641,
+      "grad_norm": 1.0842957387610836,
+      "learning_rate": 1.9663278464013524e-05,
+      "loss": 1.3069,
+      "step": 17432
+    },
+    {
+      "epoch": 0.8031049891739992,
+      "grad_norm": 1.5249392830428092,
+      "learning_rate": 1.965439421011105e-05,
+      "loss": 1.2369,
+      "step": 17433
+    },
+    {
+      "epoch": 0.8031510572626341,
+      "grad_norm": 2.191199372997976,
+      "learning_rate": 1.9645511744965728e-05,
+      "loss": 1.0396,
+      "step": 17434
+    },
+    {
+      "epoch": 0.8031971253512692,
+      "grad_norm": 1.1489990794045108,
+      "learning_rate": 1.9636631068775334e-05,
+      "loss": 1.2168,
+      "step": 17435
+    },
+    {
+      "epoch": 0.8032431934399041,
+      "grad_norm": 1.2883592250034643,
+      "learning_rate": 1.9627752181737523e-05,
+      "loss": 0.9464,
+      "step": 17436
+    },
+    {
+      "epoch": 0.8032892615285392,
+      "grad_norm": 1.097775604744147,
+      "learning_rate": 1.9618875084050037e-05,
+      "loss": 1.1924,
+      "step": 17437
+    },
+    {
+      "epoch": 0.8033353296171741,
+      "grad_norm": 0.8759057600966549,
+      "learning_rate": 1.9609999775910414e-05,
+      "loss": 1.4296,
+      "step": 17438
+    },
+    {
+      "epoch": 0.8033813977058092,
+      "grad_norm": 0.9536881260326583,
+      "learning_rate": 1.9601126257516355e-05,
+      "loss": 1.1617,
+      "step": 17439
+    },
+    {
+      "epoch": 0.8034274657944442,
+      "grad_norm": 1.0608144810638793,
+      "learning_rate": 1.9592254529065347e-05,
+      "loss": 1.0633,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8034735338830792,
+      "grad_norm": 1.7868929628463883,
+      "learning_rate": 1.958338459075495e-05,
+      "loss": 1.2344,
+      "step": 17441
+    },
+    {
+      "epoch": 0.8035196019717142,
+      "grad_norm": 0.7077203921977682,
+      "learning_rate": 1.9574516442782554e-05,
+      "loss": 1.2986,
+      "step": 17442
+    },
+    {
+      "epoch": 0.8035656700603492,
+      "grad_norm": 1.5356436053522857,
+      "learning_rate": 1.95656500853457e-05,
+      "loss": 0.7026,
+      "step": 17443
+    },
+    {
+      "epoch": 0.8036117381489842,
+      "grad_norm": 0.8629650370443408,
+      "learning_rate": 1.955678551864173e-05,
+      "loss": 1.2887,
+      "step": 17444
+    },
+    {
+      "epoch": 0.8036578062376192,
+      "grad_norm": 1.0820289973319293,
+      "learning_rate": 1.9547922742867976e-05,
+      "loss": 1.5823,
+      "step": 17445
+    },
+    {
+      "epoch": 0.8037038743262542,
+      "grad_norm": 0.8419483697776823,
+      "learning_rate": 1.9539061758221777e-05,
+      "loss": 1.4298,
+      "step": 17446
+    },
+    {
+      "epoch": 0.8037499424148892,
+      "grad_norm": 0.8584514962335179,
+      "learning_rate": 1.953020256490041e-05,
+      "loss": 1.2391,
+      "step": 17447
+    },
+    {
+      "epoch": 0.8037960105035242,
+      "grad_norm": 2.4950991981678126,
+      "learning_rate": 1.9521345163101136e-05,
+      "loss": 1.3094,
+      "step": 17448
+    },
+    {
+      "epoch": 0.8038420785921592,
+      "grad_norm": 1.2198004790626027,
+      "learning_rate": 1.9512489553021052e-05,
+      "loss": 1.087,
+      "step": 17449
+    },
+    {
+      "epoch": 0.8038881466807942,
+      "grad_norm": 0.8394682729645762,
+      "learning_rate": 1.950363573485745e-05,
+      "loss": 1.1326,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8039342147694292,
+      "grad_norm": 1.3530247687289267,
+      "learning_rate": 1.949478370880734e-05,
+      "loss": 1.3789,
+      "step": 17451
+    },
+    {
+      "epoch": 0.8039802828580642,
+      "grad_norm": 1.146662988812335,
+      "learning_rate": 1.9485933475067864e-05,
+      "loss": 1.0281,
+      "step": 17452
+    },
+    {
+      "epoch": 0.8040263509466993,
+      "grad_norm": 1.4585497765811,
+      "learning_rate": 1.9477085033836008e-05,
+      "loss": 1.5246,
+      "step": 17453
+    },
+    {
+      "epoch": 0.8040724190353342,
+      "grad_norm": 0.9157638825692783,
+      "learning_rate": 1.9468238385308777e-05,
+      "loss": 0.9154,
+      "step": 17454
+    },
+    {
+      "epoch": 0.8041184871239693,
+      "grad_norm": 1.552612621145514,
+      "learning_rate": 1.9459393529683166e-05,
+      "loss": 1.2098,
+      "step": 17455
+    },
+    {
+      "epoch": 0.8041645552126042,
+      "grad_norm": 0.9720553583268526,
+      "learning_rate": 1.9450550467156038e-05,
+      "loss": 1.1129,
+      "step": 17456
+    },
+    {
+      "epoch": 0.8042106233012393,
+      "grad_norm": 1.004180570196498,
+      "learning_rate": 1.9441709197924287e-05,
+      "loss": 1.454,
+      "step": 17457
+    },
+    {
+      "epoch": 0.8042566913898742,
+      "grad_norm": 0.7577316108527177,
+      "learning_rate": 1.9432869722184753e-05,
+      "loss": 0.908,
+      "step": 17458
+    },
+    {
+      "epoch": 0.8043027594785093,
+      "grad_norm": 1.4889742373811932,
+      "learning_rate": 1.9424032040134254e-05,
+      "loss": 0.9812,
+      "step": 17459
+    },
+    {
+      "epoch": 0.8043488275671442,
+      "grad_norm": 1.1750812669041157,
+      "learning_rate": 1.9415196151969507e-05,
+      "loss": 1.2748,
+      "step": 17460
+    },
+    {
+      "epoch": 0.8043948956557793,
+      "grad_norm": 0.8054807012442177,
+      "learning_rate": 1.940636205788724e-05,
+      "loss": 1.4125,
+      "step": 17461
+    },
+    {
+      "epoch": 0.8044409637444142,
+      "grad_norm": 0.8761853717347275,
+      "learning_rate": 1.9397529758084155e-05,
+      "loss": 1.4966,
+      "step": 17462
+    },
+    {
+      "epoch": 0.8044870318330493,
+      "grad_norm": 0.8306066348352165,
+      "learning_rate": 1.9388699252756837e-05,
+      "loss": 1.12,
+      "step": 17463
+    },
+    {
+      "epoch": 0.8045330999216842,
+      "grad_norm": 2.1816418053096838,
+      "learning_rate": 1.9379870542101918e-05,
+      "loss": 1.4704,
+      "step": 17464
+    },
+    {
+      "epoch": 0.8045791680103193,
+      "grad_norm": 4.813831990135977,
+      "learning_rate": 1.9371043626315954e-05,
+      "loss": 1.1698,
+      "step": 17465
+    },
+    {
+      "epoch": 0.8046252360989542,
+      "grad_norm": 0.9337952762384999,
+      "learning_rate": 1.9362218505595476e-05,
+      "loss": 1.244,
+      "step": 17466
+    },
+    {
+      "epoch": 0.8046713041875893,
+      "grad_norm": 1.1546770544467888,
+      "learning_rate": 1.93533951801369e-05,
+      "loss": 1.479,
+      "step": 17467
+    },
+    {
+      "epoch": 0.8047173722762243,
+      "grad_norm": 0.8808422785312509,
+      "learning_rate": 1.934457365013672e-05,
+      "loss": 1.2087,
+      "step": 17468
+    },
+    {
+      "epoch": 0.8047634403648593,
+      "grad_norm": 0.9315960580312361,
+      "learning_rate": 1.9335753915791298e-05,
+      "loss": 1.3354,
+      "step": 17469
+    },
+    {
+      "epoch": 0.8048095084534943,
+      "grad_norm": 1.2813261478735076,
+      "learning_rate": 1.9326935977297045e-05,
+      "loss": 1.1545,
+      "step": 17470
+    },
+    {
+      "epoch": 0.8048555765421292,
+      "grad_norm": 1.0551750882030013,
+      "learning_rate": 1.93181198348502e-05,
+      "loss": 1.1129,
+      "step": 17471
+    },
+    {
+      "epoch": 0.8049016446307643,
+      "grad_norm": 0.647527144467827,
+      "learning_rate": 1.930930548864709e-05,
+      "loss": 1.1039,
+      "step": 17472
+    },
+    {
+      "epoch": 0.8049477127193992,
+      "grad_norm": 0.9187339732922944,
+      "learning_rate": 1.930049293888395e-05,
+      "loss": 1.4484,
+      "step": 17473
+    },
+    {
+      "epoch": 0.8049937808080343,
+      "grad_norm": 0.8282027618189653,
+      "learning_rate": 1.9291682185756942e-05,
+      "loss": 1.3082,
+      "step": 17474
+    },
+    {
+      "epoch": 0.8050398488966692,
+      "grad_norm": 1.0926386784985347,
+      "learning_rate": 1.928287322946225e-05,
+      "loss": 1.305,
+      "step": 17475
+    },
+    {
+      "epoch": 0.8050859169853043,
+      "grad_norm": 2.2112481716303316,
+      "learning_rate": 1.9274066070195985e-05,
+      "loss": 1.0824,
+      "step": 17476
+    },
+    {
+      "epoch": 0.8051319850739392,
+      "grad_norm": 1.0770811938688314,
+      "learning_rate": 1.9265260708154253e-05,
+      "loss": 1.3012,
+      "step": 17477
+    },
+    {
+      "epoch": 0.8051780531625743,
+      "grad_norm": 1.3303111669035865,
+      "learning_rate": 1.9256457143533023e-05,
+      "loss": 1.1767,
+      "step": 17478
+    },
+    {
+      "epoch": 0.8052241212512092,
+      "grad_norm": 2.4130923810687843,
+      "learning_rate": 1.924765537652834e-05,
+      "loss": 1.1477,
+      "step": 17479
+    },
+    {
+      "epoch": 0.8052701893398443,
+      "grad_norm": 0.8304691318567776,
+      "learning_rate": 1.9238855407336176e-05,
+      "loss": 0.9745,
+      "step": 17480
+    },
+    {
+      "epoch": 0.8053162574284793,
+      "grad_norm": 0.8518420461826273,
+      "learning_rate": 1.923005723615239e-05,
+      "loss": 1.3137,
+      "step": 17481
+    },
+    {
+      "epoch": 0.8053623255171143,
+      "grad_norm": 0.7901997720406738,
+      "learning_rate": 1.9221260863172885e-05,
+      "loss": 1.3752,
+      "step": 17482
+    },
+    {
+      "epoch": 0.8054083936057493,
+      "grad_norm": 1.0941133913107954,
+      "learning_rate": 1.921246628859351e-05,
+      "loss": 1.301,
+      "step": 17483
+    },
+    {
+      "epoch": 0.8054544616943843,
+      "grad_norm": 1.011771285900202,
+      "learning_rate": 1.9203673512610076e-05,
+      "loss": 1.261,
+      "step": 17484
+    },
+    {
+      "epoch": 0.8055005297830193,
+      "grad_norm": 1.125422795985411,
+      "learning_rate": 1.919488253541829e-05,
+      "loss": 1.2033,
+      "step": 17485
+    },
+    {
+      "epoch": 0.8055465978716543,
+      "grad_norm": 0.7907357830084816,
+      "learning_rate": 1.9186093357213885e-05,
+      "loss": 1.1773,
+      "step": 17486
+    },
+    {
+      "epoch": 0.8055926659602893,
+      "grad_norm": 1.299611150916429,
+      "learning_rate": 1.917730597819255e-05,
+      "loss": 1.2997,
+      "step": 17487
+    },
+    {
+      "epoch": 0.8056387340489243,
+      "grad_norm": 0.964543394723912,
+      "learning_rate": 1.916852039854995e-05,
+      "loss": 1.276,
+      "step": 17488
+    },
+    {
+      "epoch": 0.8056848021375593,
+      "grad_norm": 0.8051684729279245,
+      "learning_rate": 1.9159736618481605e-05,
+      "loss": 1.4447,
+      "step": 17489
+    },
+    {
+      "epoch": 0.8057308702261943,
+      "grad_norm": 0.8697281970595495,
+      "learning_rate": 1.9150954638183128e-05,
+      "loss": 0.897,
+      "step": 17490
+    },
+    {
+      "epoch": 0.8057769383148293,
+      "grad_norm": 1.1852075821053327,
+      "learning_rate": 1.9142174457850036e-05,
+      "loss": 1.4752,
+      "step": 17491
+    },
+    {
+      "epoch": 0.8058230064034643,
+      "grad_norm": 0.9630189037686815,
+      "learning_rate": 1.9133396077677768e-05,
+      "loss": 1.2545,
+      "step": 17492
+    },
+    {
+      "epoch": 0.8058690744920993,
+      "grad_norm": 0.8097889094981543,
+      "learning_rate": 1.9124619497861774e-05,
+      "loss": 1.2276,
+      "step": 17493
+    },
+    {
+      "epoch": 0.8059151425807344,
+      "grad_norm": 1.1904199006125873,
+      "learning_rate": 1.9115844718597465e-05,
+      "loss": 0.9241,
+      "step": 17494
+    },
+    {
+      "epoch": 0.8059612106693693,
+      "grad_norm": 1.311152253078576,
+      "learning_rate": 1.9107071740080207e-05,
+      "loss": 1.1946,
+      "step": 17495
+    },
+    {
+      "epoch": 0.8060072787580044,
+      "grad_norm": 1.5819145549480362,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 1.4024,
+      "step": 17496
+    },
+    {
+      "epoch": 0.8060533468466393,
+      "grad_norm": 0.8354893328127523,
+      "learning_rate": 1.9089531186067943e-05,
+      "loss": 1.5789,
+      "step": 17497
+    },
+    {
+      "epoch": 0.8060994149352744,
+      "grad_norm": 2.1899302246174606,
+      "learning_rate": 1.9080763610963514e-05,
+      "loss": 1.29,
+      "step": 17498
+    },
+    {
+      "epoch": 0.8061454830239093,
+      "grad_norm": 0.8364651218560321,
+      "learning_rate": 1.90719978373871e-05,
+      "loss": 1.1614,
+      "step": 17499
+    },
+    {
+      "epoch": 0.8061915511125444,
+      "grad_norm": 1.2350910341561547,
+      "learning_rate": 1.9063233865533903e-05,
+      "loss": 1.2726,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8062376192011793,
+      "grad_norm": 0.7828258379614608,
+      "learning_rate": 1.9054471695599017e-05,
+      "loss": 1.3973,
+      "step": 17501
+    },
+    {
+      "epoch": 0.8062836872898144,
+      "grad_norm": 1.0411467630055797,
+      "learning_rate": 1.904571132777756e-05,
+      "loss": 1.0724,
+      "step": 17502
+    },
+    {
+      "epoch": 0.8063297553784493,
+      "grad_norm": 1.9739274024848668,
+      "learning_rate": 1.9036952762264505e-05,
+      "loss": 1.356,
+      "step": 17503
+    },
+    {
+      "epoch": 0.8063758234670844,
+      "grad_norm": 1.142778227710434,
+      "learning_rate": 1.9028195999254873e-05,
+      "loss": 1.0397,
+      "step": 17504
+    },
+    {
+      "epoch": 0.8064218915557193,
+      "grad_norm": 1.4520792454308786,
+      "learning_rate": 1.901944103894362e-05,
+      "loss": 1.1604,
+      "step": 17505
+    },
+    {
+      "epoch": 0.8064679596443544,
+      "grad_norm": 0.6423675769137223,
+      "learning_rate": 1.901068788152569e-05,
+      "loss": 1.3805,
+      "step": 17506
+    },
+    {
+      "epoch": 0.8065140277329894,
+      "grad_norm": 1.0424648955765525,
+      "learning_rate": 1.9001936527195895e-05,
+      "loss": 1.5163,
+      "step": 17507
+    },
+    {
+      "epoch": 0.8065600958216244,
+      "grad_norm": 1.029694262522422,
+      "learning_rate": 1.899318697614911e-05,
+      "loss": 0.9359,
+      "step": 17508
+    },
+    {
+      "epoch": 0.8066061639102594,
+      "grad_norm": 1.0610354644344093,
+      "learning_rate": 1.8984439228580142e-05,
+      "loss": 0.835,
+      "step": 17509
+    },
+    {
+      "epoch": 0.8066522319988944,
+      "grad_norm": 0.7372469124212094,
+      "learning_rate": 1.8975693284683704e-05,
+      "loss": 1.4099,
+      "step": 17510
+    },
+    {
+      "epoch": 0.8066983000875294,
+      "grad_norm": 1.4948506701646862,
+      "learning_rate": 1.896694914465451e-05,
+      "loss": 1.4851,
+      "step": 17511
+    },
+    {
+      "epoch": 0.8067443681761644,
+      "grad_norm": 0.8476462404338596,
+      "learning_rate": 1.895820680868727e-05,
+      "loss": 1.222,
+      "step": 17512
+    },
+    {
+      "epoch": 0.8067904362647994,
+      "grad_norm": 0.8529254148725007,
+      "learning_rate": 1.894946627697661e-05,
+      "loss": 1.0921,
+      "step": 17513
+    },
+    {
+      "epoch": 0.8068365043534343,
+      "grad_norm": 1.256447607062256,
+      "learning_rate": 1.8940727549717095e-05,
+      "loss": 1.1405,
+      "step": 17514
+    },
+    {
+      "epoch": 0.8068825724420694,
+      "grad_norm": 1.1869895507052666,
+      "learning_rate": 1.8931990627103292e-05,
+      "loss": 1.0915,
+      "step": 17515
+    },
+    {
+      "epoch": 0.8069286405307043,
+      "grad_norm": 0.7024340785158164,
+      "learning_rate": 1.892325550932972e-05,
+      "loss": 1.2085,
+      "step": 17516
+    },
+    {
+      "epoch": 0.8069747086193394,
+      "grad_norm": 2.0614158186460037,
+      "learning_rate": 1.891452219659087e-05,
+      "loss": 1.5422,
+      "step": 17517
+    },
+    {
+      "epoch": 0.8070207767079743,
+      "grad_norm": 0.8222529445447612,
+      "learning_rate": 1.8905790689081147e-05,
+      "loss": 1.1649,
+      "step": 17518
+    },
+    {
+      "epoch": 0.8070668447966094,
+      "grad_norm": 0.9065852954340771,
+      "learning_rate": 1.8897060986994896e-05,
+      "loss": 1.055,
+      "step": 17519
+    },
+    {
+      "epoch": 0.8071129128852443,
+      "grad_norm": 1.4034638706412113,
+      "learning_rate": 1.888833309052658e-05,
+      "loss": 1.2894,
+      "step": 17520
+    },
+    {
+      "epoch": 0.8071589809738794,
+      "grad_norm": 0.9602523100704605,
+      "learning_rate": 1.8879606999870415e-05,
+      "loss": 1.0968,
+      "step": 17521
+    },
+    {
+      "epoch": 0.8072050490625144,
+      "grad_norm": 1.3552882690682926,
+      "learning_rate": 1.8870882715220716e-05,
+      "loss": 1.1737,
+      "step": 17522
+    },
+    {
+      "epoch": 0.8072511171511494,
+      "grad_norm": 4.462038296420384,
+      "learning_rate": 1.886216023677171e-05,
+      "loss": 1.1995,
+      "step": 17523
+    },
+    {
+      "epoch": 0.8072971852397844,
+      "grad_norm": 1.1242131869443477,
+      "learning_rate": 1.8853439564717603e-05,
+      "loss": 1.1191,
+      "step": 17524
+    },
+    {
+      "epoch": 0.8073432533284194,
+      "grad_norm": 0.8950679002484927,
+      "learning_rate": 1.8844720699252506e-05,
+      "loss": 1.3192,
+      "step": 17525
+    },
+    {
+      "epoch": 0.8073893214170544,
+      "grad_norm": 2.08915961591158,
+      "learning_rate": 1.8836003640570542e-05,
+      "loss": 1.192,
+      "step": 17526
+    },
+    {
+      "epoch": 0.8074353895056894,
+      "grad_norm": 0.8461587930200388,
+      "learning_rate": 1.882728838886583e-05,
+      "loss": 1.3031,
+      "step": 17527
+    },
+    {
+      "epoch": 0.8074814575943244,
+      "grad_norm": 0.7507213296773958,
+      "learning_rate": 1.8818574944332322e-05,
+      "loss": 1.4158,
+      "step": 17528
+    },
+    {
+      "epoch": 0.8075275256829594,
+      "grad_norm": 0.8397064250282006,
+      "learning_rate": 1.8809863307164077e-05,
+      "loss": 1.1903,
+      "step": 17529
+    },
+    {
+      "epoch": 0.8075735937715944,
+      "grad_norm": 1.0480333237643968,
+      "learning_rate": 1.8801153477554967e-05,
+      "loss": 1.2006,
+      "step": 17530
+    },
+    {
+      "epoch": 0.8076196618602294,
+      "grad_norm": 1.0421060348575926,
+      "learning_rate": 1.8792445455698992e-05,
+      "loss": 1.3233,
+      "step": 17531
+    },
+    {
+      "epoch": 0.8076657299488644,
+      "grad_norm": 0.7345374813104992,
+      "learning_rate": 1.8783739241789967e-05,
+      "loss": 1.2248,
+      "step": 17532
+    },
+    {
+      "epoch": 0.8077117980374994,
+      "grad_norm": 0.9067260915227399,
+      "learning_rate": 1.877503483602172e-05,
+      "loss": 1.2164,
+      "step": 17533
+    },
+    {
+      "epoch": 0.8077578661261344,
+      "grad_norm": 1.8841517709182274,
+      "learning_rate": 1.8766332238588047e-05,
+      "loss": 1.2133,
+      "step": 17534
+    },
+    {
+      "epoch": 0.8078039342147695,
+      "grad_norm": 0.9469454590101131,
+      "learning_rate": 1.8757631449682744e-05,
+      "loss": 1.4769,
+      "step": 17535
+    },
+    {
+      "epoch": 0.8078500023034044,
+      "grad_norm": 6.091463591842624,
+      "learning_rate": 1.874893246949946e-05,
+      "loss": 1.3066,
+      "step": 17536
+    },
+    {
+      "epoch": 0.8078960703920395,
+      "grad_norm": 0.7637958607667067,
+      "learning_rate": 1.8740235298231833e-05,
+      "loss": 1.2797,
+      "step": 17537
+    },
+    {
+      "epoch": 0.8079421384806744,
+      "grad_norm": 1.0269070934115236,
+      "learning_rate": 1.8731539936073584e-05,
+      "loss": 1.0642,
+      "step": 17538
+    },
+    {
+      "epoch": 0.8079882065693095,
+      "grad_norm": 1.010772895132591,
+      "learning_rate": 1.872284638321823e-05,
+      "loss": 1.2718,
+      "step": 17539
+    },
+    {
+      "epoch": 0.8080342746579444,
+      "grad_norm": 0.7325345461426261,
+      "learning_rate": 1.871415463985937e-05,
+      "loss": 1.0919,
+      "step": 17540
+    },
+    {
+      "epoch": 0.8080803427465795,
+      "grad_norm": 1.757955135200994,
+      "learning_rate": 1.8705464706190424e-05,
+      "loss": 1.1657,
+      "step": 17541
+    },
+    {
+      "epoch": 0.8081264108352144,
+      "grad_norm": 4.772209527285096,
+      "learning_rate": 1.8696776582404972e-05,
+      "loss": 0.9606,
+      "step": 17542
+    },
+    {
+      "epoch": 0.8081724789238495,
+      "grad_norm": 0.6663419783080835,
+      "learning_rate": 1.8688090268696358e-05,
+      "loss": 1.3534,
+      "step": 17543
+    },
+    {
+      "epoch": 0.8082185470124844,
+      "grad_norm": 0.8720060558740058,
+      "learning_rate": 1.8679405765258018e-05,
+      "loss": 1.3567,
+      "step": 17544
+    },
+    {
+      "epoch": 0.8082646151011195,
+      "grad_norm": 2.858684548927998,
+      "learning_rate": 1.8670723072283258e-05,
+      "loss": 1.1133,
+      "step": 17545
+    },
+    {
+      "epoch": 0.8083106831897544,
+      "grad_norm": 1.0056009516081843,
+      "learning_rate": 1.866204218996539e-05,
+      "loss": 1.1372,
+      "step": 17546
+    },
+    {
+      "epoch": 0.8083567512783895,
+      "grad_norm": 0.9154290435989474,
+      "learning_rate": 1.8653363118497713e-05,
+      "loss": 1.2316,
+      "step": 17547
+    },
+    {
+      "epoch": 0.8084028193670245,
+      "grad_norm": 0.9008858351523906,
+      "learning_rate": 1.8644685858073375e-05,
+      "loss": 1.0021,
+      "step": 17548
+    },
+    {
+      "epoch": 0.8084488874556595,
+      "grad_norm": 1.0124373515978304,
+      "learning_rate": 1.8636010408885663e-05,
+      "loss": 1.2079,
+      "step": 17549
+    },
+    {
+      "epoch": 0.8084949555442945,
+      "grad_norm": 1.4435497451594064,
+      "learning_rate": 1.8627336771127647e-05,
+      "loss": 1.0382,
+      "step": 17550
+    },
+    {
+      "epoch": 0.8085410236329295,
+      "grad_norm": 1.8010877378866363,
+      "learning_rate": 1.861866494499248e-05,
+      "loss": 1.1008,
+      "step": 17551
+    },
+    {
+      "epoch": 0.8085870917215645,
+      "grad_norm": 4.0615651526557786,
+      "learning_rate": 1.8609994930673134e-05,
+      "loss": 1.3664,
+      "step": 17552
+    },
+    {
+      "epoch": 0.8086331598101995,
+      "grad_norm": 1.543920221501174,
+      "learning_rate": 1.8601326728362766e-05,
+      "loss": 1.0458,
+      "step": 17553
+    },
+    {
+      "epoch": 0.8086792278988345,
+      "grad_norm": 1.3010120821122242,
+      "learning_rate": 1.8592660338254286e-05,
+      "loss": 1.391,
+      "step": 17554
+    },
+    {
+      "epoch": 0.8087252959874695,
+      "grad_norm": 1.3815369871519896,
+      "learning_rate": 1.8583995760540607e-05,
+      "loss": 0.9298,
+      "step": 17555
+    },
+    {
+      "epoch": 0.8087713640761045,
+      "grad_norm": 1.2574683285020534,
+      "learning_rate": 1.8575332995414674e-05,
+      "loss": 1.0461,
+      "step": 17556
+    },
+    {
+      "epoch": 0.8088174321647394,
+      "grad_norm": 1.1465124005499918,
+      "learning_rate": 1.8566672043069335e-05,
+      "loss": 1.3144,
+      "step": 17557
+    },
+    {
+      "epoch": 0.8088635002533745,
+      "grad_norm": 1.6022167835064676,
+      "learning_rate": 1.855801290369743e-05,
+      "loss": 1.0712,
+      "step": 17558
+    },
+    {
+      "epoch": 0.8089095683420094,
+      "grad_norm": 0.8089471495369912,
+      "learning_rate": 1.854935557749171e-05,
+      "loss": 1.3941,
+      "step": 17559
+    },
+    {
+      "epoch": 0.8089556364306445,
+      "grad_norm": 2.1461747798541304,
+      "learning_rate": 1.8540700064644935e-05,
+      "loss": 0.8334,
+      "step": 17560
+    },
+    {
+      "epoch": 0.8090017045192794,
+      "grad_norm": 1.0103846199159758,
+      "learning_rate": 1.8532046365349787e-05,
+      "loss": 1.1633,
+      "step": 17561
+    },
+    {
+      "epoch": 0.8090477726079145,
+      "grad_norm": 1.499461569305844,
+      "learning_rate": 1.8523394479798974e-05,
+      "loss": 1.52,
+      "step": 17562
+    },
+    {
+      "epoch": 0.8090938406965495,
+      "grad_norm": 0.8499062268092961,
+      "learning_rate": 1.8514744408185046e-05,
+      "loss": 1.1832,
+      "step": 17563
+    },
+    {
+      "epoch": 0.8091399087851845,
+      "grad_norm": 2.263594131234423,
+      "learning_rate": 1.8506096150700624e-05,
+      "loss": 1.2716,
+      "step": 17564
+    },
+    {
+      "epoch": 0.8091859768738195,
+      "grad_norm": 1.1841989478170059,
+      "learning_rate": 1.849744970753826e-05,
+      "loss": 1.233,
+      "step": 17565
+    },
+    {
+      "epoch": 0.8092320449624545,
+      "grad_norm": 0.7629264752432452,
+      "learning_rate": 1.8488805078890414e-05,
+      "loss": 1.4111,
+      "step": 17566
+    },
+    {
+      "epoch": 0.8092781130510895,
+      "grad_norm": 0.7234812465191315,
+      "learning_rate": 1.8480162264949554e-05,
+      "loss": 1.4461,
+      "step": 17567
+    },
+    {
+      "epoch": 0.8093241811397245,
+      "grad_norm": 1.8405173686535006,
+      "learning_rate": 1.847152126590811e-05,
+      "loss": 0.9832,
+      "step": 17568
+    },
+    {
+      "epoch": 0.8093702492283595,
+      "grad_norm": 0.8232268390454047,
+      "learning_rate": 1.8462882081958467e-05,
+      "loss": 1.3064,
+      "step": 17569
+    },
+    {
+      "epoch": 0.8094163173169945,
+      "grad_norm": 0.8125459574055977,
+      "learning_rate": 1.8454244713292934e-05,
+      "loss": 0.9978,
+      "step": 17570
+    },
+    {
+      "epoch": 0.8094623854056295,
+      "grad_norm": 3.1074146563251155,
+      "learning_rate": 1.8445609160103816e-05,
+      "loss": 1.0259,
+      "step": 17571
+    },
+    {
+      "epoch": 0.8095084534942645,
+      "grad_norm": 1.229673445507364,
+      "learning_rate": 1.8436975422583404e-05,
+      "loss": 1.1801,
+      "step": 17572
+    },
+    {
+      "epoch": 0.8095545215828995,
+      "grad_norm": 0.822064265119227,
+      "learning_rate": 1.8428343500923862e-05,
+      "loss": 1.3463,
+      "step": 17573
+    },
+    {
+      "epoch": 0.8096005896715345,
+      "grad_norm": 1.0995514146818486,
+      "learning_rate": 1.8419713395317373e-05,
+      "loss": 1.0959,
+      "step": 17574
+    },
+    {
+      "epoch": 0.8096466577601695,
+      "grad_norm": 1.475420354683699,
+      "learning_rate": 1.8411085105956093e-05,
+      "loss": 1.5464,
+      "step": 17575
+    },
+    {
+      "epoch": 0.8096927258488046,
+      "grad_norm": 0.7940369180270055,
+      "learning_rate": 1.8402458633032137e-05,
+      "loss": 1.0976,
+      "step": 17576
+    },
+    {
+      "epoch": 0.8097387939374395,
+      "grad_norm": 1.5617439911883835,
+      "learning_rate": 1.8393833976737507e-05,
+      "loss": 1.1959,
+      "step": 17577
+    },
+    {
+      "epoch": 0.8097848620260746,
+      "grad_norm": 1.6249947490759782,
+      "learning_rate": 1.8385211137264225e-05,
+      "loss": 0.8633,
+      "step": 17578
+    },
+    {
+      "epoch": 0.8098309301147095,
+      "grad_norm": 1.0357845243779817,
+      "learning_rate": 1.8376590114804293e-05,
+      "loss": 1.2917,
+      "step": 17579
+    },
+    {
+      "epoch": 0.8098769982033446,
+      "grad_norm": 1.1568271314632428,
+      "learning_rate": 1.8367970909549636e-05,
+      "loss": 1.1575,
+      "step": 17580
+    },
+    {
+      "epoch": 0.8099230662919795,
+      "grad_norm": 1.3360170670529348,
+      "learning_rate": 1.8359353521692123e-05,
+      "loss": 1.4162,
+      "step": 17581
+    },
+    {
+      "epoch": 0.8099691343806146,
+      "grad_norm": 3.4886157871084813,
+      "learning_rate": 1.8350737951423624e-05,
+      "loss": 0.9159,
+      "step": 17582
+    },
+    {
+      "epoch": 0.8100152024692495,
+      "grad_norm": 1.0621183511389316,
+      "learning_rate": 1.834212419893597e-05,
+      "loss": 1.1147,
+      "step": 17583
+    },
+    {
+      "epoch": 0.8100612705578846,
+      "grad_norm": 2.939328671111355,
+      "learning_rate": 1.833351226442087e-05,
+      "loss": 1.0545,
+      "step": 17584
+    },
+    {
+      "epoch": 0.8101073386465195,
+      "grad_norm": 1.8726320330405337,
+      "learning_rate": 1.832490214807011e-05,
+      "loss": 1.2535,
+      "step": 17585
+    },
+    {
+      "epoch": 0.8101534067351546,
+      "grad_norm": 0.7169567497625662,
+      "learning_rate": 1.831629385007535e-05,
+      "loss": 1.094,
+      "step": 17586
+    },
+    {
+      "epoch": 0.8101994748237895,
+      "grad_norm": 0.7905700460785368,
+      "learning_rate": 1.830768737062828e-05,
+      "loss": 1.1786,
+      "step": 17587
+    },
+    {
+      "epoch": 0.8102455429124246,
+      "grad_norm": 1.5630163292522516,
+      "learning_rate": 1.8299082709920444e-05,
+      "loss": 1.3404,
+      "step": 17588
+    },
+    {
+      "epoch": 0.8102916110010596,
+      "grad_norm": 1.2386015819196927,
+      "learning_rate": 1.829047986814345e-05,
+      "loss": 0.9692,
+      "step": 17589
+    },
+    {
+      "epoch": 0.8103376790896946,
+      "grad_norm": 0.8387141786298398,
+      "learning_rate": 1.8281878845488842e-05,
+      "loss": 1.4629,
+      "step": 17590
+    },
+    {
+      "epoch": 0.8103837471783296,
+      "grad_norm": 0.9120882255122001,
+      "learning_rate": 1.8273279642148067e-05,
+      "loss": 1.1022,
+      "step": 17591
+    },
+    {
+      "epoch": 0.8104298152669646,
+      "grad_norm": 0.7362466368202274,
+      "learning_rate": 1.8264682258312583e-05,
+      "loss": 1.1383,
+      "step": 17592
+    },
+    {
+      "epoch": 0.8104758833555996,
+      "grad_norm": 0.9253244834542973,
+      "learning_rate": 1.8256086694173812e-05,
+      "loss": 1.2441,
+      "step": 17593
+    },
+    {
+      "epoch": 0.8105219514442346,
+      "grad_norm": 0.8773750352975181,
+      "learning_rate": 1.824749294992313e-05,
+      "loss": 1.29,
+      "step": 17594
+    },
+    {
+      "epoch": 0.8105680195328696,
+      "grad_norm": 0.9888838393187587,
+      "learning_rate": 1.823890102575181e-05,
+      "loss": 1.2116,
+      "step": 17595
+    },
+    {
+      "epoch": 0.8106140876215046,
+      "grad_norm": 0.9318977077287248,
+      "learning_rate": 1.823031092185117e-05,
+      "loss": 1.3174,
+      "step": 17596
+    },
+    {
+      "epoch": 0.8106601557101396,
+      "grad_norm": 0.7618089987553537,
+      "learning_rate": 1.822172263841245e-05,
+      "loss": 1.0525,
+      "step": 17597
+    },
+    {
+      "epoch": 0.8107062237987746,
+      "grad_norm": 0.9889992279202618,
+      "learning_rate": 1.82131361756269e-05,
+      "loss": 1.2185,
+      "step": 17598
+    },
+    {
+      "epoch": 0.8107522918874096,
+      "grad_norm": 0.8270695259021924,
+      "learning_rate": 1.8204551533685598e-05,
+      "loss": 1.2525,
+      "step": 17599
+    },
+    {
+      "epoch": 0.8107983599760445,
+      "grad_norm": 1.2511738022154821,
+      "learning_rate": 1.8195968712779708e-05,
+      "loss": 1.1944,
+      "step": 17600
+    },
+    {
+      "epoch": 0.8108444280646796,
+      "grad_norm": 0.8133809743454593,
+      "learning_rate": 1.8187387713100334e-05,
+      "loss": 1.1788,
+      "step": 17601
+    },
+    {
+      "epoch": 0.8108904961533145,
+      "grad_norm": 0.992079186932813,
+      "learning_rate": 1.8178808534838475e-05,
+      "loss": 0.9654,
+      "step": 17602
+    },
+    {
+      "epoch": 0.8109365642419496,
+      "grad_norm": 1.040652863916448,
+      "learning_rate": 1.817023117818515e-05,
+      "loss": 1.2467,
+      "step": 17603
+    },
+    {
+      "epoch": 0.8109826323305847,
+      "grad_norm": 1.358094515945456,
+      "learning_rate": 1.816165564333131e-05,
+      "loss": 1.2828,
+      "step": 17604
+    },
+    {
+      "epoch": 0.8110287004192196,
+      "grad_norm": 9.741290488274574,
+      "learning_rate": 1.8153081930467917e-05,
+      "loss": 0.9081,
+      "step": 17605
+    },
+    {
+      "epoch": 0.8110747685078546,
+      "grad_norm": 2.2644672332988716,
+      "learning_rate": 1.81445100397858e-05,
+      "loss": 1.3766,
+      "step": 17606
+    },
+    {
+      "epoch": 0.8111208365964896,
+      "grad_norm": 0.9237671751845389,
+      "learning_rate": 1.8135939971475803e-05,
+      "loss": 0.9375,
+      "step": 17607
+    },
+    {
+      "epoch": 0.8111669046851246,
+      "grad_norm": 0.7489485428220066,
+      "learning_rate": 1.812737172572877e-05,
+      "loss": 1.5846,
+      "step": 17608
+    },
+    {
+      "epoch": 0.8112129727737596,
+      "grad_norm": 1.1279235815989441,
+      "learning_rate": 1.8118805302735386e-05,
+      "loss": 1.2407,
+      "step": 17609
+    },
+    {
+      "epoch": 0.8112590408623946,
+      "grad_norm": 1.1322657175244029,
+      "learning_rate": 1.811024070268642e-05,
+      "loss": 1.5942,
+      "step": 17610
+    },
+    {
+      "epoch": 0.8113051089510296,
+      "grad_norm": 1.4340879710942047,
+      "learning_rate": 1.810167792577253e-05,
+      "loss": 1.1542,
+      "step": 17611
+    },
+    {
+      "epoch": 0.8113511770396646,
+      "grad_norm": 0.5854540268959001,
+      "learning_rate": 1.8093116972184378e-05,
+      "loss": 1.0055,
+      "step": 17612
+    },
+    {
+      "epoch": 0.8113972451282996,
+      "grad_norm": 1.1643126703092568,
+      "learning_rate": 1.8084557842112515e-05,
+      "loss": 1.3937,
+      "step": 17613
+    },
+    {
+      "epoch": 0.8114433132169346,
+      "grad_norm": 0.9388311437495397,
+      "learning_rate": 1.8076000535747507e-05,
+      "loss": 1.2832,
+      "step": 17614
+    },
+    {
+      "epoch": 0.8114893813055696,
+      "grad_norm": 0.9072679400823567,
+      "learning_rate": 1.8067445053279886e-05,
+      "loss": 1.138,
+      "step": 17615
+    },
+    {
+      "epoch": 0.8115354493942046,
+      "grad_norm": 0.7299541651269849,
+      "learning_rate": 1.8058891394900136e-05,
+      "loss": 1.4654,
+      "step": 17616
+    },
+    {
+      "epoch": 0.8115815174828397,
+      "grad_norm": 3.360916809402002,
+      "learning_rate": 1.805033956079868e-05,
+      "loss": 0.9793,
+      "step": 17617
+    },
+    {
+      "epoch": 0.8116275855714746,
+      "grad_norm": 1.4341446849912884,
+      "learning_rate": 1.8041789551165844e-05,
+      "loss": 1.0317,
+      "step": 17618
+    },
+    {
+      "epoch": 0.8116736536601097,
+      "grad_norm": 1.6104074158773998,
+      "learning_rate": 1.803324136619209e-05,
+      "loss": 1.0385,
+      "step": 17619
+    },
+    {
+      "epoch": 0.8117197217487446,
+      "grad_norm": 0.9043561444557456,
+      "learning_rate": 1.802469500606764e-05,
+      "loss": 1.2599,
+      "step": 17620
+    },
+    {
+      "epoch": 0.8117657898373797,
+      "grad_norm": 0.8302001706634325,
+      "learning_rate": 1.8016150470982806e-05,
+      "loss": 1.0794,
+      "step": 17621
+    },
+    {
+      "epoch": 0.8118118579260146,
+      "grad_norm": 0.7708428168562447,
+      "learning_rate": 1.8007607761127797e-05,
+      "loss": 1.4205,
+      "step": 17622
+    },
+    {
+      "epoch": 0.8118579260146497,
+      "grad_norm": 13.632074748231616,
+      "learning_rate": 1.7999066876692848e-05,
+      "loss": 1.1253,
+      "step": 17623
+    },
+    {
+      "epoch": 0.8119039941032846,
+      "grad_norm": 1.4208614582463848,
+      "learning_rate": 1.799052781786805e-05,
+      "loss": 1.1218,
+      "step": 17624
+    },
+    {
+      "epoch": 0.8119500621919197,
+      "grad_norm": 1.0664954842326773,
+      "learning_rate": 1.7981990584843545e-05,
+      "loss": 1.471,
+      "step": 17625
+    },
+    {
+      "epoch": 0.8119961302805546,
+      "grad_norm": 1.9219531057749275,
+      "learning_rate": 1.7973455177809394e-05,
+      "loss": 1.2382,
+      "step": 17626
+    },
+    {
+      "epoch": 0.8120421983691897,
+      "grad_norm": 1.0539713916490983,
+      "learning_rate": 1.796492159695561e-05,
+      "loss": 1.0278,
+      "step": 17627
+    },
+    {
+      "epoch": 0.8120882664578246,
+      "grad_norm": 0.9382394431026584,
+      "learning_rate": 1.79563898424722e-05,
+      "loss": 1.0518,
+      "step": 17628
+    },
+    {
+      "epoch": 0.8121343345464597,
+      "grad_norm": 0.8450806316987446,
+      "learning_rate": 1.7947859914549048e-05,
+      "loss": 1.1724,
+      "step": 17629
+    },
+    {
+      "epoch": 0.8121804026350947,
+      "grad_norm": 0.9314476253517469,
+      "learning_rate": 1.793933181337616e-05,
+      "loss": 0.9964,
+      "step": 17630
+    },
+    {
+      "epoch": 0.8122264707237297,
+      "grad_norm": 0.7971308028777903,
+      "learning_rate": 1.7930805539143313e-05,
+      "loss": 1.2138,
+      "step": 17631
+    },
+    {
+      "epoch": 0.8122725388123647,
+      "grad_norm": 0.735928411431847,
+      "learning_rate": 1.7922281092040393e-05,
+      "loss": 1.103,
+      "step": 17632
+    },
+    {
+      "epoch": 0.8123186069009997,
+      "grad_norm": 0.7875458624704534,
+      "learning_rate": 1.791375847225709e-05,
+      "loss": 1.6082,
+      "step": 17633
+    },
+    {
+      "epoch": 0.8123646749896347,
+      "grad_norm": 0.8332337441492615,
+      "learning_rate": 1.7905237679983278e-05,
+      "loss": 1.4539,
+      "step": 17634
+    },
+    {
+      "epoch": 0.8124107430782697,
+      "grad_norm": 1.242758687300329,
+      "learning_rate": 1.7896718715408544e-05,
+      "loss": 0.8323,
+      "step": 17635
+    },
+    {
+      "epoch": 0.8124568111669047,
+      "grad_norm": 0.9868234780695078,
+      "learning_rate": 1.7888201578722595e-05,
+      "loss": 1.2809,
+      "step": 17636
+    },
+    {
+      "epoch": 0.8125028792555397,
+      "grad_norm": 1.1412589451661326,
+      "learning_rate": 1.7879686270115083e-05,
+      "loss": 1.2346,
+      "step": 17637
+    },
+    {
+      "epoch": 0.8125489473441747,
+      "grad_norm": 1.0094583829725914,
+      "learning_rate": 1.7871172789775505e-05,
+      "loss": 0.8493,
+      "step": 17638
+    },
+    {
+      "epoch": 0.8125950154328097,
+      "grad_norm": 1.3626128713612875,
+      "learning_rate": 1.786266113789349e-05,
+      "loss": 1.2883,
+      "step": 17639
+    },
+    {
+      "epoch": 0.8126410835214447,
+      "grad_norm": 0.8897222478217591,
+      "learning_rate": 1.7854151314658417e-05,
+      "loss": 1.2136,
+      "step": 17640
+    },
+    {
+      "epoch": 0.8126871516100796,
+      "grad_norm": 0.8773977811878004,
+      "learning_rate": 1.7845643320259886e-05,
+      "loss": 1.3877,
+      "step": 17641
+    },
+    {
+      "epoch": 0.8127332196987147,
+      "grad_norm": 1.0182606158269527,
+      "learning_rate": 1.783713715488721e-05,
+      "loss": 1.1643,
+      "step": 17642
+    },
+    {
+      "epoch": 0.8127792877873498,
+      "grad_norm": 1.116096507748918,
+      "learning_rate": 1.7828632818729807e-05,
+      "loss": 1.3525,
+      "step": 17643
+    },
+    {
+      "epoch": 0.8128253558759847,
+      "grad_norm": 1.0923948911712396,
+      "learning_rate": 1.7820130311976967e-05,
+      "loss": 0.9526,
+      "step": 17644
+    },
+    {
+      "epoch": 0.8128714239646198,
+      "grad_norm": 1.6505302972444063,
+      "learning_rate": 1.781162963481806e-05,
+      "loss": 1.0979,
+      "step": 17645
+    },
+    {
+      "epoch": 0.8129174920532547,
+      "grad_norm": 0.8788682639732064,
+      "learning_rate": 1.78031307874423e-05,
+      "loss": 1.2744,
+      "step": 17646
+    },
+    {
+      "epoch": 0.8129635601418898,
+      "grad_norm": 0.7412495370943849,
+      "learning_rate": 1.779463377003885e-05,
+      "loss": 1.4401,
+      "step": 17647
+    },
+    {
+      "epoch": 0.8130096282305247,
+      "grad_norm": 0.8660386502567108,
+      "learning_rate": 1.7786138582796974e-05,
+      "loss": 1.3105,
+      "step": 17648
+    },
+    {
+      "epoch": 0.8130556963191597,
+      "grad_norm": 0.8887085100487901,
+      "learning_rate": 1.7777645225905727e-05,
+      "loss": 1.2895,
+      "step": 17649
+    },
+    {
+      "epoch": 0.8131017644077947,
+      "grad_norm": 1.8467796548370043,
+      "learning_rate": 1.7769153699554253e-05,
+      "loss": 1.3795,
+      "step": 17650
+    },
+    {
+      "epoch": 0.8131478324964297,
+      "grad_norm": 0.8142662552363781,
+      "learning_rate": 1.7760664003931526e-05,
+      "loss": 1.1696,
+      "step": 17651
+    },
+    {
+      "epoch": 0.8131939005850647,
+      "grad_norm": 0.804434025310505,
+      "learning_rate": 1.775217613922666e-05,
+      "loss": 1.3351,
+      "step": 17652
+    },
+    {
+      "epoch": 0.8132399686736997,
+      "grad_norm": 0.7830797044357624,
+      "learning_rate": 1.7743690105628543e-05,
+      "loss": 1.3723,
+      "step": 17653
+    },
+    {
+      "epoch": 0.8132860367623347,
+      "grad_norm": 2.37216966725491,
+      "learning_rate": 1.7735205903326147e-05,
+      "loss": 1.0116,
+      "step": 17654
+    },
+    {
+      "epoch": 0.8133321048509697,
+      "grad_norm": 2.1482970011004956,
+      "learning_rate": 1.7726723532508306e-05,
+      "loss": 1.2133,
+      "step": 17655
+    },
+    {
+      "epoch": 0.8133781729396047,
+      "grad_norm": 7.917129882764806,
+      "learning_rate": 1.771824299336391e-05,
+      "loss": 1.1412,
+      "step": 17656
+    },
+    {
+      "epoch": 0.8134242410282397,
+      "grad_norm": 0.9561784684919151,
+      "learning_rate": 1.7709764286081765e-05,
+      "loss": 1.4279,
+      "step": 17657
+    },
+    {
+      "epoch": 0.8134703091168748,
+      "grad_norm": 1.0681066050178853,
+      "learning_rate": 1.7701287410850597e-05,
+      "loss": 1.2289,
+      "step": 17658
+    },
+    {
+      "epoch": 0.8135163772055097,
+      "grad_norm": 0.7911906413655887,
+      "learning_rate": 1.769281236785916e-05,
+      "loss": 1.3996,
+      "step": 17659
+    },
+    {
+      "epoch": 0.8135624452941448,
+      "grad_norm": 0.6564324632187534,
+      "learning_rate": 1.7684339157296125e-05,
+      "loss": 1.4119,
+      "step": 17660
+    },
+    {
+      "epoch": 0.8136085133827797,
+      "grad_norm": 1.2280442832950942,
+      "learning_rate": 1.7675867779350164e-05,
+      "loss": 1.2056,
+      "step": 17661
+    },
+    {
+      "epoch": 0.8136545814714148,
+      "grad_norm": 1.146908986178407,
+      "learning_rate": 1.7667398234209798e-05,
+      "loss": 1.1695,
+      "step": 17662
+    },
+    {
+      "epoch": 0.8137006495600497,
+      "grad_norm": 1.1672303443985397,
+      "learning_rate": 1.76589305220637e-05,
+      "loss": 1.2868,
+      "step": 17663
+    },
+    {
+      "epoch": 0.8137467176486848,
+      "grad_norm": 0.5916281371002895,
+      "learning_rate": 1.7650464643100317e-05,
+      "loss": 1.4014,
+      "step": 17664
+    },
+    {
+      "epoch": 0.8137927857373197,
+      "grad_norm": 1.1618565735770257,
+      "learning_rate": 1.7642000597508125e-05,
+      "loss": 1.2372,
+      "step": 17665
+    },
+    {
+      "epoch": 0.8138388538259548,
+      "grad_norm": 1.7030591182629329,
+      "learning_rate": 1.7633538385475567e-05,
+      "loss": 0.9929,
+      "step": 17666
+    },
+    {
+      "epoch": 0.8138849219145897,
+      "grad_norm": 0.859752160441629,
+      "learning_rate": 1.7625078007191054e-05,
+      "loss": 1.1848,
+      "step": 17667
+    },
+    {
+      "epoch": 0.8139309900032248,
+      "grad_norm": 1.1115430196813114,
+      "learning_rate": 1.7616619462842955e-05,
+      "loss": 1.4267,
+      "step": 17668
+    },
+    {
+      "epoch": 0.8139770580918597,
+      "grad_norm": 0.7611838123531467,
+      "learning_rate": 1.760816275261955e-05,
+      "loss": 1.1717,
+      "step": 17669
+    },
+    {
+      "epoch": 0.8140231261804948,
+      "grad_norm": 1.056874302072817,
+      "learning_rate": 1.7599707876709125e-05,
+      "loss": 1.4183,
+      "step": 17670
+    },
+    {
+      "epoch": 0.8140691942691298,
+      "grad_norm": 1.2736972640770303,
+      "learning_rate": 1.759125483529992e-05,
+      "loss": 1.3303,
+      "step": 17671
+    },
+    {
+      "epoch": 0.8141152623577648,
+      "grad_norm": 0.6383795030842182,
+      "learning_rate": 1.7582803628580147e-05,
+      "loss": 1.3696,
+      "step": 17672
+    },
+    {
+      "epoch": 0.8141613304463998,
+      "grad_norm": 0.919758724982064,
+      "learning_rate": 1.7574354256737912e-05,
+      "loss": 1.3791,
+      "step": 17673
+    },
+    {
+      "epoch": 0.8142073985350348,
+      "grad_norm": 0.959917240339847,
+      "learning_rate": 1.7565906719961356e-05,
+      "loss": 0.9994,
+      "step": 17674
+    },
+    {
+      "epoch": 0.8142534666236698,
+      "grad_norm": 1.5709667556738098,
+      "learning_rate": 1.7557461018438558e-05,
+      "loss": 1.1742,
+      "step": 17675
+    },
+    {
+      "epoch": 0.8142995347123048,
+      "grad_norm": 1.3815909584600194,
+      "learning_rate": 1.7549017152357518e-05,
+      "loss": 0.8748,
+      "step": 17676
+    },
+    {
+      "epoch": 0.8143456028009398,
+      "grad_norm": 0.7316235653461303,
+      "learning_rate": 1.7540575121906223e-05,
+      "loss": 1.4857,
+      "step": 17677
+    },
+    {
+      "epoch": 0.8143916708895748,
+      "grad_norm": 1.0259874942648417,
+      "learning_rate": 1.7532134927272647e-05,
+      "loss": 1.1079,
+      "step": 17678
+    },
+    {
+      "epoch": 0.8144377389782098,
+      "grad_norm": 0.7026264341352259,
+      "learning_rate": 1.752369656864471e-05,
+      "loss": 1.3087,
+      "step": 17679
+    },
+    {
+      "epoch": 0.8144838070668448,
+      "grad_norm": 1.236264236786914,
+      "learning_rate": 1.7515260046210223e-05,
+      "loss": 1.3277,
+      "step": 17680
+    },
+    {
+      "epoch": 0.8145298751554798,
+      "grad_norm": 1.2687891719822912,
+      "learning_rate": 1.7506825360157042e-05,
+      "loss": 1.0453,
+      "step": 17681
+    },
+    {
+      "epoch": 0.8145759432441148,
+      "grad_norm": 0.9761716551301524,
+      "learning_rate": 1.749839251067297e-05,
+      "loss": 1.3305,
+      "step": 17682
+    },
+    {
+      "epoch": 0.8146220113327498,
+      "grad_norm": 0.924100241158924,
+      "learning_rate": 1.7489961497945706e-05,
+      "loss": 1.1738,
+      "step": 17683
+    },
+    {
+      "epoch": 0.8146680794213849,
+      "grad_norm": 0.6347072854228933,
+      "learning_rate": 1.7481532322162974e-05,
+      "loss": 1.5674,
+      "step": 17684
+    },
+    {
+      "epoch": 0.8147141475100198,
+      "grad_norm": 0.8832838382337833,
+      "learning_rate": 1.7473104983512433e-05,
+      "loss": 1.2251,
+      "step": 17685
+    },
+    {
+      "epoch": 0.8147602155986549,
+      "grad_norm": 0.8107411033694933,
+      "learning_rate": 1.7464679482181743e-05,
+      "loss": 1.1843,
+      "step": 17686
+    },
+    {
+      "epoch": 0.8148062836872898,
+      "grad_norm": 0.8250472538484065,
+      "learning_rate": 1.7456255818358413e-05,
+      "loss": 0.804,
+      "step": 17687
+    },
+    {
+      "epoch": 0.8148523517759249,
+      "grad_norm": 1.4863306562619742,
+      "learning_rate": 1.7447833992230013e-05,
+      "loss": 1.0755,
+      "step": 17688
+    },
+    {
+      "epoch": 0.8148984198645598,
+      "grad_norm": 0.7466737748973332,
+      "learning_rate": 1.7439414003984046e-05,
+      "loss": 0.9487,
+      "step": 17689
+    },
+    {
+      "epoch": 0.8149444879531949,
+      "grad_norm": 3.575336094212681,
+      "learning_rate": 1.7430995853807997e-05,
+      "loss": 1.2389,
+      "step": 17690
+    },
+    {
+      "epoch": 0.8149905560418298,
+      "grad_norm": 0.6818698703695258,
+      "learning_rate": 1.7422579541889217e-05,
+      "loss": 1.0615,
+      "step": 17691
+    },
+    {
+      "epoch": 0.8150366241304648,
+      "grad_norm": 1.1603403307269133,
+      "learning_rate": 1.7414165068415113e-05,
+      "loss": 1.214,
+      "step": 17692
+    },
+    {
+      "epoch": 0.8150826922190998,
+      "grad_norm": 1.3180289331257362,
+      "learning_rate": 1.7405752433573042e-05,
+      "loss": 0.9679,
+      "step": 17693
+    },
+    {
+      "epoch": 0.8151287603077348,
+      "grad_norm": 0.9896602060843784,
+      "learning_rate": 1.7397341637550257e-05,
+      "loss": 1.0513,
+      "step": 17694
+    },
+    {
+      "epoch": 0.8151748283963698,
+      "grad_norm": 0.9864897994018692,
+      "learning_rate": 1.7388932680534022e-05,
+      "loss": 1.1928,
+      "step": 17695
+    },
+    {
+      "epoch": 0.8152208964850048,
+      "grad_norm": 0.8663983021936134,
+      "learning_rate": 1.738052556271156e-05,
+      "loss": 1.4466,
+      "step": 17696
+    },
+    {
+      "epoch": 0.8152669645736398,
+      "grad_norm": 1.9351755498133685,
+      "learning_rate": 1.7372120284270055e-05,
+      "loss": 1.1014,
+      "step": 17697
+    },
+    {
+      "epoch": 0.8153130326622748,
+      "grad_norm": 0.9870470693466636,
+      "learning_rate": 1.7363716845396583e-05,
+      "loss": 1.3433,
+      "step": 17698
+    },
+    {
+      "epoch": 0.8153591007509099,
+      "grad_norm": 1.088580978400275,
+      "learning_rate": 1.735531524627827e-05,
+      "loss": 1.4213,
+      "step": 17699
+    },
+    {
+      "epoch": 0.8154051688395448,
+      "grad_norm": 2.747288676799858,
+      "learning_rate": 1.734691548710219e-05,
+      "loss": 1.1205,
+      "step": 17700
+    },
+    {
+      "epoch": 0.8154512369281799,
+      "grad_norm": 2.178192987236719,
+      "learning_rate": 1.733851756805528e-05,
+      "loss": 1.1588,
+      "step": 17701
+    },
+    {
+      "epoch": 0.8154973050168148,
+      "grad_norm": 1.2789045427633219,
+      "learning_rate": 1.7330121489324546e-05,
+      "loss": 1.265,
+      "step": 17702
+    },
+    {
+      "epoch": 0.8155433731054499,
+      "grad_norm": 1.015585288087597,
+      "learning_rate": 1.732172725109692e-05,
+      "loss": 1.4082,
+      "step": 17703
+    },
+    {
+      "epoch": 0.8155894411940848,
+      "grad_norm": 1.6017020777278175,
+      "learning_rate": 1.731333485355928e-05,
+      "loss": 0.8723,
+      "step": 17704
+    },
+    {
+      "epoch": 0.8156355092827199,
+      "grad_norm": 1.2279731358622947,
+      "learning_rate": 1.730494429689845e-05,
+      "loss": 1.1669,
+      "step": 17705
+    },
+    {
+      "epoch": 0.8156815773713548,
+      "grad_norm": 1.058074549948451,
+      "learning_rate": 1.7296555581301242e-05,
+      "loss": 0.9565,
+      "step": 17706
+    },
+    {
+      "epoch": 0.8157276454599899,
+      "grad_norm": 1.8559646108453065,
+      "learning_rate": 1.728816870695442e-05,
+      "loss": 0.9817,
+      "step": 17707
+    },
+    {
+      "epoch": 0.8157737135486248,
+      "grad_norm": 1.7169724479116675,
+      "learning_rate": 1.7279783674044723e-05,
+      "loss": 1.0736,
+      "step": 17708
+    },
+    {
+      "epoch": 0.8158197816372599,
+      "grad_norm": 0.8976453522540728,
+      "learning_rate": 1.7271400482758783e-05,
+      "loss": 1.1835,
+      "step": 17709
+    },
+    {
+      "epoch": 0.8158658497258948,
+      "grad_norm": 0.9519636815606207,
+      "learning_rate": 1.7263019133283252e-05,
+      "loss": 1.3888,
+      "step": 17710
+    },
+    {
+      "epoch": 0.8159119178145299,
+      "grad_norm": 2.972964050589725,
+      "learning_rate": 1.725463962580478e-05,
+      "loss": 1.1495,
+      "step": 17711
+    },
+    {
+      "epoch": 0.8159579859031649,
+      "grad_norm": 1.2357356436211309,
+      "learning_rate": 1.7246261960509847e-05,
+      "loss": 0.9432,
+      "step": 17712
+    },
+    {
+      "epoch": 0.8160040539917999,
+      "grad_norm": 0.786538608921022,
+      "learning_rate": 1.723788613758499e-05,
+      "loss": 1.2048,
+      "step": 17713
+    },
+    {
+      "epoch": 0.8160501220804349,
+      "grad_norm": 1.7762643469240036,
+      "learning_rate": 1.72295121572167e-05,
+      "loss": 1.1627,
+      "step": 17714
+    },
+    {
+      "epoch": 0.8160961901690699,
+      "grad_norm": 4.111139180337779,
+      "learning_rate": 1.722114001959143e-05,
+      "loss": 0.9636,
+      "step": 17715
+    },
+    {
+      "epoch": 0.8161422582577049,
+      "grad_norm": 0.7840801742988874,
+      "learning_rate": 1.7212769724895507e-05,
+      "loss": 1.0499,
+      "step": 17716
+    },
+    {
+      "epoch": 0.8161883263463399,
+      "grad_norm": 0.9280947307293218,
+      "learning_rate": 1.720440127331532e-05,
+      "loss": 1.0877,
+      "step": 17717
+    },
+    {
+      "epoch": 0.8162343944349749,
+      "grad_norm": 1.75625650145578,
+      "learning_rate": 1.7196034665037208e-05,
+      "loss": 1.134,
+      "step": 17718
+    },
+    {
+      "epoch": 0.8162804625236099,
+      "grad_norm": 0.7798236263337887,
+      "learning_rate": 1.7187669900247372e-05,
+      "loss": 0.9772,
+      "step": 17719
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 1.0218688965525247,
+      "learning_rate": 1.7179306979132072e-05,
+      "loss": 1.0078,
+      "step": 17720
+    },
+    {
+      "epoch": 0.8163725987008799,
+      "grad_norm": 0.8455136119756853,
+      "learning_rate": 1.71709459018775e-05,
+      "loss": 1.1878,
+      "step": 17721
+    },
+    {
+      "epoch": 0.8164186667895149,
+      "grad_norm": 1.0797296414202997,
+      "learning_rate": 1.7162586668669824e-05,
+      "loss": 1.0224,
+      "step": 17722
+    },
+    {
+      "epoch": 0.8164647348781499,
+      "grad_norm": 25.36181735694919,
+      "learning_rate": 1.715422927969509e-05,
+      "loss": 0.9887,
+      "step": 17723
+    },
+    {
+      "epoch": 0.8165108029667849,
+      "grad_norm": 1.062345600714272,
+      "learning_rate": 1.7145873735139384e-05,
+      "loss": 1.3745,
+      "step": 17724
+    },
+    {
+      "epoch": 0.81655687105542,
+      "grad_norm": 1.5250049016372549,
+      "learning_rate": 1.713752003518875e-05,
+      "loss": 1.0956,
+      "step": 17725
+    },
+    {
+      "epoch": 0.8166029391440549,
+      "grad_norm": 1.009092322549828,
+      "learning_rate": 1.7129168180029166e-05,
+      "loss": 1.1596,
+      "step": 17726
+    },
+    {
+      "epoch": 0.81664900723269,
+      "grad_norm": 0.9457629046216649,
+      "learning_rate": 1.7120818169846565e-05,
+      "loss": 1.1013,
+      "step": 17727
+    },
+    {
+      "epoch": 0.8166950753213249,
+      "grad_norm": 0.9228050224629861,
+      "learning_rate": 1.7112470004826787e-05,
+      "loss": 1.0351,
+      "step": 17728
+    },
+    {
+      "epoch": 0.81674114340996,
+      "grad_norm": 1.1098273158305685,
+      "learning_rate": 1.7104123685155794e-05,
+      "loss": 1.1601,
+      "step": 17729
+    },
+    {
+      "epoch": 0.8167872114985949,
+      "grad_norm": 1.3009607483393064,
+      "learning_rate": 1.7095779211019325e-05,
+      "loss": 1.4011,
+      "step": 17730
+    },
+    {
+      "epoch": 0.81683327958723,
+      "grad_norm": 0.7597717691242292,
+      "learning_rate": 1.708743658260321e-05,
+      "loss": 1.1948,
+      "step": 17731
+    },
+    {
+      "epoch": 0.8168793476758649,
+      "grad_norm": 0.8213192800758294,
+      "learning_rate": 1.7079095800093102e-05,
+      "loss": 1.3865,
+      "step": 17732
+    },
+    {
+      "epoch": 0.8169254157645,
+      "grad_norm": 1.929402457510708,
+      "learning_rate": 1.7070756863674796e-05,
+      "loss": 1.0427,
+      "step": 17733
+    },
+    {
+      "epoch": 0.8169714838531349,
+      "grad_norm": 0.8052888239500388,
+      "learning_rate": 1.706241977353388e-05,
+      "loss": 1.3809,
+      "step": 17734
+    },
+    {
+      "epoch": 0.81701755194177,
+      "grad_norm": 1.257949251127912,
+      "learning_rate": 1.7054084529855974e-05,
+      "loss": 1.2811,
+      "step": 17735
+    },
+    {
+      "epoch": 0.8170636200304049,
+      "grad_norm": 1.4392901155875382,
+      "learning_rate": 1.704575113282668e-05,
+      "loss": 1.1865,
+      "step": 17736
+    },
+    {
+      "epoch": 0.81710968811904,
+      "grad_norm": 0.8457206808951416,
+      "learning_rate": 1.7037419582631476e-05,
+      "loss": 1.1838,
+      "step": 17737
+    },
+    {
+      "epoch": 0.8171557562076749,
+      "grad_norm": 1.17975272688211,
+      "learning_rate": 1.7029089879455894e-05,
+      "loss": 1.2247,
+      "step": 17738
+    },
+    {
+      "epoch": 0.8172018242963099,
+      "grad_norm": 1.6234780081437425,
+      "learning_rate": 1.7020762023485314e-05,
+      "loss": 1.2103,
+      "step": 17739
+    },
+    {
+      "epoch": 0.817247892384945,
+      "grad_norm": 0.9001751461922042,
+      "learning_rate": 1.701243601490524e-05,
+      "loss": 1.4426,
+      "step": 17740
+    },
+    {
+      "epoch": 0.8172939604735799,
+      "grad_norm": 1.6459509242737735,
+      "learning_rate": 1.700411185390096e-05,
+      "loss": 1.5187,
+      "step": 17741
+    },
+    {
+      "epoch": 0.817340028562215,
+      "grad_norm": 0.6981580445660043,
+      "learning_rate": 1.6995789540657848e-05,
+      "loss": 1.4205,
+      "step": 17742
+    },
+    {
+      "epoch": 0.8173860966508499,
+      "grad_norm": 2.7835948983323613,
+      "learning_rate": 1.6987469075361106e-05,
+      "loss": 1.1749,
+      "step": 17743
+    },
+    {
+      "epoch": 0.817432164739485,
+      "grad_norm": 0.783441937306683,
+      "learning_rate": 1.697915045819609e-05,
+      "loss": 1.4018,
+      "step": 17744
+    },
+    {
+      "epoch": 0.8174782328281199,
+      "grad_norm": 1.0413289970859847,
+      "learning_rate": 1.6970833689347932e-05,
+      "loss": 1.0207,
+      "step": 17745
+    },
+    {
+      "epoch": 0.817524300916755,
+      "grad_norm": 0.9242284595484945,
+      "learning_rate": 1.6962518769001746e-05,
+      "loss": 0.9455,
+      "step": 17746
+    },
+    {
+      "epoch": 0.8175703690053899,
+      "grad_norm": 0.8740222481796505,
+      "learning_rate": 1.6954205697342752e-05,
+      "loss": 1.6855,
+      "step": 17747
+    },
+    {
+      "epoch": 0.817616437094025,
+      "grad_norm": 1.8184015039277126,
+      "learning_rate": 1.6945894474555945e-05,
+      "loss": 1.087,
+      "step": 17748
+    },
+    {
+      "epoch": 0.8176625051826599,
+      "grad_norm": 1.658450643206302,
+      "learning_rate": 1.6937585100826425e-05,
+      "loss": 1.0002,
+      "step": 17749
+    },
+    {
+      "epoch": 0.817708573271295,
+      "grad_norm": 0.742236939961151,
+      "learning_rate": 1.6929277576339098e-05,
+      "loss": 1.4308,
+      "step": 17750
+    },
+    {
+      "epoch": 0.8177546413599299,
+      "grad_norm": 2.7359709845966176,
+      "learning_rate": 1.6920971901279014e-05,
+      "loss": 1.102,
+      "step": 17751
+    },
+    {
+      "epoch": 0.817800709448565,
+      "grad_norm": 1.6787042642922492,
+      "learning_rate": 1.691266807583102e-05,
+      "loss": 1.4434,
+      "step": 17752
+    },
+    {
+      "epoch": 0.8178467775372,
+      "grad_norm": 0.9070877089033765,
+      "learning_rate": 1.6904366100180037e-05,
+      "loss": 1.3757,
+      "step": 17753
+    },
+    {
+      "epoch": 0.817892845625835,
+      "grad_norm": 0.8700914433111865,
+      "learning_rate": 1.689606597451081e-05,
+      "loss": 1.3434,
+      "step": 17754
+    },
+    {
+      "epoch": 0.81793891371447,
+      "grad_norm": 2.073162477645497,
+      "learning_rate": 1.6887767699008228e-05,
+      "loss": 1.2207,
+      "step": 17755
+    },
+    {
+      "epoch": 0.817984981803105,
+      "grad_norm": 1.0507836218291666,
+      "learning_rate": 1.6879471273856994e-05,
+      "loss": 1.2141,
+      "step": 17756
+    },
+    {
+      "epoch": 0.81803104989174,
+      "grad_norm": 0.8023223486228913,
+      "learning_rate": 1.6871176699241788e-05,
+      "loss": 1.5823,
+      "step": 17757
+    },
+    {
+      "epoch": 0.818077117980375,
+      "grad_norm": 1.0222631027347546,
+      "learning_rate": 1.686288397534729e-05,
+      "loss": 1.1526,
+      "step": 17758
+    },
+    {
+      "epoch": 0.81812318606901,
+      "grad_norm": 0.8652017594734168,
+      "learning_rate": 1.6854593102358142e-05,
+      "loss": 1.2578,
+      "step": 17759
+    },
+    {
+      "epoch": 0.818169254157645,
+      "grad_norm": 0.8313706964012922,
+      "learning_rate": 1.684630408045893e-05,
+      "loss": 1.274,
+      "step": 17760
+    },
+    {
+      "epoch": 0.81821532224628,
+      "grad_norm": 1.3488561333169784,
+      "learning_rate": 1.6838016909834132e-05,
+      "loss": 1.1431,
+      "step": 17761
+    },
+    {
+      "epoch": 0.818261390334915,
+      "grad_norm": 0.8454626843338628,
+      "learning_rate": 1.682973159066835e-05,
+      "loss": 0.874,
+      "step": 17762
+    },
+    {
+      "epoch": 0.81830745842355,
+      "grad_norm": 0.7125192285732351,
+      "learning_rate": 1.6821448123145955e-05,
+      "loss": 1.5653,
+      "step": 17763
+    },
+    {
+      "epoch": 0.818353526512185,
+      "grad_norm": 0.9964121009774991,
+      "learning_rate": 1.681316650745143e-05,
+      "loss": 1.0954,
+      "step": 17764
+    },
+    {
+      "epoch": 0.81839959460082,
+      "grad_norm": 1.005310509480259,
+      "learning_rate": 1.6804886743769088e-05,
+      "loss": 1.0411,
+      "step": 17765
+    },
+    {
+      "epoch": 0.8184456626894551,
+      "grad_norm": 0.9892413087660219,
+      "learning_rate": 1.6796608832283288e-05,
+      "loss": 1.1007,
+      "step": 17766
+    },
+    {
+      "epoch": 0.81849173077809,
+      "grad_norm": 0.9757374942129584,
+      "learning_rate": 1.6788332773178373e-05,
+      "loss": 1.245,
+      "step": 17767
+    },
+    {
+      "epoch": 0.8185377988667251,
+      "grad_norm": 0.8321708037178426,
+      "learning_rate": 1.6780058566638513e-05,
+      "loss": 1.2554,
+      "step": 17768
+    },
+    {
+      "epoch": 0.81858386695536,
+      "grad_norm": 12.353366343895605,
+      "learning_rate": 1.677178621284795e-05,
+      "loss": 1.3085,
+      "step": 17769
+    },
+    {
+      "epoch": 0.8186299350439951,
+      "grad_norm": 0.7187180056776535,
+      "learning_rate": 1.6763515711990875e-05,
+      "loss": 1.3297,
+      "step": 17770
+    },
+    {
+      "epoch": 0.81867600313263,
+      "grad_norm": 0.7307708201523301,
+      "learning_rate": 1.6755247064251424e-05,
+      "loss": 1.3201,
+      "step": 17771
+    },
+    {
+      "epoch": 0.8187220712212651,
+      "grad_norm": 0.7960524107520643,
+      "learning_rate": 1.6746980269813607e-05,
+      "loss": 1.136,
+      "step": 17772
+    },
+    {
+      "epoch": 0.8187681393099,
+      "grad_norm": 0.7221998567445319,
+      "learning_rate": 1.6738715328861586e-05,
+      "loss": 1.3547,
+      "step": 17773
+    },
+    {
+      "epoch": 0.818814207398535,
+      "grad_norm": 0.9713339523171921,
+      "learning_rate": 1.6730452241579296e-05,
+      "loss": 1.2907,
+      "step": 17774
+    },
+    {
+      "epoch": 0.81886027548717,
+      "grad_norm": 1.1910871212879621,
+      "learning_rate": 1.672219100815069e-05,
+      "loss": 1.3536,
+      "step": 17775
+    },
+    {
+      "epoch": 0.818906343575805,
+      "grad_norm": 1.2685397225306019,
+      "learning_rate": 1.6713931628759717e-05,
+      "loss": 1.182,
+      "step": 17776
+    },
+    {
+      "epoch": 0.81895241166444,
+      "grad_norm": 1.1195487215798876,
+      "learning_rate": 1.670567410359024e-05,
+      "loss": 1.1644,
+      "step": 17777
+    },
+    {
+      "epoch": 0.818998479753075,
+      "grad_norm": 0.9983163361883458,
+      "learning_rate": 1.669741843282614e-05,
+      "loss": 1.3812,
+      "step": 17778
+    },
+    {
+      "epoch": 0.8190445478417101,
+      "grad_norm": 1.8571022442510012,
+      "learning_rate": 1.6689164616651154e-05,
+      "loss": 1.2212,
+      "step": 17779
+    },
+    {
+      "epoch": 0.819090615930345,
+      "grad_norm": 3.4173847385675775,
+      "learning_rate": 1.668091265524908e-05,
+      "loss": 1.001,
+      "step": 17780
+    },
+    {
+      "epoch": 0.8191366840189801,
+      "grad_norm": 0.9852224749505818,
+      "learning_rate": 1.667266254880362e-05,
+      "loss": 1.3684,
+      "step": 17781
+    },
+    {
+      "epoch": 0.819182752107615,
+      "grad_norm": 1.0414791596883233,
+      "learning_rate": 1.6664414297498477e-05,
+      "loss": 1.186,
+      "step": 17782
+    },
+    {
+      "epoch": 0.8192288201962501,
+      "grad_norm": 1.0352873882830622,
+      "learning_rate": 1.665616790151724e-05,
+      "loss": 0.9557,
+      "step": 17783
+    },
+    {
+      "epoch": 0.819274888284885,
+      "grad_norm": 0.8286042566301662,
+      "learning_rate": 1.6647923361043514e-05,
+      "loss": 1.2543,
+      "step": 17784
+    },
+    {
+      "epoch": 0.8193209563735201,
+      "grad_norm": 0.9712406580114055,
+      "learning_rate": 1.663968067626088e-05,
+      "loss": 1.2403,
+      "step": 17785
+    },
+    {
+      "epoch": 0.819367024462155,
+      "grad_norm": 0.8409709909932612,
+      "learning_rate": 1.6631439847352802e-05,
+      "loss": 1.1299,
+      "step": 17786
+    },
+    {
+      "epoch": 0.8194130925507901,
+      "grad_norm": 2.776647848282938,
+      "learning_rate": 1.6623200874502765e-05,
+      "loss": 1.303,
+      "step": 17787
+    },
+    {
+      "epoch": 0.819459160639425,
+      "grad_norm": 1.4147136469298798,
+      "learning_rate": 1.661496375789421e-05,
+      "loss": 1.1741,
+      "step": 17788
+    },
+    {
+      "epoch": 0.8195052287280601,
+      "grad_norm": 0.7920145080250562,
+      "learning_rate": 1.6606728497710534e-05,
+      "loss": 1.4255,
+      "step": 17789
+    },
+    {
+      "epoch": 0.819551296816695,
+      "grad_norm": 0.744809955269903,
+      "learning_rate": 1.6598495094135024e-05,
+      "loss": 1.3046,
+      "step": 17790
+    },
+    {
+      "epoch": 0.8195973649053301,
+      "grad_norm": 1.1683577698902678,
+      "learning_rate": 1.6590263547351025e-05,
+      "loss": 0.8513,
+      "step": 17791
+    },
+    {
+      "epoch": 0.819643432993965,
+      "grad_norm": 0.9823484610067122,
+      "learning_rate": 1.6582033857541823e-05,
+      "loss": 1.3167,
+      "step": 17792
+    },
+    {
+      "epoch": 0.8196895010826001,
+      "grad_norm": 1.6532990953744195,
+      "learning_rate": 1.657380602489057e-05,
+      "loss": 1.0139,
+      "step": 17793
+    },
+    {
+      "epoch": 0.8197355691712351,
+      "grad_norm": 1.1277458017783066,
+      "learning_rate": 1.6565580049580485e-05,
+      "loss": 1.0166,
+      "step": 17794
+    },
+    {
+      "epoch": 0.8197816372598701,
+      "grad_norm": 1.7380424449191652,
+      "learning_rate": 1.6557355931794706e-05,
+      "loss": 1.3285,
+      "step": 17795
+    },
+    {
+      "epoch": 0.8198277053485051,
+      "grad_norm": 0.7838136541979159,
+      "learning_rate": 1.6549133671716332e-05,
+      "loss": 1.4854,
+      "step": 17796
+    },
+    {
+      "epoch": 0.8198737734371401,
+      "grad_norm": 0.9104825364967943,
+      "learning_rate": 1.6540913269528403e-05,
+      "loss": 1.4073,
+      "step": 17797
+    },
+    {
+      "epoch": 0.8199198415257751,
+      "grad_norm": 0.8824823180720962,
+      "learning_rate": 1.6532694725413933e-05,
+      "loss": 1.5131,
+      "step": 17798
+    },
+    {
+      "epoch": 0.8199659096144101,
+      "grad_norm": 0.8406452591916284,
+      "learning_rate": 1.652447803955589e-05,
+      "loss": 1.4374,
+      "step": 17799
+    },
+    {
+      "epoch": 0.8200119777030451,
+      "grad_norm": 1.121085135446571,
+      "learning_rate": 1.6516263212137252e-05,
+      "loss": 1.0866,
+      "step": 17800
+    },
+    {
+      "epoch": 0.8200580457916801,
+      "grad_norm": 1.0636740141005088,
+      "learning_rate": 1.6508050243340835e-05,
+      "loss": 1.4199,
+      "step": 17801
+    },
+    {
+      "epoch": 0.8201041138803151,
+      "grad_norm": 1.0088506200094403,
+      "learning_rate": 1.6499839133349526e-05,
+      "loss": 1.1006,
+      "step": 17802
+    },
+    {
+      "epoch": 0.8201501819689501,
+      "grad_norm": 1.096059951581587,
+      "learning_rate": 1.649162988234615e-05,
+      "loss": 1.3876,
+      "step": 17803
+    },
+    {
+      "epoch": 0.8201962500575851,
+      "grad_norm": 1.732851079511195,
+      "learning_rate": 1.648342249051342e-05,
+      "loss": 1.2618,
+      "step": 17804
+    },
+    {
+      "epoch": 0.8202423181462201,
+      "grad_norm": 2.375952073189153,
+      "learning_rate": 1.6475216958034103e-05,
+      "loss": 1.0098,
+      "step": 17805
+    },
+    {
+      "epoch": 0.8202883862348551,
+      "grad_norm": 1.117867918482745,
+      "learning_rate": 1.646701328509085e-05,
+      "loss": 1.285,
+      "step": 17806
+    },
+    {
+      "epoch": 0.8203344543234902,
+      "grad_norm": 0.9072510828806466,
+      "learning_rate": 1.6458811471866343e-05,
+      "loss": 1.1003,
+      "step": 17807
+    },
+    {
+      "epoch": 0.8203805224121251,
+      "grad_norm": 0.7485098160380301,
+      "learning_rate": 1.6450611518543136e-05,
+      "loss": 1.1961,
+      "step": 17808
+    },
+    {
+      "epoch": 0.8204265905007602,
+      "grad_norm": 0.8483308947352108,
+      "learning_rate": 1.6442413425303814e-05,
+      "loss": 1.3136,
+      "step": 17809
+    },
+    {
+      "epoch": 0.8204726585893951,
+      "grad_norm": 2.5929756351211037,
+      "learning_rate": 1.6434217192330902e-05,
+      "loss": 0.9983,
+      "step": 17810
+    },
+    {
+      "epoch": 0.8205187266780302,
+      "grad_norm": 1.4815585701292882,
+      "learning_rate": 1.642602281980684e-05,
+      "loss": 0.9792,
+      "step": 17811
+    },
+    {
+      "epoch": 0.8205647947666651,
+      "grad_norm": 0.7681821284380529,
+      "learning_rate": 1.6417830307914074e-05,
+      "loss": 1.4389,
+      "step": 17812
+    },
+    {
+      "epoch": 0.8206108628553002,
+      "grad_norm": 0.6781531316532539,
+      "learning_rate": 1.640963965683501e-05,
+      "loss": 1.651,
+      "step": 17813
+    },
+    {
+      "epoch": 0.8206569309439351,
+      "grad_norm": 0.6028759921549186,
+      "learning_rate": 1.6401450866752022e-05,
+      "loss": 1.3596,
+      "step": 17814
+    },
+    {
+      "epoch": 0.8207029990325702,
+      "grad_norm": 0.9561522714552846,
+      "learning_rate": 1.6393263937847358e-05,
+      "loss": 1.4915,
+      "step": 17815
+    },
+    {
+      "epoch": 0.8207490671212051,
+      "grad_norm": 0.6582915569388599,
+      "learning_rate": 1.6385078870303317e-05,
+      "loss": 1.4933,
+      "step": 17816
+    },
+    {
+      "epoch": 0.8207951352098402,
+      "grad_norm": 0.9311733779561305,
+      "learning_rate": 1.637689566430213e-05,
+      "loss": 1.3884,
+      "step": 17817
+    },
+    {
+      "epoch": 0.8208412032984751,
+      "grad_norm": 0.823233641426075,
+      "learning_rate": 1.636871432002599e-05,
+      "loss": 1.2282,
+      "step": 17818
+    },
+    {
+      "epoch": 0.8208872713871102,
+      "grad_norm": 1.072579126764794,
+      "learning_rate": 1.6360534837657014e-05,
+      "loss": 1.3116,
+      "step": 17819
+    },
+    {
+      "epoch": 0.8209333394757452,
+      "grad_norm": 0.7936710719225263,
+      "learning_rate": 1.6352357217377323e-05,
+      "loss": 1.2696,
+      "step": 17820
+    },
+    {
+      "epoch": 0.8209794075643801,
+      "grad_norm": 0.9621743614551577,
+      "learning_rate": 1.634418145936899e-05,
+      "loss": 1.4577,
+      "step": 17821
+    },
+    {
+      "epoch": 0.8210254756530152,
+      "grad_norm": 1.2316402024036608,
+      "learning_rate": 1.6336007563814003e-05,
+      "loss": 1.1671,
+      "step": 17822
+    },
+    {
+      "epoch": 0.8210715437416501,
+      "grad_norm": 1.1752336402708325,
+      "learning_rate": 1.632783553089434e-05,
+      "loss": 1.2482,
+      "step": 17823
+    },
+    {
+      "epoch": 0.8211176118302852,
+      "grad_norm": 0.8569405644710796,
+      "learning_rate": 1.631966536079197e-05,
+      "loss": 1.4453,
+      "step": 17824
+    },
+    {
+      "epoch": 0.8211636799189201,
+      "grad_norm": 0.8872457160681403,
+      "learning_rate": 1.6311497053688783e-05,
+      "loss": 1.3545,
+      "step": 17825
+    },
+    {
+      "epoch": 0.8212097480075552,
+      "grad_norm": 0.6680398876193023,
+      "learning_rate": 1.6303330609766597e-05,
+      "loss": 1.5342,
+      "step": 17826
+    },
+    {
+      "epoch": 0.8212558160961901,
+      "grad_norm": 0.9318760220457474,
+      "learning_rate": 1.6295166029207244e-05,
+      "loss": 1.055,
+      "step": 17827
+    },
+    {
+      "epoch": 0.8213018841848252,
+      "grad_norm": 1.4158196004644665,
+      "learning_rate": 1.628700331219253e-05,
+      "loss": 1.081,
+      "step": 17828
+    },
+    {
+      "epoch": 0.8213479522734601,
+      "grad_norm": 3.227447955732837,
+      "learning_rate": 1.627884245890411e-05,
+      "loss": 1.2047,
+      "step": 17829
+    },
+    {
+      "epoch": 0.8213940203620952,
+      "grad_norm": 0.9814459478359371,
+      "learning_rate": 1.627068346952374e-05,
+      "loss": 1.461,
+      "step": 17830
+    },
+    {
+      "epoch": 0.8214400884507301,
+      "grad_norm": 1.0704407844309998,
+      "learning_rate": 1.6262526344232997e-05,
+      "loss": 1.1378,
+      "step": 17831
+    },
+    {
+      "epoch": 0.8214861565393652,
+      "grad_norm": 0.7841968370465346,
+      "learning_rate": 1.6254371083213573e-05,
+      "loss": 1.3664,
+      "step": 17832
+    },
+    {
+      "epoch": 0.8215322246280001,
+      "grad_norm": 1.3491247222350566,
+      "learning_rate": 1.6246217686646957e-05,
+      "loss": 1.3189,
+      "step": 17833
+    },
+    {
+      "epoch": 0.8215782927166352,
+      "grad_norm": 1.850109096873045,
+      "learning_rate": 1.6238066154714694e-05,
+      "loss": 1.3561,
+      "step": 17834
+    },
+    {
+      "epoch": 0.8216243608052702,
+      "grad_norm": 0.8113287840800408,
+      "learning_rate": 1.6229916487598273e-05,
+      "loss": 0.9953,
+      "step": 17835
+    },
+    {
+      "epoch": 0.8216704288939052,
+      "grad_norm": 0.8064623902393534,
+      "learning_rate": 1.6221768685479143e-05,
+      "loss": 0.8689,
+      "step": 17836
+    },
+    {
+      "epoch": 0.8217164969825402,
+      "grad_norm": 1.3970956095055451,
+      "learning_rate": 1.6213622748538693e-05,
+      "loss": 1.2275,
+      "step": 17837
+    },
+    {
+      "epoch": 0.8217625650711752,
+      "grad_norm": 1.204152806772102,
+      "learning_rate": 1.6205478676958217e-05,
+      "loss": 1.163,
+      "step": 17838
+    },
+    {
+      "epoch": 0.8218086331598102,
+      "grad_norm": 0.9724289427880332,
+      "learning_rate": 1.6197336470919134e-05,
+      "loss": 1.2695,
+      "step": 17839
+    },
+    {
+      "epoch": 0.8218547012484452,
+      "grad_norm": 1.1636534387432853,
+      "learning_rate": 1.6189196130602634e-05,
+      "loss": 1.4112,
+      "step": 17840
+    },
+    {
+      "epoch": 0.8219007693370802,
+      "grad_norm": 0.8796747529426789,
+      "learning_rate": 1.6181057656190013e-05,
+      "loss": 1.5032,
+      "step": 17841
+    },
+    {
+      "epoch": 0.8219468374257152,
+      "grad_norm": 1.2441178284522432,
+      "learning_rate": 1.6172921047862366e-05,
+      "loss": 1.5556,
+      "step": 17842
+    },
+    {
+      "epoch": 0.8219929055143502,
+      "grad_norm": 1.0672349211479308,
+      "learning_rate": 1.616478630580095e-05,
+      "loss": 1.1448,
+      "step": 17843
+    },
+    {
+      "epoch": 0.8220389736029852,
+      "grad_norm": 1.0364630819594913,
+      "learning_rate": 1.6156653430186786e-05,
+      "loss": 1.3028,
+      "step": 17844
+    },
+    {
+      "epoch": 0.8220850416916202,
+      "grad_norm": 2.180652843512926,
+      "learning_rate": 1.6148522421200984e-05,
+      "loss": 1.2679,
+      "step": 17845
+    },
+    {
+      "epoch": 0.8221311097802552,
+      "grad_norm": 0.95722108287642,
+      "learning_rate": 1.614039327902458e-05,
+      "loss": 1.3679,
+      "step": 17846
+    },
+    {
+      "epoch": 0.8221771778688902,
+      "grad_norm": 0.8824547931970664,
+      "learning_rate": 1.6132266003838504e-05,
+      "loss": 1.339,
+      "step": 17847
+    },
+    {
+      "epoch": 0.8222232459575253,
+      "grad_norm": 0.8569278281947355,
+      "learning_rate": 1.6124140595823744e-05,
+      "loss": 1.2616,
+      "step": 17848
+    },
+    {
+      "epoch": 0.8222693140461602,
+      "grad_norm": 1.8273753947176987,
+      "learning_rate": 1.6116017055161136e-05,
+      "loss": 0.9513,
+      "step": 17849
+    },
+    {
+      "epoch": 0.8223153821347953,
+      "grad_norm": 0.7394014889056797,
+      "learning_rate": 1.6107895382031624e-05,
+      "loss": 1.2489,
+      "step": 17850
+    },
+    {
+      "epoch": 0.8223614502234302,
+      "grad_norm": 1.0958027046976033,
+      "learning_rate": 1.609977557661596e-05,
+      "loss": 0.9886,
+      "step": 17851
+    },
+    {
+      "epoch": 0.8224075183120653,
+      "grad_norm": 3.9570254280248855,
+      "learning_rate": 1.6091657639094947e-05,
+      "loss": 1.3585,
+      "step": 17852
+    },
+    {
+      "epoch": 0.8224535864007002,
+      "grad_norm": 0.8983974482270771,
+      "learning_rate": 1.6083541569649262e-05,
+      "loss": 1.3402,
+      "step": 17853
+    },
+    {
+      "epoch": 0.8224996544893353,
+      "grad_norm": 1.9703584349697418,
+      "learning_rate": 1.6075427368459684e-05,
+      "loss": 1.2576,
+      "step": 17854
+    },
+    {
+      "epoch": 0.8225457225779702,
+      "grad_norm": 0.7475099509470954,
+      "learning_rate": 1.6067315035706822e-05,
+      "loss": 1.29,
+      "step": 17855
+    },
+    {
+      "epoch": 0.8225917906666053,
+      "grad_norm": 0.7969744255345803,
+      "learning_rate": 1.605920457157124e-05,
+      "loss": 1.6319,
+      "step": 17856
+    },
+    {
+      "epoch": 0.8226378587552402,
+      "grad_norm": 0.8733711843647133,
+      "learning_rate": 1.6051095976233555e-05,
+      "loss": 1.2198,
+      "step": 17857
+    },
+    {
+      "epoch": 0.8226839268438753,
+      "grad_norm": 1.0089218166017135,
+      "learning_rate": 1.6042989249874274e-05,
+      "loss": 1.3034,
+      "step": 17858
+    },
+    {
+      "epoch": 0.8227299949325102,
+      "grad_norm": 1.2963342690315895,
+      "learning_rate": 1.6034884392673897e-05,
+      "loss": 1.1618,
+      "step": 17859
+    },
+    {
+      "epoch": 0.8227760630211453,
+      "grad_norm": 0.6966858275488168,
+      "learning_rate": 1.602678140481281e-05,
+      "loss": 1.218,
+      "step": 17860
+    },
+    {
+      "epoch": 0.8228221311097803,
+      "grad_norm": 1.1076530712161812,
+      "learning_rate": 1.6018680286471498e-05,
+      "loss": 1.0018,
+      "step": 17861
+    },
+    {
+      "epoch": 0.8228681991984153,
+      "grad_norm": 0.9499691547831086,
+      "learning_rate": 1.601058103783024e-05,
+      "loss": 1.1879,
+      "step": 17862
+    },
+    {
+      "epoch": 0.8229142672870503,
+      "grad_norm": 0.829416357568469,
+      "learning_rate": 1.6002483659069423e-05,
+      "loss": 1.2163,
+      "step": 17863
+    },
+    {
+      "epoch": 0.8229603353756852,
+      "grad_norm": 1.5132428010929158,
+      "learning_rate": 1.5994388150369244e-05,
+      "loss": 1.013,
+      "step": 17864
+    },
+    {
+      "epoch": 0.8230064034643203,
+      "grad_norm": 0.8257869992035574,
+      "learning_rate": 1.5986294511909983e-05,
+      "loss": 1.0989,
+      "step": 17865
+    },
+    {
+      "epoch": 0.8230524715529552,
+      "grad_norm": 1.1047000470820436,
+      "learning_rate": 1.597820274387185e-05,
+      "loss": 1.1494,
+      "step": 17866
+    },
+    {
+      "epoch": 0.8230985396415903,
+      "grad_norm": 1.0385992448111983,
+      "learning_rate": 1.5970112846434936e-05,
+      "loss": 1.172,
+      "step": 17867
+    },
+    {
+      "epoch": 0.8231446077302252,
+      "grad_norm": 1.3718039265834634,
+      "learning_rate": 1.5962024819779374e-05,
+      "loss": 1.2768,
+      "step": 17868
+    },
+    {
+      "epoch": 0.8231906758188603,
+      "grad_norm": 1.9739829527878143,
+      "learning_rate": 1.5953938664085242e-05,
+      "loss": 1.3155,
+      "step": 17869
+    },
+    {
+      "epoch": 0.8232367439074952,
+      "grad_norm": 1.0144292523697975,
+      "learning_rate": 1.5945854379532577e-05,
+      "loss": 1.2397,
+      "step": 17870
+    },
+    {
+      "epoch": 0.8232828119961303,
+      "grad_norm": 1.2436999070882422,
+      "learning_rate": 1.5937771966301308e-05,
+      "loss": 0.9141,
+      "step": 17871
+    },
+    {
+      "epoch": 0.8233288800847652,
+      "grad_norm": 0.7992058799481581,
+      "learning_rate": 1.5929691424571446e-05,
+      "loss": 1.3945,
+      "step": 17872
+    },
+    {
+      "epoch": 0.8233749481734003,
+      "grad_norm": 1.152938646259957,
+      "learning_rate": 1.592161275452285e-05,
+      "loss": 1.1344,
+      "step": 17873
+    },
+    {
+      "epoch": 0.8234210162620353,
+      "grad_norm": 0.8873371808943732,
+      "learning_rate": 1.5913535956335368e-05,
+      "loss": 1.2904,
+      "step": 17874
+    },
+    {
+      "epoch": 0.8234670843506703,
+      "grad_norm": 0.9162106692707799,
+      "learning_rate": 1.590546103018883e-05,
+      "loss": 1.1799,
+      "step": 17875
+    },
+    {
+      "epoch": 0.8235131524393053,
+      "grad_norm": 1.5720552613765462,
+      "learning_rate": 1.589738797626301e-05,
+      "loss": 1.1649,
+      "step": 17876
+    },
+    {
+      "epoch": 0.8235592205279403,
+      "grad_norm": 1.0304485987258087,
+      "learning_rate": 1.5889316794737663e-05,
+      "loss": 1.159,
+      "step": 17877
+    },
+    {
+      "epoch": 0.8236052886165753,
+      "grad_norm": 1.2693141222993884,
+      "learning_rate": 1.5881247485792438e-05,
+      "loss": 0.9542,
+      "step": 17878
+    },
+    {
+      "epoch": 0.8236513567052103,
+      "grad_norm": 1.1095565103895209,
+      "learning_rate": 1.5873180049606994e-05,
+      "loss": 1.492,
+      "step": 17879
+    },
+    {
+      "epoch": 0.8236974247938453,
+      "grad_norm": 1.458926615841886,
+      "learning_rate": 1.5865114486360965e-05,
+      "loss": 1.226,
+      "step": 17880
+    },
+    {
+      "epoch": 0.8237434928824803,
+      "grad_norm": 1.8866137872336572,
+      "learning_rate": 1.585705079623392e-05,
+      "loss": 1.2877,
+      "step": 17881
+    },
+    {
+      "epoch": 0.8237895609711153,
+      "grad_norm": 0.7111463799516391,
+      "learning_rate": 1.5848988979405343e-05,
+      "loss": 1.2586,
+      "step": 17882
+    },
+    {
+      "epoch": 0.8238356290597503,
+      "grad_norm": 1.0513727430071143,
+      "learning_rate": 1.584092903605473e-05,
+      "loss": 1.1945,
+      "step": 17883
+    },
+    {
+      "epoch": 0.8238816971483853,
+      "grad_norm": 1.315963232461727,
+      "learning_rate": 1.5832870966361556e-05,
+      "loss": 1.2067,
+      "step": 17884
+    },
+    {
+      "epoch": 0.8239277652370203,
+      "grad_norm": 1.842058076108786,
+      "learning_rate": 1.582481477050518e-05,
+      "loss": 1.0916,
+      "step": 17885
+    },
+    {
+      "epoch": 0.8239738333256553,
+      "grad_norm": 0.9845567586619928,
+      "learning_rate": 1.5816760448664968e-05,
+      "loss": 0.9261,
+      "step": 17886
+    },
+    {
+      "epoch": 0.8240199014142903,
+      "grad_norm": 0.940463936021051,
+      "learning_rate": 1.5808708001020233e-05,
+      "loss": 1.0681,
+      "step": 17887
+    },
+    {
+      "epoch": 0.8240659695029253,
+      "grad_norm": 0.809055388227448,
+      "learning_rate": 1.58006574277503e-05,
+      "loss": 1.2619,
+      "step": 17888
+    },
+    {
+      "epoch": 0.8241120375915604,
+      "grad_norm": 0.8002327595894748,
+      "learning_rate": 1.579260872903432e-05,
+      "loss": 1.3996,
+      "step": 17889
+    },
+    {
+      "epoch": 0.8241581056801953,
+      "grad_norm": 0.84742214956857,
+      "learning_rate": 1.578456190505152e-05,
+      "loss": 1.4384,
+      "step": 17890
+    },
+    {
+      "epoch": 0.8242041737688304,
+      "grad_norm": 1.0053379730415934,
+      "learning_rate": 1.5776516955981057e-05,
+      "loss": 1.1585,
+      "step": 17891
+    },
+    {
+      "epoch": 0.8242502418574653,
+      "grad_norm": 4.090221963926515,
+      "learning_rate": 1.5768473882002055e-05,
+      "loss": 0.9966,
+      "step": 17892
+    },
+    {
+      "epoch": 0.8242963099461004,
+      "grad_norm": 0.9012178450310456,
+      "learning_rate": 1.5760432683293537e-05,
+      "loss": 1.3483,
+      "step": 17893
+    },
+    {
+      "epoch": 0.8243423780347353,
+      "grad_norm": 0.9627981414646596,
+      "learning_rate": 1.5752393360034535e-05,
+      "loss": 1.255,
+      "step": 17894
+    },
+    {
+      "epoch": 0.8243884461233704,
+      "grad_norm": 1.0366715913728612,
+      "learning_rate": 1.5744355912404075e-05,
+      "loss": 1.4991,
+      "step": 17895
+    },
+    {
+      "epoch": 0.8244345142120053,
+      "grad_norm": 1.1164774703869504,
+      "learning_rate": 1.5736320340581033e-05,
+      "loss": 1.3758,
+      "step": 17896
+    },
+    {
+      "epoch": 0.8244805823006404,
+      "grad_norm": 1.7195197128588027,
+      "learning_rate": 1.572828664474434e-05,
+      "loss": 1.3267,
+      "step": 17897
+    },
+    {
+      "epoch": 0.8245266503892753,
+      "grad_norm": 0.9422546723774725,
+      "learning_rate": 1.5720254825072846e-05,
+      "loss": 1.1316,
+      "step": 17898
+    },
+    {
+      "epoch": 0.8245727184779104,
+      "grad_norm": 1.2411368176831843,
+      "learning_rate": 1.5712224881745395e-05,
+      "loss": 1.1512,
+      "step": 17899
+    },
+    {
+      "epoch": 0.8246187865665453,
+      "grad_norm": 5.175635831622304,
+      "learning_rate": 1.570419681494071e-05,
+      "loss": 0.8836,
+      "step": 17900
+    },
+    {
+      "epoch": 0.8246648546551804,
+      "grad_norm": 0.9070816198607347,
+      "learning_rate": 1.569617062483755e-05,
+      "loss": 1.6478,
+      "step": 17901
+    },
+    {
+      "epoch": 0.8247109227438154,
+      "grad_norm": 1.3539244951479024,
+      "learning_rate": 1.5688146311614628e-05,
+      "loss": 1.3292,
+      "step": 17902
+    },
+    {
+      "epoch": 0.8247569908324504,
+      "grad_norm": 0.9730365991783054,
+      "learning_rate": 1.5680123875450535e-05,
+      "loss": 1.2127,
+      "step": 17903
+    },
+    {
+      "epoch": 0.8248030589210854,
+      "grad_norm": 2.1647954221490235,
+      "learning_rate": 1.56721033165239e-05,
+      "loss": 1.3377,
+      "step": 17904
+    },
+    {
+      "epoch": 0.8248491270097204,
+      "grad_norm": 1.1408175686140751,
+      "learning_rate": 1.566408463501331e-05,
+      "loss": 1.2133,
+      "step": 17905
+    },
+    {
+      "epoch": 0.8248951950983554,
+      "grad_norm": 1.0290327453166614,
+      "learning_rate": 1.5656067831097276e-05,
+      "loss": 1.3282,
+      "step": 17906
+    },
+    {
+      "epoch": 0.8249412631869903,
+      "grad_norm": 1.4221523014284527,
+      "learning_rate": 1.564805290495426e-05,
+      "loss": 1.0861,
+      "step": 17907
+    },
+    {
+      "epoch": 0.8249873312756254,
+      "grad_norm": 0.9449194471695906,
+      "learning_rate": 1.5640039856762712e-05,
+      "loss": 1.3483,
+      "step": 17908
+    },
+    {
+      "epoch": 0.8250333993642603,
+      "grad_norm": 0.9493331651688892,
+      "learning_rate": 1.5632028686701027e-05,
+      "loss": 1.1916,
+      "step": 17909
+    },
+    {
+      "epoch": 0.8250794674528954,
+      "grad_norm": 0.8245024111380532,
+      "learning_rate": 1.5624019394947597e-05,
+      "loss": 1.063,
+      "step": 17910
+    },
+    {
+      "epoch": 0.8251255355415303,
+      "grad_norm": 0.893183269634169,
+      "learning_rate": 1.5616011981680667e-05,
+      "loss": 1.1196,
+      "step": 17911
+    },
+    {
+      "epoch": 0.8251716036301654,
+      "grad_norm": 0.9975491441992518,
+      "learning_rate": 1.560800644707855e-05,
+      "loss": 0.9257,
+      "step": 17912
+    },
+    {
+      "epoch": 0.8252176717188003,
+      "grad_norm": 0.9782946230527856,
+      "learning_rate": 1.560000279131948e-05,
+      "loss": 1.4211,
+      "step": 17913
+    },
+    {
+      "epoch": 0.8252637398074354,
+      "grad_norm": 1.0136440385759724,
+      "learning_rate": 1.5592001014581627e-05,
+      "loss": 1.3121,
+      "step": 17914
+    },
+    {
+      "epoch": 0.8253098078960704,
+      "grad_norm": 0.848919171179573,
+      "learning_rate": 1.5584001117043122e-05,
+      "loss": 1.4576,
+      "step": 17915
+    },
+    {
+      "epoch": 0.8253558759847054,
+      "grad_norm": 1.5796291289072975,
+      "learning_rate": 1.5576003098882107e-05,
+      "loss": 1.2557,
+      "step": 17916
+    },
+    {
+      "epoch": 0.8254019440733404,
+      "grad_norm": 1.2244400838064395,
+      "learning_rate": 1.5568006960276638e-05,
+      "loss": 1.0379,
+      "step": 17917
+    },
+    {
+      "epoch": 0.8254480121619754,
+      "grad_norm": 1.1019098608369178,
+      "learning_rate": 1.5560012701404714e-05,
+      "loss": 1.2286,
+      "step": 17918
+    },
+    {
+      "epoch": 0.8254940802506104,
+      "grad_norm": 0.862897673909533,
+      "learning_rate": 1.5552020322444316e-05,
+      "loss": 1.4047,
+      "step": 17919
+    },
+    {
+      "epoch": 0.8255401483392454,
+      "grad_norm": 0.8968195184384332,
+      "learning_rate": 1.5544029823573413e-05,
+      "loss": 1.0801,
+      "step": 17920
+    },
+    {
+      "epoch": 0.8255862164278804,
+      "grad_norm": 0.8165068541210179,
+      "learning_rate": 1.5536041204969854e-05,
+      "loss": 1.3078,
+      "step": 17921
+    },
+    {
+      "epoch": 0.8256322845165154,
+      "grad_norm": 1.3590475432266702,
+      "learning_rate": 1.5528054466811504e-05,
+      "loss": 0.917,
+      "step": 17922
+    },
+    {
+      "epoch": 0.8256783526051504,
+      "grad_norm": 1.4335977602406913,
+      "learning_rate": 1.552006960927619e-05,
+      "loss": 1.1123,
+      "step": 17923
+    },
+    {
+      "epoch": 0.8257244206937854,
+      "grad_norm": 0.8562808583906294,
+      "learning_rate": 1.55120866325417e-05,
+      "loss": 1.1164,
+      "step": 17924
+    },
+    {
+      "epoch": 0.8257704887824204,
+      "grad_norm": 0.8074658333662132,
+      "learning_rate": 1.5504105536785707e-05,
+      "loss": 1.4885,
+      "step": 17925
+    },
+    {
+      "epoch": 0.8258165568710554,
+      "grad_norm": 1.0997539361520412,
+      "learning_rate": 1.5496126322185922e-05,
+      "loss": 1.2031,
+      "step": 17926
+    },
+    {
+      "epoch": 0.8258626249596904,
+      "grad_norm": 0.9123341101819646,
+      "learning_rate": 1.548814898891999e-05,
+      "loss": 1.304,
+      "step": 17927
+    },
+    {
+      "epoch": 0.8259086930483254,
+      "grad_norm": 2.0250071159331133,
+      "learning_rate": 1.5480173537165545e-05,
+      "loss": 1.2683,
+      "step": 17928
+    },
+    {
+      "epoch": 0.8259547611369604,
+      "grad_norm": 0.78489478192871,
+      "learning_rate": 1.547219996710012e-05,
+      "loss": 1.106,
+      "step": 17929
+    },
+    {
+      "epoch": 0.8260008292255955,
+      "grad_norm": 0.886731142869337,
+      "learning_rate": 1.5464228278901173e-05,
+      "loss": 1.1203,
+      "step": 17930
+    },
+    {
+      "epoch": 0.8260468973142304,
+      "grad_norm": 0.6684868838432828,
+      "learning_rate": 1.5456258472746288e-05,
+      "loss": 1.2981,
+      "step": 17931
+    },
+    {
+      "epoch": 0.8260929654028655,
+      "grad_norm": 0.8450655300529305,
+      "learning_rate": 1.5448290548812814e-05,
+      "loss": 1.2603,
+      "step": 17932
+    },
+    {
+      "epoch": 0.8261390334915004,
+      "grad_norm": 0.9771013581959651,
+      "learning_rate": 1.544032450727818e-05,
+      "loss": 1.0307,
+      "step": 17933
+    },
+    {
+      "epoch": 0.8261851015801355,
+      "grad_norm": 0.7738574459734749,
+      "learning_rate": 1.543236034831974e-05,
+      "loss": 1.1648,
+      "step": 17934
+    },
+    {
+      "epoch": 0.8262311696687704,
+      "grad_norm": 1.4254288931343257,
+      "learning_rate": 1.542439807211481e-05,
+      "loss": 0.8975,
+      "step": 17935
+    },
+    {
+      "epoch": 0.8262772377574055,
+      "grad_norm": 0.9406221493452261,
+      "learning_rate": 1.5416437678840622e-05,
+      "loss": 1.0827,
+      "step": 17936
+    },
+    {
+      "epoch": 0.8263233058460404,
+      "grad_norm": 1.1346842375671586,
+      "learning_rate": 1.5408479168674417e-05,
+      "loss": 1.2989,
+      "step": 17937
+    },
+    {
+      "epoch": 0.8263693739346755,
+      "grad_norm": 1.025704370724635,
+      "learning_rate": 1.5400522541793406e-05,
+      "loss": 1.1879,
+      "step": 17938
+    },
+    {
+      "epoch": 0.8264154420233104,
+      "grad_norm": 0.8336181405500759,
+      "learning_rate": 1.5392567798374667e-05,
+      "loss": 1.2452,
+      "step": 17939
+    },
+    {
+      "epoch": 0.8264615101119455,
+      "grad_norm": 0.7799723579785718,
+      "learning_rate": 1.5384614938595376e-05,
+      "loss": 1.0765,
+      "step": 17940
+    },
+    {
+      "epoch": 0.8265075782005804,
+      "grad_norm": 0.8787993497085248,
+      "learning_rate": 1.5376663962632497e-05,
+      "loss": 1.2967,
+      "step": 17941
+    },
+    {
+      "epoch": 0.8265536462892155,
+      "grad_norm": 1.5615897378991699,
+      "learning_rate": 1.5368714870663136e-05,
+      "loss": 1.4774,
+      "step": 17942
+    },
+    {
+      "epoch": 0.8265997143778505,
+      "grad_norm": 0.8366938219414344,
+      "learning_rate": 1.5360767662864207e-05,
+      "loss": 1.2751,
+      "step": 17943
+    },
+    {
+      "epoch": 0.8266457824664855,
+      "grad_norm": 0.8372873643362734,
+      "learning_rate": 1.5352822339412665e-05,
+      "loss": 1.2916,
+      "step": 17944
+    },
+    {
+      "epoch": 0.8266918505551205,
+      "grad_norm": 0.8818525909690282,
+      "learning_rate": 1.5344878900485394e-05,
+      "loss": 1.1006,
+      "step": 17945
+    },
+    {
+      "epoch": 0.8267379186437555,
+      "grad_norm": 1.0335933298417574,
+      "learning_rate": 1.5336937346259263e-05,
+      "loss": 1.0491,
+      "step": 17946
+    },
+    {
+      "epoch": 0.8267839867323905,
+      "grad_norm": 0.9001789500424257,
+      "learning_rate": 1.5328997676911062e-05,
+      "loss": 1.2073,
+      "step": 17947
+    },
+    {
+      "epoch": 0.8268300548210255,
+      "grad_norm": 0.9698764677349244,
+      "learning_rate": 1.5321059892617495e-05,
+      "loss": 1.1726,
+      "step": 17948
+    },
+    {
+      "epoch": 0.8268761229096605,
+      "grad_norm": 1.2736515383002545,
+      "learning_rate": 1.531312399355539e-05,
+      "loss": 1.1775,
+      "step": 17949
+    },
+    {
+      "epoch": 0.8269221909982954,
+      "grad_norm": 0.8589373444821284,
+      "learning_rate": 1.5305189979901334e-05,
+      "loss": 1.4379,
+      "step": 17950
+    },
+    {
+      "epoch": 0.8269682590869305,
+      "grad_norm": 0.9725590061882038,
+      "learning_rate": 1.5297257851832035e-05,
+      "loss": 1.4625,
+      "step": 17951
+    },
+    {
+      "epoch": 0.8270143271755654,
+      "grad_norm": 1.248126697223945,
+      "learning_rate": 1.5289327609524018e-05,
+      "loss": 1.3332,
+      "step": 17952
+    },
+    {
+      "epoch": 0.8270603952642005,
+      "grad_norm": 0.8345895912521214,
+      "learning_rate": 1.5281399253153916e-05,
+      "loss": 1.3061,
+      "step": 17953
+    },
+    {
+      "epoch": 0.8271064633528354,
+      "grad_norm": 1.2461978974330312,
+      "learning_rate": 1.527347278289817e-05,
+      "loss": 1.4269,
+      "step": 17954
+    },
+    {
+      "epoch": 0.8271525314414705,
+      "grad_norm": 1.1028183902434467,
+      "learning_rate": 1.52655481989333e-05,
+      "loss": 1.5342,
+      "step": 17955
+    },
+    {
+      "epoch": 0.8271985995301055,
+      "grad_norm": 0.9780845251560951,
+      "learning_rate": 1.5257625501435691e-05,
+      "loss": 1.323,
+      "step": 17956
+    },
+    {
+      "epoch": 0.8272446676187405,
+      "grad_norm": 0.8811321213355913,
+      "learning_rate": 1.5249704690581745e-05,
+      "loss": 1.1151,
+      "step": 17957
+    },
+    {
+      "epoch": 0.8272907357073755,
+      "grad_norm": 0.7961671897279622,
+      "learning_rate": 1.5241785766547834e-05,
+      "loss": 1.3222,
+      "step": 17958
+    },
+    {
+      "epoch": 0.8273368037960105,
+      "grad_norm": 0.7653985091038464,
+      "learning_rate": 1.5233868729510193e-05,
+      "loss": 1.2542,
+      "step": 17959
+    },
+    {
+      "epoch": 0.8273828718846455,
+      "grad_norm": 0.9002163861183963,
+      "learning_rate": 1.5225953579645159e-05,
+      "loss": 1.5541,
+      "step": 17960
+    },
+    {
+      "epoch": 0.8274289399732805,
+      "grad_norm": 0.7226445642015564,
+      "learning_rate": 1.521804031712889e-05,
+      "loss": 1.618,
+      "step": 17961
+    },
+    {
+      "epoch": 0.8274750080619155,
+      "grad_norm": 1.0802786127700847,
+      "learning_rate": 1.5210128942137614e-05,
+      "loss": 1.1146,
+      "step": 17962
+    },
+    {
+      "epoch": 0.8275210761505505,
+      "grad_norm": 1.2292435001414295,
+      "learning_rate": 1.5202219454847377e-05,
+      "loss": 1.4234,
+      "step": 17963
+    },
+    {
+      "epoch": 0.8275671442391855,
+      "grad_norm": 0.9973770215148059,
+      "learning_rate": 1.5194311855434384e-05,
+      "loss": 1.2299,
+      "step": 17964
+    },
+    {
+      "epoch": 0.8276132123278205,
+      "grad_norm": 0.8251087462108531,
+      "learning_rate": 1.5186406144074628e-05,
+      "loss": 1.3543,
+      "step": 17965
+    },
+    {
+      "epoch": 0.8276592804164555,
+      "grad_norm": 0.7730819661095101,
+      "learning_rate": 1.5178502320944076e-05,
+      "loss": 1.2919,
+      "step": 17966
+    },
+    {
+      "epoch": 0.8277053485050905,
+      "grad_norm": 0.866672670823875,
+      "learning_rate": 1.5170600386218747e-05,
+      "loss": 1.2094,
+      "step": 17967
+    },
+    {
+      "epoch": 0.8277514165937255,
+      "grad_norm": 1.109949624008971,
+      "learning_rate": 1.5162700340074543e-05,
+      "loss": 1.3374,
+      "step": 17968
+    },
+    {
+      "epoch": 0.8277974846823605,
+      "grad_norm": 2.736397419879904,
+      "learning_rate": 1.5154802182687378e-05,
+      "loss": 1.3965,
+      "step": 17969
+    },
+    {
+      "epoch": 0.8278435527709955,
+      "grad_norm": 1.038501073527632,
+      "learning_rate": 1.5146905914233012e-05,
+      "loss": 1.0649,
+      "step": 17970
+    },
+    {
+      "epoch": 0.8278896208596306,
+      "grad_norm": 0.8495962249714939,
+      "learning_rate": 1.5139011534887349e-05,
+      "loss": 1.0769,
+      "step": 17971
+    },
+    {
+      "epoch": 0.8279356889482655,
+      "grad_norm": 0.9025098978243615,
+      "learning_rate": 1.513111904482606e-05,
+      "loss": 1.2289,
+      "step": 17972
+    },
+    {
+      "epoch": 0.8279817570369006,
+      "grad_norm": 0.7847344281092363,
+      "learning_rate": 1.5123228444224902e-05,
+      "loss": 1.3487,
+      "step": 17973
+    },
+    {
+      "epoch": 0.8280278251255355,
+      "grad_norm": 0.7445866807642179,
+      "learning_rate": 1.5115339733259515e-05,
+      "loss": 1.559,
+      "step": 17974
+    },
+    {
+      "epoch": 0.8280738932141706,
+      "grad_norm": 0.6842652231085586,
+      "learning_rate": 1.5107452912105546e-05,
+      "loss": 1.4212,
+      "step": 17975
+    },
+    {
+      "epoch": 0.8281199613028055,
+      "grad_norm": 0.7477621379146713,
+      "learning_rate": 1.50995679809386e-05,
+      "loss": 1.4536,
+      "step": 17976
+    },
+    {
+      "epoch": 0.8281660293914406,
+      "grad_norm": 1.636488279979364,
+      "learning_rate": 1.509168493993417e-05,
+      "loss": 1.0592,
+      "step": 17977
+    },
+    {
+      "epoch": 0.8282120974800755,
+      "grad_norm": 1.0778441658058255,
+      "learning_rate": 1.5083803789267803e-05,
+      "loss": 1.5692,
+      "step": 17978
+    },
+    {
+      "epoch": 0.8282581655687106,
+      "grad_norm": 0.7702771184613406,
+      "learning_rate": 1.5075924529114926e-05,
+      "loss": 1.1376,
+      "step": 17979
+    },
+    {
+      "epoch": 0.8283042336573455,
+      "grad_norm": 1.8930375861205164,
+      "learning_rate": 1.506804715965101e-05,
+      "loss": 1.3993,
+      "step": 17980
+    },
+    {
+      "epoch": 0.8283503017459806,
+      "grad_norm": 1.6624637157502338,
+      "learning_rate": 1.5060171681051372e-05,
+      "loss": 0.9887,
+      "step": 17981
+    },
+    {
+      "epoch": 0.8283963698346155,
+      "grad_norm": 1.5873793072630804,
+      "learning_rate": 1.5052298093491368e-05,
+      "loss": 1.5019,
+      "step": 17982
+    },
+    {
+      "epoch": 0.8284424379232506,
+      "grad_norm": 0.7638320503786881,
+      "learning_rate": 1.5044426397146328e-05,
+      "loss": 1.1665,
+      "step": 17983
+    },
+    {
+      "epoch": 0.8284885060118856,
+      "grad_norm": 1.3218193298562024,
+      "learning_rate": 1.503655659219143e-05,
+      "loss": 1.1588,
+      "step": 17984
+    },
+    {
+      "epoch": 0.8285345741005206,
+      "grad_norm": 1.4447219328487695,
+      "learning_rate": 1.5028688678801927e-05,
+      "loss": 1.4742,
+      "step": 17985
+    },
+    {
+      "epoch": 0.8285806421891556,
+      "grad_norm": 1.2346108723063425,
+      "learning_rate": 1.5020822657152967e-05,
+      "loss": 1.1797,
+      "step": 17986
+    },
+    {
+      "epoch": 0.8286267102777906,
+      "grad_norm": 0.8858272565188023,
+      "learning_rate": 1.5012958527419707e-05,
+      "loss": 1.201,
+      "step": 17987
+    },
+    {
+      "epoch": 0.8286727783664256,
+      "grad_norm": 0.8074363619163817,
+      "learning_rate": 1.5005096289777187e-05,
+      "loss": 1.2001,
+      "step": 17988
+    },
+    {
+      "epoch": 0.8287188464550606,
+      "grad_norm": 1.166015002660082,
+      "learning_rate": 1.4997235944400457e-05,
+      "loss": 1.0246,
+      "step": 17989
+    },
+    {
+      "epoch": 0.8287649145436956,
+      "grad_norm": 1.2592546468245582,
+      "learning_rate": 1.498937749146453e-05,
+      "loss": 1.4307,
+      "step": 17990
+    },
+    {
+      "epoch": 0.8288109826323306,
+      "grad_norm": 0.9711564948831496,
+      "learning_rate": 1.4981520931144366e-05,
+      "loss": 1.251,
+      "step": 17991
+    },
+    {
+      "epoch": 0.8288570507209656,
+      "grad_norm": 0.824155281419541,
+      "learning_rate": 1.4973666263614838e-05,
+      "loss": 1.2269,
+      "step": 17992
+    },
+    {
+      "epoch": 0.8289031188096005,
+      "grad_norm": 0.7749069440950845,
+      "learning_rate": 1.496581348905085e-05,
+      "loss": 1.1841,
+      "step": 17993
+    },
+    {
+      "epoch": 0.8289491868982356,
+      "grad_norm": 1.4892629420874723,
+      "learning_rate": 1.495796260762724e-05,
+      "loss": 1.3387,
+      "step": 17994
+    },
+    {
+      "epoch": 0.8289952549868705,
+      "grad_norm": 1.957721092104052,
+      "learning_rate": 1.4950113619518757e-05,
+      "loss": 1.1942,
+      "step": 17995
+    },
+    {
+      "epoch": 0.8290413230755056,
+      "grad_norm": 0.9963347560317667,
+      "learning_rate": 1.494226652490016e-05,
+      "loss": 1.2407,
+      "step": 17996
+    },
+    {
+      "epoch": 0.8290873911641407,
+      "grad_norm": 1.1457287619845709,
+      "learning_rate": 1.4934421323946158e-05,
+      "loss": 1.1374,
+      "step": 17997
+    },
+    {
+      "epoch": 0.8291334592527756,
+      "grad_norm": 0.7649192294491616,
+      "learning_rate": 1.4926578016831427e-05,
+      "loss": 1.2832,
+      "step": 17998
+    },
+    {
+      "epoch": 0.8291795273414106,
+      "grad_norm": 0.9033202061726435,
+      "learning_rate": 1.4918736603730555e-05,
+      "loss": 1.2788,
+      "step": 17999
+    },
+    {
+      "epoch": 0.8292255954300456,
+      "grad_norm": 0.9106490218157504,
+      "learning_rate": 1.491089708481812e-05,
+      "loss": 1.1727,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8292716635186806,
+      "grad_norm": 0.9555409397584473,
+      "learning_rate": 1.4903059460268665e-05,
+      "loss": 1.0476,
+      "step": 18001
+    },
+    {
+      "epoch": 0.8293177316073156,
+      "grad_norm": 2.1454669698584543,
+      "learning_rate": 1.4895223730256713e-05,
+      "loss": 1.4358,
+      "step": 18002
+    },
+    {
+      "epoch": 0.8293637996959506,
+      "grad_norm": 0.989695050062823,
+      "learning_rate": 1.4887389894956649e-05,
+      "loss": 1.6907,
+      "step": 18003
+    },
+    {
+      "epoch": 0.8294098677845856,
+      "grad_norm": 1.3557178739432267,
+      "learning_rate": 1.4879557954542922e-05,
+      "loss": 1.3656,
+      "step": 18004
+    },
+    {
+      "epoch": 0.8294559358732206,
+      "grad_norm": 0.775073615927543,
+      "learning_rate": 1.4871727909189903e-05,
+      "loss": 1.3561,
+      "step": 18005
+    },
+    {
+      "epoch": 0.8295020039618556,
+      "grad_norm": 0.9721425671791537,
+      "learning_rate": 1.4863899759071875e-05,
+      "loss": 1.2019,
+      "step": 18006
+    },
+    {
+      "epoch": 0.8295480720504906,
+      "grad_norm": 1.3446746225499748,
+      "learning_rate": 1.4856073504363144e-05,
+      "loss": 1.0276,
+      "step": 18007
+    },
+    {
+      "epoch": 0.8295941401391256,
+      "grad_norm": 0.9818281933661924,
+      "learning_rate": 1.4848249145237946e-05,
+      "loss": 1.335,
+      "step": 18008
+    },
+    {
+      "epoch": 0.8296402082277606,
+      "grad_norm": 0.8345841184335158,
+      "learning_rate": 1.48404266818705e-05,
+      "loss": 1.2841,
+      "step": 18009
+    },
+    {
+      "epoch": 0.8296862763163957,
+      "grad_norm": 0.7090236651361799,
+      "learning_rate": 1.483260611443491e-05,
+      "loss": 1.1937,
+      "step": 18010
+    },
+    {
+      "epoch": 0.8297323444050306,
+      "grad_norm": 0.8976911968369006,
+      "learning_rate": 1.4824787443105315e-05,
+      "loss": 1.4773,
+      "step": 18011
+    },
+    {
+      "epoch": 0.8297784124936657,
+      "grad_norm": 1.1920239412323452,
+      "learning_rate": 1.4816970668055807e-05,
+      "loss": 1.3352,
+      "step": 18012
+    },
+    {
+      "epoch": 0.8298244805823006,
+      "grad_norm": 1.1017259958575416,
+      "learning_rate": 1.4809155789460371e-05,
+      "loss": 1.0978,
+      "step": 18013
+    },
+    {
+      "epoch": 0.8298705486709357,
+      "grad_norm": 0.8970423868338063,
+      "learning_rate": 1.4801342807493002e-05,
+      "loss": 1.045,
+      "step": 18014
+    },
+    {
+      "epoch": 0.8299166167595706,
+      "grad_norm": 0.6249601054319908,
+      "learning_rate": 1.4793531722327657e-05,
+      "loss": 1.3209,
+      "step": 18015
+    },
+    {
+      "epoch": 0.8299626848482057,
+      "grad_norm": 0.8606343056214795,
+      "learning_rate": 1.4785722534138258e-05,
+      "loss": 1.3471,
+      "step": 18016
+    },
+    {
+      "epoch": 0.8300087529368406,
+      "grad_norm": 0.8321226837911175,
+      "learning_rate": 1.4777915243098606e-05,
+      "loss": 1.4081,
+      "step": 18017
+    },
+    {
+      "epoch": 0.8300548210254757,
+      "grad_norm": 0.8720615551301305,
+      "learning_rate": 1.4770109849382541e-05,
+      "loss": 1.1791,
+      "step": 18018
+    },
+    {
+      "epoch": 0.8301008891141106,
+      "grad_norm": 0.8528945465221505,
+      "learning_rate": 1.4762306353163857e-05,
+      "loss": 1.2849,
+      "step": 18019
+    },
+    {
+      "epoch": 0.8301469572027457,
+      "grad_norm": 3.739521574648207,
+      "learning_rate": 1.4754504754616283e-05,
+      "loss": 1.0945,
+      "step": 18020
+    },
+    {
+      "epoch": 0.8301930252913806,
+      "grad_norm": 1.1271258380373128,
+      "learning_rate": 1.4746705053913479e-05,
+      "loss": 1.3208,
+      "step": 18021
+    },
+    {
+      "epoch": 0.8302390933800157,
+      "grad_norm": 1.1969700205700329,
+      "learning_rate": 1.4738907251229106e-05,
+      "loss": 1.1244,
+      "step": 18022
+    },
+    {
+      "epoch": 0.8302851614686506,
+      "grad_norm": 1.0444688268306646,
+      "learning_rate": 1.4731111346736793e-05,
+      "loss": 0.9814,
+      "step": 18023
+    },
+    {
+      "epoch": 0.8303312295572857,
+      "grad_norm": 0.7293822325640481,
+      "learning_rate": 1.4723317340610054e-05,
+      "loss": 1.2291,
+      "step": 18024
+    },
+    {
+      "epoch": 0.8303772976459207,
+      "grad_norm": 0.6718576881258257,
+      "learning_rate": 1.4715525233022442e-05,
+      "loss": 1.4016,
+      "step": 18025
+    },
+    {
+      "epoch": 0.8304233657345557,
+      "grad_norm": 0.9394381116389844,
+      "learning_rate": 1.470773502414743e-05,
+      "loss": 1.1227,
+      "step": 18026
+    },
+    {
+      "epoch": 0.8304694338231907,
+      "grad_norm": 0.7281848600583825,
+      "learning_rate": 1.4699946714158474e-05,
+      "loss": 1.3356,
+      "step": 18027
+    },
+    {
+      "epoch": 0.8305155019118257,
+      "grad_norm": 1.3231287011036472,
+      "learning_rate": 1.4692160303228919e-05,
+      "loss": 1.1706,
+      "step": 18028
+    },
+    {
+      "epoch": 0.8305615700004607,
+      "grad_norm": 0.9837826448333834,
+      "learning_rate": 1.4684375791532146e-05,
+      "loss": 1.2266,
+      "step": 18029
+    },
+    {
+      "epoch": 0.8306076380890957,
+      "grad_norm": 1.1400426994444846,
+      "learning_rate": 1.4676593179241493e-05,
+      "loss": 0.9125,
+      "step": 18030
+    },
+    {
+      "epoch": 0.8306537061777307,
+      "grad_norm": 1.3574830882496844,
+      "learning_rate": 1.4668812466530157e-05,
+      "loss": 1.3609,
+      "step": 18031
+    },
+    {
+      "epoch": 0.8306997742663657,
+      "grad_norm": 1.3796032735513526,
+      "learning_rate": 1.466103365357141e-05,
+      "loss": 1.0079,
+      "step": 18032
+    },
+    {
+      "epoch": 0.8307458423550007,
+      "grad_norm": 0.9255355121595632,
+      "learning_rate": 1.4653256740538424e-05,
+      "loss": 0.8587,
+      "step": 18033
+    },
+    {
+      "epoch": 0.8307919104436357,
+      "grad_norm": 3.221570340675116,
+      "learning_rate": 1.464548172760436e-05,
+      "loss": 1.3676,
+      "step": 18034
+    },
+    {
+      "epoch": 0.8308379785322707,
+      "grad_norm": 0.7125019525262725,
+      "learning_rate": 1.463770861494227e-05,
+      "loss": 1.2605,
+      "step": 18035
+    },
+    {
+      "epoch": 0.8308840466209056,
+      "grad_norm": 0.9816617006629497,
+      "learning_rate": 1.4629937402725225e-05,
+      "loss": 1.0705,
+      "step": 18036
+    },
+    {
+      "epoch": 0.8309301147095407,
+      "grad_norm": 0.8587862642149322,
+      "learning_rate": 1.4622168091126254e-05,
+      "loss": 1.4069,
+      "step": 18037
+    },
+    {
+      "epoch": 0.8309761827981758,
+      "grad_norm": 0.776189353346371,
+      "learning_rate": 1.461440068031834e-05,
+      "loss": 1.3874,
+      "step": 18038
+    },
+    {
+      "epoch": 0.8310222508868107,
+      "grad_norm": 0.7333200560569536,
+      "learning_rate": 1.4606635170474392e-05,
+      "loss": 1.2583,
+      "step": 18039
+    },
+    {
+      "epoch": 0.8310683189754458,
+      "grad_norm": 0.8438796072372584,
+      "learning_rate": 1.4598871561767247e-05,
+      "loss": 1.2739,
+      "step": 18040
+    },
+    {
+      "epoch": 0.8311143870640807,
+      "grad_norm": 0.8004957440793681,
+      "learning_rate": 1.4591109854369845e-05,
+      "loss": 1.3196,
+      "step": 18041
+    },
+    {
+      "epoch": 0.8311604551527157,
+      "grad_norm": 6.986061318509331,
+      "learning_rate": 1.4583350048454902e-05,
+      "loss": 1.1616,
+      "step": 18042
+    },
+    {
+      "epoch": 0.8312065232413507,
+      "grad_norm": 1.0797851464809618,
+      "learning_rate": 1.4575592144195216e-05,
+      "loss": 1.2651,
+      "step": 18043
+    },
+    {
+      "epoch": 0.8312525913299857,
+      "grad_norm": 0.7578266483694008,
+      "learning_rate": 1.4567836141763503e-05,
+      "loss": 1.1585,
+      "step": 18044
+    },
+    {
+      "epoch": 0.8312986594186207,
+      "grad_norm": 1.1490631697412468,
+      "learning_rate": 1.4560082041332457e-05,
+      "loss": 1.3901,
+      "step": 18045
+    },
+    {
+      "epoch": 0.8313447275072557,
+      "grad_norm": 0.7904451362426925,
+      "learning_rate": 1.4552329843074663e-05,
+      "loss": 1.2968,
+      "step": 18046
+    },
+    {
+      "epoch": 0.8313907955958907,
+      "grad_norm": 0.9981646785080655,
+      "learning_rate": 1.454457954716274e-05,
+      "loss": 0.9499,
+      "step": 18047
+    },
+    {
+      "epoch": 0.8314368636845257,
+      "grad_norm": 0.8800506412314861,
+      "learning_rate": 1.4536831153769237e-05,
+      "loss": 1.2242,
+      "step": 18048
+    },
+    {
+      "epoch": 0.8314829317731607,
+      "grad_norm": 0.7551481734457741,
+      "learning_rate": 1.4529084663066638e-05,
+      "loss": 0.805,
+      "step": 18049
+    },
+    {
+      "epoch": 0.8315289998617957,
+      "grad_norm": 1.7186068458500046,
+      "learning_rate": 1.4521340075227441e-05,
+      "loss": 1.3002,
+      "step": 18050
+    },
+    {
+      "epoch": 0.8315750679504308,
+      "grad_norm": 1.1226785977638507,
+      "learning_rate": 1.4513597390423994e-05,
+      "loss": 1.1545,
+      "step": 18051
+    },
+    {
+      "epoch": 0.8316211360390657,
+      "grad_norm": 0.8819259119719715,
+      "learning_rate": 1.4505856608828772e-05,
+      "loss": 1.1075,
+      "step": 18052
+    },
+    {
+      "epoch": 0.8316672041277008,
+      "grad_norm": 1.531992847184891,
+      "learning_rate": 1.4498117730614036e-05,
+      "loss": 1.1863,
+      "step": 18053
+    },
+    {
+      "epoch": 0.8317132722163357,
+      "grad_norm": 0.8788884952102645,
+      "learning_rate": 1.4490380755952126e-05,
+      "loss": 0.9529,
+      "step": 18054
+    },
+    {
+      "epoch": 0.8317593403049708,
+      "grad_norm": 1.170785210285662,
+      "learning_rate": 1.4482645685015228e-05,
+      "loss": 1.2752,
+      "step": 18055
+    },
+    {
+      "epoch": 0.8318054083936057,
+      "grad_norm": 0.7160294922311584,
+      "learning_rate": 1.4474912517975637e-05,
+      "loss": 1.2269,
+      "step": 18056
+    },
+    {
+      "epoch": 0.8318514764822408,
+      "grad_norm": 1.025941309092586,
+      "learning_rate": 1.4467181255005468e-05,
+      "loss": 1.5251,
+      "step": 18057
+    },
+    {
+      "epoch": 0.8318975445708757,
+      "grad_norm": 0.9618569896499877,
+      "learning_rate": 1.445945189627681e-05,
+      "loss": 1.1296,
+      "step": 18058
+    },
+    {
+      "epoch": 0.8319436126595108,
+      "grad_norm": 1.2162339398241435,
+      "learning_rate": 1.4451724441961834e-05,
+      "loss": 1.4786,
+      "step": 18059
+    },
+    {
+      "epoch": 0.8319896807481457,
+      "grad_norm": 1.2924352235701866,
+      "learning_rate": 1.4443998892232513e-05,
+      "loss": 1.1104,
+      "step": 18060
+    },
+    {
+      "epoch": 0.8320357488367808,
+      "grad_norm": 0.8534289804235482,
+      "learning_rate": 1.4436275247260867e-05,
+      "loss": 1.5275,
+      "step": 18061
+    },
+    {
+      "epoch": 0.8320818169254157,
+      "grad_norm": 1.0204318745160461,
+      "learning_rate": 1.4428553507218812e-05,
+      "loss": 0.9163,
+      "step": 18062
+    },
+    {
+      "epoch": 0.8321278850140508,
+      "grad_norm": 0.7555853274052379,
+      "learning_rate": 1.4420833672278322e-05,
+      "loss": 1.5477,
+      "step": 18063
+    },
+    {
+      "epoch": 0.8321739531026857,
+      "grad_norm": 1.1337409234678153,
+      "learning_rate": 1.4413115742611215e-05,
+      "loss": 1.1685,
+      "step": 18064
+    },
+    {
+      "epoch": 0.8322200211913208,
+      "grad_norm": 0.8611449051037915,
+      "learning_rate": 1.4405399718389368e-05,
+      "loss": 0.9614,
+      "step": 18065
+    },
+    {
+      "epoch": 0.8322660892799558,
+      "grad_norm": 2.3977058065489603,
+      "learning_rate": 1.4397685599784494e-05,
+      "loss": 1.1274,
+      "step": 18066
+    },
+    {
+      "epoch": 0.8323121573685908,
+      "grad_norm": 0.8894550378135452,
+      "learning_rate": 1.4389973386968392e-05,
+      "loss": 1.404,
+      "step": 18067
+    },
+    {
+      "epoch": 0.8323582254572258,
+      "grad_norm": 0.6720178019459406,
+      "learning_rate": 1.4382263080112745e-05,
+      "loss": 1.2746,
+      "step": 18068
+    },
+    {
+      "epoch": 0.8324042935458608,
+      "grad_norm": 0.9887983203882952,
+      "learning_rate": 1.4374554679389174e-05,
+      "loss": 1.1037,
+      "step": 18069
+    },
+    {
+      "epoch": 0.8324503616344958,
+      "grad_norm": 4.318111173776777,
+      "learning_rate": 1.4366848184969372e-05,
+      "loss": 1.0091,
+      "step": 18070
+    },
+    {
+      "epoch": 0.8324964297231308,
+      "grad_norm": 0.9722748073725862,
+      "learning_rate": 1.435914359702485e-05,
+      "loss": 1.4415,
+      "step": 18071
+    },
+    {
+      "epoch": 0.8325424978117658,
+      "grad_norm": 0.9939400543785604,
+      "learning_rate": 1.4351440915727166e-05,
+      "loss": 1.3185,
+      "step": 18072
+    },
+    {
+      "epoch": 0.8325885659004008,
+      "grad_norm": 0.8059543385503561,
+      "learning_rate": 1.4343740141247752e-05,
+      "loss": 1.3341,
+      "step": 18073
+    },
+    {
+      "epoch": 0.8326346339890358,
+      "grad_norm": 0.9236058077506858,
+      "learning_rate": 1.4336041273758138e-05,
+      "loss": 1.0798,
+      "step": 18074
+    },
+    {
+      "epoch": 0.8326807020776708,
+      "grad_norm": 0.6715682138719523,
+      "learning_rate": 1.4328344313429675e-05,
+      "loss": 1.6233,
+      "step": 18075
+    },
+    {
+      "epoch": 0.8327267701663058,
+      "grad_norm": 0.7851112145568564,
+      "learning_rate": 1.4320649260433728e-05,
+      "loss": 1.1959,
+      "step": 18076
+    },
+    {
+      "epoch": 0.8327728382549408,
+      "grad_norm": 0.9096318682702399,
+      "learning_rate": 1.43129561149416e-05,
+      "loss": 1.0915,
+      "step": 18077
+    },
+    {
+      "epoch": 0.8328189063435758,
+      "grad_norm": 0.8096243601322061,
+      "learning_rate": 1.430526487712459e-05,
+      "loss": 1.3201,
+      "step": 18078
+    },
+    {
+      "epoch": 0.8328649744322109,
+      "grad_norm": 0.8146725031718991,
+      "learning_rate": 1.4297575547153941e-05,
+      "loss": 1.2551,
+      "step": 18079
+    },
+    {
+      "epoch": 0.8329110425208458,
+      "grad_norm": 0.8463539652170333,
+      "learning_rate": 1.4289888125200801e-05,
+      "loss": 1.4176,
+      "step": 18080
+    },
+    {
+      "epoch": 0.8329571106094809,
+      "grad_norm": 2.0017734130572418,
+      "learning_rate": 1.4282202611436346e-05,
+      "loss": 0.9371,
+      "step": 18081
+    },
+    {
+      "epoch": 0.8330031786981158,
+      "grad_norm": 0.8013541578825351,
+      "learning_rate": 1.4274519006031673e-05,
+      "loss": 1.109,
+      "step": 18082
+    },
+    {
+      "epoch": 0.8330492467867509,
+      "grad_norm": 1.0544899699108954,
+      "learning_rate": 1.4266837309157866e-05,
+      "loss": 1.2879,
+      "step": 18083
+    },
+    {
+      "epoch": 0.8330953148753858,
+      "grad_norm": 0.8505559885353272,
+      "learning_rate": 1.4259157520985911e-05,
+      "loss": 1.2801,
+      "step": 18084
+    },
+    {
+      "epoch": 0.8331413829640208,
+      "grad_norm": 0.6501186902562676,
+      "learning_rate": 1.4251479641686794e-05,
+      "loss": 1.4518,
+      "step": 18085
+    },
+    {
+      "epoch": 0.8331874510526558,
+      "grad_norm": 0.8536924636998147,
+      "learning_rate": 1.4243803671431478e-05,
+      "loss": 1.3617,
+      "step": 18086
+    },
+    {
+      "epoch": 0.8332335191412908,
+      "grad_norm": 0.8441789827748415,
+      "learning_rate": 1.4236129610390802e-05,
+      "loss": 1.1398,
+      "step": 18087
+    },
+    {
+      "epoch": 0.8332795872299258,
+      "grad_norm": 1.1469719308579784,
+      "learning_rate": 1.4228457458735656e-05,
+      "loss": 1.0059,
+      "step": 18088
+    },
+    {
+      "epoch": 0.8333256553185608,
+      "grad_norm": 1.047948538188556,
+      "learning_rate": 1.4220787216636844e-05,
+      "loss": 1.2812,
+      "step": 18089
+    },
+    {
+      "epoch": 0.8333717234071958,
+      "grad_norm": 1.0179451189753022,
+      "learning_rate": 1.4213118884265141e-05,
+      "loss": 1.3038,
+      "step": 18090
+    },
+    {
+      "epoch": 0.8334177914958308,
+      "grad_norm": 0.8931949548216318,
+      "learning_rate": 1.4205452461791224e-05,
+      "loss": 1.0923,
+      "step": 18091
+    },
+    {
+      "epoch": 0.8334638595844659,
+      "grad_norm": 0.7284481355741061,
+      "learning_rate": 1.4197787949385811e-05,
+      "loss": 1.15,
+      "step": 18092
+    },
+    {
+      "epoch": 0.8335099276731008,
+      "grad_norm": 1.225671282789625,
+      "learning_rate": 1.4190125347219552e-05,
+      "loss": 1.2392,
+      "step": 18093
+    },
+    {
+      "epoch": 0.8335559957617359,
+      "grad_norm": 0.9401321727194458,
+      "learning_rate": 1.418246465546299e-05,
+      "loss": 1.146,
+      "step": 18094
+    },
+    {
+      "epoch": 0.8336020638503708,
+      "grad_norm": 0.9835837146484102,
+      "learning_rate": 1.4174805874286712e-05,
+      "loss": 1.3515,
+      "step": 18095
+    },
+    {
+      "epoch": 0.8336481319390059,
+      "grad_norm": 0.6577020878985587,
+      "learning_rate": 1.4167149003861224e-05,
+      "loss": 1.4734,
+      "step": 18096
+    },
+    {
+      "epoch": 0.8336942000276408,
+      "grad_norm": 0.8743038615303028,
+      "learning_rate": 1.4159494044357014e-05,
+      "loss": 1.1259,
+      "step": 18097
+    },
+    {
+      "epoch": 0.8337402681162759,
+      "grad_norm": 1.5655566981527207,
+      "learning_rate": 1.4151840995944465e-05,
+      "loss": 1.2822,
+      "step": 18098
+    },
+    {
+      "epoch": 0.8337863362049108,
+      "grad_norm": 1.4978978691962335,
+      "learning_rate": 1.4144189858793967e-05,
+      "loss": 1.2928,
+      "step": 18099
+    },
+    {
+      "epoch": 0.8338324042935459,
+      "grad_norm": 1.8537279480204085,
+      "learning_rate": 1.413654063307588e-05,
+      "loss": 1.2341,
+      "step": 18100
+    },
+    {
+      "epoch": 0.8338784723821808,
+      "grad_norm": 0.9982537403913705,
+      "learning_rate": 1.4128893318960513e-05,
+      "loss": 0.8986,
+      "step": 18101
+    },
+    {
+      "epoch": 0.8339245404708159,
+      "grad_norm": 1.0939171767358213,
+      "learning_rate": 1.4121247916618074e-05,
+      "loss": 1.2191,
+      "step": 18102
+    },
+    {
+      "epoch": 0.8339706085594508,
+      "grad_norm": 0.9571272289455853,
+      "learning_rate": 1.4113604426218796e-05,
+      "loss": 1.3332,
+      "step": 18103
+    },
+    {
+      "epoch": 0.8340166766480859,
+      "grad_norm": 0.8440441045210234,
+      "learning_rate": 1.4105962847932885e-05,
+      "loss": 1.3343,
+      "step": 18104
+    },
+    {
+      "epoch": 0.8340627447367208,
+      "grad_norm": 0.8972424425888935,
+      "learning_rate": 1.4098323181930407e-05,
+      "loss": 1.1742,
+      "step": 18105
+    },
+    {
+      "epoch": 0.8341088128253559,
+      "grad_norm": 0.9787393974012192,
+      "learning_rate": 1.4090685428381468e-05,
+      "loss": 1.2401,
+      "step": 18106
+    },
+    {
+      "epoch": 0.8341548809139909,
+      "grad_norm": 1.218514100563163,
+      "learning_rate": 1.408304958745612e-05,
+      "loss": 1.3577,
+      "step": 18107
+    },
+    {
+      "epoch": 0.8342009490026259,
+      "grad_norm": 0.9035873149840068,
+      "learning_rate": 1.4075415659324376e-05,
+      "loss": 1.5335,
+      "step": 18108
+    },
+    {
+      "epoch": 0.8342470170912609,
+      "grad_norm": 0.9443347271341948,
+      "learning_rate": 1.4067783644156152e-05,
+      "loss": 1.1857,
+      "step": 18109
+    },
+    {
+      "epoch": 0.8342930851798959,
+      "grad_norm": 0.8995171276243318,
+      "learning_rate": 1.406015354212139e-05,
+      "loss": 0.7683,
+      "step": 18110
+    },
+    {
+      "epoch": 0.8343391532685309,
+      "grad_norm": 1.103608026392614,
+      "learning_rate": 1.4052525353389967e-05,
+      "loss": 0.9509,
+      "step": 18111
+    },
+    {
+      "epoch": 0.8343852213571659,
+      "grad_norm": 1.1048546323793584,
+      "learning_rate": 1.404489907813168e-05,
+      "loss": 1.2954,
+      "step": 18112
+    },
+    {
+      "epoch": 0.8344312894458009,
+      "grad_norm": 1.762635173209368,
+      "learning_rate": 1.4037274716516335e-05,
+      "loss": 1.3126,
+      "step": 18113
+    },
+    {
+      "epoch": 0.8344773575344359,
+      "grad_norm": 1.0442905861189462,
+      "learning_rate": 1.4029652268713678e-05,
+      "loss": 1.2612,
+      "step": 18114
+    },
+    {
+      "epoch": 0.8345234256230709,
+      "grad_norm": 0.6596778820666304,
+      "learning_rate": 1.4022031734893416e-05,
+      "loss": 1.4368,
+      "step": 18115
+    },
+    {
+      "epoch": 0.8345694937117059,
+      "grad_norm": 1.8694629285977182,
+      "learning_rate": 1.4014413115225188e-05,
+      "loss": 1.2299,
+      "step": 18116
+    },
+    {
+      "epoch": 0.8346155618003409,
+      "grad_norm": 0.9351954263923307,
+      "learning_rate": 1.4006796409878608e-05,
+      "loss": 1.2429,
+      "step": 18117
+    },
+    {
+      "epoch": 0.8346616298889759,
+      "grad_norm": 0.7666180489440422,
+      "learning_rate": 1.3999181619023261e-05,
+      "loss": 1.4664,
+      "step": 18118
+    },
+    {
+      "epoch": 0.8347076979776109,
+      "grad_norm": 1.00989014748853,
+      "learning_rate": 1.3991568742828698e-05,
+      "loss": 1.2782,
+      "step": 18119
+    },
+    {
+      "epoch": 0.834753766066246,
+      "grad_norm": 0.82964731633246,
+      "learning_rate": 1.3983957781464363e-05,
+      "loss": 1.4333,
+      "step": 18120
+    },
+    {
+      "epoch": 0.8347998341548809,
+      "grad_norm": 1.2498287403394368,
+      "learning_rate": 1.3976348735099732e-05,
+      "loss": 1.192,
+      "step": 18121
+    },
+    {
+      "epoch": 0.834845902243516,
+      "grad_norm": 2.722524384613397,
+      "learning_rate": 1.3968741603904211e-05,
+      "loss": 0.903,
+      "step": 18122
+    },
+    {
+      "epoch": 0.8348919703321509,
+      "grad_norm": 2.023149458986499,
+      "learning_rate": 1.3961136388047114e-05,
+      "loss": 1.2284,
+      "step": 18123
+    },
+    {
+      "epoch": 0.834938038420786,
+      "grad_norm": 0.6851777525512688,
+      "learning_rate": 1.3953533087697802e-05,
+      "loss": 0.9665,
+      "step": 18124
+    },
+    {
+      "epoch": 0.8349841065094209,
+      "grad_norm": 1.2630350375353117,
+      "learning_rate": 1.3945931703025538e-05,
+      "loss": 1.1708,
+      "step": 18125
+    },
+    {
+      "epoch": 0.835030174598056,
+      "grad_norm": 1.100360355964651,
+      "learning_rate": 1.393833223419957e-05,
+      "loss": 1.0922,
+      "step": 18126
+    },
+    {
+      "epoch": 0.8350762426866909,
+      "grad_norm": 0.9193096186637075,
+      "learning_rate": 1.3930734681389058e-05,
+      "loss": 1.1592,
+      "step": 18127
+    },
+    {
+      "epoch": 0.835122310775326,
+      "grad_norm": 1.523504725931718,
+      "learning_rate": 1.3923139044763156e-05,
+      "loss": 1.1425,
+      "step": 18128
+    },
+    {
+      "epoch": 0.8351683788639609,
+      "grad_norm": 0.9797450213724751,
+      "learning_rate": 1.3915545324490964e-05,
+      "loss": 1.305,
+      "step": 18129
+    },
+    {
+      "epoch": 0.835214446952596,
+      "grad_norm": 1.6602502971235573,
+      "learning_rate": 1.390795352074159e-05,
+      "loss": 0.9802,
+      "step": 18130
+    },
+    {
+      "epoch": 0.8352605150412309,
+      "grad_norm": 0.8248147317536332,
+      "learning_rate": 1.3900363633683978e-05,
+      "loss": 1.1207,
+      "step": 18131
+    },
+    {
+      "epoch": 0.8353065831298659,
+      "grad_norm": 0.6790937709998276,
+      "learning_rate": 1.3892775663487145e-05,
+      "loss": 1.3169,
+      "step": 18132
+    },
+    {
+      "epoch": 0.835352651218501,
+      "grad_norm": 0.8180642918419748,
+      "learning_rate": 1.3885189610320048e-05,
+      "loss": 1.2339,
+      "step": 18133
+    },
+    {
+      "epoch": 0.8353987193071359,
+      "grad_norm": 0.6908822736619497,
+      "learning_rate": 1.387760547435153e-05,
+      "loss": 1.1853,
+      "step": 18134
+    },
+    {
+      "epoch": 0.835444787395771,
+      "grad_norm": 0.9292985810563524,
+      "learning_rate": 1.3870023255750442e-05,
+      "loss": 1.3582,
+      "step": 18135
+    },
+    {
+      "epoch": 0.8354908554844059,
+      "grad_norm": 0.749071456363456,
+      "learning_rate": 1.3862442954685617e-05,
+      "loss": 1.3351,
+      "step": 18136
+    },
+    {
+      "epoch": 0.835536923573041,
+      "grad_norm": 0.7569803626251678,
+      "learning_rate": 1.385486457132582e-05,
+      "loss": 1.4088,
+      "step": 18137
+    },
+    {
+      "epoch": 0.8355829916616759,
+      "grad_norm": 0.7472092838612626,
+      "learning_rate": 1.3847288105839739e-05,
+      "loss": 1.2796,
+      "step": 18138
+    },
+    {
+      "epoch": 0.835629059750311,
+      "grad_norm": 1.1344427871127332,
+      "learning_rate": 1.3839713558396072e-05,
+      "loss": 1.088,
+      "step": 18139
+    },
+    {
+      "epoch": 0.8356751278389459,
+      "grad_norm": 1.7119428590331225,
+      "learning_rate": 1.3832140929163461e-05,
+      "loss": 1.0686,
+      "step": 18140
+    },
+    {
+      "epoch": 0.835721195927581,
+      "grad_norm": 2.706506416578185,
+      "learning_rate": 1.3824570218310473e-05,
+      "loss": 1.3588,
+      "step": 18141
+    },
+    {
+      "epoch": 0.8357672640162159,
+      "grad_norm": 0.8830392093802332,
+      "learning_rate": 1.381700142600567e-05,
+      "loss": 1.1816,
+      "step": 18142
+    },
+    {
+      "epoch": 0.835813332104851,
+      "grad_norm": 2.7592851460620325,
+      "learning_rate": 1.3809434552417554e-05,
+      "loss": 1.2213,
+      "step": 18143
+    },
+    {
+      "epoch": 0.8358594001934859,
+      "grad_norm": 2.6102754061220703,
+      "learning_rate": 1.3801869597714613e-05,
+      "loss": 0.8283,
+      "step": 18144
+    },
+    {
+      "epoch": 0.835905468282121,
+      "grad_norm": 0.9492748087077518,
+      "learning_rate": 1.3794306562065228e-05,
+      "loss": 0.9954,
+      "step": 18145
+    },
+    {
+      "epoch": 0.835951536370756,
+      "grad_norm": 0.925643458282194,
+      "learning_rate": 1.3786745445637794e-05,
+      "loss": 1.3583,
+      "step": 18146
+    },
+    {
+      "epoch": 0.835997604459391,
+      "grad_norm": 0.8841488616543384,
+      "learning_rate": 1.377918624860065e-05,
+      "loss": 0.9478,
+      "step": 18147
+    },
+    {
+      "epoch": 0.836043672548026,
+      "grad_norm": 0.8117969333325632,
+      "learning_rate": 1.3771628971122109e-05,
+      "loss": 1.0058,
+      "step": 18148
+    },
+    {
+      "epoch": 0.836089740636661,
+      "grad_norm": 0.8437358500325111,
+      "learning_rate": 1.3764073613370398e-05,
+      "loss": 1.1562,
+      "step": 18149
+    },
+    {
+      "epoch": 0.836135808725296,
+      "grad_norm": 0.7124870083629583,
+      "learning_rate": 1.3756520175513688e-05,
+      "loss": 1.2432,
+      "step": 18150
+    },
+    {
+      "epoch": 0.836181876813931,
+      "grad_norm": 1.7500416331498803,
+      "learning_rate": 1.3748968657720219e-05,
+      "loss": 1.3,
+      "step": 18151
+    },
+    {
+      "epoch": 0.836227944902566,
+      "grad_norm": 1.067269525114876,
+      "learning_rate": 1.3741419060158056e-05,
+      "loss": 1.3473,
+      "step": 18152
+    },
+    {
+      "epoch": 0.836274012991201,
+      "grad_norm": 0.918815244298874,
+      "learning_rate": 1.3733871382995312e-05,
+      "loss": 1.3689,
+      "step": 18153
+    },
+    {
+      "epoch": 0.836320081079836,
+      "grad_norm": 1.4203974399489117,
+      "learning_rate": 1.3726325626399961e-05,
+      "loss": 1.3244,
+      "step": 18154
+    },
+    {
+      "epoch": 0.836366149168471,
+      "grad_norm": 0.8401418414288082,
+      "learning_rate": 1.3718781790540091e-05,
+      "loss": 1.3572,
+      "step": 18155
+    },
+    {
+      "epoch": 0.836412217257106,
+      "grad_norm": 1.1195992097663503,
+      "learning_rate": 1.3711239875583582e-05,
+      "loss": 1.235,
+      "step": 18156
+    },
+    {
+      "epoch": 0.836458285345741,
+      "grad_norm": 1.067966757064091,
+      "learning_rate": 1.3703699881698361e-05,
+      "loss": 1.2603,
+      "step": 18157
+    },
+    {
+      "epoch": 0.836504353434376,
+      "grad_norm": 0.7550125839162509,
+      "learning_rate": 1.3696161809052321e-05,
+      "loss": 1.3287,
+      "step": 18158
+    },
+    {
+      "epoch": 0.836550421523011,
+      "grad_norm": 1.0027190420353826,
+      "learning_rate": 1.3688625657813237e-05,
+      "loss": 1.1707,
+      "step": 18159
+    },
+    {
+      "epoch": 0.836596489611646,
+      "grad_norm": 1.0299141489155257,
+      "learning_rate": 1.368109142814893e-05,
+      "loss": 1.3578,
+      "step": 18160
+    },
+    {
+      "epoch": 0.8366425577002811,
+      "grad_norm": 1.8233709948197652,
+      "learning_rate": 1.3673559120227076e-05,
+      "loss": 1.1558,
+      "step": 18161
+    },
+    {
+      "epoch": 0.836688625788916,
+      "grad_norm": 0.8736562914025465,
+      "learning_rate": 1.3666028734215453e-05,
+      "loss": 1.3484,
+      "step": 18162
+    },
+    {
+      "epoch": 0.8367346938775511,
+      "grad_norm": 0.8415314055547908,
+      "learning_rate": 1.3658500270281649e-05,
+      "loss": 1.2792,
+      "step": 18163
+    },
+    {
+      "epoch": 0.836780761966186,
+      "grad_norm": 0.8249778163290973,
+      "learning_rate": 1.3650973728593307e-05,
+      "loss": 1.217,
+      "step": 18164
+    },
+    {
+      "epoch": 0.8368268300548211,
+      "grad_norm": 1.2733863192599164,
+      "learning_rate": 1.3643449109317941e-05,
+      "loss": 1.2006,
+      "step": 18165
+    },
+    {
+      "epoch": 0.836872898143456,
+      "grad_norm": 1.3299264113857765,
+      "learning_rate": 1.3635926412623156e-05,
+      "loss": 1.4386,
+      "step": 18166
+    },
+    {
+      "epoch": 0.836918966232091,
+      "grad_norm": 1.5566296795725032,
+      "learning_rate": 1.3628405638676378e-05,
+      "loss": 1.5061,
+      "step": 18167
+    },
+    {
+      "epoch": 0.836965034320726,
+      "grad_norm": 1.2136623012488654,
+      "learning_rate": 1.3620886787645015e-05,
+      "loss": 0.8387,
+      "step": 18168
+    },
+    {
+      "epoch": 0.837011102409361,
+      "grad_norm": 1.1272794540372681,
+      "learning_rate": 1.3613369859696545e-05,
+      "loss": 1.1417,
+      "step": 18169
+    },
+    {
+      "epoch": 0.837057170497996,
+      "grad_norm": 7.806772202637202,
+      "learning_rate": 1.3605854854998258e-05,
+      "loss": 1.1374,
+      "step": 18170
+    },
+    {
+      "epoch": 0.837103238586631,
+      "grad_norm": 0.8119566395380428,
+      "learning_rate": 1.3598341773717482e-05,
+      "loss": 1.3611,
+      "step": 18171
+    },
+    {
+      "epoch": 0.837149306675266,
+      "grad_norm": 0.9209284614941997,
+      "learning_rate": 1.3590830616021454e-05,
+      "loss": 1.1226,
+      "step": 18172
+    },
+    {
+      "epoch": 0.837195374763901,
+      "grad_norm": 0.7714650906714832,
+      "learning_rate": 1.3583321382077452e-05,
+      "loss": 1.1801,
+      "step": 18173
+    },
+    {
+      "epoch": 0.8372414428525361,
+      "grad_norm": 0.8731951150887073,
+      "learning_rate": 1.3575814072052617e-05,
+      "loss": 1.0695,
+      "step": 18174
+    },
+    {
+      "epoch": 0.837287510941171,
+      "grad_norm": 0.7904987976246404,
+      "learning_rate": 1.3568308686114106e-05,
+      "loss": 1.0364,
+      "step": 18175
+    },
+    {
+      "epoch": 0.8373335790298061,
+      "grad_norm": 0.8574006414632382,
+      "learning_rate": 1.3560805224428985e-05,
+      "loss": 1.2905,
+      "step": 18176
+    },
+    {
+      "epoch": 0.837379647118441,
+      "grad_norm": 0.777001908659116,
+      "learning_rate": 1.3553303687164331e-05,
+      "loss": 1.0684,
+      "step": 18177
+    },
+    {
+      "epoch": 0.8374257152070761,
+      "grad_norm": 0.9078361158848121,
+      "learning_rate": 1.3545804074487167e-05,
+      "loss": 1.196,
+      "step": 18178
+    },
+    {
+      "epoch": 0.837471783295711,
+      "grad_norm": 1.0302101740169292,
+      "learning_rate": 1.3538306386564404e-05,
+      "loss": 1.2909,
+      "step": 18179
+    },
+    {
+      "epoch": 0.8375178513843461,
+      "grad_norm": 0.7805549464548903,
+      "learning_rate": 1.3530810623563006e-05,
+      "loss": 1.2806,
+      "step": 18180
+    },
+    {
+      "epoch": 0.837563919472981,
+      "grad_norm": 1.071571360533416,
+      "learning_rate": 1.3523316785649854e-05,
+      "loss": 1.492,
+      "step": 18181
+    },
+    {
+      "epoch": 0.8376099875616161,
+      "grad_norm": 0.9175815710277553,
+      "learning_rate": 1.3515824872991789e-05,
+      "loss": 1.3657,
+      "step": 18182
+    },
+    {
+      "epoch": 0.837656055650251,
+      "grad_norm": 1.890259579839662,
+      "learning_rate": 1.3508334885755559e-05,
+      "loss": 1.2092,
+      "step": 18183
+    },
+    {
+      "epoch": 0.8377021237388861,
+      "grad_norm": 2.4011536679575545,
+      "learning_rate": 1.3500846824107992e-05,
+      "loss": 1.1536,
+      "step": 18184
+    },
+    {
+      "epoch": 0.837748191827521,
+      "grad_norm": 0.9617421668686579,
+      "learning_rate": 1.3493360688215761e-05,
+      "loss": 1.4335,
+      "step": 18185
+    },
+    {
+      "epoch": 0.8377942599161561,
+      "grad_norm": 0.9637127336588467,
+      "learning_rate": 1.3485876478245497e-05,
+      "loss": 1.122,
+      "step": 18186
+    },
+    {
+      "epoch": 0.8378403280047911,
+      "grad_norm": 0.7140285178483209,
+      "learning_rate": 1.3478394194363863e-05,
+      "loss": 1.2836,
+      "step": 18187
+    },
+    {
+      "epoch": 0.8378863960934261,
+      "grad_norm": 2.2747808040209794,
+      "learning_rate": 1.347091383673742e-05,
+      "loss": 1.2886,
+      "step": 18188
+    },
+    {
+      "epoch": 0.8379324641820611,
+      "grad_norm": 0.6395905208058569,
+      "learning_rate": 1.3463435405532742e-05,
+      "loss": 1.3289,
+      "step": 18189
+    },
+    {
+      "epoch": 0.8379785322706961,
+      "grad_norm": 2.5863458965260095,
+      "learning_rate": 1.3455958900916277e-05,
+      "loss": 1.4699,
+      "step": 18190
+    },
+    {
+      "epoch": 0.8380246003593311,
+      "grad_norm": 2.0711713527986064,
+      "learning_rate": 1.3448484323054488e-05,
+      "loss": 1.0877,
+      "step": 18191
+    },
+    {
+      "epoch": 0.8380706684479661,
+      "grad_norm": 0.7953209228977055,
+      "learning_rate": 1.344101167211379e-05,
+      "loss": 1.2089,
+      "step": 18192
+    },
+    {
+      "epoch": 0.8381167365366011,
+      "grad_norm": 1.1770859902819617,
+      "learning_rate": 1.3433540948260581e-05,
+      "loss": 1.0723,
+      "step": 18193
+    },
+    {
+      "epoch": 0.8381628046252361,
+      "grad_norm": 1.200202396204804,
+      "learning_rate": 1.3426072151661129e-05,
+      "loss": 1.0914,
+      "step": 18194
+    },
+    {
+      "epoch": 0.8382088727138711,
+      "grad_norm": 1.3354696098851442,
+      "learning_rate": 1.3418605282481733e-05,
+      "loss": 1.1779,
+      "step": 18195
+    },
+    {
+      "epoch": 0.8382549408025061,
+      "grad_norm": 0.7061944567126893,
+      "learning_rate": 1.341114034088866e-05,
+      "loss": 1.0513,
+      "step": 18196
+    },
+    {
+      "epoch": 0.8383010088911411,
+      "grad_norm": 1.2725463984254364,
+      "learning_rate": 1.340367732704806e-05,
+      "loss": 0.954,
+      "step": 18197
+    },
+    {
+      "epoch": 0.8383470769797761,
+      "grad_norm": 1.0305877760452427,
+      "learning_rate": 1.3396216241126103e-05,
+      "loss": 0.7575,
+      "step": 18198
+    },
+    {
+      "epoch": 0.8383931450684111,
+      "grad_norm": 0.9102954409128007,
+      "learning_rate": 1.3388757083288894e-05,
+      "loss": 0.9514,
+      "step": 18199
+    },
+    {
+      "epoch": 0.8384392131570461,
+      "grad_norm": 0.762642796363896,
+      "learning_rate": 1.3381299853702533e-05,
+      "loss": 1.4439,
+      "step": 18200
+    },
+    {
+      "epoch": 0.8384852812456811,
+      "grad_norm": 1.2435744320988864,
+      "learning_rate": 1.3373844552532977e-05,
+      "loss": 1.5053,
+      "step": 18201
+    },
+    {
+      "epoch": 0.8385313493343162,
+      "grad_norm": 2.3034014795338824,
+      "learning_rate": 1.3366391179946248e-05,
+      "loss": 1.2986,
+      "step": 18202
+    },
+    {
+      "epoch": 0.8385774174229511,
+      "grad_norm": 0.9217552728087286,
+      "learning_rate": 1.33589397361083e-05,
+      "loss": 1.3397,
+      "step": 18203
+    },
+    {
+      "epoch": 0.8386234855115862,
+      "grad_norm": 1.1718608374578152,
+      "learning_rate": 1.335149022118497e-05,
+      "loss": 1.3977,
+      "step": 18204
+    },
+    {
+      "epoch": 0.8386695536002211,
+      "grad_norm": 1.0511221926478698,
+      "learning_rate": 1.3344042635342147e-05,
+      "loss": 1.5409,
+      "step": 18205
+    },
+    {
+      "epoch": 0.8387156216888562,
+      "grad_norm": 0.785510141108479,
+      "learning_rate": 1.333659697874562e-05,
+      "loss": 0.9636,
+      "step": 18206
+    },
+    {
+      "epoch": 0.8387616897774911,
+      "grad_norm": 0.8655485254457777,
+      "learning_rate": 1.33291532515612e-05,
+      "loss": 1.4017,
+      "step": 18207
+    },
+    {
+      "epoch": 0.8388077578661262,
+      "grad_norm": 2.3460721965359874,
+      "learning_rate": 1.3321711453954543e-05,
+      "loss": 1.0418,
+      "step": 18208
+    },
+    {
+      "epoch": 0.8388538259547611,
+      "grad_norm": 0.7782558843976721,
+      "learning_rate": 1.3314271586091364e-05,
+      "loss": 1.3239,
+      "step": 18209
+    },
+    {
+      "epoch": 0.8388998940433962,
+      "grad_norm": 1.2934767849956084,
+      "learning_rate": 1.3306833648137295e-05,
+      "loss": 1.0371,
+      "step": 18210
+    },
+    {
+      "epoch": 0.8389459621320311,
+      "grad_norm": 0.9378092780685585,
+      "learning_rate": 1.329939764025795e-05,
+      "loss": 1.0791,
+      "step": 18211
+    },
+    {
+      "epoch": 0.8389920302206662,
+      "grad_norm": 0.7628873859020304,
+      "learning_rate": 1.3291963562618836e-05,
+      "loss": 1.1815,
+      "step": 18212
+    },
+    {
+      "epoch": 0.8390380983093011,
+      "grad_norm": 0.9793636281592201,
+      "learning_rate": 1.3284531415385482e-05,
+      "loss": 1.4153,
+      "step": 18213
+    },
+    {
+      "epoch": 0.8390841663979361,
+      "grad_norm": 0.86895678447572,
+      "learning_rate": 1.3277101198723363e-05,
+      "loss": 0.8932,
+      "step": 18214
+    },
+    {
+      "epoch": 0.8391302344865712,
+      "grad_norm": 0.8495514935490425,
+      "learning_rate": 1.326967291279788e-05,
+      "loss": 1.231,
+      "step": 18215
+    },
+    {
+      "epoch": 0.8391763025752061,
+      "grad_norm": 1.0444307839013698,
+      "learning_rate": 1.3262246557774405e-05,
+      "loss": 1.1426,
+      "step": 18216
+    },
+    {
+      "epoch": 0.8392223706638412,
+      "grad_norm": 1.5303454973214146,
+      "learning_rate": 1.3254822133818301e-05,
+      "loss": 0.9831,
+      "step": 18217
+    },
+    {
+      "epoch": 0.8392684387524761,
+      "grad_norm": 1.3100362902573488,
+      "learning_rate": 1.324739964109486e-05,
+      "loss": 1.2648,
+      "step": 18218
+    },
+    {
+      "epoch": 0.8393145068411112,
+      "grad_norm": 0.7461761257496579,
+      "learning_rate": 1.3239979079769293e-05,
+      "loss": 1.0708,
+      "step": 18219
+    },
+    {
+      "epoch": 0.8393605749297461,
+      "grad_norm": 1.0220418220466023,
+      "learning_rate": 1.3232560450006836e-05,
+      "loss": 1.1463,
+      "step": 18220
+    },
+    {
+      "epoch": 0.8394066430183812,
+      "grad_norm": 0.8784732152452809,
+      "learning_rate": 1.3225143751972658e-05,
+      "loss": 1.3513,
+      "step": 18221
+    },
+    {
+      "epoch": 0.8394527111070161,
+      "grad_norm": 0.9619607450441544,
+      "learning_rate": 1.3217728985831856e-05,
+      "loss": 1.3333,
+      "step": 18222
+    },
+    {
+      "epoch": 0.8394987791956512,
+      "grad_norm": 0.8368689660425648,
+      "learning_rate": 1.3210316151749513e-05,
+      "loss": 1.4282,
+      "step": 18223
+    },
+    {
+      "epoch": 0.8395448472842861,
+      "grad_norm": 1.0644294056860975,
+      "learning_rate": 1.3202905249890673e-05,
+      "loss": 1.3042,
+      "step": 18224
+    },
+    {
+      "epoch": 0.8395909153729212,
+      "grad_norm": 1.1424826504602714,
+      "learning_rate": 1.3195496280420339e-05,
+      "loss": 1.064,
+      "step": 18225
+    },
+    {
+      "epoch": 0.8396369834615561,
+      "grad_norm": 0.7447297796767598,
+      "learning_rate": 1.3188089243503432e-05,
+      "loss": 1.3816,
+      "step": 18226
+    },
+    {
+      "epoch": 0.8396830515501912,
+      "grad_norm": 0.8627695523515914,
+      "learning_rate": 1.3180684139304855e-05,
+      "loss": 1.3782,
+      "step": 18227
+    },
+    {
+      "epoch": 0.8397291196388262,
+      "grad_norm": 0.7092398796296864,
+      "learning_rate": 1.3173280967989488e-05,
+      "loss": 1.442,
+      "step": 18228
+    },
+    {
+      "epoch": 0.8397751877274612,
+      "grad_norm": 1.2728346681258957,
+      "learning_rate": 1.3165879729722163e-05,
+      "loss": 1.0061,
+      "step": 18229
+    },
+    {
+      "epoch": 0.8398212558160962,
+      "grad_norm": 0.6727436020730069,
+      "learning_rate": 1.3158480424667618e-05,
+      "loss": 1.2297,
+      "step": 18230
+    },
+    {
+      "epoch": 0.8398673239047312,
+      "grad_norm": 1.9511000097780242,
+      "learning_rate": 1.3151083052990598e-05,
+      "loss": 1.0257,
+      "step": 18231
+    },
+    {
+      "epoch": 0.8399133919933662,
+      "grad_norm": 1.0234521387122106,
+      "learning_rate": 1.3143687614855838e-05,
+      "loss": 1.2798,
+      "step": 18232
+    },
+    {
+      "epoch": 0.8399594600820012,
+      "grad_norm": 0.7295134362173141,
+      "learning_rate": 1.3136294110427904e-05,
+      "loss": 0.9662,
+      "step": 18233
+    },
+    {
+      "epoch": 0.8400055281706362,
+      "grad_norm": 1.0046754112877698,
+      "learning_rate": 1.3128902539871458e-05,
+      "loss": 1.3589,
+      "step": 18234
+    },
+    {
+      "epoch": 0.8400515962592712,
+      "grad_norm": 2.5132651044782515,
+      "learning_rate": 1.312151290335104e-05,
+      "loss": 1.3854,
+      "step": 18235
+    },
+    {
+      "epoch": 0.8400976643479062,
+      "grad_norm": 0.7816326379289743,
+      "learning_rate": 1.3114125201031192e-05,
+      "loss": 1.1007,
+      "step": 18236
+    },
+    {
+      "epoch": 0.8401437324365412,
+      "grad_norm": 2.2160809203146328,
+      "learning_rate": 1.3106739433076355e-05,
+      "loss": 1.2832,
+      "step": 18237
+    },
+    {
+      "epoch": 0.8401898005251762,
+      "grad_norm": 1.2514366881220862,
+      "learning_rate": 1.3099355599650965e-05,
+      "loss": 1.4498,
+      "step": 18238
+    },
+    {
+      "epoch": 0.8402358686138112,
+      "grad_norm": 1.2253641526967036,
+      "learning_rate": 1.3091973700919424e-05,
+      "loss": 1.4711,
+      "step": 18239
+    },
+    {
+      "epoch": 0.8402819367024462,
+      "grad_norm": 1.1056135964321683,
+      "learning_rate": 1.3084593737046103e-05,
+      "loss": 1.2983,
+      "step": 18240
+    },
+    {
+      "epoch": 0.8403280047910813,
+      "grad_norm": 0.9790893147922095,
+      "learning_rate": 1.3077215708195233e-05,
+      "loss": 0.9885,
+      "step": 18241
+    },
+    {
+      "epoch": 0.8403740728797162,
+      "grad_norm": 2.1531017075237524,
+      "learning_rate": 1.3069839614531133e-05,
+      "loss": 1.2095,
+      "step": 18242
+    },
+    {
+      "epoch": 0.8404201409683513,
+      "grad_norm": 0.8388425558680765,
+      "learning_rate": 1.306246545621802e-05,
+      "loss": 1.4701,
+      "step": 18243
+    },
+    {
+      "epoch": 0.8404662090569862,
+      "grad_norm": 0.9465975947445312,
+      "learning_rate": 1.3055093233420012e-05,
+      "loss": 1.2125,
+      "step": 18244
+    },
+    {
+      "epoch": 0.8405122771456213,
+      "grad_norm": 0.8608299554304486,
+      "learning_rate": 1.304772294630129e-05,
+      "loss": 1.313,
+      "step": 18245
+    },
+    {
+      "epoch": 0.8405583452342562,
+      "grad_norm": 1.095985311005434,
+      "learning_rate": 1.3040354595025916e-05,
+      "loss": 1.1165,
+      "step": 18246
+    },
+    {
+      "epoch": 0.8406044133228913,
+      "grad_norm": 1.708285678629219,
+      "learning_rate": 1.3032988179757977e-05,
+      "loss": 1.2683,
+      "step": 18247
+    },
+    {
+      "epoch": 0.8406504814115262,
+      "grad_norm": 1.6886578503534282,
+      "learning_rate": 1.3025623700661404e-05,
+      "loss": 1.1947,
+      "step": 18248
+    },
+    {
+      "epoch": 0.8406965495001613,
+      "grad_norm": 0.6310721996042428,
+      "learning_rate": 1.3018261157900203e-05,
+      "loss": 1.2762,
+      "step": 18249
+    },
+    {
+      "epoch": 0.8407426175887962,
+      "grad_norm": 0.9785274924264488,
+      "learning_rate": 1.3010900551638284e-05,
+      "loss": 1.1304,
+      "step": 18250
+    },
+    {
+      "epoch": 0.8407886856774313,
+      "grad_norm": 0.9595115474381455,
+      "learning_rate": 1.3003541882039493e-05,
+      "loss": 1.0965,
+      "step": 18251
+    },
+    {
+      "epoch": 0.8408347537660662,
+      "grad_norm": 0.7463576890568777,
+      "learning_rate": 1.2996185149267703e-05,
+      "loss": 1.3482,
+      "step": 18252
+    },
+    {
+      "epoch": 0.8408808218547013,
+      "grad_norm": 0.9693802239608298,
+      "learning_rate": 1.2988830353486626e-05,
+      "loss": 1.2236,
+      "step": 18253
+    },
+    {
+      "epoch": 0.8409268899433362,
+      "grad_norm": 1.5414682198428684,
+      "learning_rate": 1.2981477494860083e-05,
+      "loss": 1.1803,
+      "step": 18254
+    },
+    {
+      "epoch": 0.8409729580319713,
+      "grad_norm": 0.9670122136186972,
+      "learning_rate": 1.2974126573551726e-05,
+      "loss": 1.1577,
+      "step": 18255
+    },
+    {
+      "epoch": 0.8410190261206063,
+      "grad_norm": 1.286074023387419,
+      "learning_rate": 1.296677758972522e-05,
+      "loss": 1.072,
+      "step": 18256
+    },
+    {
+      "epoch": 0.8410650942092412,
+      "grad_norm": 2.311290179564512,
+      "learning_rate": 1.2959430543544182e-05,
+      "loss": 1.2239,
+      "step": 18257
+    },
+    {
+      "epoch": 0.8411111622978763,
+      "grad_norm": 1.6565963574738811,
+      "learning_rate": 1.29520854351722e-05,
+      "loss": 0.8043,
+      "step": 18258
+    },
+    {
+      "epoch": 0.8411572303865112,
+      "grad_norm": 0.8488903610428794,
+      "learning_rate": 1.294474226477278e-05,
+      "loss": 1.4105,
+      "step": 18259
+    },
+    {
+      "epoch": 0.8412032984751463,
+      "grad_norm": 0.8814873773989027,
+      "learning_rate": 1.2937401032509367e-05,
+      "loss": 1.3214,
+      "step": 18260
+    },
+    {
+      "epoch": 0.8412493665637812,
+      "grad_norm": 1.7035870646377669,
+      "learning_rate": 1.2930061738545484e-05,
+      "loss": 1.1697,
+      "step": 18261
+    },
+    {
+      "epoch": 0.8412954346524163,
+      "grad_norm": 0.59940961716386,
+      "learning_rate": 1.292272438304446e-05,
+      "loss": 1.2593,
+      "step": 18262
+    },
+    {
+      "epoch": 0.8413415027410512,
+      "grad_norm": 1.508050313906535,
+      "learning_rate": 1.2915388966169695e-05,
+      "loss": 1.2385,
+      "step": 18263
+    },
+    {
+      "epoch": 0.8413875708296863,
+      "grad_norm": 0.7905766111676171,
+      "learning_rate": 1.2908055488084426e-05,
+      "loss": 1.0821,
+      "step": 18264
+    },
+    {
+      "epoch": 0.8414336389183212,
+      "grad_norm": 0.8075703928168159,
+      "learning_rate": 1.290072394895202e-05,
+      "loss": 1.0984,
+      "step": 18265
+    },
+    {
+      "epoch": 0.8414797070069563,
+      "grad_norm": 0.9488535750013617,
+      "learning_rate": 1.2893394348935629e-05,
+      "loss": 1.3663,
+      "step": 18266
+    },
+    {
+      "epoch": 0.8415257750955912,
+      "grad_norm": 0.7854275839965048,
+      "learning_rate": 1.2886066688198461e-05,
+      "loss": 1.5093,
+      "step": 18267
+    },
+    {
+      "epoch": 0.8415718431842263,
+      "grad_norm": 1.2469024839510623,
+      "learning_rate": 1.2878740966903668e-05,
+      "loss": 1.3118,
+      "step": 18268
+    },
+    {
+      "epoch": 0.8416179112728613,
+      "grad_norm": 0.9087140017982448,
+      "learning_rate": 1.2871417185214296e-05,
+      "loss": 1.1801,
+      "step": 18269
+    },
+    {
+      "epoch": 0.8416639793614963,
+      "grad_norm": 1.0994203309654473,
+      "learning_rate": 1.2864095343293459e-05,
+      "loss": 1.128,
+      "step": 18270
+    },
+    {
+      "epoch": 0.8417100474501313,
+      "grad_norm": 1.2318269704783464,
+      "learning_rate": 1.2856775441304091e-05,
+      "loss": 1.0289,
+      "step": 18271
+    },
+    {
+      "epoch": 0.8417561155387663,
+      "grad_norm": 0.9404339562942318,
+      "learning_rate": 1.2849457479409244e-05,
+      "loss": 1.4484,
+      "step": 18272
+    },
+    {
+      "epoch": 0.8418021836274013,
+      "grad_norm": 1.5712043935472162,
+      "learning_rate": 1.284214145777176e-05,
+      "loss": 1.3269,
+      "step": 18273
+    },
+    {
+      "epoch": 0.8418482517160363,
+      "grad_norm": 0.8400754146829129,
+      "learning_rate": 1.2834827376554592e-05,
+      "loss": 1.0266,
+      "step": 18274
+    },
+    {
+      "epoch": 0.8418943198046713,
+      "grad_norm": 1.519954656953144,
+      "learning_rate": 1.2827515235920485e-05,
+      "loss": 1.339,
+      "step": 18275
+    },
+    {
+      "epoch": 0.8419403878933063,
+      "grad_norm": 1.028434825828768,
+      "learning_rate": 1.2820205036032317e-05,
+      "loss": 1.3178,
+      "step": 18276
+    },
+    {
+      "epoch": 0.8419864559819413,
+      "grad_norm": 0.8464352067418253,
+      "learning_rate": 1.2812896777052818e-05,
+      "loss": 1.1199,
+      "step": 18277
+    },
+    {
+      "epoch": 0.8420325240705763,
+      "grad_norm": 1.0525407976752952,
+      "learning_rate": 1.2805590459144645e-05,
+      "loss": 1.0764,
+      "step": 18278
+    },
+    {
+      "epoch": 0.8420785921592113,
+      "grad_norm": 1.099992355562576,
+      "learning_rate": 1.2798286082470501e-05,
+      "loss": 0.9538,
+      "step": 18279
+    },
+    {
+      "epoch": 0.8421246602478463,
+      "grad_norm": 1.1551375279118048,
+      "learning_rate": 1.2790983647193e-05,
+      "loss": 1.2204,
+      "step": 18280
+    },
+    {
+      "epoch": 0.8421707283364813,
+      "grad_norm": 1.1778058624367749,
+      "learning_rate": 1.2783683153474734e-05,
+      "loss": 1.0139,
+      "step": 18281
+    },
+    {
+      "epoch": 0.8422167964251164,
+      "grad_norm": 1.2847041270462762,
+      "learning_rate": 1.2776384601478175e-05,
+      "loss": 1.2451,
+      "step": 18282
+    },
+    {
+      "epoch": 0.8422628645137513,
+      "grad_norm": 1.2016044959409633,
+      "learning_rate": 1.2769087991365902e-05,
+      "loss": 1.3731,
+      "step": 18283
+    },
+    {
+      "epoch": 0.8423089326023864,
+      "grad_norm": 1.4800598172362793,
+      "learning_rate": 1.2761793323300287e-05,
+      "loss": 1.2525,
+      "step": 18284
+    },
+    {
+      "epoch": 0.8423550006910213,
+      "grad_norm": 0.9825624970976059,
+      "learning_rate": 1.2754500597443786e-05,
+      "loss": 1.5101,
+      "step": 18285
+    },
+    {
+      "epoch": 0.8424010687796564,
+      "grad_norm": 1.4498376160941608,
+      "learning_rate": 1.2747209813958727e-05,
+      "loss": 1.1257,
+      "step": 18286
+    },
+    {
+      "epoch": 0.8424471368682913,
+      "grad_norm": 2.3051367020450786,
+      "learning_rate": 1.2739920973007424e-05,
+      "loss": 1.0519,
+      "step": 18287
+    },
+    {
+      "epoch": 0.8424932049569264,
+      "grad_norm": 0.7463888791436437,
+      "learning_rate": 1.2732634074752181e-05,
+      "loss": 1.3506,
+      "step": 18288
+    },
+    {
+      "epoch": 0.8425392730455613,
+      "grad_norm": 0.7901865874159036,
+      "learning_rate": 1.272534911935519e-05,
+      "loss": 1.4804,
+      "step": 18289
+    },
+    {
+      "epoch": 0.8425853411341964,
+      "grad_norm": 0.944580095773167,
+      "learning_rate": 1.2718066106978655e-05,
+      "loss": 1.2041,
+      "step": 18290
+    },
+    {
+      "epoch": 0.8426314092228313,
+      "grad_norm": 1.0195379727936698,
+      "learning_rate": 1.2710785037784733e-05,
+      "loss": 1.1821,
+      "step": 18291
+    },
+    {
+      "epoch": 0.8426774773114664,
+      "grad_norm": 1.3959121222706696,
+      "learning_rate": 1.270350591193552e-05,
+      "loss": 1.4179,
+      "step": 18292
+    },
+    {
+      "epoch": 0.8427235454001013,
+      "grad_norm": 0.9248280605176585,
+      "learning_rate": 1.2696228729593029e-05,
+      "loss": 1.0312,
+      "step": 18293
+    },
+    {
+      "epoch": 0.8427696134887364,
+      "grad_norm": 1.7856351298894806,
+      "learning_rate": 1.2688953490919352e-05,
+      "loss": 0.9227,
+      "step": 18294
+    },
+    {
+      "epoch": 0.8428156815773713,
+      "grad_norm": 1.5551355196294065,
+      "learning_rate": 1.268168019607643e-05,
+      "loss": 1.0926,
+      "step": 18295
+    },
+    {
+      "epoch": 0.8428617496660064,
+      "grad_norm": 0.7425302066127105,
+      "learning_rate": 1.2674408845226148e-05,
+      "loss": 1.3506,
+      "step": 18296
+    },
+    {
+      "epoch": 0.8429078177546414,
+      "grad_norm": 0.9156314595317311,
+      "learning_rate": 1.2667139438530428e-05,
+      "loss": 1.6063,
+      "step": 18297
+    },
+    {
+      "epoch": 0.8429538858432764,
+      "grad_norm": 0.7280560469313305,
+      "learning_rate": 1.2659871976151093e-05,
+      "loss": 1.2516,
+      "step": 18298
+    },
+    {
+      "epoch": 0.8429999539319114,
+      "grad_norm": 0.8717669103799466,
+      "learning_rate": 1.2652606458249982e-05,
+      "loss": 1.3231,
+      "step": 18299
+    },
+    {
+      "epoch": 0.8430460220205463,
+      "grad_norm": 0.7482554949123214,
+      "learning_rate": 1.2645342884988797e-05,
+      "loss": 1.3205,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8430920901091814,
+      "grad_norm": 1.0662734045419373,
+      "learning_rate": 1.2638081256529278e-05,
+      "loss": 1.2506,
+      "step": 18301
+    },
+    {
+      "epoch": 0.8431381581978163,
+      "grad_norm": 1.1689722256035031,
+      "learning_rate": 1.263082157303308e-05,
+      "loss": 1.1164,
+      "step": 18302
+    },
+    {
+      "epoch": 0.8431842262864514,
+      "grad_norm": 0.801060565350871,
+      "learning_rate": 1.2623563834661856e-05,
+      "loss": 1.1314,
+      "step": 18303
+    },
+    {
+      "epoch": 0.8432302943750863,
+      "grad_norm": 1.562461395024195,
+      "learning_rate": 1.261630804157714e-05,
+      "loss": 0.9773,
+      "step": 18304
+    },
+    {
+      "epoch": 0.8432763624637214,
+      "grad_norm": 1.571132380735865,
+      "learning_rate": 1.2609054193940496e-05,
+      "loss": 1.3917,
+      "step": 18305
+    },
+    {
+      "epoch": 0.8433224305523563,
+      "grad_norm": 1.0112499564731623,
+      "learning_rate": 1.2601802291913445e-05,
+      "loss": 1.0968,
+      "step": 18306
+    },
+    {
+      "epoch": 0.8433684986409914,
+      "grad_norm": 0.729762550826439,
+      "learning_rate": 1.2594552335657383e-05,
+      "loss": 1.0947,
+      "step": 18307
+    },
+    {
+      "epoch": 0.8434145667296263,
+      "grad_norm": 0.8401940671790454,
+      "learning_rate": 1.2587304325333749e-05,
+      "loss": 1.3703,
+      "step": 18308
+    },
+    {
+      "epoch": 0.8434606348182614,
+      "grad_norm": 0.9873147049317796,
+      "learning_rate": 1.25800582611039e-05,
+      "loss": 1.2002,
+      "step": 18309
+    },
+    {
+      "epoch": 0.8435067029068964,
+      "grad_norm": 0.8930521403457267,
+      "learning_rate": 1.2572814143129175e-05,
+      "loss": 1.3692,
+      "step": 18310
+    },
+    {
+      "epoch": 0.8435527709955314,
+      "grad_norm": 1.0429677616174204,
+      "learning_rate": 1.2565571971570822e-05,
+      "loss": 0.978,
+      "step": 18311
+    },
+    {
+      "epoch": 0.8435988390841664,
+      "grad_norm": 4.480757412263185,
+      "learning_rate": 1.2558331746590101e-05,
+      "loss": 1.0699,
+      "step": 18312
+    },
+    {
+      "epoch": 0.8436449071728014,
+      "grad_norm": 0.643463544799422,
+      "learning_rate": 1.2551093468348208e-05,
+      "loss": 1.132,
+      "step": 18313
+    },
+    {
+      "epoch": 0.8436909752614364,
+      "grad_norm": 1.0823572748742016,
+      "learning_rate": 1.2543857137006244e-05,
+      "loss": 1.3909,
+      "step": 18314
+    },
+    {
+      "epoch": 0.8437370433500714,
+      "grad_norm": 1.7298429136709277,
+      "learning_rate": 1.2536622752725357e-05,
+      "loss": 1.0529,
+      "step": 18315
+    },
+    {
+      "epoch": 0.8437831114387064,
+      "grad_norm": 0.8789619236048903,
+      "learning_rate": 1.2529390315666601e-05,
+      "loss": 1.3714,
+      "step": 18316
+    },
+    {
+      "epoch": 0.8438291795273414,
+      "grad_norm": 0.909824189048356,
+      "learning_rate": 1.2522159825990998e-05,
+      "loss": 1.1656,
+      "step": 18317
+    },
+    {
+      "epoch": 0.8438752476159764,
+      "grad_norm": 0.9210990593546371,
+      "learning_rate": 1.2514931283859498e-05,
+      "loss": 1.2631,
+      "step": 18318
+    },
+    {
+      "epoch": 0.8439213157046114,
+      "grad_norm": 0.666222655767388,
+      "learning_rate": 1.2507704689433053e-05,
+      "loss": 1.4654,
+      "step": 18319
+    },
+    {
+      "epoch": 0.8439673837932464,
+      "grad_norm": 1.8247596155709171,
+      "learning_rate": 1.2500480042872542e-05,
+      "loss": 1.1248,
+      "step": 18320
+    },
+    {
+      "epoch": 0.8440134518818814,
+      "grad_norm": 1.6189644109510357,
+      "learning_rate": 1.2493257344338837e-05,
+      "loss": 1.2613,
+      "step": 18321
+    },
+    {
+      "epoch": 0.8440595199705164,
+      "grad_norm": 0.9958761430631915,
+      "learning_rate": 1.248603659399269e-05,
+      "loss": 1.3022,
+      "step": 18322
+    },
+    {
+      "epoch": 0.8441055880591515,
+      "grad_norm": 2.8067134191906145,
+      "learning_rate": 1.247881779199489e-05,
+      "loss": 1.3436,
+      "step": 18323
+    },
+    {
+      "epoch": 0.8441516561477864,
+      "grad_norm": 1.0410052466964619,
+      "learning_rate": 1.2471600938506167e-05,
+      "loss": 1.4308,
+      "step": 18324
+    },
+    {
+      "epoch": 0.8441977242364215,
+      "grad_norm": 0.924153482860622,
+      "learning_rate": 1.2464386033687136e-05,
+      "loss": 1.3358,
+      "step": 18325
+    },
+    {
+      "epoch": 0.8442437923250564,
+      "grad_norm": 6.415298387280538,
+      "learning_rate": 1.2457173077698469e-05,
+      "loss": 1.1856,
+      "step": 18326
+    },
+    {
+      "epoch": 0.8442898604136915,
+      "grad_norm": 1.1222644898731773,
+      "learning_rate": 1.2449962070700738e-05,
+      "loss": 1.2308,
+      "step": 18327
+    },
+    {
+      "epoch": 0.8443359285023264,
+      "grad_norm": 1.3418455065235102,
+      "learning_rate": 1.2442753012854513e-05,
+      "loss": 1.0666,
+      "step": 18328
+    },
+    {
+      "epoch": 0.8443819965909615,
+      "grad_norm": 0.8769951878735357,
+      "learning_rate": 1.2435545904320235e-05,
+      "loss": 1.0303,
+      "step": 18329
+    },
+    {
+      "epoch": 0.8444280646795964,
+      "grad_norm": 1.8181816619505442,
+      "learning_rate": 1.2428340745258381e-05,
+      "loss": 1.1323,
+      "step": 18330
+    },
+    {
+      "epoch": 0.8444741327682315,
+      "grad_norm": 1.3137923235860156,
+      "learning_rate": 1.2421137535829397e-05,
+      "loss": 1.3453,
+      "step": 18331
+    },
+    {
+      "epoch": 0.8445202008568664,
+      "grad_norm": 0.7024557912653757,
+      "learning_rate": 1.2413936276193605e-05,
+      "loss": 1.316,
+      "step": 18332
+    },
+    {
+      "epoch": 0.8445662689455015,
+      "grad_norm": 0.795463406619809,
+      "learning_rate": 1.2406736966511335e-05,
+      "loss": 1.0401,
+      "step": 18333
+    },
+    {
+      "epoch": 0.8446123370341364,
+      "grad_norm": 1.130197075417325,
+      "learning_rate": 1.2399539606942878e-05,
+      "loss": 1.4171,
+      "step": 18334
+    },
+    {
+      "epoch": 0.8446584051227715,
+      "grad_norm": 1.0051637084895801,
+      "learning_rate": 1.2392344197648486e-05,
+      "loss": 0.969,
+      "step": 18335
+    },
+    {
+      "epoch": 0.8447044732114064,
+      "grad_norm": 1.042673169757457,
+      "learning_rate": 1.2385150738788321e-05,
+      "loss": 1.1131,
+      "step": 18336
+    },
+    {
+      "epoch": 0.8447505413000415,
+      "grad_norm": 0.8554193432848283,
+      "learning_rate": 1.237795923052254e-05,
+      "loss": 1.6036,
+      "step": 18337
+    },
+    {
+      "epoch": 0.8447966093886765,
+      "grad_norm": 1.1858196500231757,
+      "learning_rate": 1.2370769673011262e-05,
+      "loss": 0.9229,
+      "step": 18338
+    },
+    {
+      "epoch": 0.8448426774773115,
+      "grad_norm": 1.2007798240128753,
+      "learning_rate": 1.236358206641457e-05,
+      "loss": 1.0081,
+      "step": 18339
+    },
+    {
+      "epoch": 0.8448887455659465,
+      "grad_norm": 0.9199445657106087,
+      "learning_rate": 1.2356396410892435e-05,
+      "loss": 1.1154,
+      "step": 18340
+    },
+    {
+      "epoch": 0.8449348136545815,
+      "grad_norm": 0.8110715494932834,
+      "learning_rate": 1.234921270660485e-05,
+      "loss": 1.3608,
+      "step": 18341
+    },
+    {
+      "epoch": 0.8449808817432165,
+      "grad_norm": 6.6066245025360315,
+      "learning_rate": 1.234203095371178e-05,
+      "loss": 0.9154,
+      "step": 18342
+    },
+    {
+      "epoch": 0.8450269498318514,
+      "grad_norm": 0.7024962786031768,
+      "learning_rate": 1.2334851152373072e-05,
+      "loss": 1.4957,
+      "step": 18343
+    },
+    {
+      "epoch": 0.8450730179204865,
+      "grad_norm": 0.8638211100141837,
+      "learning_rate": 1.23276733027486e-05,
+      "loss": 1.0305,
+      "step": 18344
+    },
+    {
+      "epoch": 0.8451190860091214,
+      "grad_norm": 3.33985768906308,
+      "learning_rate": 1.2320497404998143e-05,
+      "loss": 1.3125,
+      "step": 18345
+    },
+    {
+      "epoch": 0.8451651540977565,
+      "grad_norm": 1.0876288792039053,
+      "learning_rate": 1.231332345928149e-05,
+      "loss": 1.3876,
+      "step": 18346
+    },
+    {
+      "epoch": 0.8452112221863914,
+      "grad_norm": 1.053117634399942,
+      "learning_rate": 1.230615146575833e-05,
+      "loss": 1.3238,
+      "step": 18347
+    },
+    {
+      "epoch": 0.8452572902750265,
+      "grad_norm": 0.7971328706673436,
+      "learning_rate": 1.2298981424588341e-05,
+      "loss": 1.1432,
+      "step": 18348
+    },
+    {
+      "epoch": 0.8453033583636614,
+      "grad_norm": 0.7282186203963553,
+      "learning_rate": 1.2291813335931168e-05,
+      "loss": 1.1818,
+      "step": 18349
+    },
+    {
+      "epoch": 0.8453494264522965,
+      "grad_norm": 1.0689058447541724,
+      "learning_rate": 1.2284647199946375e-05,
+      "loss": 1.2748,
+      "step": 18350
+    },
+    {
+      "epoch": 0.8453954945409315,
+      "grad_norm": 0.6687773583496905,
+      "learning_rate": 1.2277483016793523e-05,
+      "loss": 1.5164,
+      "step": 18351
+    },
+    {
+      "epoch": 0.8454415626295665,
+      "grad_norm": 1.6384275775905035,
+      "learning_rate": 1.2270320786632061e-05,
+      "loss": 1.0239,
+      "step": 18352
+    },
+    {
+      "epoch": 0.8454876307182015,
+      "grad_norm": 1.340687175715555,
+      "learning_rate": 1.226316050962153e-05,
+      "loss": 1.2635,
+      "step": 18353
+    },
+    {
+      "epoch": 0.8455336988068365,
+      "grad_norm": 0.8225142712940074,
+      "learning_rate": 1.2256002185921267e-05,
+      "loss": 1.4728,
+      "step": 18354
+    },
+    {
+      "epoch": 0.8455797668954715,
+      "grad_norm": 0.9670650384986605,
+      "learning_rate": 1.2248845815690679e-05,
+      "loss": 1.4362,
+      "step": 18355
+    },
+    {
+      "epoch": 0.8456258349841065,
+      "grad_norm": 1.1166551414525756,
+      "learning_rate": 1.2241691399089062e-05,
+      "loss": 1.2965,
+      "step": 18356
+    },
+    {
+      "epoch": 0.8456719030727415,
+      "grad_norm": 4.170924475858848,
+      "learning_rate": 1.223453893627574e-05,
+      "loss": 0.948,
+      "step": 18357
+    },
+    {
+      "epoch": 0.8457179711613765,
+      "grad_norm": 0.6592926242714979,
+      "learning_rate": 1.2227388427409925e-05,
+      "loss": 1.3902,
+      "step": 18358
+    },
+    {
+      "epoch": 0.8457640392500115,
+      "grad_norm": 1.368318776975953,
+      "learning_rate": 1.2220239872650763e-05,
+      "loss": 1.0998,
+      "step": 18359
+    },
+    {
+      "epoch": 0.8458101073386465,
+      "grad_norm": 0.7406746256896682,
+      "learning_rate": 1.2213093272157494e-05,
+      "loss": 1.2588,
+      "step": 18360
+    },
+    {
+      "epoch": 0.8458561754272815,
+      "grad_norm": 0.885372332089742,
+      "learning_rate": 1.2205948626089159e-05,
+      "loss": 1.28,
+      "step": 18361
+    },
+    {
+      "epoch": 0.8459022435159165,
+      "grad_norm": 0.8455266359724142,
+      "learning_rate": 1.2198805934604873e-05,
+      "loss": 1.1399,
+      "step": 18362
+    },
+    {
+      "epoch": 0.8459483116045515,
+      "grad_norm": 0.8583445342235816,
+      "learning_rate": 1.2191665197863578e-05,
+      "loss": 1.1563,
+      "step": 18363
+    },
+    {
+      "epoch": 0.8459943796931866,
+      "grad_norm": 1.4610168735024418,
+      "learning_rate": 1.2184526416024333e-05,
+      "loss": 1.1282,
+      "step": 18364
+    },
+    {
+      "epoch": 0.8460404477818215,
+      "grad_norm": 0.6867784488355567,
+      "learning_rate": 1.2177389589246025e-05,
+      "loss": 1.4093,
+      "step": 18365
+    },
+    {
+      "epoch": 0.8460865158704566,
+      "grad_norm": 0.9223061088408232,
+      "learning_rate": 1.2170254717687545e-05,
+      "loss": 0.9637,
+      "step": 18366
+    },
+    {
+      "epoch": 0.8461325839590915,
+      "grad_norm": 0.7212146767485449,
+      "learning_rate": 1.216312180150776e-05,
+      "loss": 1.1553,
+      "step": 18367
+    },
+    {
+      "epoch": 0.8461786520477266,
+      "grad_norm": 1.1173603761695088,
+      "learning_rate": 1.2155990840865473e-05,
+      "loss": 1.6064,
+      "step": 18368
+    },
+    {
+      "epoch": 0.8462247201363615,
+      "grad_norm": 0.9854174999035342,
+      "learning_rate": 1.2148861835919434e-05,
+      "loss": 1.2527,
+      "step": 18369
+    },
+    {
+      "epoch": 0.8462707882249966,
+      "grad_norm": 0.7879080608762298,
+      "learning_rate": 1.2141734786828308e-05,
+      "loss": 1.2681,
+      "step": 18370
+    },
+    {
+      "epoch": 0.8463168563136315,
+      "grad_norm": 0.8212875014815688,
+      "learning_rate": 1.2134609693750853e-05,
+      "loss": 1.2694,
+      "step": 18371
+    },
+    {
+      "epoch": 0.8463629244022666,
+      "grad_norm": 0.8594126796824478,
+      "learning_rate": 1.2127486556845635e-05,
+      "loss": 1.0568,
+      "step": 18372
+    },
+    {
+      "epoch": 0.8464089924909015,
+      "grad_norm": 1.1761125915162636,
+      "learning_rate": 1.212036537627128e-05,
+      "loss": 1.1216,
+      "step": 18373
+    },
+    {
+      "epoch": 0.8464550605795366,
+      "grad_norm": 0.8823888424437483,
+      "learning_rate": 1.2113246152186274e-05,
+      "loss": 1.2137,
+      "step": 18374
+    },
+    {
+      "epoch": 0.8465011286681715,
+      "grad_norm": 0.7910685241584251,
+      "learning_rate": 1.210612888474919e-05,
+      "loss": 1.1391,
+      "step": 18375
+    },
+    {
+      "epoch": 0.8465471967568066,
+      "grad_norm": 1.586946688971577,
+      "learning_rate": 1.2099013574118413e-05,
+      "loss": 0.9806,
+      "step": 18376
+    },
+    {
+      "epoch": 0.8465932648454416,
+      "grad_norm": 0.9468940709075894,
+      "learning_rate": 1.2091900220452412e-05,
+      "loss": 1.0679,
+      "step": 18377
+    },
+    {
+      "epoch": 0.8466393329340766,
+      "grad_norm": 1.1875993516809542,
+      "learning_rate": 1.2084788823909499e-05,
+      "loss": 1.0376,
+      "step": 18378
+    },
+    {
+      "epoch": 0.8466854010227116,
+      "grad_norm": 0.9416939947304986,
+      "learning_rate": 1.2077679384648011e-05,
+      "loss": 1.2953,
+      "step": 18379
+    },
+    {
+      "epoch": 0.8467314691113466,
+      "grad_norm": 0.9556476406873388,
+      "learning_rate": 1.2070571902826267e-05,
+      "loss": 1.2905,
+      "step": 18380
+    },
+    {
+      "epoch": 0.8467775371999816,
+      "grad_norm": 1.0344238664376537,
+      "learning_rate": 1.2063466378602429e-05,
+      "loss": 0.925,
+      "step": 18381
+    },
+    {
+      "epoch": 0.8468236052886166,
+      "grad_norm": 0.9213814962942989,
+      "learning_rate": 1.205636281213477e-05,
+      "loss": 1.2502,
+      "step": 18382
+    },
+    {
+      "epoch": 0.8468696733772516,
+      "grad_norm": 4.195306267223219,
+      "learning_rate": 1.2049261203581386e-05,
+      "loss": 1.1539,
+      "step": 18383
+    },
+    {
+      "epoch": 0.8469157414658866,
+      "grad_norm": 0.8839069195983937,
+      "learning_rate": 1.2042161553100417e-05,
+      "loss": 0.9834,
+      "step": 18384
+    },
+    {
+      "epoch": 0.8469618095545216,
+      "grad_norm": 1.138822228436878,
+      "learning_rate": 1.203506386084987e-05,
+      "loss": 1.3807,
+      "step": 18385
+    },
+    {
+      "epoch": 0.8470078776431565,
+      "grad_norm": 0.7055898854821916,
+      "learning_rate": 1.2027968126987831e-05,
+      "loss": 1.4674,
+      "step": 18386
+    },
+    {
+      "epoch": 0.8470539457317916,
+      "grad_norm": 0.7019270462074662,
+      "learning_rate": 1.2020874351672251e-05,
+      "loss": 1.3694,
+      "step": 18387
+    },
+    {
+      "epoch": 0.8471000138204265,
+      "grad_norm": 0.8125868932963718,
+      "learning_rate": 1.2013782535061025e-05,
+      "loss": 1.4302,
+      "step": 18388
+    },
+    {
+      "epoch": 0.8471460819090616,
+      "grad_norm": 2.4636614453773875,
+      "learning_rate": 1.200669267731207e-05,
+      "loss": 1.0827,
+      "step": 18389
+    },
+    {
+      "epoch": 0.8471921499976965,
+      "grad_norm": 3.834938300540425,
+      "learning_rate": 1.1999604778583217e-05,
+      "loss": 1.1227,
+      "step": 18390
+    },
+    {
+      "epoch": 0.8472382180863316,
+      "grad_norm": 1.2950652224206953,
+      "learning_rate": 1.1992518839032308e-05,
+      "loss": 1.3767,
+      "step": 18391
+    },
+    {
+      "epoch": 0.8472842861749666,
+      "grad_norm": 0.8382286420677953,
+      "learning_rate": 1.1985434858817024e-05,
+      "loss": 1.291,
+      "step": 18392
+    },
+    {
+      "epoch": 0.8473303542636016,
+      "grad_norm": 1.117112374349984,
+      "learning_rate": 1.1978352838095153e-05,
+      "loss": 1.2338,
+      "step": 18393
+    },
+    {
+      "epoch": 0.8473764223522366,
+      "grad_norm": 0.7114779999525414,
+      "learning_rate": 1.1971272777024312e-05,
+      "loss": 1.4827,
+      "step": 18394
+    },
+    {
+      "epoch": 0.8474224904408716,
+      "grad_norm": 0.9245927306913752,
+      "learning_rate": 1.1964194675762175e-05,
+      "loss": 1.3617,
+      "step": 18395
+    },
+    {
+      "epoch": 0.8474685585295066,
+      "grad_norm": 0.6285556265080787,
+      "learning_rate": 1.1957118534466272e-05,
+      "loss": 1.4364,
+      "step": 18396
+    },
+    {
+      "epoch": 0.8475146266181416,
+      "grad_norm": 1.0930139303003896,
+      "learning_rate": 1.1950044353294154e-05,
+      "loss": 1.0878,
+      "step": 18397
+    },
+    {
+      "epoch": 0.8475606947067766,
+      "grad_norm": 0.9478012825563442,
+      "learning_rate": 1.1942972132403351e-05,
+      "loss": 1.2,
+      "step": 18398
+    },
+    {
+      "epoch": 0.8476067627954116,
+      "grad_norm": 1.334083118085784,
+      "learning_rate": 1.1935901871951272e-05,
+      "loss": 1.4847,
+      "step": 18399
+    },
+    {
+      "epoch": 0.8476528308840466,
+      "grad_norm": 1.012856973495732,
+      "learning_rate": 1.1928833572095343e-05,
+      "loss": 1.2261,
+      "step": 18400
+    },
+    {
+      "epoch": 0.8476988989726816,
+      "grad_norm": 1.2244652671375345,
+      "learning_rate": 1.192176723299292e-05,
+      "loss": 0.8245,
+      "step": 18401
+    },
+    {
+      "epoch": 0.8477449670613166,
+      "grad_norm": 0.800281339773052,
+      "learning_rate": 1.191470285480134e-05,
+      "loss": 1.5029,
+      "step": 18402
+    },
+    {
+      "epoch": 0.8477910351499516,
+      "grad_norm": 0.9853852551426928,
+      "learning_rate": 1.1907640437677859e-05,
+      "loss": 1.0768,
+      "step": 18403
+    },
+    {
+      "epoch": 0.8478371032385866,
+      "grad_norm": 0.8841756410594933,
+      "learning_rate": 1.1900579981779714e-05,
+      "loss": 1.2491,
+      "step": 18404
+    },
+    {
+      "epoch": 0.8478831713272217,
+      "grad_norm": 0.9686283716696225,
+      "learning_rate": 1.1893521487264114e-05,
+      "loss": 0.8803,
+      "step": 18405
+    },
+    {
+      "epoch": 0.8479292394158566,
+      "grad_norm": 1.2705958644792938,
+      "learning_rate": 1.1886464954288168e-05,
+      "loss": 1.4056,
+      "step": 18406
+    },
+    {
+      "epoch": 0.8479753075044917,
+      "grad_norm": 0.8306040858049497,
+      "learning_rate": 1.1879410383009004e-05,
+      "loss": 1.5584,
+      "step": 18407
+    },
+    {
+      "epoch": 0.8480213755931266,
+      "grad_norm": 0.8436830613606274,
+      "learning_rate": 1.1872357773583665e-05,
+      "loss": 1.3485,
+      "step": 18408
+    },
+    {
+      "epoch": 0.8480674436817617,
+      "grad_norm": 0.9397100787535745,
+      "learning_rate": 1.18653071261692e-05,
+      "loss": 1.2268,
+      "step": 18409
+    },
+    {
+      "epoch": 0.8481135117703966,
+      "grad_norm": 0.8108556887967475,
+      "learning_rate": 1.1858258440922531e-05,
+      "loss": 1.4066,
+      "step": 18410
+    },
+    {
+      "epoch": 0.8481595798590317,
+      "grad_norm": 1.2107245947862253,
+      "learning_rate": 1.185121171800061e-05,
+      "loss": 1.1671,
+      "step": 18411
+    },
+    {
+      "epoch": 0.8482056479476666,
+      "grad_norm": 1.1433339945693908,
+      "learning_rate": 1.184416695756031e-05,
+      "loss": 1.1658,
+      "step": 18412
+    },
+    {
+      "epoch": 0.8482517160363017,
+      "grad_norm": 0.7411882191237424,
+      "learning_rate": 1.1837124159758506e-05,
+      "loss": 1.5406,
+      "step": 18413
+    },
+    {
+      "epoch": 0.8482977841249366,
+      "grad_norm": 1.211224074997606,
+      "learning_rate": 1.183008332475194e-05,
+      "loss": 1.1303,
+      "step": 18414
+    },
+    {
+      "epoch": 0.8483438522135717,
+      "grad_norm": 0.9049558393573006,
+      "learning_rate": 1.1823044452697396e-05,
+      "loss": 1.3441,
+      "step": 18415
+    },
+    {
+      "epoch": 0.8483899203022066,
+      "grad_norm": 1.0273259104321444,
+      "learning_rate": 1.1816007543751583e-05,
+      "loss": 1.261,
+      "step": 18416
+    },
+    {
+      "epoch": 0.8484359883908417,
+      "grad_norm": 1.7307289250863138,
+      "learning_rate": 1.1808972598071155e-05,
+      "loss": 1.463,
+      "step": 18417
+    },
+    {
+      "epoch": 0.8484820564794767,
+      "grad_norm": 1.0869453918633392,
+      "learning_rate": 1.1801939615812729e-05,
+      "loss": 0.9032,
+      "step": 18418
+    },
+    {
+      "epoch": 0.8485281245681117,
+      "grad_norm": 0.8394289700955597,
+      "learning_rate": 1.1794908597132893e-05,
+      "loss": 1.1629,
+      "step": 18419
+    },
+    {
+      "epoch": 0.8485741926567467,
+      "grad_norm": 0.89988767489408,
+      "learning_rate": 1.1787879542188195e-05,
+      "loss": 1.5151,
+      "step": 18420
+    },
+    {
+      "epoch": 0.8486202607453817,
+      "grad_norm": 0.9673538670764544,
+      "learning_rate": 1.1780852451135082e-05,
+      "loss": 1.3373,
+      "step": 18421
+    },
+    {
+      "epoch": 0.8486663288340167,
+      "grad_norm": 0.835558236660372,
+      "learning_rate": 1.1773827324130038e-05,
+      "loss": 1.2592,
+      "step": 18422
+    },
+    {
+      "epoch": 0.8487123969226517,
+      "grad_norm": 1.399346812984365,
+      "learning_rate": 1.176680416132947e-05,
+      "loss": 1.0719,
+      "step": 18423
+    },
+    {
+      "epoch": 0.8487584650112867,
+      "grad_norm": 1.6975319944326566,
+      "learning_rate": 1.1759782962889698e-05,
+      "loss": 1.2645,
+      "step": 18424
+    },
+    {
+      "epoch": 0.8488045330999217,
+      "grad_norm": 1.2012789466373799,
+      "learning_rate": 1.1752763728967053e-05,
+      "loss": 1.0605,
+      "step": 18425
+    },
+    {
+      "epoch": 0.8488506011885567,
+      "grad_norm": 1.1004743411039708,
+      "learning_rate": 1.174574645971782e-05,
+      "loss": 1.314,
+      "step": 18426
+    },
+    {
+      "epoch": 0.8488966692771917,
+      "grad_norm": 1.088548478996535,
+      "learning_rate": 1.1738731155298232e-05,
+      "loss": 1.2655,
+      "step": 18427
+    },
+    {
+      "epoch": 0.8489427373658267,
+      "grad_norm": 0.7778647960950705,
+      "learning_rate": 1.173171781586444e-05,
+      "loss": 1.3125,
+      "step": 18428
+    },
+    {
+      "epoch": 0.8489888054544616,
+      "grad_norm": 0.8543321197848395,
+      "learning_rate": 1.1724706441572608e-05,
+      "loss": 1.237,
+      "step": 18429
+    },
+    {
+      "epoch": 0.8490348735430967,
+      "grad_norm": 1.583133330832604,
+      "learning_rate": 1.1717697032578823e-05,
+      "loss": 1.0058,
+      "step": 18430
+    },
+    {
+      "epoch": 0.8490809416317316,
+      "grad_norm": 1.0645602580040958,
+      "learning_rate": 1.1710689589039169e-05,
+      "loss": 1.1999,
+      "step": 18431
+    },
+    {
+      "epoch": 0.8491270097203667,
+      "grad_norm": 1.428966852854669,
+      "learning_rate": 1.1703684111109603e-05,
+      "loss": 1.3591,
+      "step": 18432
+    },
+    {
+      "epoch": 0.8491730778090018,
+      "grad_norm": 3.158527753835838,
+      "learning_rate": 1.169668059894612e-05,
+      "loss": 1.0854,
+      "step": 18433
+    },
+    {
+      "epoch": 0.8492191458976367,
+      "grad_norm": 0.8633062599588397,
+      "learning_rate": 1.168967905270465e-05,
+      "loss": 1.2056,
+      "step": 18434
+    },
+    {
+      "epoch": 0.8492652139862717,
+      "grad_norm": 0.9687242646976388,
+      "learning_rate": 1.1682679472541036e-05,
+      "loss": 1.2646,
+      "step": 18435
+    },
+    {
+      "epoch": 0.8493112820749067,
+      "grad_norm": 1.0835215358397325,
+      "learning_rate": 1.167568185861113e-05,
+      "loss": 1.3109,
+      "step": 18436
+    },
+    {
+      "epoch": 0.8493573501635417,
+      "grad_norm": 0.9556429602112881,
+      "learning_rate": 1.1668686211070723e-05,
+      "loss": 1.1301,
+      "step": 18437
+    },
+    {
+      "epoch": 0.8494034182521767,
+      "grad_norm": 1.2913297118100886,
+      "learning_rate": 1.1661692530075585e-05,
+      "loss": 1.3838,
+      "step": 18438
+    },
+    {
+      "epoch": 0.8494494863408117,
+      "grad_norm": 0.8491702688709389,
+      "learning_rate": 1.1654700815781372e-05,
+      "loss": 1.22,
+      "step": 18439
+    },
+    {
+      "epoch": 0.8494955544294467,
+      "grad_norm": 4.044268335449554,
+      "learning_rate": 1.164771106834377e-05,
+      "loss": 1.1768,
+      "step": 18440
+    },
+    {
+      "epoch": 0.8495416225180817,
+      "grad_norm": 1.111506222101976,
+      "learning_rate": 1.164072328791841e-05,
+      "loss": 1.1882,
+      "step": 18441
+    },
+    {
+      "epoch": 0.8495876906067167,
+      "grad_norm": 0.9205008148926417,
+      "learning_rate": 1.1633737474660822e-05,
+      "loss": 1.1621,
+      "step": 18442
+    },
+    {
+      "epoch": 0.8496337586953517,
+      "grad_norm": 0.8668522997400518,
+      "learning_rate": 1.162675362872655e-05,
+      "loss": 1.1091,
+      "step": 18443
+    },
+    {
+      "epoch": 0.8496798267839867,
+      "grad_norm": 1.088016136907286,
+      "learning_rate": 1.161977175027108e-05,
+      "loss": 1.2233,
+      "step": 18444
+    },
+    {
+      "epoch": 0.8497258948726217,
+      "grad_norm": 1.1733357076421804,
+      "learning_rate": 1.1612791839449877e-05,
+      "loss": 1.1125,
+      "step": 18445
+    },
+    {
+      "epoch": 0.8497719629612568,
+      "grad_norm": 0.8153952670921542,
+      "learning_rate": 1.1605813896418282e-05,
+      "loss": 1.1764,
+      "step": 18446
+    },
+    {
+      "epoch": 0.8498180310498917,
+      "grad_norm": 1.1239846485453204,
+      "learning_rate": 1.1598837921331695e-05,
+      "loss": 1.2844,
+      "step": 18447
+    },
+    {
+      "epoch": 0.8498640991385268,
+      "grad_norm": 1.1306290880380416,
+      "learning_rate": 1.159186391434539e-05,
+      "loss": 1.486,
+      "step": 18448
+    },
+    {
+      "epoch": 0.8499101672271617,
+      "grad_norm": 0.7142370304267824,
+      "learning_rate": 1.1584891875614678e-05,
+      "loss": 0.9579,
+      "step": 18449
+    },
+    {
+      "epoch": 0.8499562353157968,
+      "grad_norm": 1.1266079357158683,
+      "learning_rate": 1.1577921805294745e-05,
+      "loss": 0.9497,
+      "step": 18450
+    },
+    {
+      "epoch": 0.8500023034044317,
+      "grad_norm": 0.756307999472556,
+      "learning_rate": 1.1570953703540732e-05,
+      "loss": 1.3577,
+      "step": 18451
+    },
+    {
+      "epoch": 0.8500483714930668,
+      "grad_norm": 2.0241144342033244,
+      "learning_rate": 1.156398757050785e-05,
+      "loss": 1.4131,
+      "step": 18452
+    },
+    {
+      "epoch": 0.8500944395817017,
+      "grad_norm": 1.8645997193378947,
+      "learning_rate": 1.155702340635113e-05,
+      "loss": 0.7912,
+      "step": 18453
+    },
+    {
+      "epoch": 0.8501405076703368,
+      "grad_norm": 0.9212830335489348,
+      "learning_rate": 1.1550061211225637e-05,
+      "loss": 1.1765,
+      "step": 18454
+    },
+    {
+      "epoch": 0.8501865757589717,
+      "grad_norm": 0.8037836733238741,
+      "learning_rate": 1.1543100985286371e-05,
+      "loss": 1.4061,
+      "step": 18455
+    },
+    {
+      "epoch": 0.8502326438476068,
+      "grad_norm": 0.9989523941273704,
+      "learning_rate": 1.1536142728688304e-05,
+      "loss": 1.2944,
+      "step": 18456
+    },
+    {
+      "epoch": 0.8502787119362417,
+      "grad_norm": 0.7795443736644547,
+      "learning_rate": 1.1529186441586325e-05,
+      "loss": 1.4625,
+      "step": 18457
+    },
+    {
+      "epoch": 0.8503247800248768,
+      "grad_norm": 0.8689737983516378,
+      "learning_rate": 1.1522232124135313e-05,
+      "loss": 1.4218,
+      "step": 18458
+    },
+    {
+      "epoch": 0.8503708481135118,
+      "grad_norm": 0.9651878629237554,
+      "learning_rate": 1.1515279776490118e-05,
+      "loss": 1.0027,
+      "step": 18459
+    },
+    {
+      "epoch": 0.8504169162021468,
+      "grad_norm": 0.8784823718833592,
+      "learning_rate": 1.1508329398805484e-05,
+      "loss": 1.2747,
+      "step": 18460
+    },
+    {
+      "epoch": 0.8504629842907818,
+      "grad_norm": 0.6835392751116561,
+      "learning_rate": 1.1501380991236176e-05,
+      "loss": 1.1968,
+      "step": 18461
+    },
+    {
+      "epoch": 0.8505090523794168,
+      "grad_norm": 0.9716135021292304,
+      "learning_rate": 1.149443455393684e-05,
+      "loss": 1.1338,
+      "step": 18462
+    },
+    {
+      "epoch": 0.8505551204680518,
+      "grad_norm": 0.7609456769751585,
+      "learning_rate": 1.1487490087062203e-05,
+      "loss": 1.1382,
+      "step": 18463
+    },
+    {
+      "epoch": 0.8506011885566868,
+      "grad_norm": 0.8003377886228357,
+      "learning_rate": 1.1480547590766822e-05,
+      "loss": 1.1466,
+      "step": 18464
+    },
+    {
+      "epoch": 0.8506472566453218,
+      "grad_norm": 1.1411008354484182,
+      "learning_rate": 1.1473607065205261e-05,
+      "loss": 1.3261,
+      "step": 18465
+    },
+    {
+      "epoch": 0.8506933247339568,
+      "grad_norm": 1.5240794716103174,
+      "learning_rate": 1.1466668510532052e-05,
+      "loss": 0.9405,
+      "step": 18466
+    },
+    {
+      "epoch": 0.8507393928225918,
+      "grad_norm": 0.9850054932836346,
+      "learning_rate": 1.1459731926901684e-05,
+      "loss": 1.2363,
+      "step": 18467
+    },
+    {
+      "epoch": 0.8507854609112268,
+      "grad_norm": 1.2583146034177712,
+      "learning_rate": 1.1452797314468578e-05,
+      "loss": 0.913,
+      "step": 18468
+    },
+    {
+      "epoch": 0.8508315289998618,
+      "grad_norm": 1.4423758143099585,
+      "learning_rate": 1.1445864673387063e-05,
+      "loss": 0.9622,
+      "step": 18469
+    },
+    {
+      "epoch": 0.8508775970884968,
+      "grad_norm": 0.8439445978799072,
+      "learning_rate": 1.1438934003811585e-05,
+      "loss": 1.2452,
+      "step": 18470
+    },
+    {
+      "epoch": 0.8509236651771318,
+      "grad_norm": 0.95574764203682,
+      "learning_rate": 1.1432005305896364e-05,
+      "loss": 1.1119,
+      "step": 18471
+    },
+    {
+      "epoch": 0.8509697332657667,
+      "grad_norm": 1.8975142915933767,
+      "learning_rate": 1.14250785797957e-05,
+      "loss": 1.0927,
+      "step": 18472
+    },
+    {
+      "epoch": 0.8510158013544018,
+      "grad_norm": 0.6771064711941202,
+      "learning_rate": 1.1418153825663758e-05,
+      "loss": 0.855,
+      "step": 18473
+    },
+    {
+      "epoch": 0.8510618694430369,
+      "grad_norm": 1.2934417212929754,
+      "learning_rate": 1.1411231043654768e-05,
+      "loss": 1.3239,
+      "step": 18474
+    },
+    {
+      "epoch": 0.8511079375316718,
+      "grad_norm": 1.6552709145846953,
+      "learning_rate": 1.14043102339228e-05,
+      "loss": 1.1959,
+      "step": 18475
+    },
+    {
+      "epoch": 0.8511540056203069,
+      "grad_norm": 0.8152329047837346,
+      "learning_rate": 1.139739139662198e-05,
+      "loss": 1.0132,
+      "step": 18476
+    },
+    {
+      "epoch": 0.8512000737089418,
+      "grad_norm": 1.1619605139744027,
+      "learning_rate": 1.1390474531906304e-05,
+      "loss": 0.758,
+      "step": 18477
+    },
+    {
+      "epoch": 0.8512461417975768,
+      "grad_norm": 1.5228718873996927,
+      "learning_rate": 1.1383559639929775e-05,
+      "loss": 1.274,
+      "step": 18478
+    },
+    {
+      "epoch": 0.8512922098862118,
+      "grad_norm": 1.1387860688606333,
+      "learning_rate": 1.1376646720846363e-05,
+      "loss": 0.8824,
+      "step": 18479
+    },
+    {
+      "epoch": 0.8513382779748468,
+      "grad_norm": 1.0323911820140832,
+      "learning_rate": 1.136973577480992e-05,
+      "loss": 1.1796,
+      "step": 18480
+    },
+    {
+      "epoch": 0.8513843460634818,
+      "grad_norm": 0.8472002504866247,
+      "learning_rate": 1.136282680197438e-05,
+      "loss": 1.1053,
+      "step": 18481
+    },
+    {
+      "epoch": 0.8514304141521168,
+      "grad_norm": 0.7114615054331277,
+      "learning_rate": 1.1355919802493498e-05,
+      "loss": 1.1379,
+      "step": 18482
+    },
+    {
+      "epoch": 0.8514764822407518,
+      "grad_norm": 0.8805071715224365,
+      "learning_rate": 1.1349014776521094e-05,
+      "loss": 1.4258,
+      "step": 18483
+    },
+    {
+      "epoch": 0.8515225503293868,
+      "grad_norm": 0.9704314091853459,
+      "learning_rate": 1.1342111724210835e-05,
+      "loss": 1.4579,
+      "step": 18484
+    },
+    {
+      "epoch": 0.8515686184180218,
+      "grad_norm": 0.6930427726934821,
+      "learning_rate": 1.1335210645716477e-05,
+      "loss": 1.2771,
+      "step": 18485
+    },
+    {
+      "epoch": 0.8516146865066568,
+      "grad_norm": 1.7068527249056316,
+      "learning_rate": 1.1328311541191605e-05,
+      "loss": 1.4192,
+      "step": 18486
+    },
+    {
+      "epoch": 0.8516607545952919,
+      "grad_norm": 0.7792161638372157,
+      "learning_rate": 1.1321414410789866e-05,
+      "loss": 0.9146,
+      "step": 18487
+    },
+    {
+      "epoch": 0.8517068226839268,
+      "grad_norm": 0.8652169713052709,
+      "learning_rate": 1.1314519254664757e-05,
+      "loss": 1.3057,
+      "step": 18488
+    },
+    {
+      "epoch": 0.8517528907725619,
+      "grad_norm": 1.693048152561278,
+      "learning_rate": 1.1307626072969813e-05,
+      "loss": 1.4435,
+      "step": 18489
+    },
+    {
+      "epoch": 0.8517989588611968,
+      "grad_norm": 0.9185209026130594,
+      "learning_rate": 1.130073486585852e-05,
+      "loss": 1.0585,
+      "step": 18490
+    },
+    {
+      "epoch": 0.8518450269498319,
+      "grad_norm": 0.9354961700493003,
+      "learning_rate": 1.1293845633484245e-05,
+      "loss": 1.0652,
+      "step": 18491
+    },
+    {
+      "epoch": 0.8518910950384668,
+      "grad_norm": 1.0118488306137037,
+      "learning_rate": 1.1286958376000433e-05,
+      "loss": 1.1401,
+      "step": 18492
+    },
+    {
+      "epoch": 0.8519371631271019,
+      "grad_norm": 0.8542545362359286,
+      "learning_rate": 1.1280073093560361e-05,
+      "loss": 1.2747,
+      "step": 18493
+    },
+    {
+      "epoch": 0.8519832312157368,
+      "grad_norm": 1.9582739534840141,
+      "learning_rate": 1.1273189786317362e-05,
+      "loss": 1.3235,
+      "step": 18494
+    },
+    {
+      "epoch": 0.8520292993043719,
+      "grad_norm": 1.1094287510227725,
+      "learning_rate": 1.1266308454424624e-05,
+      "loss": 1.2819,
+      "step": 18495
+    },
+    {
+      "epoch": 0.8520753673930068,
+      "grad_norm": 0.7796697811512721,
+      "learning_rate": 1.1259429098035412e-05,
+      "loss": 1.1574,
+      "step": 18496
+    },
+    {
+      "epoch": 0.8521214354816419,
+      "grad_norm": 1.5292115831095545,
+      "learning_rate": 1.1252551717302862e-05,
+      "loss": 1.5671,
+      "step": 18497
+    },
+    {
+      "epoch": 0.8521675035702768,
+      "grad_norm": 0.9519505351650117,
+      "learning_rate": 1.124567631238005e-05,
+      "loss": 1.2928,
+      "step": 18498
+    },
+    {
+      "epoch": 0.8522135716589119,
+      "grad_norm": 0.8510938890992447,
+      "learning_rate": 1.1238802883420085e-05,
+      "loss": 1.2833,
+      "step": 18499
+    },
+    {
+      "epoch": 0.8522596397475469,
+      "grad_norm": 0.874855371457064,
+      "learning_rate": 1.1231931430575971e-05,
+      "loss": 1.3079,
+      "step": 18500
+    },
+    {
+      "epoch": 0.8523057078361819,
+      "grad_norm": 0.8941106535002193,
+      "learning_rate": 1.1225061954000727e-05,
+      "loss": 1.1702,
+      "step": 18501
+    },
+    {
+      "epoch": 0.8523517759248169,
+      "grad_norm": 0.9427221972558576,
+      "learning_rate": 1.1218194453847242e-05,
+      "loss": 0.92,
+      "step": 18502
+    },
+    {
+      "epoch": 0.8523978440134519,
+      "grad_norm": 0.9354825528250323,
+      "learning_rate": 1.1211328930268417e-05,
+      "loss": 1.0051,
+      "step": 18503
+    },
+    {
+      "epoch": 0.8524439121020869,
+      "grad_norm": 1.0640756038657757,
+      "learning_rate": 1.120446538341713e-05,
+      "loss": 1.1409,
+      "step": 18504
+    },
+    {
+      "epoch": 0.8524899801907219,
+      "grad_norm": 1.5844496011823945,
+      "learning_rate": 1.1197603813446177e-05,
+      "loss": 1.2641,
+      "step": 18505
+    },
+    {
+      "epoch": 0.8525360482793569,
+      "grad_norm": 0.7820651579359118,
+      "learning_rate": 1.1190744220508299e-05,
+      "loss": 0.9574,
+      "step": 18506
+    },
+    {
+      "epoch": 0.8525821163679919,
+      "grad_norm": 0.6980830640930281,
+      "learning_rate": 1.1183886604756222e-05,
+      "loss": 1.2392,
+      "step": 18507
+    },
+    {
+      "epoch": 0.8526281844566269,
+      "grad_norm": 0.8128471313625771,
+      "learning_rate": 1.117703096634265e-05,
+      "loss": 1.2723,
+      "step": 18508
+    },
+    {
+      "epoch": 0.8526742525452619,
+      "grad_norm": 2.269215850954824,
+      "learning_rate": 1.117017730542016e-05,
+      "loss": 1.2946,
+      "step": 18509
+    },
+    {
+      "epoch": 0.8527203206338969,
+      "grad_norm": 1.009528770537553,
+      "learning_rate": 1.1163325622141351e-05,
+      "loss": 1.276,
+      "step": 18510
+    },
+    {
+      "epoch": 0.8527663887225319,
+      "grad_norm": 0.7847856260536468,
+      "learning_rate": 1.1156475916658794e-05,
+      "loss": 1.3941,
+      "step": 18511
+    },
+    {
+      "epoch": 0.8528124568111669,
+      "grad_norm": 0.8798103104001951,
+      "learning_rate": 1.1149628189124972e-05,
+      "loss": 1.3091,
+      "step": 18512
+    },
+    {
+      "epoch": 0.852858524899802,
+      "grad_norm": 1.153837255286163,
+      "learning_rate": 1.1142782439692323e-05,
+      "loss": 1.2031,
+      "step": 18513
+    },
+    {
+      "epoch": 0.8529045929884369,
+      "grad_norm": 1.5924966051606257,
+      "learning_rate": 1.1135938668513258e-05,
+      "loss": 1.1776,
+      "step": 18514
+    },
+    {
+      "epoch": 0.852950661077072,
+      "grad_norm": 1.1715736916381139,
+      "learning_rate": 1.1129096875740175e-05,
+      "loss": 1.298,
+      "step": 18515
+    },
+    {
+      "epoch": 0.8529967291657069,
+      "grad_norm": 0.8592548328152748,
+      "learning_rate": 1.1122257061525344e-05,
+      "loss": 1.2276,
+      "step": 18516
+    },
+    {
+      "epoch": 0.853042797254342,
+      "grad_norm": 1.853870792663033,
+      "learning_rate": 1.1115419226021063e-05,
+      "loss": 1.3037,
+      "step": 18517
+    },
+    {
+      "epoch": 0.8530888653429769,
+      "grad_norm": 0.751524656048468,
+      "learning_rate": 1.1108583369379577e-05,
+      "loss": 1.3788,
+      "step": 18518
+    },
+    {
+      "epoch": 0.853134933431612,
+      "grad_norm": 1.8081135056419952,
+      "learning_rate": 1.1101749491753089e-05,
+      "loss": 1.4528,
+      "step": 18519
+    },
+    {
+      "epoch": 0.8531810015202469,
+      "grad_norm": 2.207186912468498,
+      "learning_rate": 1.1094917593293685e-05,
+      "loss": 1.3072,
+      "step": 18520
+    },
+    {
+      "epoch": 0.853227069608882,
+      "grad_norm": 1.005636003935898,
+      "learning_rate": 1.1088087674153513e-05,
+      "loss": 0.9685,
+      "step": 18521
+    },
+    {
+      "epoch": 0.8532731376975169,
+      "grad_norm": 1.908018285907421,
+      "learning_rate": 1.1081259734484616e-05,
+      "loss": 1.0531,
+      "step": 18522
+    },
+    {
+      "epoch": 0.853319205786152,
+      "grad_norm": 1.4408802115696828,
+      "learning_rate": 1.1074433774439031e-05,
+      "loss": 1.0729,
+      "step": 18523
+    },
+    {
+      "epoch": 0.8533652738747869,
+      "grad_norm": 1.4172161680368132,
+      "learning_rate": 1.1067609794168676e-05,
+      "loss": 1.53,
+      "step": 18524
+    },
+    {
+      "epoch": 0.8534113419634219,
+      "grad_norm": 1.6181923683073558,
+      "learning_rate": 1.106078779382551e-05,
+      "loss": 1.4594,
+      "step": 18525
+    },
+    {
+      "epoch": 0.8534574100520569,
+      "grad_norm": 0.8720541681584588,
+      "learning_rate": 1.1053967773561424e-05,
+      "loss": 1.1091,
+      "step": 18526
+    },
+    {
+      "epoch": 0.8535034781406919,
+      "grad_norm": 0.9400201108789937,
+      "learning_rate": 1.1047149733528217e-05,
+      "loss": 1.2238,
+      "step": 18527
+    },
+    {
+      "epoch": 0.853549546229327,
+      "grad_norm": 3.953128104250402,
+      "learning_rate": 1.10403336738777e-05,
+      "loss": 1.223,
+      "step": 18528
+    },
+    {
+      "epoch": 0.8535956143179619,
+      "grad_norm": 0.7514265698383555,
+      "learning_rate": 1.103351959476161e-05,
+      "loss": 0.855,
+      "step": 18529
+    },
+    {
+      "epoch": 0.853641682406597,
+      "grad_norm": 1.1603311997911265,
+      "learning_rate": 1.102670749633169e-05,
+      "loss": 1.1202,
+      "step": 18530
+    },
+    {
+      "epoch": 0.8536877504952319,
+      "grad_norm": 0.8384612013764716,
+      "learning_rate": 1.1019897378739552e-05,
+      "loss": 1.4253,
+      "step": 18531
+    },
+    {
+      "epoch": 0.853733818583867,
+      "grad_norm": 0.9617028550804599,
+      "learning_rate": 1.101308924213683e-05,
+      "loss": 1.0235,
+      "step": 18532
+    },
+    {
+      "epoch": 0.8537798866725019,
+      "grad_norm": 1.1485756535805431,
+      "learning_rate": 1.1006283086675118e-05,
+      "loss": 1.0749,
+      "step": 18533
+    },
+    {
+      "epoch": 0.853825954761137,
+      "grad_norm": 0.8849125351363886,
+      "learning_rate": 1.09994789125059e-05,
+      "loss": 1.1909,
+      "step": 18534
+    },
+    {
+      "epoch": 0.8538720228497719,
+      "grad_norm": 0.941030042764003,
+      "learning_rate": 1.0992676719780682e-05,
+      "loss": 0.902,
+      "step": 18535
+    },
+    {
+      "epoch": 0.853918090938407,
+      "grad_norm": 1.321444671756478,
+      "learning_rate": 1.0985876508650895e-05,
+      "loss": 1.3375,
+      "step": 18536
+    },
+    {
+      "epoch": 0.8539641590270419,
+      "grad_norm": 1.274215543628614,
+      "learning_rate": 1.0979078279267973e-05,
+      "loss": 1.2242,
+      "step": 18537
+    },
+    {
+      "epoch": 0.854010227115677,
+      "grad_norm": 0.7577878665280725,
+      "learning_rate": 1.09722820317832e-05,
+      "loss": 1.4672,
+      "step": 18538
+    },
+    {
+      "epoch": 0.8540562952043119,
+      "grad_norm": 1.4578538422185114,
+      "learning_rate": 1.0965487766347926e-05,
+      "loss": 1.1681,
+      "step": 18539
+    },
+    {
+      "epoch": 0.854102363292947,
+      "grad_norm": 2.273097017649922,
+      "learning_rate": 1.0958695483113412e-05,
+      "loss": 0.9538,
+      "step": 18540
+    },
+    {
+      "epoch": 0.854148431381582,
+      "grad_norm": 3.1349885564928797,
+      "learning_rate": 1.0951905182230881e-05,
+      "loss": 1.1105,
+      "step": 18541
+    },
+    {
+      "epoch": 0.854194499470217,
+      "grad_norm": 1.0538824404651799,
+      "learning_rate": 1.094511686385148e-05,
+      "loss": 1.2516,
+      "step": 18542
+    },
+    {
+      "epoch": 0.854240567558852,
+      "grad_norm": 0.9415150630987139,
+      "learning_rate": 1.093833052812635e-05,
+      "loss": 1.491,
+      "step": 18543
+    },
+    {
+      "epoch": 0.854286635647487,
+      "grad_norm": 1.2930991326413819,
+      "learning_rate": 1.0931546175206608e-05,
+      "loss": 1.0809,
+      "step": 18544
+    },
+    {
+      "epoch": 0.854332703736122,
+      "grad_norm": 0.9541138112889223,
+      "learning_rate": 1.0924763805243255e-05,
+      "loss": 1.2029,
+      "step": 18545
+    },
+    {
+      "epoch": 0.854378771824757,
+      "grad_norm": 3.2172963924233455,
+      "learning_rate": 1.0917983418387289e-05,
+      "loss": 1.0254,
+      "step": 18546
+    },
+    {
+      "epoch": 0.854424839913392,
+      "grad_norm": 2.5885936661459197,
+      "learning_rate": 1.0911205014789693e-05,
+      "loss": 1.3703,
+      "step": 18547
+    },
+    {
+      "epoch": 0.854470908002027,
+      "grad_norm": 1.1952867829821532,
+      "learning_rate": 1.090442859460138e-05,
+      "loss": 1.3289,
+      "step": 18548
+    },
+    {
+      "epoch": 0.854516976090662,
+      "grad_norm": 3.302619300779269,
+      "learning_rate": 1.0897654157973181e-05,
+      "loss": 1.4512,
+      "step": 18549
+    },
+    {
+      "epoch": 0.854563044179297,
+      "grad_norm": 2.0877697667117054,
+      "learning_rate": 1.0890881705055933e-05,
+      "loss": 1.3913,
+      "step": 18550
+    },
+    {
+      "epoch": 0.854609112267932,
+      "grad_norm": 0.8854850464428368,
+      "learning_rate": 1.0884111236000428e-05,
+      "loss": 1.2735,
+      "step": 18551
+    },
+    {
+      "epoch": 0.854655180356567,
+      "grad_norm": 1.149072651540496,
+      "learning_rate": 1.0877342750957376e-05,
+      "loss": 1.0904,
+      "step": 18552
+    },
+    {
+      "epoch": 0.854701248445202,
+      "grad_norm": 0.8056699173208706,
+      "learning_rate": 1.0870576250077469e-05,
+      "loss": 1.4164,
+      "step": 18553
+    },
+    {
+      "epoch": 0.8547473165338371,
+      "grad_norm": 1.6812155660328247,
+      "learning_rate": 1.0863811733511353e-05,
+      "loss": 1.1145,
+      "step": 18554
+    },
+    {
+      "epoch": 0.854793384622472,
+      "grad_norm": 1.443299487585796,
+      "learning_rate": 1.0857049201409653e-05,
+      "loss": 1.1325,
+      "step": 18555
+    },
+    {
+      "epoch": 0.8548394527111071,
+      "grad_norm": 1.0713584418909032,
+      "learning_rate": 1.085028865392289e-05,
+      "loss": 1.1906,
+      "step": 18556
+    },
+    {
+      "epoch": 0.854885520799742,
+      "grad_norm": 0.983027309066266,
+      "learning_rate": 1.0843530091201592e-05,
+      "loss": 1.3342,
+      "step": 18557
+    },
+    {
+      "epoch": 0.8549315888883771,
+      "grad_norm": 0.6930034627212386,
+      "learning_rate": 1.0836773513396226e-05,
+      "loss": 1.4219,
+      "step": 18558
+    },
+    {
+      "epoch": 0.854977656977012,
+      "grad_norm": 0.7095482939018919,
+      "learning_rate": 1.0830018920657236e-05,
+      "loss": 0.9655,
+      "step": 18559
+    },
+    {
+      "epoch": 0.8550237250656471,
+      "grad_norm": 1.2635290561135735,
+      "learning_rate": 1.0823266313134983e-05,
+      "loss": 1.2447,
+      "step": 18560
+    },
+    {
+      "epoch": 0.855069793154282,
+      "grad_norm": 0.8125518175321834,
+      "learning_rate": 1.0816515690979756e-05,
+      "loss": 1.284,
+      "step": 18561
+    },
+    {
+      "epoch": 0.855115861242917,
+      "grad_norm": 0.9128728473348757,
+      "learning_rate": 1.0809767054341923e-05,
+      "loss": 0.9701,
+      "step": 18562
+    },
+    {
+      "epoch": 0.855161929331552,
+      "grad_norm": 1.032490800075365,
+      "learning_rate": 1.0803020403371688e-05,
+      "loss": 1.2526,
+      "step": 18563
+    },
+    {
+      "epoch": 0.855207997420187,
+      "grad_norm": 0.6462041162579032,
+      "learning_rate": 1.0796275738219263e-05,
+      "loss": 1.2198,
+      "step": 18564
+    },
+    {
+      "epoch": 0.855254065508822,
+      "grad_norm": 0.8379506746796365,
+      "learning_rate": 1.0789533059034806e-05,
+      "loss": 1.2294,
+      "step": 18565
+    },
+    {
+      "epoch": 0.855300133597457,
+      "grad_norm": 0.8819706286694023,
+      "learning_rate": 1.0782792365968453e-05,
+      "loss": 1.0858,
+      "step": 18566
+    },
+    {
+      "epoch": 0.855346201686092,
+      "grad_norm": 1.4768570738523392,
+      "learning_rate": 1.0776053659170226e-05,
+      "loss": 1.0148,
+      "step": 18567
+    },
+    {
+      "epoch": 0.855392269774727,
+      "grad_norm": 1.4615417272385087,
+      "learning_rate": 1.0769316938790186e-05,
+      "loss": 0.9627,
+      "step": 18568
+    },
+    {
+      "epoch": 0.8554383378633621,
+      "grad_norm": 0.7912597544534743,
+      "learning_rate": 1.0762582204978323e-05,
+      "loss": 1.219,
+      "step": 18569
+    },
+    {
+      "epoch": 0.855484405951997,
+      "grad_norm": 0.7898904574452766,
+      "learning_rate": 1.0755849457884548e-05,
+      "loss": 1.1099,
+      "step": 18570
+    },
+    {
+      "epoch": 0.8555304740406321,
+      "grad_norm": 1.5335958790635973,
+      "learning_rate": 1.0749118697658766e-05,
+      "loss": 1.17,
+      "step": 18571
+    },
+    {
+      "epoch": 0.855576542129267,
+      "grad_norm": 3.335187950763789,
+      "learning_rate": 1.0742389924450802e-05,
+      "loss": 1.118,
+      "step": 18572
+    },
+    {
+      "epoch": 0.8556226102179021,
+      "grad_norm": 0.7579574314295197,
+      "learning_rate": 1.0735663138410511e-05,
+      "loss": 1.3001,
+      "step": 18573
+    },
+    {
+      "epoch": 0.855668678306537,
+      "grad_norm": 1.6544039074924297,
+      "learning_rate": 1.072893833968761e-05,
+      "loss": 1.3745,
+      "step": 18574
+    },
+    {
+      "epoch": 0.8557147463951721,
+      "grad_norm": 0.8086035271547033,
+      "learning_rate": 1.0722215528431845e-05,
+      "loss": 1.1137,
+      "step": 18575
+    },
+    {
+      "epoch": 0.855760814483807,
+      "grad_norm": 0.7344301406357713,
+      "learning_rate": 1.0715494704792839e-05,
+      "loss": 1.2178,
+      "step": 18576
+    },
+    {
+      "epoch": 0.8558068825724421,
+      "grad_norm": 1.0538336643731703,
+      "learning_rate": 1.0708775868920285e-05,
+      "loss": 0.9463,
+      "step": 18577
+    },
+    {
+      "epoch": 0.855852950661077,
+      "grad_norm": 1.7449538983513537,
+      "learning_rate": 1.070205902096374e-05,
+      "loss": 1.215,
+      "step": 18578
+    },
+    {
+      "epoch": 0.8558990187497121,
+      "grad_norm": 0.8604983780592615,
+      "learning_rate": 1.0695344161072696e-05,
+      "loss": 1.2186,
+      "step": 18579
+    },
+    {
+      "epoch": 0.855945086838347,
+      "grad_norm": 0.735310958655395,
+      "learning_rate": 1.0688631289396734e-05,
+      "loss": 1.5997,
+      "step": 18580
+    },
+    {
+      "epoch": 0.8559911549269821,
+      "grad_norm": 1.1203520697176228,
+      "learning_rate": 1.0681920406085233e-05,
+      "loss": 1.3145,
+      "step": 18581
+    },
+    {
+      "epoch": 0.8560372230156171,
+      "grad_norm": 0.7139183645705443,
+      "learning_rate": 1.0675211511287641e-05,
+      "loss": 1.3987,
+      "step": 18582
+    },
+    {
+      "epoch": 0.8560832911042521,
+      "grad_norm": 0.8391668908200238,
+      "learning_rate": 1.0668504605153284e-05,
+      "loss": 1.0454,
+      "step": 18583
+    },
+    {
+      "epoch": 0.8561293591928871,
+      "grad_norm": 1.7091322481277085,
+      "learning_rate": 1.066179968783153e-05,
+      "loss": 1.3589,
+      "step": 18584
+    },
+    {
+      "epoch": 0.8561754272815221,
+      "grad_norm": 0.7421526292907162,
+      "learning_rate": 1.0655096759471606e-05,
+      "loss": 1.348,
+      "step": 18585
+    },
+    {
+      "epoch": 0.8562214953701571,
+      "grad_norm": 0.6622254050417498,
+      "learning_rate": 1.0648395820222778e-05,
+      "loss": 0.9587,
+      "step": 18586
+    },
+    {
+      "epoch": 0.8562675634587921,
+      "grad_norm": 0.8863379590392245,
+      "learning_rate": 1.0641696870234185e-05,
+      "loss": 1.0785,
+      "step": 18587
+    },
+    {
+      "epoch": 0.8563136315474271,
+      "grad_norm": 1.182901472434966,
+      "learning_rate": 1.0634999909655007e-05,
+      "loss": 1.6611,
+      "step": 18588
+    },
+    {
+      "epoch": 0.8563596996360621,
+      "grad_norm": 0.8364192529207131,
+      "learning_rate": 1.0628304938634336e-05,
+      "loss": 0.947,
+      "step": 18589
+    },
+    {
+      "epoch": 0.8564057677246971,
+      "grad_norm": 1.1558801576299766,
+      "learning_rate": 1.0621611957321187e-05,
+      "loss": 1.2729,
+      "step": 18590
+    },
+    {
+      "epoch": 0.8564518358133321,
+      "grad_norm": 1.112790884920199,
+      "learning_rate": 1.0614920965864638e-05,
+      "loss": 1.1528,
+      "step": 18591
+    },
+    {
+      "epoch": 0.8564979039019671,
+      "grad_norm": 0.8239378010171442,
+      "learning_rate": 1.0608231964413585e-05,
+      "loss": 1.4218,
+      "step": 18592
+    },
+    {
+      "epoch": 0.8565439719906021,
+      "grad_norm": 1.2304264082201748,
+      "learning_rate": 1.0601544953116994e-05,
+      "loss": 0.9254,
+      "step": 18593
+    },
+    {
+      "epoch": 0.8565900400792371,
+      "grad_norm": 0.9411993113065862,
+      "learning_rate": 1.059485993212368e-05,
+      "loss": 1.0101,
+      "step": 18594
+    },
+    {
+      "epoch": 0.8566361081678722,
+      "grad_norm": 2.702268119792359,
+      "learning_rate": 1.0588176901582547e-05,
+      "loss": 1.1356,
+      "step": 18595
+    },
+    {
+      "epoch": 0.8566821762565071,
+      "grad_norm": 2.079397754652966,
+      "learning_rate": 1.0581495861642355e-05,
+      "loss": 1.0963,
+      "step": 18596
+    },
+    {
+      "epoch": 0.8567282443451422,
+      "grad_norm": 1.4284090024489684,
+      "learning_rate": 1.0574816812451815e-05,
+      "loss": 1.0974,
+      "step": 18597
+    },
+    {
+      "epoch": 0.8567743124337771,
+      "grad_norm": 0.7074116823959785,
+      "learning_rate": 1.0568139754159645e-05,
+      "loss": 1.1686,
+      "step": 18598
+    },
+    {
+      "epoch": 0.8568203805224122,
+      "grad_norm": 1.0282616688774686,
+      "learning_rate": 1.0561464686914501e-05,
+      "loss": 1.086,
+      "step": 18599
+    },
+    {
+      "epoch": 0.8568664486110471,
+      "grad_norm": 1.07447625171677,
+      "learning_rate": 1.0554791610865011e-05,
+      "loss": 1.1349,
+      "step": 18600
+    },
+    {
+      "epoch": 0.8569125166996822,
+      "grad_norm": 0.8616709326586386,
+      "learning_rate": 1.0548120526159699e-05,
+      "loss": 1.1628,
+      "step": 18601
+    },
+    {
+      "epoch": 0.8569585847883171,
+      "grad_norm": 1.3475253450733222,
+      "learning_rate": 1.0541451432947113e-05,
+      "loss": 1.0213,
+      "step": 18602
+    },
+    {
+      "epoch": 0.8570046528769522,
+      "grad_norm": 0.8314541072116022,
+      "learning_rate": 1.0534784331375724e-05,
+      "loss": 1.4909,
+      "step": 18603
+    },
+    {
+      "epoch": 0.8570507209655871,
+      "grad_norm": 1.3068877347272725,
+      "learning_rate": 1.0528119221593969e-05,
+      "loss": 0.9384,
+      "step": 18604
+    },
+    {
+      "epoch": 0.8570967890542222,
+      "grad_norm": 1.0999080438089333,
+      "learning_rate": 1.0521456103750193e-05,
+      "loss": 1.3214,
+      "step": 18605
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 1.0179327513294563,
+      "learning_rate": 1.0514794977992826e-05,
+      "loss": 1.093,
+      "step": 18606
+    },
+    {
+      "epoch": 0.8571889252314921,
+      "grad_norm": 0.8677812955800348,
+      "learning_rate": 1.0508135844470101e-05,
+      "loss": 1.3987,
+      "step": 18607
+    },
+    {
+      "epoch": 0.8572349933201272,
+      "grad_norm": 0.6985855552333871,
+      "learning_rate": 1.050147870333027e-05,
+      "loss": 1.3178,
+      "step": 18608
+    },
+    {
+      "epoch": 0.8572810614087621,
+      "grad_norm": 0.8210590009019686,
+      "learning_rate": 1.0494823554721556e-05,
+      "loss": 0.9951,
+      "step": 18609
+    },
+    {
+      "epoch": 0.8573271294973972,
+      "grad_norm": 0.8104349122467479,
+      "learning_rate": 1.048817039879213e-05,
+      "loss": 1.1627,
+      "step": 18610
+    },
+    {
+      "epoch": 0.8573731975860321,
+      "grad_norm": 0.6494400379208605,
+      "learning_rate": 1.048151923569013e-05,
+      "loss": 1.0286,
+      "step": 18611
+    },
+    {
+      "epoch": 0.8574192656746672,
+      "grad_norm": 3.3529057784839673,
+      "learning_rate": 1.047487006556359e-05,
+      "loss": 1.4288,
+      "step": 18612
+    },
+    {
+      "epoch": 0.8574653337633021,
+      "grad_norm": 1.3940229083730709,
+      "learning_rate": 1.0468222888560575e-05,
+      "loss": 1.1935,
+      "step": 18613
+    },
+    {
+      "epoch": 0.8575114018519372,
+      "grad_norm": 1.1352029100593317,
+      "learning_rate": 1.0461577704829052e-05,
+      "loss": 1.4977,
+      "step": 18614
+    },
+    {
+      "epoch": 0.8575574699405721,
+      "grad_norm": 0.88330192941408,
+      "learning_rate": 1.0454934514517001e-05,
+      "loss": 1.4127,
+      "step": 18615
+    },
+    {
+      "epoch": 0.8576035380292072,
+      "grad_norm": 1.3961161622324894,
+      "learning_rate": 1.0448293317772285e-05,
+      "loss": 1.3359,
+      "step": 18616
+    },
+    {
+      "epoch": 0.8576496061178421,
+      "grad_norm": 0.7235820609258354,
+      "learning_rate": 1.044165411474276e-05,
+      "loss": 1.1561,
+      "step": 18617
+    },
+    {
+      "epoch": 0.8576956742064772,
+      "grad_norm": 0.9945748763508754,
+      "learning_rate": 1.0435016905576268e-05,
+      "loss": 1.3049,
+      "step": 18618
+    },
+    {
+      "epoch": 0.8577417422951121,
+      "grad_norm": 1.1383131950577312,
+      "learning_rate": 1.042838169042053e-05,
+      "loss": 1.333,
+      "step": 18619
+    },
+    {
+      "epoch": 0.8577878103837472,
+      "grad_norm": 0.9051733772042283,
+      "learning_rate": 1.0421748469423286e-05,
+      "loss": 1.1937,
+      "step": 18620
+    },
+    {
+      "epoch": 0.8578338784723821,
+      "grad_norm": 0.8392213989260601,
+      "learning_rate": 1.041511724273223e-05,
+      "loss": 1.3938,
+      "step": 18621
+    },
+    {
+      "epoch": 0.8578799465610172,
+      "grad_norm": 0.7372630074493274,
+      "learning_rate": 1.0408488010494998e-05,
+      "loss": 1.1329,
+      "step": 18622
+    },
+    {
+      "epoch": 0.8579260146496522,
+      "grad_norm": 1.1978834102408051,
+      "learning_rate": 1.0401860772859151e-05,
+      "loss": 1.2281,
+      "step": 18623
+    },
+    {
+      "epoch": 0.8579720827382872,
+      "grad_norm": 0.7358406250354179,
+      "learning_rate": 1.0395235529972236e-05,
+      "loss": 0.7912,
+      "step": 18624
+    },
+    {
+      "epoch": 0.8580181508269222,
+      "grad_norm": 1.1337270354774955,
+      "learning_rate": 1.0388612281981792e-05,
+      "loss": 1.2796,
+      "step": 18625
+    },
+    {
+      "epoch": 0.8580642189155572,
+      "grad_norm": 0.948345675866866,
+      "learning_rate": 1.038199102903522e-05,
+      "loss": 1.1557,
+      "step": 18626
+    },
+    {
+      "epoch": 0.8581102870041922,
+      "grad_norm": 0.6228349918075474,
+      "learning_rate": 1.0375371771279951e-05,
+      "loss": 1.5072,
+      "step": 18627
+    },
+    {
+      "epoch": 0.8581563550928272,
+      "grad_norm": 0.9128706429431218,
+      "learning_rate": 1.0368754508863365e-05,
+      "loss": 1.4128,
+      "step": 18628
+    },
+    {
+      "epoch": 0.8582024231814622,
+      "grad_norm": 1.1468523835648823,
+      "learning_rate": 1.036213924193279e-05,
+      "loss": 1.361,
+      "step": 18629
+    },
+    {
+      "epoch": 0.8582484912700972,
+      "grad_norm": 1.1473661154465427,
+      "learning_rate": 1.035552597063547e-05,
+      "loss": 1.0321,
+      "step": 18630
+    },
+    {
+      "epoch": 0.8582945593587322,
+      "grad_norm": 5.840436444730337,
+      "learning_rate": 1.034891469511866e-05,
+      "loss": 0.9604,
+      "step": 18631
+    },
+    {
+      "epoch": 0.8583406274473672,
+      "grad_norm": 0.6933813117535435,
+      "learning_rate": 1.0342305415529552e-05,
+      "loss": 1.3953,
+      "step": 18632
+    },
+    {
+      "epoch": 0.8583866955360022,
+      "grad_norm": 0.9391807088278711,
+      "learning_rate": 1.0335698132015292e-05,
+      "loss": 0.9626,
+      "step": 18633
+    },
+    {
+      "epoch": 0.8584327636246372,
+      "grad_norm": 0.964524133371621,
+      "learning_rate": 1.0329092844722953e-05,
+      "loss": 1.2397,
+      "step": 18634
+    },
+    {
+      "epoch": 0.8584788317132722,
+      "grad_norm": 0.7249107138941253,
+      "learning_rate": 1.0322489553799609e-05,
+      "loss": 1.2687,
+      "step": 18635
+    },
+    {
+      "epoch": 0.8585248998019073,
+      "grad_norm": 0.6986664867617959,
+      "learning_rate": 1.0315888259392293e-05,
+      "loss": 1.3878,
+      "step": 18636
+    },
+    {
+      "epoch": 0.8585709678905422,
+      "grad_norm": 0.7676269408537043,
+      "learning_rate": 1.0309288961647933e-05,
+      "loss": 1.3282,
+      "step": 18637
+    },
+    {
+      "epoch": 0.8586170359791773,
+      "grad_norm": 1.3922344751810778,
+      "learning_rate": 1.0302691660713459e-05,
+      "loss": 1.1804,
+      "step": 18638
+    },
+    {
+      "epoch": 0.8586631040678122,
+      "grad_norm": 0.8565029726877216,
+      "learning_rate": 1.0296096356735773e-05,
+      "loss": 1.4647,
+      "step": 18639
+    },
+    {
+      "epoch": 0.8587091721564473,
+      "grad_norm": 1.060264757466553,
+      "learning_rate": 1.0289503049861692e-05,
+      "loss": 1.0828,
+      "step": 18640
+    },
+    {
+      "epoch": 0.8587552402450822,
+      "grad_norm": 0.9222192160738538,
+      "learning_rate": 1.0282911740238e-05,
+      "loss": 1.6136,
+      "step": 18641
+    },
+    {
+      "epoch": 0.8588013083337173,
+      "grad_norm": 2.1320454529545625,
+      "learning_rate": 1.0276322428011442e-05,
+      "loss": 0.9342,
+      "step": 18642
+    },
+    {
+      "epoch": 0.8588473764223522,
+      "grad_norm": 1.0712473023162548,
+      "learning_rate": 1.0269735113328738e-05,
+      "loss": 1.2418,
+      "step": 18643
+    },
+    {
+      "epoch": 0.8588934445109873,
+      "grad_norm": 1.0000534857644323,
+      "learning_rate": 1.0263149796336513e-05,
+      "loss": 0.9179,
+      "step": 18644
+    },
+    {
+      "epoch": 0.8589395125996222,
+      "grad_norm": 1.0250887594324845,
+      "learning_rate": 1.0256566477181396e-05,
+      "loss": 1.3129,
+      "step": 18645
+    },
+    {
+      "epoch": 0.8589855806882573,
+      "grad_norm": 0.785791522538718,
+      "learning_rate": 1.0249985156009944e-05,
+      "loss": 1.0881,
+      "step": 18646
+    },
+    {
+      "epoch": 0.8590316487768922,
+      "grad_norm": 3.2626491389814127,
+      "learning_rate": 1.02434058329687e-05,
+      "loss": 0.92,
+      "step": 18647
+    },
+    {
+      "epoch": 0.8590777168655273,
+      "grad_norm": 0.8200316652514876,
+      "learning_rate": 1.023682850820411e-05,
+      "loss": 1.0463,
+      "step": 18648
+    },
+    {
+      "epoch": 0.8591237849541623,
+      "grad_norm": 0.9921467004956156,
+      "learning_rate": 1.0230253181862626e-05,
+      "loss": 1.1686,
+      "step": 18649
+    },
+    {
+      "epoch": 0.8591698530427972,
+      "grad_norm": 1.809406643439725,
+      "learning_rate": 1.0223679854090628e-05,
+      "loss": 1.2052,
+      "step": 18650
+    },
+    {
+      "epoch": 0.8592159211314323,
+      "grad_norm": 1.0323369131420044,
+      "learning_rate": 1.021710852503448e-05,
+      "loss": 1.3447,
+      "step": 18651
+    },
+    {
+      "epoch": 0.8592619892200672,
+      "grad_norm": 1.1672778196467137,
+      "learning_rate": 1.0210539194840451e-05,
+      "loss": 1.2542,
+      "step": 18652
+    },
+    {
+      "epoch": 0.8593080573087023,
+      "grad_norm": 1.1024087452771842,
+      "learning_rate": 1.0203971863654815e-05,
+      "loss": 1.0297,
+      "step": 18653
+    },
+    {
+      "epoch": 0.8593541253973372,
+      "grad_norm": 1.2368673961879575,
+      "learning_rate": 1.0197406531623788e-05,
+      "loss": 1.2844,
+      "step": 18654
+    },
+    {
+      "epoch": 0.8594001934859723,
+      "grad_norm": 0.6891109845929946,
+      "learning_rate": 1.0190843198893519e-05,
+      "loss": 1.4072,
+      "step": 18655
+    },
+    {
+      "epoch": 0.8594462615746072,
+      "grad_norm": 1.3799929700407116,
+      "learning_rate": 1.0184281865610123e-05,
+      "loss": 1.2602,
+      "step": 18656
+    },
+    {
+      "epoch": 0.8594923296632423,
+      "grad_norm": 1.7859764092868553,
+      "learning_rate": 1.0177722531919697e-05,
+      "loss": 1.1721,
+      "step": 18657
+    },
+    {
+      "epoch": 0.8595383977518772,
+      "grad_norm": 1.5436051366680545,
+      "learning_rate": 1.017116519796829e-05,
+      "loss": 1.2032,
+      "step": 18658
+    },
+    {
+      "epoch": 0.8595844658405123,
+      "grad_norm": 1.8279502593397672,
+      "learning_rate": 1.0164609863901841e-05,
+      "loss": 1.3098,
+      "step": 18659
+    },
+    {
+      "epoch": 0.8596305339291472,
+      "grad_norm": 1.0407699859685775,
+      "learning_rate": 1.0158056529866322e-05,
+      "loss": 1.0211,
+      "step": 18660
+    },
+    {
+      "epoch": 0.8596766020177823,
+      "grad_norm": 4.519675496047864,
+      "learning_rate": 1.0151505196007649e-05,
+      "loss": 1.0956,
+      "step": 18661
+    },
+    {
+      "epoch": 0.8597226701064172,
+      "grad_norm": 3.3015155081598757,
+      "learning_rate": 1.014495586247164e-05,
+      "loss": 1.311,
+      "step": 18662
+    },
+    {
+      "epoch": 0.8597687381950523,
+      "grad_norm": 0.7291138734932959,
+      "learning_rate": 1.013840852940411e-05,
+      "loss": 1.5408,
+      "step": 18663
+    },
+    {
+      "epoch": 0.8598148062836873,
+      "grad_norm": 0.6831771200974146,
+      "learning_rate": 1.0131863196950841e-05,
+      "loss": 1.5408,
+      "step": 18664
+    },
+    {
+      "epoch": 0.8598608743723223,
+      "grad_norm": 1.0552227216341474,
+      "learning_rate": 1.0125319865257566e-05,
+      "loss": 1.2308,
+      "step": 18665
+    },
+    {
+      "epoch": 0.8599069424609573,
+      "grad_norm": 0.8238521707098466,
+      "learning_rate": 1.011877853446992e-05,
+      "loss": 1.3116,
+      "step": 18666
+    },
+    {
+      "epoch": 0.8599530105495923,
+      "grad_norm": 0.8240168251438837,
+      "learning_rate": 1.0112239204733553e-05,
+      "loss": 1.3059,
+      "step": 18667
+    },
+    {
+      "epoch": 0.8599990786382273,
+      "grad_norm": 0.7356737056034899,
+      "learning_rate": 1.0105701876194062e-05,
+      "loss": 1.4122,
+      "step": 18668
+    },
+    {
+      "epoch": 0.8600451467268623,
+      "grad_norm": 1.2787711265404031,
+      "learning_rate": 1.0099166548997009e-05,
+      "loss": 1.5428,
+      "step": 18669
+    },
+    {
+      "epoch": 0.8600912148154973,
+      "grad_norm": 1.3684983339595402,
+      "learning_rate": 1.009263322328785e-05,
+      "loss": 1.4029,
+      "step": 18670
+    },
+    {
+      "epoch": 0.8601372829041323,
+      "grad_norm": 1.1636843206906733,
+      "learning_rate": 1.0086101899212031e-05,
+      "loss": 1.1374,
+      "step": 18671
+    },
+    {
+      "epoch": 0.8601833509927673,
+      "grad_norm": 0.9547099655600395,
+      "learning_rate": 1.0079572576915009e-05,
+      "loss": 1.2423,
+      "step": 18672
+    },
+    {
+      "epoch": 0.8602294190814023,
+      "grad_norm": 0.9568321122010386,
+      "learning_rate": 1.0073045256542114e-05,
+      "loss": 1.0163,
+      "step": 18673
+    },
+    {
+      "epoch": 0.8602754871700373,
+      "grad_norm": 0.6702141322986559,
+      "learning_rate": 1.0066519938238695e-05,
+      "loss": 0.8387,
+      "step": 18674
+    },
+    {
+      "epoch": 0.8603215552586723,
+      "grad_norm": 1.2753964294618532,
+      "learning_rate": 1.0059996622149958e-05,
+      "loss": 1.4123,
+      "step": 18675
+    },
+    {
+      "epoch": 0.8603676233473073,
+      "grad_norm": 1.74911681523166,
+      "learning_rate": 1.0053475308421223e-05,
+      "loss": 1.3318,
+      "step": 18676
+    },
+    {
+      "epoch": 0.8604136914359424,
+      "grad_norm": 0.8829934450641673,
+      "learning_rate": 1.0046955997197627e-05,
+      "loss": 1.212,
+      "step": 18677
+    },
+    {
+      "epoch": 0.8604597595245773,
+      "grad_norm": 0.8631949042738515,
+      "learning_rate": 1.004043868862431e-05,
+      "loss": 1.0507,
+      "step": 18678
+    },
+    {
+      "epoch": 0.8605058276132124,
+      "grad_norm": 1.1340344107960423,
+      "learning_rate": 1.00339233828464e-05,
+      "loss": 1.2765,
+      "step": 18679
+    },
+    {
+      "epoch": 0.8605518957018473,
+      "grad_norm": 0.8530639838768019,
+      "learning_rate": 1.0027410080008903e-05,
+      "loss": 1.2692,
+      "step": 18680
+    },
+    {
+      "epoch": 0.8605979637904824,
+      "grad_norm": 0.9180100855440597,
+      "learning_rate": 1.0020898780256872e-05,
+      "loss": 1.0267,
+      "step": 18681
+    },
+    {
+      "epoch": 0.8606440318791173,
+      "grad_norm": 2.0812005157031557,
+      "learning_rate": 1.0014389483735209e-05,
+      "loss": 1.1966,
+      "step": 18682
+    },
+    {
+      "epoch": 0.8606900999677524,
+      "grad_norm": 0.782630856323306,
+      "learning_rate": 1.0007882190588914e-05,
+      "loss": 1.4184,
+      "step": 18683
+    },
+    {
+      "epoch": 0.8607361680563873,
+      "grad_norm": 0.6598130569797036,
+      "learning_rate": 1.00013769009628e-05,
+      "loss": 1.419,
+      "step": 18684
+    },
+    {
+      "epoch": 0.8607822361450224,
+      "grad_norm": 0.995740589605787,
+      "learning_rate": 9.994873615001732e-06,
+      "loss": 1.1322,
+      "step": 18685
+    },
+    {
+      "epoch": 0.8608283042336573,
+      "grad_norm": 1.3553233399042857,
+      "learning_rate": 9.988372332850437e-06,
+      "loss": 1.2077,
+      "step": 18686
+    },
+    {
+      "epoch": 0.8608743723222924,
+      "grad_norm": 0.9263226340306859,
+      "learning_rate": 9.981873054653735e-06,
+      "loss": 1.3922,
+      "step": 18687
+    },
+    {
+      "epoch": 0.8609204404109273,
+      "grad_norm": 1.3151553452049738,
+      "learning_rate": 9.975375780556285e-06,
+      "loss": 1.2544,
+      "step": 18688
+    },
+    {
+      "epoch": 0.8609665084995624,
+      "grad_norm": 0.9096369634002995,
+      "learning_rate": 9.968880510702694e-06,
+      "loss": 1.1979,
+      "step": 18689
+    },
+    {
+      "epoch": 0.8610125765881974,
+      "grad_norm": 1.249943731813433,
+      "learning_rate": 9.962387245237647e-06,
+      "loss": 1.0783,
+      "step": 18690
+    },
+    {
+      "epoch": 0.8610586446768324,
+      "grad_norm": 0.9172882203109084,
+      "learning_rate": 9.955895984305641e-06,
+      "loss": 1.342,
+      "step": 18691
+    },
+    {
+      "epoch": 0.8611047127654674,
+      "grad_norm": 1.4418868380606282,
+      "learning_rate": 9.949406728051247e-06,
+      "loss": 1.2279,
+      "step": 18692
+    },
+    {
+      "epoch": 0.8611507808541023,
+      "grad_norm": 3.0551268716498106,
+      "learning_rate": 9.942919476618861e-06,
+      "loss": 1.0521,
+      "step": 18693
+    },
+    {
+      "epoch": 0.8611968489427374,
+      "grad_norm": 1.9853174911201248,
+      "learning_rate": 9.936434230153002e-06,
+      "loss": 1.0639,
+      "step": 18694
+    },
+    {
+      "epoch": 0.8612429170313723,
+      "grad_norm": 0.9674066699093662,
+      "learning_rate": 9.929950988797987e-06,
+      "loss": 1.2468,
+      "step": 18695
+    },
+    {
+      "epoch": 0.8612889851200074,
+      "grad_norm": 0.817594856849989,
+      "learning_rate": 9.9234697526982e-06,
+      "loss": 1.27,
+      "step": 18696
+    },
+    {
+      "epoch": 0.8613350532086423,
+      "grad_norm": 1.5653696750525528,
+      "learning_rate": 9.916990521997882e-06,
+      "loss": 1.0984,
+      "step": 18697
+    },
+    {
+      "epoch": 0.8613811212972774,
+      "grad_norm": 1.2540072960650732,
+      "learning_rate": 9.910513296841317e-06,
+      "loss": 1.4474,
+      "step": 18698
+    },
+    {
+      "epoch": 0.8614271893859123,
+      "grad_norm": 0.8719828499513428,
+      "learning_rate": 9.904038077372724e-06,
+      "loss": 1.1887,
+      "step": 18699
+    },
+    {
+      "epoch": 0.8614732574745474,
+      "grad_norm": 1.542917527131456,
+      "learning_rate": 9.89756486373622e-06,
+      "loss": 1.2177,
+      "step": 18700
+    },
+    {
+      "epoch": 0.8615193255631823,
+      "grad_norm": 0.8467644272051246,
+      "learning_rate": 9.891093656075944e-06,
+      "loss": 1.0697,
+      "step": 18701
+    },
+    {
+      "epoch": 0.8615653936518174,
+      "grad_norm": 0.8434729112601186,
+      "learning_rate": 9.884624454535963e-06,
+      "loss": 1.144,
+      "step": 18702
+    },
+    {
+      "epoch": 0.8616114617404523,
+      "grad_norm": 1.4569694224986276,
+      "learning_rate": 9.878157259260323e-06,
+      "loss": 1.0774,
+      "step": 18703
+    },
+    {
+      "epoch": 0.8616575298290874,
+      "grad_norm": 0.9257383088305021,
+      "learning_rate": 9.871692070392946e-06,
+      "loss": 1.2531,
+      "step": 18704
+    },
+    {
+      "epoch": 0.8617035979177224,
+      "grad_norm": 0.7420877023242715,
+      "learning_rate": 9.865228888077848e-06,
+      "loss": 1.4517,
+      "step": 18705
+    },
+    {
+      "epoch": 0.8617496660063574,
+      "grad_norm": 0.8378890949771163,
+      "learning_rate": 9.858767712458883e-06,
+      "loss": 1.3828,
+      "step": 18706
+    },
+    {
+      "epoch": 0.8617957340949924,
+      "grad_norm": 0.7420887973976387,
+      "learning_rate": 9.852308543679878e-06,
+      "loss": 1.3041,
+      "step": 18707
+    },
+    {
+      "epoch": 0.8618418021836274,
+      "grad_norm": 0.7978065769709943,
+      "learning_rate": 9.845851381884653e-06,
+      "loss": 1.4763,
+      "step": 18708
+    },
+    {
+      "epoch": 0.8618878702722624,
+      "grad_norm": 0.9756715339184555,
+      "learning_rate": 9.839396227216957e-06,
+      "loss": 0.9833,
+      "step": 18709
+    },
+    {
+      "epoch": 0.8619339383608974,
+      "grad_norm": 1.077537690783493,
+      "learning_rate": 9.832943079820534e-06,
+      "loss": 1.3474,
+      "step": 18710
+    },
+    {
+      "epoch": 0.8619800064495324,
+      "grad_norm": 0.9542311120912932,
+      "learning_rate": 9.826491939839022e-06,
+      "loss": 1.3365,
+      "step": 18711
+    },
+    {
+      "epoch": 0.8620260745381674,
+      "grad_norm": 4.342128830067666,
+      "learning_rate": 9.820042807416041e-06,
+      "loss": 1.1899,
+      "step": 18712
+    },
+    {
+      "epoch": 0.8620721426268024,
+      "grad_norm": 0.9888306534048021,
+      "learning_rate": 9.813595682695188e-06,
+      "loss": 0.9544,
+      "step": 18713
+    },
+    {
+      "epoch": 0.8621182107154374,
+      "grad_norm": 0.8806689035956096,
+      "learning_rate": 9.80715056582e-06,
+      "loss": 1.2788,
+      "step": 18714
+    },
+    {
+      "epoch": 0.8621642788040724,
+      "grad_norm": 1.475818841324923,
+      "learning_rate": 9.800707456933944e-06,
+      "loss": 1.2244,
+      "step": 18715
+    },
+    {
+      "epoch": 0.8622103468927074,
+      "grad_norm": 0.8485331814955804,
+      "learning_rate": 9.794266356180482e-06,
+      "loss": 1.2619,
+      "step": 18716
+    },
+    {
+      "epoch": 0.8622564149813424,
+      "grad_norm": 1.1447580310750882,
+      "learning_rate": 9.78782726370302e-06,
+      "loss": 1.4226,
+      "step": 18717
+    },
+    {
+      "epoch": 0.8623024830699775,
+      "grad_norm": 1.0670269687508525,
+      "learning_rate": 9.781390179644889e-06,
+      "loss": 1.4348,
+      "step": 18718
+    },
+    {
+      "epoch": 0.8623485511586124,
+      "grad_norm": 1.39278269989208,
+      "learning_rate": 9.774955104149408e-06,
+      "loss": 1.177,
+      "step": 18719
+    },
+    {
+      "epoch": 0.8623946192472475,
+      "grad_norm": 1.0901546997550262,
+      "learning_rate": 9.768522037359862e-06,
+      "loss": 1.3033,
+      "step": 18720
+    },
+    {
+      "epoch": 0.8624406873358824,
+      "grad_norm": 0.8158423809967067,
+      "learning_rate": 9.762090979419469e-06,
+      "loss": 1.3821,
+      "step": 18721
+    },
+    {
+      "epoch": 0.8624867554245175,
+      "grad_norm": 5.514364251821444,
+      "learning_rate": 9.755661930471372e-06,
+      "loss": 1.258,
+      "step": 18722
+    },
+    {
+      "epoch": 0.8625328235131524,
+      "grad_norm": 0.9460615180406046,
+      "learning_rate": 9.749234890658743e-06,
+      "loss": 1.0627,
+      "step": 18723
+    },
+    {
+      "epoch": 0.8625788916017875,
+      "grad_norm": 0.9001004131261813,
+      "learning_rate": 9.742809860124657e-06,
+      "loss": 1.5859,
+      "step": 18724
+    },
+    {
+      "epoch": 0.8626249596904224,
+      "grad_norm": 1.0921832926067834,
+      "learning_rate": 9.736386839012135e-06,
+      "loss": 1.1724,
+      "step": 18725
+    },
+    {
+      "epoch": 0.8626710277790575,
+      "grad_norm": 2.3720923975009094,
+      "learning_rate": 9.729965827464193e-06,
+      "loss": 1.1675,
+      "step": 18726
+    },
+    {
+      "epoch": 0.8627170958676924,
+      "grad_norm": 0.7476099050074527,
+      "learning_rate": 9.723546825623775e-06,
+      "loss": 1.1515,
+      "step": 18727
+    },
+    {
+      "epoch": 0.8627631639563275,
+      "grad_norm": 0.7036554580615397,
+      "learning_rate": 9.717129833633831e-06,
+      "loss": 1.2585,
+      "step": 18728
+    },
+    {
+      "epoch": 0.8628092320449624,
+      "grad_norm": 0.8264845671293473,
+      "learning_rate": 9.71071485163716e-06,
+      "loss": 1.3869,
+      "step": 18729
+    },
+    {
+      "epoch": 0.8628553001335975,
+      "grad_norm": 1.0844008871239363,
+      "learning_rate": 9.7043018797766e-06,
+      "loss": 1.3126,
+      "step": 18730
+    },
+    {
+      "epoch": 0.8629013682222325,
+      "grad_norm": 1.3562663706682416,
+      "learning_rate": 9.697890918194952e-06,
+      "loss": 1.2193,
+      "step": 18731
+    },
+    {
+      "epoch": 0.8629474363108675,
+      "grad_norm": 1.124328558203401,
+      "learning_rate": 9.691481967034942e-06,
+      "loss": 1.2007,
+      "step": 18732
+    },
+    {
+      "epoch": 0.8629935043995025,
+      "grad_norm": 0.7297845152818987,
+      "learning_rate": 9.685075026439217e-06,
+      "loss": 0.9241,
+      "step": 18733
+    },
+    {
+      "epoch": 0.8630395724881375,
+      "grad_norm": 0.9900928778588188,
+      "learning_rate": 9.678670096550435e-06,
+      "loss": 1.1173,
+      "step": 18734
+    },
+    {
+      "epoch": 0.8630856405767725,
+      "grad_norm": 1.1111876682891224,
+      "learning_rate": 9.67226717751123e-06,
+      "loss": 1.2902,
+      "step": 18735
+    },
+    {
+      "epoch": 0.8631317086654074,
+      "grad_norm": 0.8900612843582978,
+      "learning_rate": 9.665866269464086e-06,
+      "loss": 1.2653,
+      "step": 18736
+    },
+    {
+      "epoch": 0.8631777767540425,
+      "grad_norm": 1.1894394021704422,
+      "learning_rate": 9.659467372551546e-06,
+      "loss": 1.0317,
+      "step": 18737
+    },
+    {
+      "epoch": 0.8632238448426774,
+      "grad_norm": 1.712751066689196,
+      "learning_rate": 9.65307048691606e-06,
+      "loss": 1.2379,
+      "step": 18738
+    },
+    {
+      "epoch": 0.8632699129313125,
+      "grad_norm": 1.2411186847219997,
+      "learning_rate": 9.646675612700063e-06,
+      "loss": 1.4633,
+      "step": 18739
+    },
+    {
+      "epoch": 0.8633159810199474,
+      "grad_norm": 0.7407293188366085,
+      "learning_rate": 9.640282750045903e-06,
+      "loss": 1.2524,
+      "step": 18740
+    },
+    {
+      "epoch": 0.8633620491085825,
+      "grad_norm": 1.8866637661671601,
+      "learning_rate": 9.633891899095903e-06,
+      "loss": 1.1049,
+      "step": 18741
+    },
+    {
+      "epoch": 0.8634081171972174,
+      "grad_norm": 1.081906085217577,
+      "learning_rate": 9.62750305999236e-06,
+      "loss": 1.3688,
+      "step": 18742
+    },
+    {
+      "epoch": 0.8634541852858525,
+      "grad_norm": 0.8489950939414055,
+      "learning_rate": 9.621116232877525e-06,
+      "loss": 1.4909,
+      "step": 18743
+    },
+    {
+      "epoch": 0.8635002533744875,
+      "grad_norm": 1.0343311365091663,
+      "learning_rate": 9.614731417893552e-06,
+      "loss": 1.1346,
+      "step": 18744
+    },
+    {
+      "epoch": 0.8635463214631225,
+      "grad_norm": 0.852268819550718,
+      "learning_rate": 9.608348615182605e-06,
+      "loss": 1.1014,
+      "step": 18745
+    },
+    {
+      "epoch": 0.8635923895517575,
+      "grad_norm": 1.345008110721566,
+      "learning_rate": 9.601967824886815e-06,
+      "loss": 1.1696,
+      "step": 18746
+    },
+    {
+      "epoch": 0.8636384576403925,
+      "grad_norm": 0.9243710748138726,
+      "learning_rate": 9.59558904714818e-06,
+      "loss": 1.3621,
+      "step": 18747
+    },
+    {
+      "epoch": 0.8636845257290275,
+      "grad_norm": 1.3399664863370184,
+      "learning_rate": 9.58921228210875e-06,
+      "loss": 1.2093,
+      "step": 18748
+    },
+    {
+      "epoch": 0.8637305938176625,
+      "grad_norm": 0.9946624380577942,
+      "learning_rate": 9.582837529910494e-06,
+      "loss": 0.9891,
+      "step": 18749
+    },
+    {
+      "epoch": 0.8637766619062975,
+      "grad_norm": 1.105219593847067,
+      "learning_rate": 9.576464790695338e-06,
+      "loss": 1.3141,
+      "step": 18750
+    },
+    {
+      "epoch": 0.8638227299949325,
+      "grad_norm": 1.0172812019385196,
+      "learning_rate": 9.570094064605128e-06,
+      "loss": 1.4772,
+      "step": 18751
+    },
+    {
+      "epoch": 0.8638687980835675,
+      "grad_norm": 1.3993893146542506,
+      "learning_rate": 9.563725351781727e-06,
+      "loss": 1.3191,
+      "step": 18752
+    },
+    {
+      "epoch": 0.8639148661722025,
+      "grad_norm": 1.0307494805970254,
+      "learning_rate": 9.557358652366932e-06,
+      "loss": 1.2878,
+      "step": 18753
+    },
+    {
+      "epoch": 0.8639609342608375,
+      "grad_norm": 1.2844549913519492,
+      "learning_rate": 9.550993966502453e-06,
+      "loss": 1.0176,
+      "step": 18754
+    },
+    {
+      "epoch": 0.8640070023494725,
+      "grad_norm": 1.0124084585049795,
+      "learning_rate": 9.544631294330009e-06,
+      "loss": 1.3341,
+      "step": 18755
+    },
+    {
+      "epoch": 0.8640530704381075,
+      "grad_norm": 0.7896938892131947,
+      "learning_rate": 9.538270635991254e-06,
+      "loss": 1.2478,
+      "step": 18756
+    },
+    {
+      "epoch": 0.8640991385267425,
+      "grad_norm": 0.8669206752092475,
+      "learning_rate": 9.531911991627807e-06,
+      "loss": 1.1039,
+      "step": 18757
+    },
+    {
+      "epoch": 0.8641452066153775,
+      "grad_norm": 1.2314071449449318,
+      "learning_rate": 9.5255553613812e-06,
+      "loss": 1.4238,
+      "step": 18758
+    },
+    {
+      "epoch": 0.8641912747040126,
+      "grad_norm": 2.5687002183111307,
+      "learning_rate": 9.519200745392975e-06,
+      "loss": 1.029,
+      "step": 18759
+    },
+    {
+      "epoch": 0.8642373427926475,
+      "grad_norm": 0.853417627508044,
+      "learning_rate": 9.51284814380461e-06,
+      "loss": 1.4751,
+      "step": 18760
+    },
+    {
+      "epoch": 0.8642834108812826,
+      "grad_norm": 0.8746317753673799,
+      "learning_rate": 9.506497556757543e-06,
+      "loss": 1.133,
+      "step": 18761
+    },
+    {
+      "epoch": 0.8643294789699175,
+      "grad_norm": 1.2503884899794364,
+      "learning_rate": 9.50014898439312e-06,
+      "loss": 1.0508,
+      "step": 18762
+    },
+    {
+      "epoch": 0.8643755470585526,
+      "grad_norm": 1.3451211675671257,
+      "learning_rate": 9.493802426852715e-06,
+      "loss": 1.2615,
+      "step": 18763
+    },
+    {
+      "epoch": 0.8644216151471875,
+      "grad_norm": 1.1531759793058867,
+      "learning_rate": 9.487457884277629e-06,
+      "loss": 1.0352,
+      "step": 18764
+    },
+    {
+      "epoch": 0.8644676832358226,
+      "grad_norm": 0.7773992051792273,
+      "learning_rate": 9.481115356809079e-06,
+      "loss": 1.2782,
+      "step": 18765
+    },
+    {
+      "epoch": 0.8645137513244575,
+      "grad_norm": 0.970268036676458,
+      "learning_rate": 9.474774844588285e-06,
+      "loss": 1.3282,
+      "step": 18766
+    },
+    {
+      "epoch": 0.8645598194130926,
+      "grad_norm": 2.7489975170887244,
+      "learning_rate": 9.468436347756414e-06,
+      "loss": 1.0397,
+      "step": 18767
+    },
+    {
+      "epoch": 0.8646058875017275,
+      "grad_norm": 1.6997662867201744,
+      "learning_rate": 9.462099866454599e-06,
+      "loss": 1.3048,
+      "step": 18768
+    },
+    {
+      "epoch": 0.8646519555903626,
+      "grad_norm": 1.002165651009242,
+      "learning_rate": 9.45576540082388e-06,
+      "loss": 1.3456,
+      "step": 18769
+    },
+    {
+      "epoch": 0.8646980236789975,
+      "grad_norm": 1.3160145835016268,
+      "learning_rate": 9.449432951005288e-06,
+      "loss": 1.1961,
+      "step": 18770
+    },
+    {
+      "epoch": 0.8647440917676326,
+      "grad_norm": 1.1535560643348166,
+      "learning_rate": 9.443102517139823e-06,
+      "loss": 1.0665,
+      "step": 18771
+    },
+    {
+      "epoch": 0.8647901598562676,
+      "grad_norm": 1.1528336561551566,
+      "learning_rate": 9.436774099368394e-06,
+      "loss": 1.1849,
+      "step": 18772
+    },
+    {
+      "epoch": 0.8648362279449026,
+      "grad_norm": 0.8018179182736628,
+      "learning_rate": 9.430447697831923e-06,
+      "loss": 1.312,
+      "step": 18773
+    },
+    {
+      "epoch": 0.8648822960335376,
+      "grad_norm": 1.0081441766796253,
+      "learning_rate": 9.424123312671218e-06,
+      "loss": 1.2907,
+      "step": 18774
+    },
+    {
+      "epoch": 0.8649283641221726,
+      "grad_norm": 1.7635382400464379,
+      "learning_rate": 9.417800944027122e-06,
+      "loss": 0.9669,
+      "step": 18775
+    },
+    {
+      "epoch": 0.8649744322108076,
+      "grad_norm": 0.8527609374608239,
+      "learning_rate": 9.411480592040367e-06,
+      "loss": 1.271,
+      "step": 18776
+    },
+    {
+      "epoch": 0.8650205002994426,
+      "grad_norm": 0.9352211636357272,
+      "learning_rate": 9.405162256851662e-06,
+      "loss": 1.1628,
+      "step": 18777
+    },
+    {
+      "epoch": 0.8650665683880776,
+      "grad_norm": 0.9439894331024994,
+      "learning_rate": 9.398845938601687e-06,
+      "loss": 1.4776,
+      "step": 18778
+    },
+    {
+      "epoch": 0.8651126364767125,
+      "grad_norm": 1.1240317411366025,
+      "learning_rate": 9.392531637431079e-06,
+      "loss": 1.2142,
+      "step": 18779
+    },
+    {
+      "epoch": 0.8651587045653476,
+      "grad_norm": 1.6638583232794413,
+      "learning_rate": 9.386219353480397e-06,
+      "loss": 1.195,
+      "step": 18780
+    },
+    {
+      "epoch": 0.8652047726539825,
+      "grad_norm": 1.91517464638842,
+      "learning_rate": 9.379909086890126e-06,
+      "loss": 1.3761,
+      "step": 18781
+    },
+    {
+      "epoch": 0.8652508407426176,
+      "grad_norm": 1.1093776745730672,
+      "learning_rate": 9.373600837800845e-06,
+      "loss": 1.1897,
+      "step": 18782
+    },
+    {
+      "epoch": 0.8652969088312525,
+      "grad_norm": 0.7882961793794984,
+      "learning_rate": 9.367294606352927e-06,
+      "loss": 1.5225,
+      "step": 18783
+    },
+    {
+      "epoch": 0.8653429769198876,
+      "grad_norm": 0.9768248620603034,
+      "learning_rate": 9.360990392686809e-06,
+      "loss": 1.0769,
+      "step": 18784
+    },
+    {
+      "epoch": 0.8653890450085227,
+      "grad_norm": 1.8818320385664338,
+      "learning_rate": 9.354688196942796e-06,
+      "loss": 1.1252,
+      "step": 18785
+    },
+    {
+      "epoch": 0.8654351130971576,
+      "grad_norm": 3.0233751768866455,
+      "learning_rate": 9.348388019261267e-06,
+      "loss": 1.2829,
+      "step": 18786
+    },
+    {
+      "epoch": 0.8654811811857926,
+      "grad_norm": 0.8606203958910136,
+      "learning_rate": 9.342089859782421e-06,
+      "loss": 1.1147,
+      "step": 18787
+    },
+    {
+      "epoch": 0.8655272492744276,
+      "grad_norm": 0.952436281350534,
+      "learning_rate": 9.33579371864649e-06,
+      "loss": 1.1901,
+      "step": 18788
+    },
+    {
+      "epoch": 0.8655733173630626,
+      "grad_norm": 1.414313097886476,
+      "learning_rate": 9.329499595993685e-06,
+      "loss": 1.0896,
+      "step": 18789
+    },
+    {
+      "epoch": 0.8656193854516976,
+      "grad_norm": 1.749845880616534,
+      "learning_rate": 9.323207491964091e-06,
+      "loss": 1.0577,
+      "step": 18790
+    },
+    {
+      "epoch": 0.8656654535403326,
+      "grad_norm": 0.9242099708423545,
+      "learning_rate": 9.316917406697812e-06,
+      "loss": 0.906,
+      "step": 18791
+    },
+    {
+      "epoch": 0.8657115216289676,
+      "grad_norm": 1.0781401394102106,
+      "learning_rate": 9.310629340334853e-06,
+      "loss": 1.2567,
+      "step": 18792
+    },
+    {
+      "epoch": 0.8657575897176026,
+      "grad_norm": 0.9656553704277373,
+      "learning_rate": 9.30434329301526e-06,
+      "loss": 1.4042,
+      "step": 18793
+    },
+    {
+      "epoch": 0.8658036578062376,
+      "grad_norm": 1.0128336797058206,
+      "learning_rate": 9.298059264878944e-06,
+      "loss": 0.9403,
+      "step": 18794
+    },
+    {
+      "epoch": 0.8658497258948726,
+      "grad_norm": 0.867815131414964,
+      "learning_rate": 9.291777256065826e-06,
+      "loss": 1.2442,
+      "step": 18795
+    },
+    {
+      "epoch": 0.8658957939835076,
+      "grad_norm": 2.026078980221683,
+      "learning_rate": 9.285497266715726e-06,
+      "loss": 1.0981,
+      "step": 18796
+    },
+    {
+      "epoch": 0.8659418620721426,
+      "grad_norm": 1.1569837270666556,
+      "learning_rate": 9.279219296968523e-06,
+      "loss": 1.2476,
+      "step": 18797
+    },
+    {
+      "epoch": 0.8659879301607776,
+      "grad_norm": 2.3649389111230095,
+      "learning_rate": 9.272943346963958e-06,
+      "loss": 1.2216,
+      "step": 18798
+    },
+    {
+      "epoch": 0.8660339982494126,
+      "grad_norm": 0.8257668823439058,
+      "learning_rate": 9.26666941684171e-06,
+      "loss": 1.2062,
+      "step": 18799
+    },
+    {
+      "epoch": 0.8660800663380477,
+      "grad_norm": 0.7712199151997057,
+      "learning_rate": 9.260397506741503e-06,
+      "loss": 0.9691,
+      "step": 18800
+    },
+    {
+      "epoch": 0.8661261344266826,
+      "grad_norm": 0.8554812293212886,
+      "learning_rate": 9.254127616802965e-06,
+      "loss": 1.2148,
+      "step": 18801
+    },
+    {
+      "epoch": 0.8661722025153177,
+      "grad_norm": 0.6938238976872468,
+      "learning_rate": 9.247859747165699e-06,
+      "loss": 1.1803,
+      "step": 18802
+    },
+    {
+      "epoch": 0.8662182706039526,
+      "grad_norm": 4.386194421169327,
+      "learning_rate": 9.241593897969181e-06,
+      "loss": 1.3322,
+      "step": 18803
+    },
+    {
+      "epoch": 0.8662643386925877,
+      "grad_norm": 1.4251933393853726,
+      "learning_rate": 9.235330069353e-06,
+      "loss": 0.9353,
+      "step": 18804
+    },
+    {
+      "epoch": 0.8663104067812226,
+      "grad_norm": 6.398172083886999,
+      "learning_rate": 9.229068261456552e-06,
+      "loss": 1.0292,
+      "step": 18805
+    },
+    {
+      "epoch": 0.8663564748698577,
+      "grad_norm": 0.973240569244764,
+      "learning_rate": 9.222808474419275e-06,
+      "loss": 1.2164,
+      "step": 18806
+    },
+    {
+      "epoch": 0.8664025429584926,
+      "grad_norm": 1.2730479190430544,
+      "learning_rate": 9.2165507083805e-06,
+      "loss": 1.0505,
+      "step": 18807
+    },
+    {
+      "epoch": 0.8664486110471277,
+      "grad_norm": 0.9699718877713531,
+      "learning_rate": 9.21029496347956e-06,
+      "loss": 1.1281,
+      "step": 18808
+    },
+    {
+      "epoch": 0.8664946791357626,
+      "grad_norm": 1.0463499178107194,
+      "learning_rate": 9.204041239855754e-06,
+      "loss": 1.389,
+      "step": 18809
+    },
+    {
+      "epoch": 0.8665407472243977,
+      "grad_norm": 0.9345653256429336,
+      "learning_rate": 9.197789537648271e-06,
+      "loss": 1.0065,
+      "step": 18810
+    },
+    {
+      "epoch": 0.8665868153130326,
+      "grad_norm": 2.9158789041989244,
+      "learning_rate": 9.191539856996311e-06,
+      "loss": 1.2477,
+      "step": 18811
+    },
+    {
+      "epoch": 0.8666328834016677,
+      "grad_norm": 1.1269029009965656,
+      "learning_rate": 9.185292198039008e-06,
+      "loss": 1.4189,
+      "step": 18812
+    },
+    {
+      "epoch": 0.8666789514903027,
+      "grad_norm": 0.8161153662244519,
+      "learning_rate": 9.179046560915483e-06,
+      "loss": 1.4296,
+      "step": 18813
+    },
+    {
+      "epoch": 0.8667250195789377,
+      "grad_norm": 1.0910064921882212,
+      "learning_rate": 9.172802945764736e-06,
+      "loss": 1.2337,
+      "step": 18814
+    },
+    {
+      "epoch": 0.8667710876675727,
+      "grad_norm": 1.42651837882452,
+      "learning_rate": 9.166561352725812e-06,
+      "loss": 1.0487,
+      "step": 18815
+    },
+    {
+      "epoch": 0.8668171557562077,
+      "grad_norm": 1.10522703261024,
+      "learning_rate": 9.160321781937652e-06,
+      "loss": 1.2052,
+      "step": 18816
+    },
+    {
+      "epoch": 0.8668632238448427,
+      "grad_norm": 0.8051143150511891,
+      "learning_rate": 9.154084233539173e-06,
+      "loss": 0.921,
+      "step": 18817
+    },
+    {
+      "epoch": 0.8669092919334777,
+      "grad_norm": 0.9741347075875955,
+      "learning_rate": 9.147848707669226e-06,
+      "loss": 1.0589,
+      "step": 18818
+    },
+    {
+      "epoch": 0.8669553600221127,
+      "grad_norm": 5.560949140038307,
+      "learning_rate": 9.141615204466659e-06,
+      "loss": 1.1834,
+      "step": 18819
+    },
+    {
+      "epoch": 0.8670014281107477,
+      "grad_norm": 0.8146292971534851,
+      "learning_rate": 9.135383724070256e-06,
+      "loss": 1.447,
+      "step": 18820
+    },
+    {
+      "epoch": 0.8670474961993827,
+      "grad_norm": 0.9867382340957715,
+      "learning_rate": 9.12915426661871e-06,
+      "loss": 1.2336,
+      "step": 18821
+    },
+    {
+      "epoch": 0.8670935642880176,
+      "grad_norm": 0.9918048513529921,
+      "learning_rate": 9.122926832250745e-06,
+      "loss": 1.0314,
+      "step": 18822
+    },
+    {
+      "epoch": 0.8671396323766527,
+      "grad_norm": 0.9484744242590052,
+      "learning_rate": 9.116701421104979e-06,
+      "loss": 1.2226,
+      "step": 18823
+    },
+    {
+      "epoch": 0.8671857004652876,
+      "grad_norm": 0.6835265340607012,
+      "learning_rate": 9.110478033320059e-06,
+      "loss": 1.3783,
+      "step": 18824
+    },
+    {
+      "epoch": 0.8672317685539227,
+      "grad_norm": 1.3284017807465454,
+      "learning_rate": 9.104256669034472e-06,
+      "loss": 1.324,
+      "step": 18825
+    },
+    {
+      "epoch": 0.8672778366425578,
+      "grad_norm": 2.1695594242828804,
+      "learning_rate": 9.098037328386765e-06,
+      "loss": 1.3942,
+      "step": 18826
+    },
+    {
+      "epoch": 0.8673239047311927,
+      "grad_norm": 1.3514581914769839,
+      "learning_rate": 9.091820011515418e-06,
+      "loss": 1.1601,
+      "step": 18827
+    },
+    {
+      "epoch": 0.8673699728198278,
+      "grad_norm": 0.8729022502966405,
+      "learning_rate": 9.085604718558794e-06,
+      "loss": 0.9813,
+      "step": 18828
+    },
+    {
+      "epoch": 0.8674160409084627,
+      "grad_norm": 0.773761161003829,
+      "learning_rate": 9.079391449655294e-06,
+      "loss": 1.1896,
+      "step": 18829
+    },
+    {
+      "epoch": 0.8674621089970977,
+      "grad_norm": 1.1964629609012574,
+      "learning_rate": 9.073180204943254e-06,
+      "loss": 1.0038,
+      "step": 18830
+    },
+    {
+      "epoch": 0.8675081770857327,
+      "grad_norm": 0.8001648333884216,
+      "learning_rate": 9.066970984560973e-06,
+      "loss": 1.2064,
+      "step": 18831
+    },
+    {
+      "epoch": 0.8675542451743677,
+      "grad_norm": 3.0493582543589537,
+      "learning_rate": 9.060763788646643e-06,
+      "loss": 0.9912,
+      "step": 18832
+    },
+    {
+      "epoch": 0.8676003132630027,
+      "grad_norm": 0.9996610256700603,
+      "learning_rate": 9.054558617338483e-06,
+      "loss": 1.1993,
+      "step": 18833
+    },
+    {
+      "epoch": 0.8676463813516377,
+      "grad_norm": 1.8329354696148168,
+      "learning_rate": 9.04835547077465e-06,
+      "loss": 1.4287,
+      "step": 18834
+    },
+    {
+      "epoch": 0.8676924494402727,
+      "grad_norm": 0.9591444700184506,
+      "learning_rate": 9.042154349093223e-06,
+      "loss": 1.2534,
+      "step": 18835
+    },
+    {
+      "epoch": 0.8677385175289077,
+      "grad_norm": 0.7609522278543183,
+      "learning_rate": 9.03595525243227e-06,
+      "loss": 1.3831,
+      "step": 18836
+    },
+    {
+      "epoch": 0.8677845856175427,
+      "grad_norm": 1.1763075695436793,
+      "learning_rate": 9.029758180929793e-06,
+      "loss": 1.3272,
+      "step": 18837
+    },
+    {
+      "epoch": 0.8678306537061777,
+      "grad_norm": 1.0959505274947545,
+      "learning_rate": 9.0235631347238e-06,
+      "loss": 1.267,
+      "step": 18838
+    },
+    {
+      "epoch": 0.8678767217948127,
+      "grad_norm": 0.962384484236217,
+      "learning_rate": 9.017370113952161e-06,
+      "loss": 0.931,
+      "step": 18839
+    },
+    {
+      "epoch": 0.8679227898834477,
+      "grad_norm": 0.8818999723278187,
+      "learning_rate": 9.011179118752766e-06,
+      "loss": 1.1201,
+      "step": 18840
+    },
+    {
+      "epoch": 0.8679688579720828,
+      "grad_norm": 0.6610711728351424,
+      "learning_rate": 9.00499014926347e-06,
+      "loss": 1.277,
+      "step": 18841
+    },
+    {
+      "epoch": 0.8680149260607177,
+      "grad_norm": 0.7934590046411262,
+      "learning_rate": 8.998803205622053e-06,
+      "loss": 1.2147,
+      "step": 18842
+    },
+    {
+      "epoch": 0.8680609941493528,
+      "grad_norm": 1.188357856450213,
+      "learning_rate": 8.992618287966237e-06,
+      "loss": 1.3079,
+      "step": 18843
+    },
+    {
+      "epoch": 0.8681070622379877,
+      "grad_norm": 1.7669682451983058,
+      "learning_rate": 8.986435396433723e-06,
+      "loss": 1.2667,
+      "step": 18844
+    },
+    {
+      "epoch": 0.8681531303266228,
+      "grad_norm": 0.9511718826935102,
+      "learning_rate": 8.980254531162191e-06,
+      "loss": 1.0521,
+      "step": 18845
+    },
+    {
+      "epoch": 0.8681991984152577,
+      "grad_norm": 1.0020310922794426,
+      "learning_rate": 8.974075692289207e-06,
+      "loss": 1.488,
+      "step": 18846
+    },
+    {
+      "epoch": 0.8682452665038928,
+      "grad_norm": 1.4014360729179176,
+      "learning_rate": 8.967898879952363e-06,
+      "loss": 1.421,
+      "step": 18847
+    },
+    {
+      "epoch": 0.8682913345925277,
+      "grad_norm": 0.9476750227254658,
+      "learning_rate": 8.961724094289148e-06,
+      "loss": 0.9891,
+      "step": 18848
+    },
+    {
+      "epoch": 0.8683374026811628,
+      "grad_norm": 1.065311845825306,
+      "learning_rate": 8.955551335437074e-06,
+      "loss": 1.1839,
+      "step": 18849
+    },
+    {
+      "epoch": 0.8683834707697977,
+      "grad_norm": 1.0789872915059593,
+      "learning_rate": 8.949380603533531e-06,
+      "loss": 1.1134,
+      "step": 18850
+    },
+    {
+      "epoch": 0.8684295388584328,
+      "grad_norm": 2.691042823364398,
+      "learning_rate": 8.943211898715897e-06,
+      "loss": 1.1469,
+      "step": 18851
+    },
+    {
+      "epoch": 0.8684756069470677,
+      "grad_norm": 0.8800337825819682,
+      "learning_rate": 8.937045221121531e-06,
+      "loss": 1.2736,
+      "step": 18852
+    },
+    {
+      "epoch": 0.8685216750357028,
+      "grad_norm": 1.3155853746183452,
+      "learning_rate": 8.930880570887735e-06,
+      "loss": 1.5594,
+      "step": 18853
+    },
+    {
+      "epoch": 0.8685677431243378,
+      "grad_norm": 0.9435329333476893,
+      "learning_rate": 8.924717948151706e-06,
+      "loss": 1.0782,
+      "step": 18854
+    },
+    {
+      "epoch": 0.8686138112129728,
+      "grad_norm": 0.7998657385829809,
+      "learning_rate": 8.918557353050671e-06,
+      "loss": 1.2587,
+      "step": 18855
+    },
+    {
+      "epoch": 0.8686598793016078,
+      "grad_norm": 0.8517132478706226,
+      "learning_rate": 8.91239878572181e-06,
+      "loss": 1.0951,
+      "step": 18856
+    },
+    {
+      "epoch": 0.8687059473902428,
+      "grad_norm": 0.9691689297725198,
+      "learning_rate": 8.906242246302188e-06,
+      "loss": 1.1271,
+      "step": 18857
+    },
+    {
+      "epoch": 0.8687520154788778,
+      "grad_norm": 1.629756152817429,
+      "learning_rate": 8.900087734928886e-06,
+      "loss": 1.3532,
+      "step": 18858
+    },
+    {
+      "epoch": 0.8687980835675128,
+      "grad_norm": 0.9292084308107081,
+      "learning_rate": 8.893935251738928e-06,
+      "loss": 1.4995,
+      "step": 18859
+    },
+    {
+      "epoch": 0.8688441516561478,
+      "grad_norm": 1.0370189775836294,
+      "learning_rate": 8.887784796869314e-06,
+      "loss": 1.2122,
+      "step": 18860
+    },
+    {
+      "epoch": 0.8688902197447828,
+      "grad_norm": 0.7841464959375193,
+      "learning_rate": 8.881636370456914e-06,
+      "loss": 0.9911,
+      "step": 18861
+    },
+    {
+      "epoch": 0.8689362878334178,
+      "grad_norm": 0.9346628977466205,
+      "learning_rate": 8.875489972638662e-06,
+      "loss": 1.121,
+      "step": 18862
+    },
+    {
+      "epoch": 0.8689823559220528,
+      "grad_norm": 1.5342096173106836,
+      "learning_rate": 8.869345603551393e-06,
+      "loss": 1.224,
+      "step": 18863
+    },
+    {
+      "epoch": 0.8690284240106878,
+      "grad_norm": 0.8617295274964327,
+      "learning_rate": 8.863203263331866e-06,
+      "loss": 1.3639,
+      "step": 18864
+    },
+    {
+      "epoch": 0.8690744920993227,
+      "grad_norm": 5.381315972818255,
+      "learning_rate": 8.857062952116857e-06,
+      "loss": 1.1211,
+      "step": 18865
+    },
+    {
+      "epoch": 0.8691205601879578,
+      "grad_norm": 1.11283740341396,
+      "learning_rate": 8.850924670043071e-06,
+      "loss": 1.2798,
+      "step": 18866
+    },
+    {
+      "epoch": 0.8691666282765929,
+      "grad_norm": 1.2854323090825868,
+      "learning_rate": 8.844788417247164e-06,
+      "loss": 1.15,
+      "step": 18867
+    },
+    {
+      "epoch": 0.8692126963652278,
+      "grad_norm": 0.9768819000681467,
+      "learning_rate": 8.838654193865736e-06,
+      "loss": 1.084,
+      "step": 18868
+    },
+    {
+      "epoch": 0.8692587644538629,
+      "grad_norm": 2.5341474206244645,
+      "learning_rate": 8.83252200003536e-06,
+      "loss": 1.2317,
+      "step": 18869
+    },
+    {
+      "epoch": 0.8693048325424978,
+      "grad_norm": 1.1485326176494617,
+      "learning_rate": 8.826391835892566e-06,
+      "loss": 1.0789,
+      "step": 18870
+    },
+    {
+      "epoch": 0.8693509006311329,
+      "grad_norm": 1.1282939828321263,
+      "learning_rate": 8.82026370157385e-06,
+      "loss": 1.1879,
+      "step": 18871
+    },
+    {
+      "epoch": 0.8693969687197678,
+      "grad_norm": 1.9219431720523557,
+      "learning_rate": 8.814137597215621e-06,
+      "loss": 1.4059,
+      "step": 18872
+    },
+    {
+      "epoch": 0.8694430368084028,
+      "grad_norm": 1.4944733282183154,
+      "learning_rate": 8.808013522954228e-06,
+      "loss": 1.1893,
+      "step": 18873
+    },
+    {
+      "epoch": 0.8694891048970378,
+      "grad_norm": 0.8249311450117736,
+      "learning_rate": 8.801891478926094e-06,
+      "loss": 1.4932,
+      "step": 18874
+    },
+    {
+      "epoch": 0.8695351729856728,
+      "grad_norm": 0.831041800984817,
+      "learning_rate": 8.795771465267455e-06,
+      "loss": 1.1414,
+      "step": 18875
+    },
+    {
+      "epoch": 0.8695812410743078,
+      "grad_norm": 0.9726658410499756,
+      "learning_rate": 8.789653482114578e-06,
+      "loss": 1.4774,
+      "step": 18876
+    },
+    {
+      "epoch": 0.8696273091629428,
+      "grad_norm": 0.9439377312599319,
+      "learning_rate": 8.78353752960368e-06,
+      "loss": 1.432,
+      "step": 18877
+    },
+    {
+      "epoch": 0.8696733772515778,
+      "grad_norm": 0.8868375265172957,
+      "learning_rate": 8.777423607870928e-06,
+      "loss": 1.0616,
+      "step": 18878
+    },
+    {
+      "epoch": 0.8697194453402128,
+      "grad_norm": 1.3537478223310557,
+      "learning_rate": 8.771311717052411e-06,
+      "loss": 1.2776,
+      "step": 18879
+    },
+    {
+      "epoch": 0.8697655134288479,
+      "grad_norm": 0.7784570849193565,
+      "learning_rate": 8.765201857284223e-06,
+      "loss": 1.2278,
+      "step": 18880
+    },
+    {
+      "epoch": 0.8698115815174828,
+      "grad_norm": 1.0683654771281488,
+      "learning_rate": 8.759094028702386e-06,
+      "loss": 0.9447,
+      "step": 18881
+    },
+    {
+      "epoch": 0.8698576496061179,
+      "grad_norm": 1.4105733494675472,
+      "learning_rate": 8.752988231442872e-06,
+      "loss": 1.1985,
+      "step": 18882
+    },
+    {
+      "epoch": 0.8699037176947528,
+      "grad_norm": 1.0753780231793868,
+      "learning_rate": 8.746884465641625e-06,
+      "loss": 1.3029,
+      "step": 18883
+    },
+    {
+      "epoch": 0.8699497857833879,
+      "grad_norm": 0.8625386973271505,
+      "learning_rate": 8.740782731434505e-06,
+      "loss": 1.3901,
+      "step": 18884
+    },
+    {
+      "epoch": 0.8699958538720228,
+      "grad_norm": 0.8528049679648968,
+      "learning_rate": 8.734683028957402e-06,
+      "loss": 1.1361,
+      "step": 18885
+    },
+    {
+      "epoch": 0.8700419219606579,
+      "grad_norm": 1.154807999601204,
+      "learning_rate": 8.728585358346098e-06,
+      "loss": 1.0676,
+      "step": 18886
+    },
+    {
+      "epoch": 0.8700879900492928,
+      "grad_norm": 2.0054137144229043,
+      "learning_rate": 8.722489719736326e-06,
+      "loss": 1.2722,
+      "step": 18887
+    },
+    {
+      "epoch": 0.8701340581379279,
+      "grad_norm": 0.8229200734202241,
+      "learning_rate": 8.716396113263825e-06,
+      "loss": 1.2078,
+      "step": 18888
+    },
+    {
+      "epoch": 0.8701801262265628,
+      "grad_norm": 1.8785529405691002,
+      "learning_rate": 8.71030453906425e-06,
+      "loss": 1.532,
+      "step": 18889
+    },
+    {
+      "epoch": 0.8702261943151979,
+      "grad_norm": 1.2343250276519502,
+      "learning_rate": 8.70421499727323e-06,
+      "loss": 0.9735,
+      "step": 18890
+    },
+    {
+      "epoch": 0.8702722624038328,
+      "grad_norm": 1.5136923257048427,
+      "learning_rate": 8.698127488026287e-06,
+      "loss": 1.583,
+      "step": 18891
+    },
+    {
+      "epoch": 0.8703183304924679,
+      "grad_norm": 0.7695493702948855,
+      "learning_rate": 8.692042011459023e-06,
+      "loss": 1.4581,
+      "step": 18892
+    },
+    {
+      "epoch": 0.8703643985811028,
+      "grad_norm": 0.8763847701228031,
+      "learning_rate": 8.685958567706853e-06,
+      "loss": 1.4077,
+      "step": 18893
+    },
+    {
+      "epoch": 0.8704104666697379,
+      "grad_norm": 2.2866347287141915,
+      "learning_rate": 8.679877156905281e-06,
+      "loss": 1.0568,
+      "step": 18894
+    },
+    {
+      "epoch": 0.8704565347583729,
+      "grad_norm": 1.1342827956321735,
+      "learning_rate": 8.673797779189618e-06,
+      "loss": 1.0958,
+      "step": 18895
+    },
+    {
+      "epoch": 0.8705026028470079,
+      "grad_norm": 1.3835942453774694,
+      "learning_rate": 8.667720434695292e-06,
+      "loss": 0.8872,
+      "step": 18896
+    },
+    {
+      "epoch": 0.8705486709356429,
+      "grad_norm": 1.0909240467067622,
+      "learning_rate": 8.66164512355756e-06,
+      "loss": 1.0445,
+      "step": 18897
+    },
+    {
+      "epoch": 0.8705947390242779,
+      "grad_norm": 0.7059492009649412,
+      "learning_rate": 8.655571845911692e-06,
+      "loss": 1.1217,
+      "step": 18898
+    },
+    {
+      "epoch": 0.8706408071129129,
+      "grad_norm": 2.803163710589475,
+      "learning_rate": 8.64950060189289e-06,
+      "loss": 1.5239,
+      "step": 18899
+    },
+    {
+      "epoch": 0.8706868752015479,
+      "grad_norm": 1.4907085711328045,
+      "learning_rate": 8.643431391636314e-06,
+      "loss": 1.1311,
+      "step": 18900
+    },
+    {
+      "epoch": 0.8707329432901829,
+      "grad_norm": 0.9979136930488407,
+      "learning_rate": 8.637364215277121e-06,
+      "loss": 1.1773,
+      "step": 18901
+    },
+    {
+      "epoch": 0.8707790113788179,
+      "grad_norm": 1.3845064714282573,
+      "learning_rate": 8.631299072950317e-06,
+      "loss": 1.1712,
+      "step": 18902
+    },
+    {
+      "epoch": 0.8708250794674529,
+      "grad_norm": 1.0035853275908972,
+      "learning_rate": 8.625235964791012e-06,
+      "loss": 1.2815,
+      "step": 18903
+    },
+    {
+      "epoch": 0.8708711475560879,
+      "grad_norm": 0.8718081643459279,
+      "learning_rate": 8.619174890934146e-06,
+      "loss": 1.4494,
+      "step": 18904
+    },
+    {
+      "epoch": 0.8709172156447229,
+      "grad_norm": 5.478586468972247,
+      "learning_rate": 8.613115851514675e-06,
+      "loss": 1.2342,
+      "step": 18905
+    },
+    {
+      "epoch": 0.8709632837333579,
+      "grad_norm": 0.8757936499637323,
+      "learning_rate": 8.607058846667449e-06,
+      "loss": 1.3008,
+      "step": 18906
+    },
+    {
+      "epoch": 0.8710093518219929,
+      "grad_norm": 2.198280770626043,
+      "learning_rate": 8.601003876527391e-06,
+      "loss": 1.2089,
+      "step": 18907
+    },
+    {
+      "epoch": 0.871055419910628,
+      "grad_norm": 1.0707985041455654,
+      "learning_rate": 8.594950941229274e-06,
+      "loss": 1.4532,
+      "step": 18908
+    },
+    {
+      "epoch": 0.8711014879992629,
+      "grad_norm": 2.0396268357035012,
+      "learning_rate": 8.58890004090782e-06,
+      "loss": 1.5496,
+      "step": 18909
+    },
+    {
+      "epoch": 0.871147556087898,
+      "grad_norm": 2.20484340621417,
+      "learning_rate": 8.58285117569777e-06,
+      "loss": 1.6177,
+      "step": 18910
+    },
+    {
+      "epoch": 0.8711936241765329,
+      "grad_norm": 1.489244415083228,
+      "learning_rate": 8.576804345733802e-06,
+      "loss": 1.3374,
+      "step": 18911
+    },
+    {
+      "epoch": 0.871239692265168,
+      "grad_norm": 1.0733114597180087,
+      "learning_rate": 8.570759551150531e-06,
+      "loss": 1.1284,
+      "step": 18912
+    },
+    {
+      "epoch": 0.8712857603538029,
+      "grad_norm": 0.771147301019921,
+      "learning_rate": 8.564716792082528e-06,
+      "loss": 1.2007,
+      "step": 18913
+    },
+    {
+      "epoch": 0.871331828442438,
+      "grad_norm": 0.9276796940589146,
+      "learning_rate": 8.558676068664317e-06,
+      "loss": 1.2433,
+      "step": 18914
+    },
+    {
+      "epoch": 0.8713778965310729,
+      "grad_norm": 0.8433538529139628,
+      "learning_rate": 8.552637381030393e-06,
+      "loss": 1.1318,
+      "step": 18915
+    },
+    {
+      "epoch": 0.871423964619708,
+      "grad_norm": 1.3094227580052291,
+      "learning_rate": 8.546600729315224e-06,
+      "loss": 1.0639,
+      "step": 18916
+    },
+    {
+      "epoch": 0.8714700327083429,
+      "grad_norm": 0.8171153349285754,
+      "learning_rate": 8.54056611365316e-06,
+      "loss": 1.2997,
+      "step": 18917
+    },
+    {
+      "epoch": 0.8715161007969779,
+      "grad_norm": 1.0159403275428915,
+      "learning_rate": 8.53453353417858e-06,
+      "loss": 0.9538,
+      "step": 18918
+    },
+    {
+      "epoch": 0.8715621688856129,
+      "grad_norm": 1.0267451658346833,
+      "learning_rate": 8.528502991025789e-06,
+      "loss": 1.2378,
+      "step": 18919
+    },
+    {
+      "epoch": 0.8716082369742479,
+      "grad_norm": 2.14868770253607,
+      "learning_rate": 8.522474484329024e-06,
+      "loss": 1.2125,
+      "step": 18920
+    },
+    {
+      "epoch": 0.871654305062883,
+      "grad_norm": 1.22909015209701,
+      "learning_rate": 8.516448014222522e-06,
+      "loss": 0.8796,
+      "step": 18921
+    },
+    {
+      "epoch": 0.8717003731515179,
+      "grad_norm": 1.1352557581864897,
+      "learning_rate": 8.510423580840431e-06,
+      "loss": 1.0879,
+      "step": 18922
+    },
+    {
+      "epoch": 0.871746441240153,
+      "grad_norm": 3.5481831574183893,
+      "learning_rate": 8.504401184316913e-06,
+      "loss": 1.1329,
+      "step": 18923
+    },
+    {
+      "epoch": 0.8717925093287879,
+      "grad_norm": 1.4617576559370573,
+      "learning_rate": 8.498380824786001e-06,
+      "loss": 1.0682,
+      "step": 18924
+    },
+    {
+      "epoch": 0.871838577417423,
+      "grad_norm": 4.191223710993098,
+      "learning_rate": 8.492362502381756e-06,
+      "loss": 1.3135,
+      "step": 18925
+    },
+    {
+      "epoch": 0.8718846455060579,
+      "grad_norm": 1.2011279941841253,
+      "learning_rate": 8.486346217238162e-06,
+      "loss": 1.2684,
+      "step": 18926
+    },
+    {
+      "epoch": 0.871930713594693,
+      "grad_norm": 0.8092192105897875,
+      "learning_rate": 8.480331969489153e-06,
+      "loss": 1.3814,
+      "step": 18927
+    },
+    {
+      "epoch": 0.8719767816833279,
+      "grad_norm": 0.7299963861836978,
+      "learning_rate": 8.474319759268623e-06,
+      "loss": 1.2606,
+      "step": 18928
+    },
+    {
+      "epoch": 0.872022849771963,
+      "grad_norm": 0.7373352445794737,
+      "learning_rate": 8.468309586710432e-06,
+      "loss": 1.2486,
+      "step": 18929
+    },
+    {
+      "epoch": 0.8720689178605979,
+      "grad_norm": 0.7153596762869237,
+      "learning_rate": 8.462301451948406e-06,
+      "loss": 0.981,
+      "step": 18930
+    },
+    {
+      "epoch": 0.872114985949233,
+      "grad_norm": 1.5673569370525626,
+      "learning_rate": 8.456295355116261e-06,
+      "loss": 1.522,
+      "step": 18931
+    },
+    {
+      "epoch": 0.8721610540378679,
+      "grad_norm": 1.3922901054269115,
+      "learning_rate": 8.450291296347734e-06,
+      "loss": 0.7732,
+      "step": 18932
+    },
+    {
+      "epoch": 0.872207122126503,
+      "grad_norm": 0.9474074687506644,
+      "learning_rate": 8.444289275776507e-06,
+      "loss": 1.1258,
+      "step": 18933
+    },
+    {
+      "epoch": 0.8722531902151379,
+      "grad_norm": 0.9263799623784909,
+      "learning_rate": 8.438289293536205e-06,
+      "loss": 1.0746,
+      "step": 18934
+    },
+    {
+      "epoch": 0.872299258303773,
+      "grad_norm": 0.8190705423529686,
+      "learning_rate": 8.43229134976038e-06,
+      "loss": 1.2334,
+      "step": 18935
+    },
+    {
+      "epoch": 0.872345326392408,
+      "grad_norm": 1.2774314186125213,
+      "learning_rate": 8.426295444582589e-06,
+      "loss": 1.4262,
+      "step": 18936
+    },
+    {
+      "epoch": 0.872391394481043,
+      "grad_norm": 0.8136014484255775,
+      "learning_rate": 8.420301578136325e-06,
+      "loss": 1.369,
+      "step": 18937
+    },
+    {
+      "epoch": 0.872437462569678,
+      "grad_norm": 0.7970488391366148,
+      "learning_rate": 8.414309750555005e-06,
+      "loss": 1.2675,
+      "step": 18938
+    },
+    {
+      "epoch": 0.872483530658313,
+      "grad_norm": 1.944051170782675,
+      "learning_rate": 8.408319961972044e-06,
+      "loss": 0.9792,
+      "step": 18939
+    },
+    {
+      "epoch": 0.872529598746948,
+      "grad_norm": 1.2260970148725487,
+      "learning_rate": 8.402332212520792e-06,
+      "loss": 1.2812,
+      "step": 18940
+    },
+    {
+      "epoch": 0.872575666835583,
+      "grad_norm": 1.9137789067143929,
+      "learning_rate": 8.396346502334584e-06,
+      "loss": 1.0421,
+      "step": 18941
+    },
+    {
+      "epoch": 0.872621734924218,
+      "grad_norm": 0.8303774184764805,
+      "learning_rate": 8.390362831546628e-06,
+      "loss": 1.3241,
+      "step": 18942
+    },
+    {
+      "epoch": 0.872667803012853,
+      "grad_norm": 0.7947642365946249,
+      "learning_rate": 8.38438120029016e-06,
+      "loss": 1.1591,
+      "step": 18943
+    },
+    {
+      "epoch": 0.872713871101488,
+      "grad_norm": 0.9387919054060663,
+      "learning_rate": 8.378401608698383e-06,
+      "loss": 1.2645,
+      "step": 18944
+    },
+    {
+      "epoch": 0.872759939190123,
+      "grad_norm": 1.0114058933823136,
+      "learning_rate": 8.372424056904382e-06,
+      "loss": 1.1097,
+      "step": 18945
+    },
+    {
+      "epoch": 0.872806007278758,
+      "grad_norm": 1.034337787462599,
+      "learning_rate": 8.36644854504125e-06,
+      "loss": 1.1947,
+      "step": 18946
+    },
+    {
+      "epoch": 0.872852075367393,
+      "grad_norm": 1.0259548876837243,
+      "learning_rate": 8.360475073242025e-06,
+      "loss": 1.2453,
+      "step": 18947
+    },
+    {
+      "epoch": 0.872898143456028,
+      "grad_norm": 1.0319315092214019,
+      "learning_rate": 8.3545036416397e-06,
+      "loss": 1.2469,
+      "step": 18948
+    },
+    {
+      "epoch": 0.8729442115446631,
+      "grad_norm": 1.003720001038521,
+      "learning_rate": 8.348534250367213e-06,
+      "loss": 0.8649,
+      "step": 18949
+    },
+    {
+      "epoch": 0.872990279633298,
+      "grad_norm": 0.7853185889985352,
+      "learning_rate": 8.342566899557447e-06,
+      "loss": 1.2679,
+      "step": 18950
+    },
+    {
+      "epoch": 0.8730363477219331,
+      "grad_norm": 0.7502723767186342,
+      "learning_rate": 8.336601589343273e-06,
+      "loss": 1.3375,
+      "step": 18951
+    },
+    {
+      "epoch": 0.873082415810568,
+      "grad_norm": 1.349733635692586,
+      "learning_rate": 8.33063831985752e-06,
+      "loss": 1.2447,
+      "step": 18952
+    },
+    {
+      "epoch": 0.8731284838992031,
+      "grad_norm": 1.4588178289024663,
+      "learning_rate": 8.324677091232901e-06,
+      "loss": 1.3392,
+      "step": 18953
+    },
+    {
+      "epoch": 0.873174551987838,
+      "grad_norm": 3.4360910452067226,
+      "learning_rate": 8.318717903602168e-06,
+      "loss": 1.3493,
+      "step": 18954
+    },
+    {
+      "epoch": 0.873220620076473,
+      "grad_norm": 0.8963183375833983,
+      "learning_rate": 8.312760757097993e-06,
+      "loss": 1.2293,
+      "step": 18955
+    },
+    {
+      "epoch": 0.873266688165108,
+      "grad_norm": 1.2434155443980834,
+      "learning_rate": 8.306805651852967e-06,
+      "loss": 1.5282,
+      "step": 18956
+    },
+    {
+      "epoch": 0.873312756253743,
+      "grad_norm": 1.7797019664469347,
+      "learning_rate": 8.300852587999697e-06,
+      "loss": 1.3074,
+      "step": 18957
+    },
+    {
+      "epoch": 0.873358824342378,
+      "grad_norm": 1.464057598894806,
+      "learning_rate": 8.29490156567071e-06,
+      "loss": 1.0744,
+      "step": 18958
+    },
+    {
+      "epoch": 0.873404892431013,
+      "grad_norm": 2.221051157189762,
+      "learning_rate": 8.288952584998521e-06,
+      "loss": 1.2559,
+      "step": 18959
+    },
+    {
+      "epoch": 0.873450960519648,
+      "grad_norm": 0.7813524332702259,
+      "learning_rate": 8.283005646115538e-06,
+      "loss": 1.414,
+      "step": 18960
+    },
+    {
+      "epoch": 0.873497028608283,
+      "grad_norm": 1.2679691738720764,
+      "learning_rate": 8.277060749154175e-06,
+      "loss": 1.0211,
+      "step": 18961
+    },
+    {
+      "epoch": 0.8735430966969181,
+      "grad_norm": 1.2509364397143865,
+      "learning_rate": 8.271117894246793e-06,
+      "loss": 1.1048,
+      "step": 18962
+    },
+    {
+      "epoch": 0.873589164785553,
+      "grad_norm": 1.1394183648301273,
+      "learning_rate": 8.265177081525677e-06,
+      "loss": 1.4148,
+      "step": 18963
+    },
+    {
+      "epoch": 0.8736352328741881,
+      "grad_norm": 0.8615783180128505,
+      "learning_rate": 8.259238311123097e-06,
+      "loss": 1.2956,
+      "step": 18964
+    },
+    {
+      "epoch": 0.873681300962823,
+      "grad_norm": 0.7035263791813291,
+      "learning_rate": 8.25330158317127e-06,
+      "loss": 1.2894,
+      "step": 18965
+    },
+    {
+      "epoch": 0.8737273690514581,
+      "grad_norm": 1.1536215107357912,
+      "learning_rate": 8.2473668978024e-06,
+      "loss": 1.4277,
+      "step": 18966
+    },
+    {
+      "epoch": 0.873773437140093,
+      "grad_norm": 1.2711450779560969,
+      "learning_rate": 8.241434255148562e-06,
+      "loss": 1.1738,
+      "step": 18967
+    },
+    {
+      "epoch": 0.8738195052287281,
+      "grad_norm": 0.9162192493048427,
+      "learning_rate": 8.235503655341859e-06,
+      "loss": 1.1213,
+      "step": 18968
+    },
+    {
+      "epoch": 0.873865573317363,
+      "grad_norm": 2.806967582075807,
+      "learning_rate": 8.22957509851432e-06,
+      "loss": 1.4482,
+      "step": 18969
+    },
+    {
+      "epoch": 0.8739116414059981,
+      "grad_norm": 1.011780205156484,
+      "learning_rate": 8.22364858479796e-06,
+      "loss": 1.2712,
+      "step": 18970
+    },
+    {
+      "epoch": 0.873957709494633,
+      "grad_norm": 1.2184038995458384,
+      "learning_rate": 8.217724114324687e-06,
+      "loss": 1.5454,
+      "step": 18971
+    },
+    {
+      "epoch": 0.8740037775832681,
+      "grad_norm": 0.8749417230305876,
+      "learning_rate": 8.211801687226407e-06,
+      "loss": 1.2901,
+      "step": 18972
+    },
+    {
+      "epoch": 0.874049845671903,
+      "grad_norm": 1.2007254382180994,
+      "learning_rate": 8.205881303635e-06,
+      "loss": 1.2195,
+      "step": 18973
+    },
+    {
+      "epoch": 0.8740959137605381,
+      "grad_norm": 0.8685848874324226,
+      "learning_rate": 8.199962963682239e-06,
+      "loss": 1.2512,
+      "step": 18974
+    },
+    {
+      "epoch": 0.8741419818491731,
+      "grad_norm": 1.4789142592637958,
+      "learning_rate": 8.1940466674999e-06,
+      "loss": 1.2875,
+      "step": 18975
+    },
+    {
+      "epoch": 0.8741880499378081,
+      "grad_norm": 0.7433547078371161,
+      "learning_rate": 8.188132415219696e-06,
+      "loss": 1.4049,
+      "step": 18976
+    },
+    {
+      "epoch": 0.8742341180264431,
+      "grad_norm": 0.7409268675682937,
+      "learning_rate": 8.182220206973313e-06,
+      "loss": 1.4062,
+      "step": 18977
+    },
+    {
+      "epoch": 0.8742801861150781,
+      "grad_norm": 1.0813351645616842,
+      "learning_rate": 8.176310042892343e-06,
+      "loss": 1.1294,
+      "step": 18978
+    },
+    {
+      "epoch": 0.8743262542037131,
+      "grad_norm": 1.676422506432868,
+      "learning_rate": 8.170401923108395e-06,
+      "loss": 1.4174,
+      "step": 18979
+    },
+    {
+      "epoch": 0.8743723222923481,
+      "grad_norm": 1.083891340546044,
+      "learning_rate": 8.164495847752996e-06,
+      "loss": 0.996,
+      "step": 18980
+    },
+    {
+      "epoch": 0.8744183903809831,
+      "grad_norm": 0.8899872667905364,
+      "learning_rate": 8.158591816957639e-06,
+      "loss": 1.1403,
+      "step": 18981
+    },
+    {
+      "epoch": 0.8744644584696181,
+      "grad_norm": 1.168921229115018,
+      "learning_rate": 8.152689830853777e-06,
+      "loss": 1.4142,
+      "step": 18982
+    },
+    {
+      "epoch": 0.8745105265582531,
+      "grad_norm": 1.2931503081083258,
+      "learning_rate": 8.146789889572747e-06,
+      "loss": 0.8117,
+      "step": 18983
+    },
+    {
+      "epoch": 0.8745565946468881,
+      "grad_norm": 1.5120639953022839,
+      "learning_rate": 8.14089199324598e-06,
+      "loss": 1.2251,
+      "step": 18984
+    },
+    {
+      "epoch": 0.8746026627355231,
+      "grad_norm": 1.5342580595055053,
+      "learning_rate": 8.134996142004725e-06,
+      "loss": 1.1,
+      "step": 18985
+    },
+    {
+      "epoch": 0.8746487308241581,
+      "grad_norm": 2.467656463576376,
+      "learning_rate": 8.129102335980276e-06,
+      "loss": 1.2608,
+      "step": 18986
+    },
+    {
+      "epoch": 0.8746947989127931,
+      "grad_norm": 0.9663799575244127,
+      "learning_rate": 8.12321057530384e-06,
+      "loss": 1.1062,
+      "step": 18987
+    },
+    {
+      "epoch": 0.8747408670014281,
+      "grad_norm": 1.0109995737068558,
+      "learning_rate": 8.117320860106592e-06,
+      "loss": 0.947,
+      "step": 18988
+    },
+    {
+      "epoch": 0.8747869350900631,
+      "grad_norm": 0.8556840773992417,
+      "learning_rate": 8.111433190519647e-06,
+      "loss": 1.0287,
+      "step": 18989
+    },
+    {
+      "epoch": 0.8748330031786982,
+      "grad_norm": 0.8451902057523986,
+      "learning_rate": 8.105547566674077e-06,
+      "loss": 1.3137,
+      "step": 18990
+    },
+    {
+      "epoch": 0.8748790712673331,
+      "grad_norm": 1.3369142166243226,
+      "learning_rate": 8.099663988700944e-06,
+      "loss": 1.2282,
+      "step": 18991
+    },
+    {
+      "epoch": 0.8749251393559682,
+      "grad_norm": 1.0439606492434945,
+      "learning_rate": 8.093782456731203e-06,
+      "loss": 1.1309,
+      "step": 18992
+    },
+    {
+      "epoch": 0.8749712074446031,
+      "grad_norm": 1.4909896317435203,
+      "learning_rate": 8.087902970895822e-06,
+      "loss": 1.2165,
+      "step": 18993
+    },
+    {
+      "epoch": 0.8750172755332382,
+      "grad_norm": 0.6796060934832512,
+      "learning_rate": 8.082025531325655e-06,
+      "loss": 1.258,
+      "step": 18994
+    },
+    {
+      "epoch": 0.8750633436218731,
+      "grad_norm": 0.8490007810242495,
+      "learning_rate": 8.076150138151616e-06,
+      "loss": 1.215,
+      "step": 18995
+    },
+    {
+      "epoch": 0.8751094117105082,
+      "grad_norm": 1.1377416071529898,
+      "learning_rate": 8.070276791504461e-06,
+      "loss": 1.0249,
+      "step": 18996
+    },
+    {
+      "epoch": 0.8751554797991431,
+      "grad_norm": 0.9780261206251102,
+      "learning_rate": 8.064405491514992e-06,
+      "loss": 1.2926,
+      "step": 18997
+    },
+    {
+      "epoch": 0.8752015478877782,
+      "grad_norm": 0.9468296661076026,
+      "learning_rate": 8.058536238313852e-06,
+      "loss": 1.0341,
+      "step": 18998
+    },
+    {
+      "epoch": 0.8752476159764131,
+      "grad_norm": 1.5830236777768942,
+      "learning_rate": 8.052669032031801e-06,
+      "loss": 1.2229,
+      "step": 18999
+    },
+    {
+      "epoch": 0.8752936840650481,
+      "grad_norm": 2.0792648451979705,
+      "learning_rate": 8.046803872799414e-06,
+      "loss": 1.3161,
+      "step": 19000
+    },
+    {
+      "epoch": 0.8753397521536831,
+      "grad_norm": 0.7917715696555427,
+      "learning_rate": 8.04094076074724e-06,
+      "loss": 0.9414,
+      "step": 19001
+    },
+    {
+      "epoch": 0.8753858202423181,
+      "grad_norm": 0.915780785822683,
+      "learning_rate": 8.035079696005886e-06,
+      "loss": 1.0054,
+      "step": 19002
+    },
+    {
+      "epoch": 0.8754318883309532,
+      "grad_norm": 0.8379621507541192,
+      "learning_rate": 8.029220678705761e-06,
+      "loss": 1.4453,
+      "step": 19003
+    },
+    {
+      "epoch": 0.8754779564195881,
+      "grad_norm": 1.0974800969013025,
+      "learning_rate": 8.023363708977372e-06,
+      "loss": 1.1589,
+      "step": 19004
+    },
+    {
+      "epoch": 0.8755240245082232,
+      "grad_norm": 0.8462428269664642,
+      "learning_rate": 8.017508786951056e-06,
+      "loss": 0.9879,
+      "step": 19005
+    },
+    {
+      "epoch": 0.8755700925968581,
+      "grad_norm": 0.7478135314837413,
+      "learning_rate": 8.01165591275721e-06,
+      "loss": 1.3179,
+      "step": 19006
+    },
+    {
+      "epoch": 0.8756161606854932,
+      "grad_norm": 0.7726816260813056,
+      "learning_rate": 8.005805086526109e-06,
+      "loss": 1.206,
+      "step": 19007
+    },
+    {
+      "epoch": 0.8756622287741281,
+      "grad_norm": 1.7601746544161434,
+      "learning_rate": 7.999956308388034e-06,
+      "loss": 1.0499,
+      "step": 19008
+    },
+    {
+      "epoch": 0.8757082968627632,
+      "grad_norm": 0.8802018042415765,
+      "learning_rate": 7.994109578473174e-06,
+      "loss": 1.2591,
+      "step": 19009
+    },
+    {
+      "epoch": 0.8757543649513981,
+      "grad_norm": 1.2327534622729563,
+      "learning_rate": 7.988264896911713e-06,
+      "loss": 1.4595,
+      "step": 19010
+    },
+    {
+      "epoch": 0.8758004330400332,
+      "grad_norm": 0.9262999334689785,
+      "learning_rate": 7.982422263833777e-06,
+      "loss": 1.3796,
+      "step": 19011
+    },
+    {
+      "epoch": 0.8758465011286681,
+      "grad_norm": 1.2875356012139743,
+      "learning_rate": 7.97658167936941e-06,
+      "loss": 1.3581,
+      "step": 19012
+    },
+    {
+      "epoch": 0.8758925692173032,
+      "grad_norm": 0.7157818872949849,
+      "learning_rate": 7.970743143648674e-06,
+      "loss": 1.415,
+      "step": 19013
+    },
+    {
+      "epoch": 0.8759386373059381,
+      "grad_norm": 1.1707937523796141,
+      "learning_rate": 7.964906656801552e-06,
+      "loss": 1.1233,
+      "step": 19014
+    },
+    {
+      "epoch": 0.8759847053945732,
+      "grad_norm": 1.0299515386979043,
+      "learning_rate": 7.959072218957986e-06,
+      "loss": 1.0331,
+      "step": 19015
+    },
+    {
+      "epoch": 0.8760307734832082,
+      "grad_norm": 1.7159216328476985,
+      "learning_rate": 7.953239830247816e-06,
+      "loss": 0.9767,
+      "step": 19016
+    },
+    {
+      "epoch": 0.8760768415718432,
+      "grad_norm": 0.7724810812180554,
+      "learning_rate": 7.947409490800984e-06,
+      "loss": 1.3653,
+      "step": 19017
+    },
+    {
+      "epoch": 0.8761229096604782,
+      "grad_norm": 0.7698013721556205,
+      "learning_rate": 7.941581200747239e-06,
+      "loss": 1.1538,
+      "step": 19018
+    },
+    {
+      "epoch": 0.8761689777491132,
+      "grad_norm": 0.8590871783805217,
+      "learning_rate": 7.935754960216324e-06,
+      "loss": 1.1799,
+      "step": 19019
+    },
+    {
+      "epoch": 0.8762150458377482,
+      "grad_norm": 0.6376302836842825,
+      "learning_rate": 7.929930769337968e-06,
+      "loss": 1.4158,
+      "step": 19020
+    },
+    {
+      "epoch": 0.8762611139263832,
+      "grad_norm": 1.1629754962688685,
+      "learning_rate": 7.924108628241844e-06,
+      "loss": 1.1662,
+      "step": 19021
+    },
+    {
+      "epoch": 0.8763071820150182,
+      "grad_norm": 0.9253801920469403,
+      "learning_rate": 7.918288537057571e-06,
+      "loss": 1.3663,
+      "step": 19022
+    },
+    {
+      "epoch": 0.8763532501036532,
+      "grad_norm": 1.0173935303925592,
+      "learning_rate": 7.912470495914715e-06,
+      "loss": 1.1616,
+      "step": 19023
+    },
+    {
+      "epoch": 0.8763993181922882,
+      "grad_norm": 0.8026857943510952,
+      "learning_rate": 7.90665450494279e-06,
+      "loss": 1.3643,
+      "step": 19024
+    },
+    {
+      "epoch": 0.8764453862809232,
+      "grad_norm": 1.0221505973225953,
+      "learning_rate": 7.900840564271306e-06,
+      "loss": 1.0902,
+      "step": 19025
+    },
+    {
+      "epoch": 0.8764914543695582,
+      "grad_norm": 0.7553485754293021,
+      "learning_rate": 7.895028674029703e-06,
+      "loss": 1.1205,
+      "step": 19026
+    },
+    {
+      "epoch": 0.8765375224581932,
+      "grad_norm": 0.7985370868894822,
+      "learning_rate": 7.889218834347356e-06,
+      "loss": 1.2886,
+      "step": 19027
+    },
+    {
+      "epoch": 0.8765835905468282,
+      "grad_norm": 1.1564424994274853,
+      "learning_rate": 7.883411045353607e-06,
+      "loss": 1.137,
+      "step": 19028
+    },
+    {
+      "epoch": 0.8766296586354632,
+      "grad_norm": 1.7366294512460054,
+      "learning_rate": 7.877605307177772e-06,
+      "loss": 1.2783,
+      "step": 19029
+    },
+    {
+      "epoch": 0.8766757267240982,
+      "grad_norm": 1.3796208639317769,
+      "learning_rate": 7.871801619949093e-06,
+      "loss": 0.9686,
+      "step": 19030
+    },
+    {
+      "epoch": 0.8767217948127333,
+      "grad_norm": 0.8017736716746267,
+      "learning_rate": 7.865999983796779e-06,
+      "loss": 1.277,
+      "step": 19031
+    },
+    {
+      "epoch": 0.8767678629013682,
+      "grad_norm": 0.9343965693500136,
+      "learning_rate": 7.860200398850003e-06,
+      "loss": 1.1351,
+      "step": 19032
+    },
+    {
+      "epoch": 0.8768139309900033,
+      "grad_norm": 0.9426678859913549,
+      "learning_rate": 7.854402865237897e-06,
+      "loss": 1.2949,
+      "step": 19033
+    },
+    {
+      "epoch": 0.8768599990786382,
+      "grad_norm": 0.9431790337852377,
+      "learning_rate": 7.8486073830895e-06,
+      "loss": 1.45,
+      "step": 19034
+    },
+    {
+      "epoch": 0.8769060671672733,
+      "grad_norm": 0.8714820762608607,
+      "learning_rate": 7.842813952533856e-06,
+      "loss": 0.8975,
+      "step": 19035
+    },
+    {
+      "epoch": 0.8769521352559082,
+      "grad_norm": 0.8469886121709544,
+      "learning_rate": 7.837022573699959e-06,
+      "loss": 1.1907,
+      "step": 19036
+    },
+    {
+      "epoch": 0.8769982033445433,
+      "grad_norm": 1.688760254538161,
+      "learning_rate": 7.831233246716707e-06,
+      "loss": 1.3625,
+      "step": 19037
+    },
+    {
+      "epoch": 0.8770442714331782,
+      "grad_norm": 0.9418033339371991,
+      "learning_rate": 7.825445971713008e-06,
+      "loss": 1.3806,
+      "step": 19038
+    },
+    {
+      "epoch": 0.8770903395218133,
+      "grad_norm": 0.9241932579503054,
+      "learning_rate": 7.819660748817715e-06,
+      "loss": 0.9221,
+      "step": 19039
+    },
+    {
+      "epoch": 0.8771364076104482,
+      "grad_norm": 0.9812143874123749,
+      "learning_rate": 7.813877578159645e-06,
+      "loss": 1.1583,
+      "step": 19040
+    },
+    {
+      "epoch": 0.8771824756990833,
+      "grad_norm": 0.7952953704742731,
+      "learning_rate": 7.808096459867497e-06,
+      "loss": 1.3178,
+      "step": 19041
+    },
+    {
+      "epoch": 0.8772285437877182,
+      "grad_norm": 1.0155161072229268,
+      "learning_rate": 7.802317394070003e-06,
+      "loss": 1.1628,
+      "step": 19042
+    },
+    {
+      "epoch": 0.8772746118763532,
+      "grad_norm": 1.1732841058071748,
+      "learning_rate": 7.796540380895834e-06,
+      "loss": 0.8332,
+      "step": 19043
+    },
+    {
+      "epoch": 0.8773206799649883,
+      "grad_norm": 1.0322131992047703,
+      "learning_rate": 7.790765420473622e-06,
+      "loss": 1.1593,
+      "step": 19044
+    },
+    {
+      "epoch": 0.8773667480536232,
+      "grad_norm": 1.5390747773324622,
+      "learning_rate": 7.784992512931888e-06,
+      "loss": 1.0342,
+      "step": 19045
+    },
+    {
+      "epoch": 0.8774128161422583,
+      "grad_norm": 1.713679424227072,
+      "learning_rate": 7.779221658399171e-06,
+      "loss": 1.3162,
+      "step": 19046
+    },
+    {
+      "epoch": 0.8774588842308932,
+      "grad_norm": 0.8591872163319553,
+      "learning_rate": 7.773452857003994e-06,
+      "loss": 1.2423,
+      "step": 19047
+    },
+    {
+      "epoch": 0.8775049523195283,
+      "grad_norm": 1.1023884036430271,
+      "learning_rate": 7.767686108874727e-06,
+      "loss": 1.4443,
+      "step": 19048
+    },
+    {
+      "epoch": 0.8775510204081632,
+      "grad_norm": 0.7061969044896843,
+      "learning_rate": 7.761921414139783e-06,
+      "loss": 1.4147,
+      "step": 19049
+    },
+    {
+      "epoch": 0.8775970884967983,
+      "grad_norm": 0.7733215083099422,
+      "learning_rate": 7.756158772927512e-06,
+      "loss": 1.2011,
+      "step": 19050
+    },
+    {
+      "epoch": 0.8776431565854332,
+      "grad_norm": 1.311926548823953,
+      "learning_rate": 7.750398185366214e-06,
+      "loss": 1.3149,
+      "step": 19051
+    },
+    {
+      "epoch": 0.8776892246740683,
+      "grad_norm": 1.4455674518350379,
+      "learning_rate": 7.744639651584118e-06,
+      "loss": 1.0579,
+      "step": 19052
+    },
+    {
+      "epoch": 0.8777352927627032,
+      "grad_norm": 1.0458936654170532,
+      "learning_rate": 7.738883171709421e-06,
+      "loss": 1.1044,
+      "step": 19053
+    },
+    {
+      "epoch": 0.8777813608513383,
+      "grad_norm": 1.1042598753740258,
+      "learning_rate": 7.733128745870322e-06,
+      "loss": 1.0863,
+      "step": 19054
+    },
+    {
+      "epoch": 0.8778274289399732,
+      "grad_norm": 0.6828803296978254,
+      "learning_rate": 7.727376374194895e-06,
+      "loss": 1.6223,
+      "step": 19055
+    },
+    {
+      "epoch": 0.8778734970286083,
+      "grad_norm": 3.787209672218706,
+      "learning_rate": 7.721626056811215e-06,
+      "loss": 1.3259,
+      "step": 19056
+    },
+    {
+      "epoch": 0.8779195651172433,
+      "grad_norm": 0.9622040445849478,
+      "learning_rate": 7.715877793847303e-06,
+      "loss": 1.3752,
+      "step": 19057
+    },
+    {
+      "epoch": 0.8779656332058783,
+      "grad_norm": 1.5509955601045806,
+      "learning_rate": 7.710131585431157e-06,
+      "loss": 1.3241,
+      "step": 19058
+    },
+    {
+      "epoch": 0.8780117012945133,
+      "grad_norm": 0.9670403869741458,
+      "learning_rate": 7.704387431690673e-06,
+      "loss": 1.5288,
+      "step": 19059
+    },
+    {
+      "epoch": 0.8780577693831483,
+      "grad_norm": 1.7009461751051989,
+      "learning_rate": 7.698645332753752e-06,
+      "loss": 1.2086,
+      "step": 19060
+    },
+    {
+      "epoch": 0.8781038374717833,
+      "grad_norm": 0.7691946402770816,
+      "learning_rate": 7.69290528874822e-06,
+      "loss": 1.4928,
+      "step": 19061
+    },
+    {
+      "epoch": 0.8781499055604183,
+      "grad_norm": 1.3606400712054119,
+      "learning_rate": 7.687167299801912e-06,
+      "loss": 1.1289,
+      "step": 19062
+    },
+    {
+      "epoch": 0.8781959736490533,
+      "grad_norm": 2.0894830270133826,
+      "learning_rate": 7.681431366042514e-06,
+      "loss": 1.1328,
+      "step": 19063
+    },
+    {
+      "epoch": 0.8782420417376883,
+      "grad_norm": 0.9141696651842466,
+      "learning_rate": 7.675697487597755e-06,
+      "loss": 1.2102,
+      "step": 19064
+    },
+    {
+      "epoch": 0.8782881098263233,
+      "grad_norm": 1.097734451780372,
+      "learning_rate": 7.669965664595302e-06,
+      "loss": 1.3467,
+      "step": 19065
+    },
+    {
+      "epoch": 0.8783341779149583,
+      "grad_norm": 0.8352039759526965,
+      "learning_rate": 7.664235897162742e-06,
+      "loss": 0.898,
+      "step": 19066
+    },
+    {
+      "epoch": 0.8783802460035933,
+      "grad_norm": 0.8256446544973228,
+      "learning_rate": 7.658508185427648e-06,
+      "loss": 1.2128,
+      "step": 19067
+    },
+    {
+      "epoch": 0.8784263140922283,
+      "grad_norm": 0.8170328804022414,
+      "learning_rate": 7.652782529517532e-06,
+      "loss": 1.3017,
+      "step": 19068
+    },
+    {
+      "epoch": 0.8784723821808633,
+      "grad_norm": 2.104851584830508,
+      "learning_rate": 7.647058929559891e-06,
+      "loss": 1.146,
+      "step": 19069
+    },
+    {
+      "epoch": 0.8785184502694983,
+      "grad_norm": 1.604999132046053,
+      "learning_rate": 7.641337385682123e-06,
+      "loss": 1.162,
+      "step": 19070
+    },
+    {
+      "epoch": 0.8785645183581333,
+      "grad_norm": 1.5050290916002855,
+      "learning_rate": 7.635617898011605e-06,
+      "loss": 1.4857,
+      "step": 19071
+    },
+    {
+      "epoch": 0.8786105864467684,
+      "grad_norm": 1.8642566618508574,
+      "learning_rate": 7.62990046667571e-06,
+      "loss": 1.1193,
+      "step": 19072
+    },
+    {
+      "epoch": 0.8786566545354033,
+      "grad_norm": 0.6942552081906038,
+      "learning_rate": 7.624185091801672e-06,
+      "loss": 1.4058,
+      "step": 19073
+    },
+    {
+      "epoch": 0.8787027226240384,
+      "grad_norm": 1.5191563557857808,
+      "learning_rate": 7.618471773516766e-06,
+      "loss": 1.3853,
+      "step": 19074
+    },
+    {
+      "epoch": 0.8787487907126733,
+      "grad_norm": 1.3601318553881103,
+      "learning_rate": 7.612760511948191e-06,
+      "loss": 1.4344,
+      "step": 19075
+    },
+    {
+      "epoch": 0.8787948588013084,
+      "grad_norm": 1.083589379108934,
+      "learning_rate": 7.607051307223112e-06,
+      "loss": 1.3249,
+      "step": 19076
+    },
+    {
+      "epoch": 0.8788409268899433,
+      "grad_norm": 1.0475986334672296,
+      "learning_rate": 7.601344159468593e-06,
+      "loss": 1.3476,
+      "step": 19077
+    },
+    {
+      "epoch": 0.8788869949785784,
+      "grad_norm": 1.283122044315514,
+      "learning_rate": 7.59563906881171e-06,
+      "loss": 0.9467,
+      "step": 19078
+    },
+    {
+      "epoch": 0.8789330630672133,
+      "grad_norm": 1.765883236136538,
+      "learning_rate": 7.5899360353794965e-06,
+      "loss": 1.0824,
+      "step": 19079
+    },
+    {
+      "epoch": 0.8789791311558484,
+      "grad_norm": 1.6512575529256213,
+      "learning_rate": 7.584235059298917e-06,
+      "loss": 1.1764,
+      "step": 19080
+    },
+    {
+      "epoch": 0.8790251992444833,
+      "grad_norm": 0.7844633380959264,
+      "learning_rate": 7.57853614069689e-06,
+      "loss": 1.4572,
+      "step": 19081
+    },
+    {
+      "epoch": 0.8790712673331184,
+      "grad_norm": 1.6264146973542062,
+      "learning_rate": 7.57283927970025e-06,
+      "loss": 1.5722,
+      "step": 19082
+    },
+    {
+      "epoch": 0.8791173354217533,
+      "grad_norm": 0.9451630584412025,
+      "learning_rate": 7.567144476435895e-06,
+      "loss": 1.0956,
+      "step": 19083
+    },
+    {
+      "epoch": 0.8791634035103884,
+      "grad_norm": 1.485386456847931,
+      "learning_rate": 7.561451731030567e-06,
+      "loss": 1.5372,
+      "step": 19084
+    },
+    {
+      "epoch": 0.8792094715990234,
+      "grad_norm": 2.951322938049177,
+      "learning_rate": 7.555761043611043e-06,
+      "loss": 1.1166,
+      "step": 19085
+    },
+    {
+      "epoch": 0.8792555396876583,
+      "grad_norm": 0.7697850695179205,
+      "learning_rate": 7.550072414303943e-06,
+      "loss": 0.8748,
+      "step": 19086
+    },
+    {
+      "epoch": 0.8793016077762934,
+      "grad_norm": 1.022547524434503,
+      "learning_rate": 7.544385843236002e-06,
+      "loss": 1.0869,
+      "step": 19087
+    },
+    {
+      "epoch": 0.8793476758649283,
+      "grad_norm": 1.261858418116768,
+      "learning_rate": 7.53870133053377e-06,
+      "loss": 1.0193,
+      "step": 19088
+    },
+    {
+      "epoch": 0.8793937439535634,
+      "grad_norm": 1.2917492095183973,
+      "learning_rate": 7.5330188763238165e-06,
+      "loss": 0.8865,
+      "step": 19089
+    },
+    {
+      "epoch": 0.8794398120421983,
+      "grad_norm": 0.851255290643403,
+      "learning_rate": 7.5273384807326595e-06,
+      "loss": 1.4824,
+      "step": 19090
+    },
+    {
+      "epoch": 0.8794858801308334,
+      "grad_norm": 0.9280552688606802,
+      "learning_rate": 7.521660143886744e-06,
+      "loss": 1.4205,
+      "step": 19091
+    },
+    {
+      "epoch": 0.8795319482194683,
+      "grad_norm": 2.6493873704599347,
+      "learning_rate": 7.515983865912513e-06,
+      "loss": 0.9348,
+      "step": 19092
+    },
+    {
+      "epoch": 0.8795780163081034,
+      "grad_norm": 0.893399449918441,
+      "learning_rate": 7.510309646936298e-06,
+      "loss": 1.3142,
+      "step": 19093
+    },
+    {
+      "epoch": 0.8796240843967383,
+      "grad_norm": 0.9998003499781712,
+      "learning_rate": 7.504637487084476e-06,
+      "loss": 1.2546,
+      "step": 19094
+    },
+    {
+      "epoch": 0.8796701524853734,
+      "grad_norm": 0.7534717698948147,
+      "learning_rate": 7.498967386483302e-06,
+      "loss": 1.3742,
+      "step": 19095
+    },
+    {
+      "epoch": 0.8797162205740083,
+      "grad_norm": 0.8517711381828353,
+      "learning_rate": 7.49329934525902e-06,
+      "loss": 1.2805,
+      "step": 19096
+    },
+    {
+      "epoch": 0.8797622886626434,
+      "grad_norm": 1.2819817607300374,
+      "learning_rate": 7.487633363537783e-06,
+      "loss": 1.0485,
+      "step": 19097
+    },
+    {
+      "epoch": 0.8798083567512784,
+      "grad_norm": 0.9748456382480838,
+      "learning_rate": 7.481969441445791e-06,
+      "loss": 1.214,
+      "step": 19098
+    },
+    {
+      "epoch": 0.8798544248399134,
+      "grad_norm": 0.723267858643906,
+      "learning_rate": 7.47630757910911e-06,
+      "loss": 1.2216,
+      "step": 19099
+    },
+    {
+      "epoch": 0.8799004929285484,
+      "grad_norm": 1.692471586886002,
+      "learning_rate": 7.470647776653783e-06,
+      "loss": 1.3277,
+      "step": 19100
+    },
+    {
+      "epoch": 0.8799465610171834,
+      "grad_norm": 0.977028839859673,
+      "learning_rate": 7.464990034205854e-06,
+      "loss": 1.3236,
+      "step": 19101
+    },
+    {
+      "epoch": 0.8799926291058184,
+      "grad_norm": 0.8291396125539711,
+      "learning_rate": 7.459334351891245e-06,
+      "loss": 1.3491,
+      "step": 19102
+    },
+    {
+      "epoch": 0.8800386971944534,
+      "grad_norm": 1.0867822840957133,
+      "learning_rate": 7.453680729835899e-06,
+      "loss": 1.2734,
+      "step": 19103
+    },
+    {
+      "epoch": 0.8800847652830884,
+      "grad_norm": 0.9753416556526066,
+      "learning_rate": 7.448029168165627e-06,
+      "loss": 1.2238,
+      "step": 19104
+    },
+    {
+      "epoch": 0.8801308333717234,
+      "grad_norm": 1.1927096953789234,
+      "learning_rate": 7.4423796670063385e-06,
+      "loss": 1.5323,
+      "step": 19105
+    },
+    {
+      "epoch": 0.8801769014603584,
+      "grad_norm": 0.8443787821059924,
+      "learning_rate": 7.436732226483745e-06,
+      "loss": 1.2095,
+      "step": 19106
+    },
+    {
+      "epoch": 0.8802229695489934,
+      "grad_norm": 0.8615297460282745,
+      "learning_rate": 7.431086846723601e-06,
+      "loss": 1.282,
+      "step": 19107
+    },
+    {
+      "epoch": 0.8802690376376284,
+      "grad_norm": 1.0846125757861196,
+      "learning_rate": 7.425443527851561e-06,
+      "loss": 1.6361,
+      "step": 19108
+    },
+    {
+      "epoch": 0.8803151057262634,
+      "grad_norm": 1.5402494400251678,
+      "learning_rate": 7.419802269993326e-06,
+      "loss": 0.9304,
+      "step": 19109
+    },
+    {
+      "epoch": 0.8803611738148984,
+      "grad_norm": 1.4698441158898687,
+      "learning_rate": 7.414163073274449e-06,
+      "loss": 1.0875,
+      "step": 19110
+    },
+    {
+      "epoch": 0.8804072419035335,
+      "grad_norm": 1.2817203405364768,
+      "learning_rate": 7.408525937820466e-06,
+      "loss": 1.0266,
+      "step": 19111
+    },
+    {
+      "epoch": 0.8804533099921684,
+      "grad_norm": 0.8054561743294967,
+      "learning_rate": 7.402890863756895e-06,
+      "loss": 1.0856,
+      "step": 19112
+    },
+    {
+      "epoch": 0.8804993780808035,
+      "grad_norm": 0.9828603800238822,
+      "learning_rate": 7.397257851209182e-06,
+      "loss": 1.3284,
+      "step": 19113
+    },
+    {
+      "epoch": 0.8805454461694384,
+      "grad_norm": 1.688145650427158,
+      "learning_rate": 7.391626900302773e-06,
+      "loss": 0.9763,
+      "step": 19114
+    },
+    {
+      "epoch": 0.8805915142580735,
+      "grad_norm": 1.415298454655443,
+      "learning_rate": 7.385998011162954e-06,
+      "loss": 1.1152,
+      "step": 19115
+    },
+    {
+      "epoch": 0.8806375823467084,
+      "grad_norm": 0.8312616970881598,
+      "learning_rate": 7.380371183915125e-06,
+      "loss": 1.04,
+      "step": 19116
+    },
+    {
+      "epoch": 0.8806836504353435,
+      "grad_norm": 0.8433989223699151,
+      "learning_rate": 7.374746418684508e-06,
+      "loss": 1.5205,
+      "step": 19117
+    },
+    {
+      "epoch": 0.8807297185239784,
+      "grad_norm": 1.0064719262836364,
+      "learning_rate": 7.369123715596371e-06,
+      "loss": 1.1603,
+      "step": 19118
+    },
+    {
+      "epoch": 0.8807757866126135,
+      "grad_norm": 0.9421028676506513,
+      "learning_rate": 7.3635030747758346e-06,
+      "loss": 1.2721,
+      "step": 19119
+    },
+    {
+      "epoch": 0.8808218547012484,
+      "grad_norm": 0.9551648865810451,
+      "learning_rate": 7.357884496348077e-06,
+      "loss": 0.9331,
+      "step": 19120
+    },
+    {
+      "epoch": 0.8808679227898835,
+      "grad_norm": 1.3413434547866192,
+      "learning_rate": 7.352267980438188e-06,
+      "loss": 1.3542,
+      "step": 19121
+    },
+    {
+      "epoch": 0.8809139908785184,
+      "grad_norm": 1.7079849822519735,
+      "learning_rate": 7.346653527171177e-06,
+      "loss": 0.953,
+      "step": 19122
+    },
+    {
+      "epoch": 0.8809600589671535,
+      "grad_norm": 1.2158213623985235,
+      "learning_rate": 7.341041136672055e-06,
+      "loss": 1.2369,
+      "step": 19123
+    },
+    {
+      "epoch": 0.8810061270557884,
+      "grad_norm": 1.1212771263602999,
+      "learning_rate": 7.33543080906578e-06,
+      "loss": 1.0887,
+      "step": 19124
+    },
+    {
+      "epoch": 0.8810521951444235,
+      "grad_norm": 0.7620705119723872,
+      "learning_rate": 7.329822544477261e-06,
+      "loss": 1.2179,
+      "step": 19125
+    },
+    {
+      "epoch": 0.8810982632330585,
+      "grad_norm": 0.7203333304321364,
+      "learning_rate": 7.3242163430313205e-06,
+      "loss": 1.241,
+      "step": 19126
+    },
+    {
+      "epoch": 0.8811443313216935,
+      "grad_norm": 0.7959130704562273,
+      "learning_rate": 7.318612204852837e-06,
+      "loss": 1.0336,
+      "step": 19127
+    },
+    {
+      "epoch": 0.8811903994103285,
+      "grad_norm": 0.7930802672050189,
+      "learning_rate": 7.313010130066533e-06,
+      "loss": 1.3215,
+      "step": 19128
+    },
+    {
+      "epoch": 0.8812364674989634,
+      "grad_norm": 1.6357394664548066,
+      "learning_rate": 7.30741011879712e-06,
+      "loss": 1.5108,
+      "step": 19129
+    },
+    {
+      "epoch": 0.8812825355875985,
+      "grad_norm": 1.2922538722868755,
+      "learning_rate": 7.301812171169287e-06,
+      "loss": 1.2729,
+      "step": 19130
+    },
+    {
+      "epoch": 0.8813286036762334,
+      "grad_norm": 1.3316565626707488,
+      "learning_rate": 7.296216287307656e-06,
+      "loss": 1.3117,
+      "step": 19131
+    },
+    {
+      "epoch": 0.8813746717648685,
+      "grad_norm": 1.2721181748473405,
+      "learning_rate": 7.290622467336838e-06,
+      "loss": 1.1646,
+      "step": 19132
+    },
+    {
+      "epoch": 0.8814207398535034,
+      "grad_norm": 0.7244658788080998,
+      "learning_rate": 7.285030711381336e-06,
+      "loss": 1.2258,
+      "step": 19133
+    },
+    {
+      "epoch": 0.8814668079421385,
+      "grad_norm": 1.5193756948815313,
+      "learning_rate": 7.279441019565647e-06,
+      "loss": 0.9852,
+      "step": 19134
+    },
+    {
+      "epoch": 0.8815128760307734,
+      "grad_norm": 0.8524443098442501,
+      "learning_rate": 7.2738533920142186e-06,
+      "loss": 1.2196,
+      "step": 19135
+    },
+    {
+      "epoch": 0.8815589441194085,
+      "grad_norm": 1.4615033446566155,
+      "learning_rate": 7.2682678288514825e-06,
+      "loss": 1.3898,
+      "step": 19136
+    },
+    {
+      "epoch": 0.8816050122080434,
+      "grad_norm": 0.6410264852524274,
+      "learning_rate": 7.262684330201741e-06,
+      "loss": 1.1792,
+      "step": 19137
+    },
+    {
+      "epoch": 0.8816510802966785,
+      "grad_norm": 0.9487921463424305,
+      "learning_rate": 7.257102896189316e-06,
+      "loss": 1.1552,
+      "step": 19138
+    },
+    {
+      "epoch": 0.8816971483853135,
+      "grad_norm": 0.9703920258009883,
+      "learning_rate": 7.251523526938497e-06,
+      "loss": 1.2274,
+      "step": 19139
+    },
+    {
+      "epoch": 0.8817432164739485,
+      "grad_norm": 0.9184305450331751,
+      "learning_rate": 7.245946222573452e-06,
+      "loss": 1.0953,
+      "step": 19140
+    },
+    {
+      "epoch": 0.8817892845625835,
+      "grad_norm": 1.079254068742195,
+      "learning_rate": 7.240370983218392e-06,
+      "loss": 1.1405,
+      "step": 19141
+    },
+    {
+      "epoch": 0.8818353526512185,
+      "grad_norm": 1.1711295457486086,
+      "learning_rate": 7.234797808997407e-06,
+      "loss": 1.3868,
+      "step": 19142
+    },
+    {
+      "epoch": 0.8818814207398535,
+      "grad_norm": 1.3116642644703278,
+      "learning_rate": 7.229226700034619e-06,
+      "loss": 1.4945,
+      "step": 19143
+    },
+    {
+      "epoch": 0.8819274888284885,
+      "grad_norm": 0.9268174890819472,
+      "learning_rate": 7.2236576564540085e-06,
+      "loss": 1.1646,
+      "step": 19144
+    },
+    {
+      "epoch": 0.8819735569171235,
+      "grad_norm": 0.821477422694507,
+      "learning_rate": 7.218090678379586e-06,
+      "loss": 1.3218,
+      "step": 19145
+    },
+    {
+      "epoch": 0.8820196250057585,
+      "grad_norm": 0.868460865925489,
+      "learning_rate": 7.2125257659352965e-06,
+      "loss": 1.3287,
+      "step": 19146
+    },
+    {
+      "epoch": 0.8820656930943935,
+      "grad_norm": 1.1437630543611934,
+      "learning_rate": 7.206962919245008e-06,
+      "loss": 1.3614,
+      "step": 19147
+    },
+    {
+      "epoch": 0.8821117611830285,
+      "grad_norm": 0.9735387612599933,
+      "learning_rate": 7.201402138432589e-06,
+      "loss": 1.1089,
+      "step": 19148
+    },
+    {
+      "epoch": 0.8821578292716635,
+      "grad_norm": 1.0863908657052412,
+      "learning_rate": 7.195843423621839e-06,
+      "loss": 1.2887,
+      "step": 19149
+    },
+    {
+      "epoch": 0.8822038973602985,
+      "grad_norm": 1.8562950174409407,
+      "learning_rate": 7.190286774936528e-06,
+      "loss": 1.0261,
+      "step": 19150
+    },
+    {
+      "epoch": 0.8822499654489335,
+      "grad_norm": 0.932089636841939,
+      "learning_rate": 7.184732192500321e-06,
+      "loss": 1.2285,
+      "step": 19151
+    },
+    {
+      "epoch": 0.8822960335375686,
+      "grad_norm": 1.5721054402315973,
+      "learning_rate": 7.17917967643692e-06,
+      "loss": 0.9692,
+      "step": 19152
+    },
+    {
+      "epoch": 0.8823421016262035,
+      "grad_norm": 0.8942908039774936,
+      "learning_rate": 7.173629226869916e-06,
+      "loss": 1.3725,
+      "step": 19153
+    },
+    {
+      "epoch": 0.8823881697148386,
+      "grad_norm": 1.0553717090979158,
+      "learning_rate": 7.16808084392292e-06,
+      "loss": 1.3749,
+      "step": 19154
+    },
+    {
+      "epoch": 0.8824342378034735,
+      "grad_norm": 2.4061147088210095,
+      "learning_rate": 7.162534527719411e-06,
+      "loss": 1.2994,
+      "step": 19155
+    },
+    {
+      "epoch": 0.8824803058921086,
+      "grad_norm": 1.460820261694473,
+      "learning_rate": 7.156990278382891e-06,
+      "loss": 0.9306,
+      "step": 19156
+    },
+    {
+      "epoch": 0.8825263739807435,
+      "grad_norm": 0.9898235748707983,
+      "learning_rate": 7.151448096036806e-06,
+      "loss": 1.2569,
+      "step": 19157
+    },
+    {
+      "epoch": 0.8825724420693786,
+      "grad_norm": 2.71516167405548,
+      "learning_rate": 7.145907980804522e-06,
+      "loss": 1.0222,
+      "step": 19158
+    },
+    {
+      "epoch": 0.8826185101580135,
+      "grad_norm": 5.107019497366828,
+      "learning_rate": 7.1403699328093765e-06,
+      "loss": 1.2121,
+      "step": 19159
+    },
+    {
+      "epoch": 0.8826645782466486,
+      "grad_norm": 1.707142887117788,
+      "learning_rate": 7.1348339521746686e-06,
+      "loss": 1.143,
+      "step": 19160
+    },
+    {
+      "epoch": 0.8827106463352835,
+      "grad_norm": 0.9353751625415815,
+      "learning_rate": 7.129300039023678e-06,
+      "loss": 1.1835,
+      "step": 19161
+    },
+    {
+      "epoch": 0.8827567144239186,
+      "grad_norm": 0.8148203793798038,
+      "learning_rate": 7.123768193479563e-06,
+      "loss": 1.1533,
+      "step": 19162
+    },
+    {
+      "epoch": 0.8828027825125535,
+      "grad_norm": 0.8033943032662052,
+      "learning_rate": 7.118238415665501e-06,
+      "loss": 1.1515,
+      "step": 19163
+    },
+    {
+      "epoch": 0.8828488506011886,
+      "grad_norm": 0.9448901070578396,
+      "learning_rate": 7.112710705704617e-06,
+      "loss": 1.1778,
+      "step": 19164
+    },
+    {
+      "epoch": 0.8828949186898235,
+      "grad_norm": 0.8087948215987362,
+      "learning_rate": 7.107185063719946e-06,
+      "loss": 1.4158,
+      "step": 19165
+    },
+    {
+      "epoch": 0.8829409867784586,
+      "grad_norm": 1.0489292289503847,
+      "learning_rate": 7.101661489834532e-06,
+      "loss": 1.4078,
+      "step": 19166
+    },
+    {
+      "epoch": 0.8829870548670936,
+      "grad_norm": 0.8671867633207279,
+      "learning_rate": 7.096139984171324e-06,
+      "loss": 1.4175,
+      "step": 19167
+    },
+    {
+      "epoch": 0.8830331229557286,
+      "grad_norm": 1.0310368060996449,
+      "learning_rate": 7.090620546853288e-06,
+      "loss": 1.1416,
+      "step": 19168
+    },
+    {
+      "epoch": 0.8830791910443636,
+      "grad_norm": 0.7630791659305637,
+      "learning_rate": 7.085103178003261e-06,
+      "loss": 1.1687,
+      "step": 19169
+    },
+    {
+      "epoch": 0.8831252591329986,
+      "grad_norm": 1.1013533571392917,
+      "learning_rate": 7.079587877744098e-06,
+      "loss": 1.2782,
+      "step": 19170
+    },
+    {
+      "epoch": 0.8831713272216336,
+      "grad_norm": 1.6365893840754562,
+      "learning_rate": 7.074074646198591e-06,
+      "loss": 1.0779,
+      "step": 19171
+    },
+    {
+      "epoch": 0.8832173953102685,
+      "grad_norm": 0.905002077889476,
+      "learning_rate": 7.068563483489488e-06,
+      "loss": 1.5493,
+      "step": 19172
+    },
+    {
+      "epoch": 0.8832634633989036,
+      "grad_norm": 1.376357312759017,
+      "learning_rate": 7.0630543897394655e-06,
+      "loss": 0.9166,
+      "step": 19173
+    },
+    {
+      "epoch": 0.8833095314875385,
+      "grad_norm": 21.302113253515593,
+      "learning_rate": 7.057547365071171e-06,
+      "loss": 0.9376,
+      "step": 19174
+    },
+    {
+      "epoch": 0.8833555995761736,
+      "grad_norm": 0.9927373328436419,
+      "learning_rate": 7.052042409607251e-06,
+      "loss": 1.3614,
+      "step": 19175
+    },
+    {
+      "epoch": 0.8834016676648085,
+      "grad_norm": 1.3587971358940216,
+      "learning_rate": 7.046539523470219e-06,
+      "loss": 1.2716,
+      "step": 19176
+    },
+    {
+      "epoch": 0.8834477357534436,
+      "grad_norm": 1.561832164095412,
+      "learning_rate": 7.041038706782599e-06,
+      "loss": 1.2605,
+      "step": 19177
+    },
+    {
+      "epoch": 0.8834938038420785,
+      "grad_norm": 1.0431344178780086,
+      "learning_rate": 7.035539959666859e-06,
+      "loss": 1.0986,
+      "step": 19178
+    },
+    {
+      "epoch": 0.8835398719307136,
+      "grad_norm": 0.8743035532875655,
+      "learning_rate": 7.030043282245446e-06,
+      "loss": 0.9373,
+      "step": 19179
+    },
+    {
+      "epoch": 0.8835859400193486,
+      "grad_norm": 1.108804374402267,
+      "learning_rate": 7.024548674640685e-06,
+      "loss": 1.351,
+      "step": 19180
+    },
+    {
+      "epoch": 0.8836320081079836,
+      "grad_norm": 1.1246208890290041,
+      "learning_rate": 7.019056136974944e-06,
+      "loss": 1.1446,
+      "step": 19181
+    },
+    {
+      "epoch": 0.8836780761966186,
+      "grad_norm": 1.606444753898425,
+      "learning_rate": 7.013565669370492e-06,
+      "loss": 1.2079,
+      "step": 19182
+    },
+    {
+      "epoch": 0.8837241442852536,
+      "grad_norm": 0.9002350807713363,
+      "learning_rate": 7.008077271949554e-06,
+      "loss": 1.0184,
+      "step": 19183
+    },
+    {
+      "epoch": 0.8837702123738886,
+      "grad_norm": 0.9722412150408747,
+      "learning_rate": 7.002590944834342e-06,
+      "loss": 1.3215,
+      "step": 19184
+    },
+    {
+      "epoch": 0.8838162804625236,
+      "grad_norm": 1.1989893076103344,
+      "learning_rate": 6.997106688146948e-06,
+      "loss": 1.1557,
+      "step": 19185
+    },
+    {
+      "epoch": 0.8838623485511586,
+      "grad_norm": 0.7535378919233395,
+      "learning_rate": 6.991624502009553e-06,
+      "loss": 1.3198,
+      "step": 19186
+    },
+    {
+      "epoch": 0.8839084166397936,
+      "grad_norm": 0.9917514458934076,
+      "learning_rate": 6.9861443865441354e-06,
+      "loss": 1.2157,
+      "step": 19187
+    },
+    {
+      "epoch": 0.8839544847284286,
+      "grad_norm": 1.7740831124201795,
+      "learning_rate": 6.9806663418727325e-06,
+      "loss": 1.1179,
+      "step": 19188
+    },
+    {
+      "epoch": 0.8840005528170636,
+      "grad_norm": 1.192199826954937,
+      "learning_rate": 6.975190368117302e-06,
+      "loss": 1.1079,
+      "step": 19189
+    },
+    {
+      "epoch": 0.8840466209056986,
+      "grad_norm": 1.170240313476684,
+      "learning_rate": 6.969716465399767e-06,
+      "loss": 0.9173,
+      "step": 19190
+    },
+    {
+      "epoch": 0.8840926889943336,
+      "grad_norm": 1.0265845958015751,
+      "learning_rate": 6.964244633841988e-06,
+      "loss": 1.0538,
+      "step": 19191
+    },
+    {
+      "epoch": 0.8841387570829686,
+      "grad_norm": 1.0514379051691374,
+      "learning_rate": 6.9587748735657435e-06,
+      "loss": 1.2643,
+      "step": 19192
+    },
+    {
+      "epoch": 0.8841848251716037,
+      "grad_norm": 1.0053308965930536,
+      "learning_rate": 6.95330718469287e-06,
+      "loss": 1.2146,
+      "step": 19193
+    },
+    {
+      "epoch": 0.8842308932602386,
+      "grad_norm": 1.6072753740076962,
+      "learning_rate": 6.9478415673450595e-06,
+      "loss": 1.2706,
+      "step": 19194
+    },
+    {
+      "epoch": 0.8842769613488737,
+      "grad_norm": 0.7676453614411103,
+      "learning_rate": 6.942378021644036e-06,
+      "loss": 1.1553,
+      "step": 19195
+    },
+    {
+      "epoch": 0.8843230294375086,
+      "grad_norm": 1.0689137012160894,
+      "learning_rate": 6.936916547711358e-06,
+      "loss": 0.988,
+      "step": 19196
+    },
+    {
+      "epoch": 0.8843690975261437,
+      "grad_norm": 0.8813769092858794,
+      "learning_rate": 6.9314571456686964e-06,
+      "loss": 1.3305,
+      "step": 19197
+    },
+    {
+      "epoch": 0.8844151656147786,
+      "grad_norm": 0.7269197900261757,
+      "learning_rate": 6.925999815637552e-06,
+      "loss": 1.3267,
+      "step": 19198
+    },
+    {
+      "epoch": 0.8844612337034137,
+      "grad_norm": 0.9144357371214881,
+      "learning_rate": 6.920544557739439e-06,
+      "loss": 1.2306,
+      "step": 19199
+    },
+    {
+      "epoch": 0.8845073017920486,
+      "grad_norm": 1.3616117162816188,
+      "learning_rate": 6.915091372095817e-06,
+      "loss": 1.2907,
+      "step": 19200
+    },
+    {
+      "epoch": 0.8845533698806837,
+      "grad_norm": 1.0209380299972783,
+      "learning_rate": 6.909640258828065e-06,
+      "loss": 1.0221,
+      "step": 19201
+    },
+    {
+      "epoch": 0.8845994379693186,
+      "grad_norm": 2.0208411651315425,
+      "learning_rate": 6.904191218057565e-06,
+      "loss": 0.9419,
+      "step": 19202
+    },
+    {
+      "epoch": 0.8846455060579537,
+      "grad_norm": 1.2546214988210473,
+      "learning_rate": 6.898744249905597e-06,
+      "loss": 1.2933,
+      "step": 19203
+    },
+    {
+      "epoch": 0.8846915741465886,
+      "grad_norm": 2.034467268582896,
+      "learning_rate": 6.893299354493488e-06,
+      "loss": 1.2103,
+      "step": 19204
+    },
+    {
+      "epoch": 0.8847376422352237,
+      "grad_norm": 1.439234128993256,
+      "learning_rate": 6.887856531942405e-06,
+      "loss": 1.2052,
+      "step": 19205
+    },
+    {
+      "epoch": 0.8847837103238586,
+      "grad_norm": 1.5983038625652446,
+      "learning_rate": 6.882415782373553e-06,
+      "loss": 0.95,
+      "step": 19206
+    },
+    {
+      "epoch": 0.8848297784124937,
+      "grad_norm": 0.9591264139300819,
+      "learning_rate": 6.8769771059080225e-06,
+      "loss": 1.3126,
+      "step": 19207
+    },
+    {
+      "epoch": 0.8848758465011287,
+      "grad_norm": 0.9169161582792968,
+      "learning_rate": 6.871540502666962e-06,
+      "loss": 0.8657,
+      "step": 19208
+    },
+    {
+      "epoch": 0.8849219145897637,
+      "grad_norm": 0.8697573350652503,
+      "learning_rate": 6.866105972771375e-06,
+      "loss": 1.6091,
+      "step": 19209
+    },
+    {
+      "epoch": 0.8849679826783987,
+      "grad_norm": 0.7519192596325166,
+      "learning_rate": 6.860673516342219e-06,
+      "loss": 1.2344,
+      "step": 19210
+    },
+    {
+      "epoch": 0.8850140507670337,
+      "grad_norm": 1.614155459907136,
+      "learning_rate": 6.8552431335004776e-06,
+      "loss": 1.4344,
+      "step": 19211
+    },
+    {
+      "epoch": 0.8850601188556687,
+      "grad_norm": 0.6672946161742053,
+      "learning_rate": 6.84981482436704e-06,
+      "loss": 1.502,
+      "step": 19212
+    },
+    {
+      "epoch": 0.8851061869443037,
+      "grad_norm": 0.7913407803980621,
+      "learning_rate": 6.844388589062767e-06,
+      "loss": 1.3268,
+      "step": 19213
+    },
+    {
+      "epoch": 0.8851522550329387,
+      "grad_norm": 1.0328894478579804,
+      "learning_rate": 6.838964427708428e-06,
+      "loss": 1.0144,
+      "step": 19214
+    },
+    {
+      "epoch": 0.8851983231215736,
+      "grad_norm": 0.9444357769101334,
+      "learning_rate": 6.833542340424837e-06,
+      "loss": 1.215,
+      "step": 19215
+    },
+    {
+      "epoch": 0.8852443912102087,
+      "grad_norm": 0.8055653008773245,
+      "learning_rate": 6.828122327332665e-06,
+      "loss": 1.1456,
+      "step": 19216
+    },
+    {
+      "epoch": 0.8852904592988436,
+      "grad_norm": 2.5758489085121723,
+      "learning_rate": 6.822704388552604e-06,
+      "loss": 1.0307,
+      "step": 19217
+    },
+    {
+      "epoch": 0.8853365273874787,
+      "grad_norm": 1.8468448623457445,
+      "learning_rate": 6.817288524205235e-06,
+      "loss": 1.1227,
+      "step": 19218
+    },
+    {
+      "epoch": 0.8853825954761136,
+      "grad_norm": 0.8134324005281143,
+      "learning_rate": 6.811874734411206e-06,
+      "loss": 1.3037,
+      "step": 19219
+    },
+    {
+      "epoch": 0.8854286635647487,
+      "grad_norm": 1.178473811303439,
+      "learning_rate": 6.806463019290987e-06,
+      "loss": 1.2332,
+      "step": 19220
+    },
+    {
+      "epoch": 0.8854747316533838,
+      "grad_norm": 1.1327766428802402,
+      "learning_rate": 6.801053378965072e-06,
+      "loss": 1.167,
+      "step": 19221
+    },
+    {
+      "epoch": 0.8855207997420187,
+      "grad_norm": 0.9374846428343419,
+      "learning_rate": 6.795645813553897e-06,
+      "loss": 1.2115,
+      "step": 19222
+    },
+    {
+      "epoch": 0.8855668678306537,
+      "grad_norm": 0.9876972594269201,
+      "learning_rate": 6.7902403231778545e-06,
+      "loss": 0.9488,
+      "step": 19223
+    },
+    {
+      "epoch": 0.8856129359192887,
+      "grad_norm": 1.1355806588797648,
+      "learning_rate": 6.784836907957315e-06,
+      "loss": 1.2901,
+      "step": 19224
+    },
+    {
+      "epoch": 0.8856590040079237,
+      "grad_norm": 1.000013447459749,
+      "learning_rate": 6.779435568012516e-06,
+      "loss": 1.1438,
+      "step": 19225
+    },
+    {
+      "epoch": 0.8857050720965587,
+      "grad_norm": 0.8467888276357575,
+      "learning_rate": 6.774036303463771e-06,
+      "loss": 1.3502,
+      "step": 19226
+    },
+    {
+      "epoch": 0.8857511401851937,
+      "grad_norm": 0.8744864587274123,
+      "learning_rate": 6.7686391144312525e-06,
+      "loss": 0.8844,
+      "step": 19227
+    },
+    {
+      "epoch": 0.8857972082738287,
+      "grad_norm": 1.096321111237468,
+      "learning_rate": 6.76324400103513e-06,
+      "loss": 1.1163,
+      "step": 19228
+    },
+    {
+      "epoch": 0.8858432763624637,
+      "grad_norm": 1.0369343605555141,
+      "learning_rate": 6.757850963395507e-06,
+      "loss": 1.1991,
+      "step": 19229
+    },
+    {
+      "epoch": 0.8858893444510987,
+      "grad_norm": 5.247032080191576,
+      "learning_rate": 6.752460001632454e-06,
+      "loss": 1.1519,
+      "step": 19230
+    },
+    {
+      "epoch": 0.8859354125397337,
+      "grad_norm": 1.6851795839187402,
+      "learning_rate": 6.74707111586601e-06,
+      "loss": 1.3787,
+      "step": 19231
+    },
+    {
+      "epoch": 0.8859814806283687,
+      "grad_norm": 1.1625533994362243,
+      "learning_rate": 6.741684306216123e-06,
+      "loss": 1.1225,
+      "step": 19232
+    },
+    {
+      "epoch": 0.8860275487170037,
+      "grad_norm": 0.5837555066055262,
+      "learning_rate": 6.736299572802718e-06,
+      "loss": 1.4853,
+      "step": 19233
+    },
+    {
+      "epoch": 0.8860736168056388,
+      "grad_norm": 1.1015496899830368,
+      "learning_rate": 6.7309169157457e-06,
+      "loss": 1.2931,
+      "step": 19234
+    },
+    {
+      "epoch": 0.8861196848942737,
+      "grad_norm": 0.7497156006381597,
+      "learning_rate": 6.725536335164906e-06,
+      "loss": 1.2598,
+      "step": 19235
+    },
+    {
+      "epoch": 0.8861657529829088,
+      "grad_norm": 1.7221574394676362,
+      "learning_rate": 6.720157831180107e-06,
+      "loss": 0.997,
+      "step": 19236
+    },
+    {
+      "epoch": 0.8862118210715437,
+      "grad_norm": 1.332806078333403,
+      "learning_rate": 6.714781403911042e-06,
+      "loss": 1.3279,
+      "step": 19237
+    },
+    {
+      "epoch": 0.8862578891601788,
+      "grad_norm": 2.0565593073581137,
+      "learning_rate": 6.709407053477446e-06,
+      "loss": 1.0983,
+      "step": 19238
+    },
+    {
+      "epoch": 0.8863039572488137,
+      "grad_norm": 1.180451755327135,
+      "learning_rate": 6.704034779998913e-06,
+      "loss": 1.1631,
+      "step": 19239
+    },
+    {
+      "epoch": 0.8863500253374488,
+      "grad_norm": 0.8504574771830503,
+      "learning_rate": 6.698664583595093e-06,
+      "loss": 1.4002,
+      "step": 19240
+    },
+    {
+      "epoch": 0.8863960934260837,
+      "grad_norm": 1.3027554604295641,
+      "learning_rate": 6.6932964643855125e-06,
+      "loss": 1.6107,
+      "step": 19241
+    },
+    {
+      "epoch": 0.8864421615147188,
+      "grad_norm": 0.832489486355368,
+      "learning_rate": 6.68793042248973e-06,
+      "loss": 1.5205,
+      "step": 19242
+    },
+    {
+      "epoch": 0.8864882296033537,
+      "grad_norm": 0.9599868724846029,
+      "learning_rate": 6.682566458027151e-06,
+      "loss": 1.3861,
+      "step": 19243
+    },
+    {
+      "epoch": 0.8865342976919888,
+      "grad_norm": 0.9915048572746336,
+      "learning_rate": 6.677204571117235e-06,
+      "loss": 1.5771,
+      "step": 19244
+    },
+    {
+      "epoch": 0.8865803657806237,
+      "grad_norm": 1.0842462121721463,
+      "learning_rate": 6.6718447618793425e-06,
+      "loss": 0.9792,
+      "step": 19245
+    },
+    {
+      "epoch": 0.8866264338692588,
+      "grad_norm": 0.7768290674048284,
+      "learning_rate": 6.666487030432822e-06,
+      "loss": 1.0687,
+      "step": 19246
+    },
+    {
+      "epoch": 0.8866725019578938,
+      "grad_norm": 1.0786235783909903,
+      "learning_rate": 6.661131376896912e-06,
+      "loss": 1.1937,
+      "step": 19247
+    },
+    {
+      "epoch": 0.8867185700465288,
+      "grad_norm": 0.8570831117461399,
+      "learning_rate": 6.655777801390872e-06,
+      "loss": 0.9559,
+      "step": 19248
+    },
+    {
+      "epoch": 0.8867646381351638,
+      "grad_norm": 1.2144332656961352,
+      "learning_rate": 6.650426304033908e-06,
+      "loss": 0.956,
+      "step": 19249
+    },
+    {
+      "epoch": 0.8868107062237988,
+      "grad_norm": 0.7759656738118103,
+      "learning_rate": 6.6450768849451225e-06,
+      "loss": 1.2534,
+      "step": 19250
+    },
+    {
+      "epoch": 0.8868567743124338,
+      "grad_norm": 0.8958652860184129,
+      "learning_rate": 6.6397295442436335e-06,
+      "loss": 1.1372,
+      "step": 19251
+    },
+    {
+      "epoch": 0.8869028424010688,
+      "grad_norm": 0.7375385278530255,
+      "learning_rate": 6.634384282048489e-06,
+      "loss": 1.6335,
+      "step": 19252
+    },
+    {
+      "epoch": 0.8869489104897038,
+      "grad_norm": 0.7946651373755552,
+      "learning_rate": 6.629041098478705e-06,
+      "loss": 1.19,
+      "step": 19253
+    },
+    {
+      "epoch": 0.8869949785783388,
+      "grad_norm": 1.4893684456159855,
+      "learning_rate": 6.62369999365321e-06,
+      "loss": 1.1599,
+      "step": 19254
+    },
+    {
+      "epoch": 0.8870410466669738,
+      "grad_norm": 1.8643391100854536,
+      "learning_rate": 6.6183609676909176e-06,
+      "loss": 1.6559,
+      "step": 19255
+    },
+    {
+      "epoch": 0.8870871147556088,
+      "grad_norm": 0.8662753317832799,
+      "learning_rate": 6.613024020710734e-06,
+      "loss": 1.2583,
+      "step": 19256
+    },
+    {
+      "epoch": 0.8871331828442438,
+      "grad_norm": 1.1936227545238738,
+      "learning_rate": 6.60768915283142e-06,
+      "loss": 1.3831,
+      "step": 19257
+    },
+    {
+      "epoch": 0.8871792509328787,
+      "grad_norm": 1.3463061202042728,
+      "learning_rate": 6.60235636417178e-06,
+      "loss": 1.3297,
+      "step": 19258
+    },
+    {
+      "epoch": 0.8872253190215138,
+      "grad_norm": 0.7804709068169535,
+      "learning_rate": 6.59702565485053e-06,
+      "loss": 1.1902,
+      "step": 19259
+    },
+    {
+      "epoch": 0.8872713871101487,
+      "grad_norm": 0.9995651544542018,
+      "learning_rate": 6.591697024986365e-06,
+      "loss": 1.2012,
+      "step": 19260
+    },
+    {
+      "epoch": 0.8873174551987838,
+      "grad_norm": 3.198058851524536,
+      "learning_rate": 6.5863704746978895e-06,
+      "loss": 0.9737,
+      "step": 19261
+    },
+    {
+      "epoch": 0.8873635232874189,
+      "grad_norm": 1.0454041881034608,
+      "learning_rate": 6.581046004103708e-06,
+      "loss": 1.404,
+      "step": 19262
+    },
+    {
+      "epoch": 0.8874095913760538,
+      "grad_norm": 1.2243268727693266,
+      "learning_rate": 6.57572361332236e-06,
+      "loss": 1.0674,
+      "step": 19263
+    },
+    {
+      "epoch": 0.8874556594646889,
+      "grad_norm": 1.2275924752674732,
+      "learning_rate": 6.57040330247235e-06,
+      "loss": 0.9313,
+      "step": 19264
+    },
+    {
+      "epoch": 0.8875017275533238,
+      "grad_norm": 0.9045667935375256,
+      "learning_rate": 6.565085071672106e-06,
+      "loss": 1.1847,
+      "step": 19265
+    },
+    {
+      "epoch": 0.8875477956419588,
+      "grad_norm": 0.9842054991585604,
+      "learning_rate": 6.559768921040021e-06,
+      "loss": 0.8107,
+      "step": 19266
+    },
+    {
+      "epoch": 0.8875938637305938,
+      "grad_norm": 1.795207764173871,
+      "learning_rate": 6.554454850694492e-06,
+      "loss": 0.8398,
+      "step": 19267
+    },
+    {
+      "epoch": 0.8876399318192288,
+      "grad_norm": 1.105209324308165,
+      "learning_rate": 6.549142860753776e-06,
+      "loss": 1.3061,
+      "step": 19268
+    },
+    {
+      "epoch": 0.8876859999078638,
+      "grad_norm": 0.8422696284123181,
+      "learning_rate": 6.543832951336171e-06,
+      "loss": 1.0047,
+      "step": 19269
+    },
+    {
+      "epoch": 0.8877320679964988,
+      "grad_norm": 0.729358395325074,
+      "learning_rate": 6.53852512255988e-06,
+      "loss": 1.3893,
+      "step": 19270
+    },
+    {
+      "epoch": 0.8877781360851338,
+      "grad_norm": 1.538026407988377,
+      "learning_rate": 6.533219374543087e-06,
+      "loss": 1.1621,
+      "step": 19271
+    },
+    {
+      "epoch": 0.8878242041737688,
+      "grad_norm": 1.1601170282799915,
+      "learning_rate": 6.527915707403887e-06,
+      "loss": 1.2607,
+      "step": 19272
+    },
+    {
+      "epoch": 0.8878702722624038,
+      "grad_norm": 1.154570061508441,
+      "learning_rate": 6.522614121260385e-06,
+      "loss": 1.1959,
+      "step": 19273
+    },
+    {
+      "epoch": 0.8879163403510388,
+      "grad_norm": 0.9537978141781089,
+      "learning_rate": 6.517314616230596e-06,
+      "loss": 1.3384,
+      "step": 19274
+    },
+    {
+      "epoch": 0.8879624084396739,
+      "grad_norm": 1.0250846668664229,
+      "learning_rate": 6.512017192432507e-06,
+      "loss": 1.394,
+      "step": 19275
+    },
+    {
+      "epoch": 0.8880084765283088,
+      "grad_norm": 0.7190203826705532,
+      "learning_rate": 6.506721849984054e-06,
+      "loss": 1.2119,
+      "step": 19276
+    },
+    {
+      "epoch": 0.8880545446169439,
+      "grad_norm": 0.8479581868038719,
+      "learning_rate": 6.5014285890031336e-06,
+      "loss": 0.9722,
+      "step": 19277
+    },
+    {
+      "epoch": 0.8881006127055788,
+      "grad_norm": 1.5057967614481544,
+      "learning_rate": 6.4961374096076055e-06,
+      "loss": 1.1666,
+      "step": 19278
+    },
+    {
+      "epoch": 0.8881466807942139,
+      "grad_norm": 1.1109865561290242,
+      "learning_rate": 6.490848311915243e-06,
+      "loss": 1.1381,
+      "step": 19279
+    },
+    {
+      "epoch": 0.8881927488828488,
+      "grad_norm": 0.9722367917610506,
+      "learning_rate": 6.485561296043796e-06,
+      "loss": 1.0556,
+      "step": 19280
+    },
+    {
+      "epoch": 0.8882388169714839,
+      "grad_norm": 0.8807960083956687,
+      "learning_rate": 6.480276362110993e-06,
+      "loss": 1.4291,
+      "step": 19281
+    },
+    {
+      "epoch": 0.8882848850601188,
+      "grad_norm": 1.6788175208011393,
+      "learning_rate": 6.474993510234495e-06,
+      "loss": 1.3603,
+      "step": 19282
+    },
+    {
+      "epoch": 0.8883309531487539,
+      "grad_norm": 0.770046930668198,
+      "learning_rate": 6.469712740531908e-06,
+      "loss": 1.5837,
+      "step": 19283
+    },
+    {
+      "epoch": 0.8883770212373888,
+      "grad_norm": 1.2623879178465665,
+      "learning_rate": 6.464434053120772e-06,
+      "loss": 0.9938,
+      "step": 19284
+    },
+    {
+      "epoch": 0.8884230893260239,
+      "grad_norm": 0.7778587240769279,
+      "learning_rate": 6.459157448118658e-06,
+      "loss": 1.2496,
+      "step": 19285
+    },
+    {
+      "epoch": 0.8884691574146588,
+      "grad_norm": 1.267911823482378,
+      "learning_rate": 6.453882925642995e-06,
+      "loss": 0.8964,
+      "step": 19286
+    },
+    {
+      "epoch": 0.8885152255032939,
+      "grad_norm": 0.8369585093030877,
+      "learning_rate": 6.448610485811246e-06,
+      "loss": 1.3442,
+      "step": 19287
+    },
+    {
+      "epoch": 0.8885612935919289,
+      "grad_norm": 1.176661879689042,
+      "learning_rate": 6.4433401287407605e-06,
+      "loss": 1.3288,
+      "step": 19288
+    },
+    {
+      "epoch": 0.8886073616805639,
+      "grad_norm": 0.8814912865558061,
+      "learning_rate": 6.438071854548921e-06,
+      "loss": 1.163,
+      "step": 19289
+    },
+    {
+      "epoch": 0.8886534297691989,
+      "grad_norm": 5.874970556735866,
+      "learning_rate": 6.432805663352959e-06,
+      "loss": 1.0103,
+      "step": 19290
+    },
+    {
+      "epoch": 0.8886994978578339,
+      "grad_norm": 1.5538091426441218,
+      "learning_rate": 6.427541555270156e-06,
+      "loss": 0.9764,
+      "step": 19291
+    },
+    {
+      "epoch": 0.8887455659464689,
+      "grad_norm": 2.0119521482060994,
+      "learning_rate": 6.422279530417707e-06,
+      "loss": 1.2128,
+      "step": 19292
+    },
+    {
+      "epoch": 0.8887916340351039,
+      "grad_norm": 0.8939141954884787,
+      "learning_rate": 6.417019588912742e-06,
+      "loss": 1.5557,
+      "step": 19293
+    },
+    {
+      "epoch": 0.8888377021237389,
+      "grad_norm": 1.2293417138323144,
+      "learning_rate": 6.41176173087239e-06,
+      "loss": 1.3604,
+      "step": 19294
+    },
+    {
+      "epoch": 0.8888837702123739,
+      "grad_norm": 1.5581080210560325,
+      "learning_rate": 6.406505956413667e-06,
+      "loss": 1.1145,
+      "step": 19295
+    },
+    {
+      "epoch": 0.8889298383010089,
+      "grad_norm": 1.1968451965297469,
+      "learning_rate": 6.401252265653634e-06,
+      "loss": 1.4141,
+      "step": 19296
+    },
+    {
+      "epoch": 0.8889759063896439,
+      "grad_norm": 1.582200621049692,
+      "learning_rate": 6.396000658709222e-06,
+      "loss": 1.1627,
+      "step": 19297
+    },
+    {
+      "epoch": 0.8890219744782789,
+      "grad_norm": 0.7206150505949256,
+      "learning_rate": 6.390751135697359e-06,
+      "loss": 1.1043,
+      "step": 19298
+    },
+    {
+      "epoch": 0.8890680425669139,
+      "grad_norm": 1.0350652015275499,
+      "learning_rate": 6.385503696734907e-06,
+      "loss": 1.3536,
+      "step": 19299
+    },
+    {
+      "epoch": 0.8891141106555489,
+      "grad_norm": 0.7694912623772724,
+      "learning_rate": 6.380258341938728e-06,
+      "loss": 1.3815,
+      "step": 19300
+    },
+    {
+      "epoch": 0.8891601787441838,
+      "grad_norm": 0.764893706294047,
+      "learning_rate": 6.3750150714255715e-06,
+      "loss": 1.2302,
+      "step": 19301
+    },
+    {
+      "epoch": 0.8892062468328189,
+      "grad_norm": 1.0122916520277412,
+      "learning_rate": 6.3697738853121355e-06,
+      "loss": 1.1918,
+      "step": 19302
+    },
+    {
+      "epoch": 0.889252314921454,
+      "grad_norm": 0.9844797066376526,
+      "learning_rate": 6.3645347837151705e-06,
+      "loss": 1.4581,
+      "step": 19303
+    },
+    {
+      "epoch": 0.8892983830100889,
+      "grad_norm": 1.1208460553310902,
+      "learning_rate": 6.359297766751282e-06,
+      "loss": 1.1675,
+      "step": 19304
+    },
+    {
+      "epoch": 0.889344451098724,
+      "grad_norm": 1.0356792098525187,
+      "learning_rate": 6.354062834537078e-06,
+      "loss": 1.4997,
+      "step": 19305
+    },
+    {
+      "epoch": 0.8893905191873589,
+      "grad_norm": 0.6232483184549423,
+      "learning_rate": 6.348829987189075e-06,
+      "loss": 1.3046,
+      "step": 19306
+    },
+    {
+      "epoch": 0.889436587275994,
+      "grad_norm": 1.297232729518307,
+      "learning_rate": 6.343599224823815e-06,
+      "loss": 1.2997,
+      "step": 19307
+    },
+    {
+      "epoch": 0.8894826553646289,
+      "grad_norm": 0.8217744149791951,
+      "learning_rate": 6.338370547557726e-06,
+      "loss": 1.3486,
+      "step": 19308
+    },
+    {
+      "epoch": 0.889528723453264,
+      "grad_norm": 1.0066080822556787,
+      "learning_rate": 6.333143955507226e-06,
+      "loss": 1.1038,
+      "step": 19309
+    },
+    {
+      "epoch": 0.8895747915418989,
+      "grad_norm": 0.6761071821865932,
+      "learning_rate": 6.327919448788666e-06,
+      "loss": 1.396,
+      "step": 19310
+    },
+    {
+      "epoch": 0.8896208596305339,
+      "grad_norm": 1.6982579501356043,
+      "learning_rate": 6.322697027518365e-06,
+      "loss": 1.2258,
+      "step": 19311
+    },
+    {
+      "epoch": 0.8896669277191689,
+      "grad_norm": 1.2987198485973872,
+      "learning_rate": 6.317476691812596e-06,
+      "loss": 1.3041,
+      "step": 19312
+    },
+    {
+      "epoch": 0.8897129958078039,
+      "grad_norm": 1.357056123764165,
+      "learning_rate": 6.312258441787555e-06,
+      "loss": 1.0372,
+      "step": 19313
+    },
+    {
+      "epoch": 0.8897590638964389,
+      "grad_norm": 0.6739673105181269,
+      "learning_rate": 6.307042277559483e-06,
+      "loss": 1.2325,
+      "step": 19314
+    },
+    {
+      "epoch": 0.8898051319850739,
+      "grad_norm": 1.2895607735004433,
+      "learning_rate": 6.301828199244431e-06,
+      "loss": 1.0561,
+      "step": 19315
+    },
+    {
+      "epoch": 0.889851200073709,
+      "grad_norm": 0.9069080458686375,
+      "learning_rate": 6.296616206958539e-06,
+      "loss": 1.3965,
+      "step": 19316
+    },
+    {
+      "epoch": 0.8898972681623439,
+      "grad_norm": 0.8934220608881857,
+      "learning_rate": 6.2914063008177925e-06,
+      "loss": 1.0464,
+      "step": 19317
+    },
+    {
+      "epoch": 0.889943336250979,
+      "grad_norm": 0.7409299839031273,
+      "learning_rate": 6.286198480938233e-06,
+      "loss": 1.2164,
+      "step": 19318
+    },
+    {
+      "epoch": 0.8899894043396139,
+      "grad_norm": 0.9664863668291018,
+      "learning_rate": 6.280992747435788e-06,
+      "loss": 1.1236,
+      "step": 19319
+    },
+    {
+      "epoch": 0.890035472428249,
+      "grad_norm": 1.2232391397849904,
+      "learning_rate": 6.275789100426333e-06,
+      "loss": 1.455,
+      "step": 19320
+    },
+    {
+      "epoch": 0.8900815405168839,
+      "grad_norm": 0.8763875570321912,
+      "learning_rate": 6.270587540025718e-06,
+      "loss": 1.3438,
+      "step": 19321
+    },
+    {
+      "epoch": 0.890127608605519,
+      "grad_norm": 0.8878876517176513,
+      "learning_rate": 6.265388066349764e-06,
+      "loss": 0.9155,
+      "step": 19322
+    },
+    {
+      "epoch": 0.8901736766941539,
+      "grad_norm": 0.9069613468176477,
+      "learning_rate": 6.260190679514244e-06,
+      "loss": 1.121,
+      "step": 19323
+    },
+    {
+      "epoch": 0.890219744782789,
+      "grad_norm": 1.9919946156114747,
+      "learning_rate": 6.25499537963482e-06,
+      "loss": 1.096,
+      "step": 19324
+    },
+    {
+      "epoch": 0.8902658128714239,
+      "grad_norm": 0.8431378378048401,
+      "learning_rate": 6.249802166827212e-06,
+      "loss": 1.2777,
+      "step": 19325
+    },
+    {
+      "epoch": 0.890311880960059,
+      "grad_norm": 1.2556739528848586,
+      "learning_rate": 6.244611041206994e-06,
+      "loss": 1.1511,
+      "step": 19326
+    },
+    {
+      "epoch": 0.8903579490486939,
+      "grad_norm": 1.024957570180839,
+      "learning_rate": 6.2394220028897724e-06,
+      "loss": 1.137,
+      "step": 19327
+    },
+    {
+      "epoch": 0.890404017137329,
+      "grad_norm": 0.9485331873285495,
+      "learning_rate": 6.234235051991044e-06,
+      "loss": 0.9719,
+      "step": 19328
+    },
+    {
+      "epoch": 0.890450085225964,
+      "grad_norm": 0.9616450882765953,
+      "learning_rate": 6.229050188626296e-06,
+      "loss": 1.1232,
+      "step": 19329
+    },
+    {
+      "epoch": 0.890496153314599,
+      "grad_norm": 0.9570797776326804,
+      "learning_rate": 6.223867412910978e-06,
+      "loss": 1.1743,
+      "step": 19330
+    },
+    {
+      "epoch": 0.890542221403234,
+      "grad_norm": 1.3829506955349409,
+      "learning_rate": 6.218686724960443e-06,
+      "loss": 1.1407,
+      "step": 19331
+    },
+    {
+      "epoch": 0.890588289491869,
+      "grad_norm": 1.0330166073565143,
+      "learning_rate": 6.213508124890044e-06,
+      "loss": 1.1609,
+      "step": 19332
+    },
+    {
+      "epoch": 0.890634357580504,
+      "grad_norm": 0.808496472921775,
+      "learning_rate": 6.208331612815088e-06,
+      "loss": 1.424,
+      "step": 19333
+    },
+    {
+      "epoch": 0.890680425669139,
+      "grad_norm": 1.0933448915053412,
+      "learning_rate": 6.203157188850817e-06,
+      "loss": 1.2923,
+      "step": 19334
+    },
+    {
+      "epoch": 0.890726493757774,
+      "grad_norm": 1.6085942756641318,
+      "learning_rate": 6.1979848531124155e-06,
+      "loss": 1.2636,
+      "step": 19335
+    },
+    {
+      "epoch": 0.890772561846409,
+      "grad_norm": 1.5787163666779014,
+      "learning_rate": 6.192814605715036e-06,
+      "loss": 1.1535,
+      "step": 19336
+    },
+    {
+      "epoch": 0.890818629935044,
+      "grad_norm": 1.5909861689124916,
+      "learning_rate": 6.1876464467738095e-06,
+      "loss": 1.4594,
+      "step": 19337
+    },
+    {
+      "epoch": 0.890864698023679,
+      "grad_norm": 0.8698013792786204,
+      "learning_rate": 6.182480376403788e-06,
+      "loss": 1.2434,
+      "step": 19338
+    },
+    {
+      "epoch": 0.890910766112314,
+      "grad_norm": 0.7379117011513903,
+      "learning_rate": 6.177316394719956e-06,
+      "loss": 1.0254,
+      "step": 19339
+    },
+    {
+      "epoch": 0.890956834200949,
+      "grad_norm": 0.8080147950409627,
+      "learning_rate": 6.172154501837313e-06,
+      "loss": 1.3105,
+      "step": 19340
+    },
+    {
+      "epoch": 0.891002902289584,
+      "grad_norm": 0.8885844700879143,
+      "learning_rate": 6.166994697870776e-06,
+      "loss": 1.3707,
+      "step": 19341
+    },
+    {
+      "epoch": 0.8910489703782191,
+      "grad_norm": 0.9302181545308938,
+      "learning_rate": 6.16183698293521e-06,
+      "loss": 1.092,
+      "step": 19342
+    },
+    {
+      "epoch": 0.891095038466854,
+      "grad_norm": 1.358916133550397,
+      "learning_rate": 6.156681357145433e-06,
+      "loss": 1.3512,
+      "step": 19343
+    },
+    {
+      "epoch": 0.8911411065554891,
+      "grad_norm": 1.338016588242218,
+      "learning_rate": 6.151527820616243e-06,
+      "loss": 1.5177,
+      "step": 19344
+    },
+    {
+      "epoch": 0.891187174644124,
+      "grad_norm": 0.7000713359707961,
+      "learning_rate": 6.146376373462393e-06,
+      "loss": 1.3636,
+      "step": 19345
+    },
+    {
+      "epoch": 0.8912332427327591,
+      "grad_norm": 3.8178011250625503,
+      "learning_rate": 6.141227015798523e-06,
+      "loss": 1.1585,
+      "step": 19346
+    },
+    {
+      "epoch": 0.891279310821394,
+      "grad_norm": 0.9033469451472403,
+      "learning_rate": 6.13607974773931e-06,
+      "loss": 1.5136,
+      "step": 19347
+    },
+    {
+      "epoch": 0.891325378910029,
+      "grad_norm": 1.0173211642046731,
+      "learning_rate": 6.130934569399349e-06,
+      "loss": 1.342,
+      "step": 19348
+    },
+    {
+      "epoch": 0.891371446998664,
+      "grad_norm": 1.587133475923193,
+      "learning_rate": 6.125791480893162e-06,
+      "loss": 1.1178,
+      "step": 19349
+    },
+    {
+      "epoch": 0.891417515087299,
+      "grad_norm": 1.596380694125229,
+      "learning_rate": 6.120650482335266e-06,
+      "loss": 1.5421,
+      "step": 19350
+    },
+    {
+      "epoch": 0.891463583175934,
+      "grad_norm": 0.9168455153381397,
+      "learning_rate": 6.115511573840127e-06,
+      "loss": 1.018,
+      "step": 19351
+    },
+    {
+      "epoch": 0.891509651264569,
+      "grad_norm": 1.0112146451973776,
+      "learning_rate": 6.110374755522153e-06,
+      "loss": 1.0502,
+      "step": 19352
+    },
+    {
+      "epoch": 0.891555719353204,
+      "grad_norm": 1.3455593674899597,
+      "learning_rate": 6.1052400274956846e-06,
+      "loss": 1.1223,
+      "step": 19353
+    },
+    {
+      "epoch": 0.891601787441839,
+      "grad_norm": 1.1360695152234968,
+      "learning_rate": 6.1001073898750535e-06,
+      "loss": 1.3118,
+      "step": 19354
+    },
+    {
+      "epoch": 0.891647855530474,
+      "grad_norm": 0.7866337420854663,
+      "learning_rate": 6.094976842774536e-06,
+      "loss": 1.092,
+      "step": 19355
+    },
+    {
+      "epoch": 0.891693923619109,
+      "grad_norm": 1.2154602881339578,
+      "learning_rate": 6.08984838630835e-06,
+      "loss": 1.1742,
+      "step": 19356
+    },
+    {
+      "epoch": 0.8917399917077441,
+      "grad_norm": 0.8801448899332721,
+      "learning_rate": 6.084722020590661e-06,
+      "loss": 1.3079,
+      "step": 19357
+    },
+    {
+      "epoch": 0.891786059796379,
+      "grad_norm": 2.3085274912518168,
+      "learning_rate": 6.079597745735599e-06,
+      "loss": 1.2274,
+      "step": 19358
+    },
+    {
+      "epoch": 0.8918321278850141,
+      "grad_norm": 0.9872478891214599,
+      "learning_rate": 6.074475561857273e-06,
+      "loss": 1.1863,
+      "step": 19359
+    },
+    {
+      "epoch": 0.891878195973649,
+      "grad_norm": 1.1272676011389633,
+      "learning_rate": 6.069355469069682e-06,
+      "loss": 1.1835,
+      "step": 19360
+    },
+    {
+      "epoch": 0.8919242640622841,
+      "grad_norm": 1.6168039640520218,
+      "learning_rate": 6.064237467486833e-06,
+      "loss": 1.3432,
+      "step": 19361
+    },
+    {
+      "epoch": 0.891970332150919,
+      "grad_norm": 1.2073228816549517,
+      "learning_rate": 6.05912155722268e-06,
+      "loss": 1.2419,
+      "step": 19362
+    },
+    {
+      "epoch": 0.8920164002395541,
+      "grad_norm": 0.9393163440355239,
+      "learning_rate": 6.054007738391121e-06,
+      "loss": 0.8381,
+      "step": 19363
+    },
+    {
+      "epoch": 0.892062468328189,
+      "grad_norm": 1.402567810289162,
+      "learning_rate": 6.0488960111059865e-06,
+      "loss": 1.2856,
+      "step": 19364
+    },
+    {
+      "epoch": 0.8921085364168241,
+      "grad_norm": 0.8685920338851278,
+      "learning_rate": 6.043786375481087e-06,
+      "loss": 1.0493,
+      "step": 19365
+    },
+    {
+      "epoch": 0.892154604505459,
+      "grad_norm": 3.067908641809089,
+      "learning_rate": 6.038678831630206e-06,
+      "loss": 1.3664,
+      "step": 19366
+    },
+    {
+      "epoch": 0.8922006725940941,
+      "grad_norm": 1.0222244188152172,
+      "learning_rate": 6.033573379667001e-06,
+      "loss": 1.1443,
+      "step": 19367
+    },
+    {
+      "epoch": 0.892246740682729,
+      "grad_norm": 1.1663923935071847,
+      "learning_rate": 6.028470019705179e-06,
+      "loss": 1.1348,
+      "step": 19368
+    },
+    {
+      "epoch": 0.8922928087713641,
+      "grad_norm": 1.4156294805367675,
+      "learning_rate": 6.023368751858349e-06,
+      "loss": 1.1164,
+      "step": 19369
+    },
+    {
+      "epoch": 0.8923388768599991,
+      "grad_norm": 1.0227643906324435,
+      "learning_rate": 6.0182695762400875e-06,
+      "loss": 1.2498,
+      "step": 19370
+    },
+    {
+      "epoch": 0.8923849449486341,
+      "grad_norm": 1.1988308246340673,
+      "learning_rate": 6.013172492963892e-06,
+      "loss": 1.3725,
+      "step": 19371
+    },
+    {
+      "epoch": 0.8924310130372691,
+      "grad_norm": 2.376876931555602,
+      "learning_rate": 6.008077502143261e-06,
+      "loss": 0.9203,
+      "step": 19372
+    },
+    {
+      "epoch": 0.8924770811259041,
+      "grad_norm": 1.142837474740697,
+      "learning_rate": 6.002984603891615e-06,
+      "loss": 1.1603,
+      "step": 19373
+    },
+    {
+      "epoch": 0.8925231492145391,
+      "grad_norm": 1.646736091933934,
+      "learning_rate": 5.9978937983223735e-06,
+      "loss": 0.8642,
+      "step": 19374
+    },
+    {
+      "epoch": 0.8925692173031741,
+      "grad_norm": 0.9706131758600172,
+      "learning_rate": 5.992805085548825e-06,
+      "loss": 1.0615,
+      "step": 19375
+    },
+    {
+      "epoch": 0.8926152853918091,
+      "grad_norm": 1.1310802929184702,
+      "learning_rate": 5.987718465684289e-06,
+      "loss": 1.0561,
+      "step": 19376
+    },
+    {
+      "epoch": 0.8926613534804441,
+      "grad_norm": 0.8945756052819333,
+      "learning_rate": 5.9826339388420196e-06,
+      "loss": 1.0469,
+      "step": 19377
+    },
+    {
+      "epoch": 0.8927074215690791,
+      "grad_norm": 0.9156589546803736,
+      "learning_rate": 5.977551505135181e-06,
+      "loss": 1.1838,
+      "step": 19378
+    },
+    {
+      "epoch": 0.8927534896577141,
+      "grad_norm": 1.2269742224444622,
+      "learning_rate": 5.97247116467694e-06,
+      "loss": 0.8846,
+      "step": 19379
+    },
+    {
+      "epoch": 0.8927995577463491,
+      "grad_norm": 1.196424833895624,
+      "learning_rate": 5.967392917580417e-06,
+      "loss": 1.4148,
+      "step": 19380
+    },
+    {
+      "epoch": 0.8928456258349841,
+      "grad_norm": 0.8205347023289654,
+      "learning_rate": 5.962316763958664e-06,
+      "loss": 1.1262,
+      "step": 19381
+    },
+    {
+      "epoch": 0.8928916939236191,
+      "grad_norm": 0.9389060755887575,
+      "learning_rate": 5.957242703924681e-06,
+      "loss": 1.1942,
+      "step": 19382
+    },
+    {
+      "epoch": 0.8929377620122542,
+      "grad_norm": 1.0270962132418813,
+      "learning_rate": 5.9521707375914336e-06,
+      "loss": 1.2937,
+      "step": 19383
+    },
+    {
+      "epoch": 0.8929838301008891,
+      "grad_norm": 1.0509894990510922,
+      "learning_rate": 5.947100865071864e-06,
+      "loss": 1.1896,
+      "step": 19384
+    },
+    {
+      "epoch": 0.8930298981895242,
+      "grad_norm": 0.751485382826227,
+      "learning_rate": 5.942033086478815e-06,
+      "loss": 1.2289,
+      "step": 19385
+    },
+    {
+      "epoch": 0.8930759662781591,
+      "grad_norm": 0.823730476515597,
+      "learning_rate": 5.936967401925131e-06,
+      "loss": 1.5373,
+      "step": 19386
+    },
+    {
+      "epoch": 0.8931220343667942,
+      "grad_norm": 0.9991135909934222,
+      "learning_rate": 5.931903811523576e-06,
+      "loss": 1.3372,
+      "step": 19387
+    },
+    {
+      "epoch": 0.8931681024554291,
+      "grad_norm": 1.049896492328246,
+      "learning_rate": 5.926842315386904e-06,
+      "loss": 0.9996,
+      "step": 19388
+    },
+    {
+      "epoch": 0.8932141705440642,
+      "grad_norm": 0.7219646761621487,
+      "learning_rate": 5.921782913627783e-06,
+      "loss": 1.1975,
+      "step": 19389
+    },
+    {
+      "epoch": 0.8932602386326991,
+      "grad_norm": 1.642319611752476,
+      "learning_rate": 5.916725606358842e-06,
+      "loss": 1.1367,
+      "step": 19390
+    },
+    {
+      "epoch": 0.8933063067213342,
+      "grad_norm": 1.5076104720920775,
+      "learning_rate": 5.911670393692692e-06,
+      "loss": 0.9803,
+      "step": 19391
+    },
+    {
+      "epoch": 0.8933523748099691,
+      "grad_norm": 1.1256647241417834,
+      "learning_rate": 5.906617275741888e-06,
+      "loss": 1.2114,
+      "step": 19392
+    },
+    {
+      "epoch": 0.8933984428986042,
+      "grad_norm": 1.0212489993115133,
+      "learning_rate": 5.901566252618917e-06,
+      "loss": 0.8722,
+      "step": 19393
+    },
+    {
+      "epoch": 0.8934445109872391,
+      "grad_norm": 1.0231172009578537,
+      "learning_rate": 5.896517324436202e-06,
+      "loss": 1.0665,
+      "step": 19394
+    },
+    {
+      "epoch": 0.8934905790758741,
+      "grad_norm": 1.3221226826773615,
+      "learning_rate": 5.891470491306206e-06,
+      "loss": 0.8876,
+      "step": 19395
+    },
+    {
+      "epoch": 0.8935366471645091,
+      "grad_norm": 1.245845791721294,
+      "learning_rate": 5.886425753341241e-06,
+      "loss": 0.9962,
+      "step": 19396
+    },
+    {
+      "epoch": 0.8935827152531441,
+      "grad_norm": 1.1445603736699481,
+      "learning_rate": 5.881383110653627e-06,
+      "loss": 1.0725,
+      "step": 19397
+    },
+    {
+      "epoch": 0.8936287833417792,
+      "grad_norm": 0.987944575229541,
+      "learning_rate": 5.8763425633556525e-06,
+      "loss": 1.6204,
+      "step": 19398
+    },
+    {
+      "epoch": 0.8936748514304141,
+      "grad_norm": 2.6549921092184454,
+      "learning_rate": 5.871304111559528e-06,
+      "loss": 1.404,
+      "step": 19399
+    },
+    {
+      "epoch": 0.8937209195190492,
+      "grad_norm": 2.3461589580360696,
+      "learning_rate": 5.866267755377419e-06,
+      "loss": 1.239,
+      "step": 19400
+    },
+    {
+      "epoch": 0.8937669876076841,
+      "grad_norm": 1.0931995802084438,
+      "learning_rate": 5.861233494921448e-06,
+      "loss": 1.0038,
+      "step": 19401
+    },
+    {
+      "epoch": 0.8938130556963192,
+      "grad_norm": 0.7236926013041033,
+      "learning_rate": 5.8562013303037124e-06,
+      "loss": 1.5687,
+      "step": 19402
+    },
+    {
+      "epoch": 0.8938591237849541,
+      "grad_norm": 0.9294337654506362,
+      "learning_rate": 5.851171261636212e-06,
+      "loss": 1.3186,
+      "step": 19403
+    },
+    {
+      "epoch": 0.8939051918735892,
+      "grad_norm": 1.1541184958712725,
+      "learning_rate": 5.84614328903097e-06,
+      "loss": 1.28,
+      "step": 19404
+    },
+    {
+      "epoch": 0.8939512599622241,
+      "grad_norm": 1.503725344797063,
+      "learning_rate": 5.841117412599884e-06,
+      "loss": 1.2791,
+      "step": 19405
+    },
+    {
+      "epoch": 0.8939973280508592,
+      "grad_norm": 1.0859635401561836,
+      "learning_rate": 5.836093632454898e-06,
+      "loss": 1.0005,
+      "step": 19406
+    },
+    {
+      "epoch": 0.8940433961394941,
+      "grad_norm": 1.0704695925872798,
+      "learning_rate": 5.8310719487078225e-06,
+      "loss": 1.1934,
+      "step": 19407
+    },
+    {
+      "epoch": 0.8940894642281292,
+      "grad_norm": 2.851460616431456,
+      "learning_rate": 5.826052361470469e-06,
+      "loss": 1.3158,
+      "step": 19408
+    },
+    {
+      "epoch": 0.8941355323167641,
+      "grad_norm": 0.9707675072527758,
+      "learning_rate": 5.821034870854569e-06,
+      "loss": 1.2056,
+      "step": 19409
+    },
+    {
+      "epoch": 0.8941816004053992,
+      "grad_norm": 1.2009476563909312,
+      "learning_rate": 5.816019476971868e-06,
+      "loss": 0.8938,
+      "step": 19410
+    },
+    {
+      "epoch": 0.8942276684940342,
+      "grad_norm": 1.1846367036296717,
+      "learning_rate": 5.811006179933997e-06,
+      "loss": 1.4819,
+      "step": 19411
+    },
+    {
+      "epoch": 0.8942737365826692,
+      "grad_norm": 1.1871523989523152,
+      "learning_rate": 5.805994979852547e-06,
+      "loss": 1.3303,
+      "step": 19412
+    },
+    {
+      "epoch": 0.8943198046713042,
+      "grad_norm": 1.2030472739785818,
+      "learning_rate": 5.800985876839149e-06,
+      "loss": 1.4131,
+      "step": 19413
+    },
+    {
+      "epoch": 0.8943658727599392,
+      "grad_norm": 0.7890280693475392,
+      "learning_rate": 5.7959788710052586e-06,
+      "loss": 1.3414,
+      "step": 19414
+    },
+    {
+      "epoch": 0.8944119408485742,
+      "grad_norm": 1.0991133659617687,
+      "learning_rate": 5.790973962462387e-06,
+      "loss": 1.2115,
+      "step": 19415
+    },
+    {
+      "epoch": 0.8944580089372092,
+      "grad_norm": 1.6552851091559175,
+      "learning_rate": 5.785971151321923e-06,
+      "loss": 1.303,
+      "step": 19416
+    },
+    {
+      "epoch": 0.8945040770258442,
+      "grad_norm": 0.6864608390475644,
+      "learning_rate": 5.780970437695299e-06,
+      "loss": 0.7193,
+      "step": 19417
+    },
+    {
+      "epoch": 0.8945501451144792,
+      "grad_norm": 0.6401471580705093,
+      "learning_rate": 5.775971821693804e-06,
+      "loss": 1.2081,
+      "step": 19418
+    },
+    {
+      "epoch": 0.8945962132031142,
+      "grad_norm": 1.3233129638061079,
+      "learning_rate": 5.77097530342875e-06,
+      "loss": 1.1605,
+      "step": 19419
+    },
+    {
+      "epoch": 0.8946422812917492,
+      "grad_norm": 1.085228664400832,
+      "learning_rate": 5.765980883011346e-06,
+      "loss": 1.523,
+      "step": 19420
+    },
+    {
+      "epoch": 0.8946883493803842,
+      "grad_norm": 0.7106932417660095,
+      "learning_rate": 5.760988560552794e-06,
+      "loss": 1.5241,
+      "step": 19421
+    },
+    {
+      "epoch": 0.8947344174690192,
+      "grad_norm": 1.7488123820212416,
+      "learning_rate": 5.75599833616427e-06,
+      "loss": 1.2601,
+      "step": 19422
+    },
+    {
+      "epoch": 0.8947804855576542,
+      "grad_norm": 2.1228024564891643,
+      "learning_rate": 5.75101020995682e-06,
+      "loss": 1.1208,
+      "step": 19423
+    },
+    {
+      "epoch": 0.8948265536462893,
+      "grad_norm": 1.0303177986553536,
+      "learning_rate": 5.746024182041554e-06,
+      "loss": 0.9497,
+      "step": 19424
+    },
+    {
+      "epoch": 0.8948726217349242,
+      "grad_norm": 0.8569706280436376,
+      "learning_rate": 5.741040252529428e-06,
+      "loss": 0.9739,
+      "step": 19425
+    },
+    {
+      "epoch": 0.8949186898235593,
+      "grad_norm": 8.621788935888633,
+      "learning_rate": 5.736058421531443e-06,
+      "loss": 1.2244,
+      "step": 19426
+    },
+    {
+      "epoch": 0.8949647579121942,
+      "grad_norm": 1.3567604025155595,
+      "learning_rate": 5.731078689158465e-06,
+      "loss": 1.2266,
+      "step": 19427
+    },
+    {
+      "epoch": 0.8950108260008293,
+      "grad_norm": 1.367615539207059,
+      "learning_rate": 5.726101055521405e-06,
+      "loss": 1.2191,
+      "step": 19428
+    },
+    {
+      "epoch": 0.8950568940894642,
+      "grad_norm": 1.3933040115189246,
+      "learning_rate": 5.7211255207310525e-06,
+      "loss": 1.4767,
+      "step": 19429
+    },
+    {
+      "epoch": 0.8951029621780993,
+      "grad_norm": 0.8619461015534823,
+      "learning_rate": 5.716152084898185e-06,
+      "loss": 1.5632,
+      "step": 19430
+    },
+    {
+      "epoch": 0.8951490302667342,
+      "grad_norm": 0.8953913829295758,
+      "learning_rate": 5.711180748133527e-06,
+      "loss": 1.016,
+      "step": 19431
+    },
+    {
+      "epoch": 0.8951950983553693,
+      "grad_norm": 0.8102394819963048,
+      "learning_rate": 5.706211510547754e-06,
+      "loss": 1.2851,
+      "step": 19432
+    },
+    {
+      "epoch": 0.8952411664440042,
+      "grad_norm": 1.4473224041954056,
+      "learning_rate": 5.701244372251513e-06,
+      "loss": 1.1767,
+      "step": 19433
+    },
+    {
+      "epoch": 0.8952872345326393,
+      "grad_norm": 1.3699466520077075,
+      "learning_rate": 5.69627933335537e-06,
+      "loss": 1.0708,
+      "step": 19434
+    },
+    {
+      "epoch": 0.8953333026212742,
+      "grad_norm": 1.2984128864537972,
+      "learning_rate": 5.691316393969881e-06,
+      "loss": 0.91,
+      "step": 19435
+    },
+    {
+      "epoch": 0.8953793707099093,
+      "grad_norm": 1.2833152289250829,
+      "learning_rate": 5.686355554205514e-06,
+      "loss": 1.4753,
+      "step": 19436
+    },
+    {
+      "epoch": 0.8954254387985442,
+      "grad_norm": 0.8796340501938708,
+      "learning_rate": 5.681396814172746e-06,
+      "loss": 1.2011,
+      "step": 19437
+    },
+    {
+      "epoch": 0.8954715068871792,
+      "grad_norm": 0.790180526635122,
+      "learning_rate": 5.676440173981945e-06,
+      "loss": 1.2991,
+      "step": 19438
+    },
+    {
+      "epoch": 0.8955175749758143,
+      "grad_norm": 1.0907745126892359,
+      "learning_rate": 5.671485633743468e-06,
+      "loss": 1.124,
+      "step": 19439
+    },
+    {
+      "epoch": 0.8955636430644492,
+      "grad_norm": 1.2372988328704038,
+      "learning_rate": 5.666533193567636e-06,
+      "loss": 1.0316,
+      "step": 19440
+    },
+    {
+      "epoch": 0.8956097111530843,
+      "grad_norm": 0.7597502188312407,
+      "learning_rate": 5.661582853564695e-06,
+      "loss": 1.3221,
+      "step": 19441
+    },
+    {
+      "epoch": 0.8956557792417192,
+      "grad_norm": 1.9868686907375304,
+      "learning_rate": 5.656634613844847e-06,
+      "loss": 1.2444,
+      "step": 19442
+    },
+    {
+      "epoch": 0.8957018473303543,
+      "grad_norm": 1.8010013690690188,
+      "learning_rate": 5.651688474518257e-06,
+      "loss": 1.2153,
+      "step": 19443
+    },
+    {
+      "epoch": 0.8957479154189892,
+      "grad_norm": 1.2034034681564392,
+      "learning_rate": 5.646744435695073e-06,
+      "loss": 1.337,
+      "step": 19444
+    },
+    {
+      "epoch": 0.8957939835076243,
+      "grad_norm": 1.0057729620832936,
+      "learning_rate": 5.641802497485338e-06,
+      "loss": 1.3916,
+      "step": 19445
+    },
+    {
+      "epoch": 0.8958400515962592,
+      "grad_norm": 0.788296267671791,
+      "learning_rate": 5.636862659999065e-06,
+      "loss": 1.0639,
+      "step": 19446
+    },
+    {
+      "epoch": 0.8958861196848943,
+      "grad_norm": 1.4757388798830753,
+      "learning_rate": 5.6319249233462655e-06,
+      "loss": 1.1226,
+      "step": 19447
+    },
+    {
+      "epoch": 0.8959321877735292,
+      "grad_norm": 1.0399947202612623,
+      "learning_rate": 5.626989287636841e-06,
+      "loss": 1.1472,
+      "step": 19448
+    },
+    {
+      "epoch": 0.8959782558621643,
+      "grad_norm": 1.3247689701063372,
+      "learning_rate": 5.6220557529806815e-06,
+      "loss": 0.9606,
+      "step": 19449
+    },
+    {
+      "epoch": 0.8960243239507992,
+      "grad_norm": 2.2317039553718994,
+      "learning_rate": 5.617124319487632e-06,
+      "loss": 1.1416,
+      "step": 19450
+    },
+    {
+      "epoch": 0.8960703920394343,
+      "grad_norm": 0.8522625146817092,
+      "learning_rate": 5.612194987267483e-06,
+      "loss": 1.3395,
+      "step": 19451
+    },
+    {
+      "epoch": 0.8961164601280693,
+      "grad_norm": 1.5940877928098198,
+      "learning_rate": 5.607267756429968e-06,
+      "loss": 1.1889,
+      "step": 19452
+    },
+    {
+      "epoch": 0.8961625282167043,
+      "grad_norm": 0.9386954991828437,
+      "learning_rate": 5.602342627084778e-06,
+      "loss": 1.1013,
+      "step": 19453
+    },
+    {
+      "epoch": 0.8962085963053393,
+      "grad_norm": 1.1242255952896327,
+      "learning_rate": 5.59741959934158e-06,
+      "loss": 1.2325,
+      "step": 19454
+    },
+    {
+      "epoch": 0.8962546643939743,
+      "grad_norm": 1.236369860665068,
+      "learning_rate": 5.592498673309987e-06,
+      "loss": 1.0676,
+      "step": 19455
+    },
+    {
+      "epoch": 0.8963007324826093,
+      "grad_norm": 0.9882202928931231,
+      "learning_rate": 5.587579849099511e-06,
+      "loss": 1.2882,
+      "step": 19456
+    },
+    {
+      "epoch": 0.8963468005712443,
+      "grad_norm": 1.077938635694208,
+      "learning_rate": 5.5826631268196875e-06,
+      "loss": 1.2519,
+      "step": 19457
+    },
+    {
+      "epoch": 0.8963928686598793,
+      "grad_norm": 1.2752156171092532,
+      "learning_rate": 5.577748506580005e-06,
+      "loss": 1.4178,
+      "step": 19458
+    },
+    {
+      "epoch": 0.8964389367485143,
+      "grad_norm": 0.9692724076764229,
+      "learning_rate": 5.572835988489822e-06,
+      "loss": 0.9587,
+      "step": 19459
+    },
+    {
+      "epoch": 0.8964850048371493,
+      "grad_norm": 0.8946690991456024,
+      "learning_rate": 5.56792557265855e-06,
+      "loss": 1.1416,
+      "step": 19460
+    },
+    {
+      "epoch": 0.8965310729257843,
+      "grad_norm": 2.633978372752452,
+      "learning_rate": 5.563017259195491e-06,
+      "loss": 1.2253,
+      "step": 19461
+    },
+    {
+      "epoch": 0.8965771410144193,
+      "grad_norm": 1.5095670711316538,
+      "learning_rate": 5.558111048209946e-06,
+      "loss": 1.3146,
+      "step": 19462
+    },
+    {
+      "epoch": 0.8966232091030543,
+      "grad_norm": 0.8573787147745726,
+      "learning_rate": 5.553206939811117e-06,
+      "loss": 1.4024,
+      "step": 19463
+    },
+    {
+      "epoch": 0.8966692771916893,
+      "grad_norm": 1.3584508359394794,
+      "learning_rate": 5.548304934108184e-06,
+      "loss": 1.3498,
+      "step": 19464
+    },
+    {
+      "epoch": 0.8967153452803244,
+      "grad_norm": 0.6650623101821727,
+      "learning_rate": 5.5434050312103025e-06,
+      "loss": 0.9957,
+      "step": 19465
+    },
+    {
+      "epoch": 0.8967614133689593,
+      "grad_norm": 0.8562629449883237,
+      "learning_rate": 5.538507231226553e-06,
+      "loss": 1.2746,
+      "step": 19466
+    },
+    {
+      "epoch": 0.8968074814575944,
+      "grad_norm": 1.1902349494504683,
+      "learning_rate": 5.5336115342659704e-06,
+      "loss": 0.9674,
+      "step": 19467
+    },
+    {
+      "epoch": 0.8968535495462293,
+      "grad_norm": 1.2873888988065434,
+      "learning_rate": 5.5287179404375446e-06,
+      "loss": 1.1273,
+      "step": 19468
+    },
+    {
+      "epoch": 0.8968996176348644,
+      "grad_norm": 0.7107802978273836,
+      "learning_rate": 5.523826449850245e-06,
+      "loss": 1.0805,
+      "step": 19469
+    },
+    {
+      "epoch": 0.8969456857234993,
+      "grad_norm": 0.8971770244437702,
+      "learning_rate": 5.518937062612939e-06,
+      "loss": 1.4341,
+      "step": 19470
+    },
+    {
+      "epoch": 0.8969917538121344,
+      "grad_norm": 0.6601138882507394,
+      "learning_rate": 5.514049778834507e-06,
+      "loss": 1.5137,
+      "step": 19471
+    },
+    {
+      "epoch": 0.8970378219007693,
+      "grad_norm": 0.8990010601443622,
+      "learning_rate": 5.509164598623739e-06,
+      "loss": 1.3241,
+      "step": 19472
+    },
+    {
+      "epoch": 0.8970838899894044,
+      "grad_norm": 1.051403281246282,
+      "learning_rate": 5.504281522089428e-06,
+      "loss": 1.2674,
+      "step": 19473
+    },
+    {
+      "epoch": 0.8971299580780393,
+      "grad_norm": 1.5162673298692637,
+      "learning_rate": 5.499400549340239e-06,
+      "loss": 1.3026,
+      "step": 19474
+    },
+    {
+      "epoch": 0.8971760261666744,
+      "grad_norm": 0.9413645009672068,
+      "learning_rate": 5.494521680484877e-06,
+      "loss": 1.0466,
+      "step": 19475
+    },
+    {
+      "epoch": 0.8972220942553093,
+      "grad_norm": 1.6468176790789704,
+      "learning_rate": 5.489644915631953e-06,
+      "loss": 1.143,
+      "step": 19476
+    },
+    {
+      "epoch": 0.8972681623439444,
+      "grad_norm": 1.9898651860938925,
+      "learning_rate": 5.484770254890015e-06,
+      "loss": 1.2987,
+      "step": 19477
+    },
+    {
+      "epoch": 0.8973142304325794,
+      "grad_norm": 2.2140766372658556,
+      "learning_rate": 5.47989769836762e-06,
+      "loss": 1.012,
+      "step": 19478
+    },
+    {
+      "epoch": 0.8973602985212144,
+      "grad_norm": 0.8257319306887694,
+      "learning_rate": 5.475027246173226e-06,
+      "loss": 1.2678,
+      "step": 19479
+    },
+    {
+      "epoch": 0.8974063666098494,
+      "grad_norm": 0.7316639822212065,
+      "learning_rate": 5.470158898415301e-06,
+      "loss": 1.138,
+      "step": 19480
+    },
+    {
+      "epoch": 0.8974524346984843,
+      "grad_norm": 0.8119943084374527,
+      "learning_rate": 5.465292655202181e-06,
+      "loss": 1.2795,
+      "step": 19481
+    },
+    {
+      "epoch": 0.8974985027871194,
+      "grad_norm": 0.886694529431873,
+      "learning_rate": 5.4604285166422246e-06,
+      "loss": 0.9715,
+      "step": 19482
+    },
+    {
+      "epoch": 0.8975445708757543,
+      "grad_norm": 0.7677491969747443,
+      "learning_rate": 5.455566482843732e-06,
+      "loss": 1.365,
+      "step": 19483
+    },
+    {
+      "epoch": 0.8975906389643894,
+      "grad_norm": 0.8720032333318158,
+      "learning_rate": 5.450706553914953e-06,
+      "loss": 1.1973,
+      "step": 19484
+    },
+    {
+      "epoch": 0.8976367070530243,
+      "grad_norm": 1.3046982815121915,
+      "learning_rate": 5.445848729964065e-06,
+      "loss": 1.0032,
+      "step": 19485
+    },
+    {
+      "epoch": 0.8976827751416594,
+      "grad_norm": 0.7750749960927369,
+      "learning_rate": 5.440993011099227e-06,
+      "loss": 1.3215,
+      "step": 19486
+    },
+    {
+      "epoch": 0.8977288432302943,
+      "grad_norm": 1.6931425912533864,
+      "learning_rate": 5.436139397428563e-06,
+      "loss": 1.2001,
+      "step": 19487
+    },
+    {
+      "epoch": 0.8977749113189294,
+      "grad_norm": 1.1311861753155865,
+      "learning_rate": 5.431287889060099e-06,
+      "loss": 0.9745,
+      "step": 19488
+    },
+    {
+      "epoch": 0.8978209794075643,
+      "grad_norm": 1.1562644990581932,
+      "learning_rate": 5.426438486101859e-06,
+      "loss": 1.1021,
+      "step": 19489
+    },
+    {
+      "epoch": 0.8978670474961994,
+      "grad_norm": 1.0903711676743941,
+      "learning_rate": 5.421591188661801e-06,
+      "loss": 1.1595,
+      "step": 19490
+    },
+    {
+      "epoch": 0.8979131155848343,
+      "grad_norm": 1.0772957525068496,
+      "learning_rate": 5.416745996847872e-06,
+      "loss": 1.1319,
+      "step": 19491
+    },
+    {
+      "epoch": 0.8979591836734694,
+      "grad_norm": 1.3633949728409007,
+      "learning_rate": 5.411902910767897e-06,
+      "loss": 1.2304,
+      "step": 19492
+    },
+    {
+      "epoch": 0.8980052517621044,
+      "grad_norm": 0.7534881004111982,
+      "learning_rate": 5.407061930529722e-06,
+      "loss": 1.1549,
+      "step": 19493
+    },
+    {
+      "epoch": 0.8980513198507394,
+      "grad_norm": 1.0179453542185135,
+      "learning_rate": 5.4022230562411294e-06,
+      "loss": 1.2554,
+      "step": 19494
+    },
+    {
+      "epoch": 0.8980973879393744,
+      "grad_norm": 1.0972537135335625,
+      "learning_rate": 5.397386288009831e-06,
+      "loss": 1.143,
+      "step": 19495
+    },
+    {
+      "epoch": 0.8981434560280094,
+      "grad_norm": 0.9413395285057052,
+      "learning_rate": 5.39255162594352e-06,
+      "loss": 1.3817,
+      "step": 19496
+    },
+    {
+      "epoch": 0.8981895241166444,
+      "grad_norm": 1.4149510139477406,
+      "learning_rate": 5.3877190701498325e-06,
+      "loss": 1.2792,
+      "step": 19497
+    },
+    {
+      "epoch": 0.8982355922052794,
+      "grad_norm": 1.0325383707178561,
+      "learning_rate": 5.382888620736359e-06,
+      "loss": 0.9801,
+      "step": 19498
+    },
+    {
+      "epoch": 0.8982816602939144,
+      "grad_norm": 0.8365769759542164,
+      "learning_rate": 5.378060277810637e-06,
+      "loss": 1.2179,
+      "step": 19499
+    },
+    {
+      "epoch": 0.8983277283825494,
+      "grad_norm": 0.8813116769948887,
+      "learning_rate": 5.373234041480146e-06,
+      "loss": 1.2366,
+      "step": 19500
+    },
+    {
+      "epoch": 0.8983737964711844,
+      "grad_norm": 0.9061950139470897,
+      "learning_rate": 5.368409911852357e-06,
+      "loss": 1.3783,
+      "step": 19501
+    },
+    {
+      "epoch": 0.8984198645598194,
+      "grad_norm": 0.853051294124185,
+      "learning_rate": 5.363587889034683e-06,
+      "loss": 1.1167,
+      "step": 19502
+    },
+    {
+      "epoch": 0.8984659326484544,
+      "grad_norm": 1.0868457822436393,
+      "learning_rate": 5.358767973134449e-06,
+      "loss": 1.098,
+      "step": 19503
+    },
+    {
+      "epoch": 0.8985120007370894,
+      "grad_norm": 1.0244565672230248,
+      "learning_rate": 5.353950164258936e-06,
+      "loss": 1.2115,
+      "step": 19504
+    },
+    {
+      "epoch": 0.8985580688257244,
+      "grad_norm": 0.8171700557843491,
+      "learning_rate": 5.349134462515481e-06,
+      "loss": 1.2612,
+      "step": 19505
+    },
+    {
+      "epoch": 0.8986041369143595,
+      "grad_norm": 1.5158545105384866,
+      "learning_rate": 5.344320868011232e-06,
+      "loss": 1.253,
+      "step": 19506
+    },
+    {
+      "epoch": 0.8986502050029944,
+      "grad_norm": 1.0635960809585048,
+      "learning_rate": 5.339509380853403e-06,
+      "loss": 1.2601,
+      "step": 19507
+    },
+    {
+      "epoch": 0.8986962730916295,
+      "grad_norm": 1.238762448340952,
+      "learning_rate": 5.334700001149051e-06,
+      "loss": 1.2654,
+      "step": 19508
+    },
+    {
+      "epoch": 0.8987423411802644,
+      "grad_norm": 1.972704567178239,
+      "learning_rate": 5.329892729005326e-06,
+      "loss": 1.2201,
+      "step": 19509
+    },
+    {
+      "epoch": 0.8987884092688995,
+      "grad_norm": 0.9007096856031381,
+      "learning_rate": 5.325087564529197e-06,
+      "loss": 1.3664,
+      "step": 19510
+    },
+    {
+      "epoch": 0.8988344773575344,
+      "grad_norm": 1.3103796834954484,
+      "learning_rate": 5.320284507827656e-06,
+      "loss": 1.1777,
+      "step": 19511
+    },
+    {
+      "epoch": 0.8988805454461695,
+      "grad_norm": 1.3701738970328632,
+      "learning_rate": 5.315483559007661e-06,
+      "loss": 1.1771,
+      "step": 19512
+    },
+    {
+      "epoch": 0.8989266135348044,
+      "grad_norm": 0.9701793984155472,
+      "learning_rate": 5.310684718176062e-06,
+      "loss": 1.3231,
+      "step": 19513
+    },
+    {
+      "epoch": 0.8989726816234395,
+      "grad_norm": 1.062739431995019,
+      "learning_rate": 5.305887985439717e-06,
+      "loss": 1.4754,
+      "step": 19514
+    },
+    {
+      "epoch": 0.8990187497120744,
+      "grad_norm": 1.5690448246751112,
+      "learning_rate": 5.301093360905385e-06,
+      "loss": 1.0365,
+      "step": 19515
+    },
+    {
+      "epoch": 0.8990648178007095,
+      "grad_norm": 1.0721358439139133,
+      "learning_rate": 5.296300844679858e-06,
+      "loss": 0.8667,
+      "step": 19516
+    },
+    {
+      "epoch": 0.8991108858893444,
+      "grad_norm": 0.9876902242232891,
+      "learning_rate": 5.2915104368698065e-06,
+      "loss": 1.4706,
+      "step": 19517
+    },
+    {
+      "epoch": 0.8991569539779795,
+      "grad_norm": 0.6857033185743616,
+      "learning_rate": 5.286722137581901e-06,
+      "loss": 1.1809,
+      "step": 19518
+    },
+    {
+      "epoch": 0.8992030220666145,
+      "grad_norm": 1.5069964520132784,
+      "learning_rate": 5.281935946922689e-06,
+      "loss": 1.1359,
+      "step": 19519
+    },
+    {
+      "epoch": 0.8992490901552495,
+      "grad_norm": 1.1274362800345683,
+      "learning_rate": 5.2771518649987975e-06,
+      "loss": 1.0183,
+      "step": 19520
+    },
+    {
+      "epoch": 0.8992951582438845,
+      "grad_norm": 0.8536561197252565,
+      "learning_rate": 5.272369891916706e-06,
+      "loss": 1.0758,
+      "step": 19521
+    },
+    {
+      "epoch": 0.8993412263325195,
+      "grad_norm": 1.0341287008868627,
+      "learning_rate": 5.267590027782854e-06,
+      "loss": 1.2465,
+      "step": 19522
+    },
+    {
+      "epoch": 0.8993872944211545,
+      "grad_norm": 2.514332565113744,
+      "learning_rate": 5.262812272703699e-06,
+      "loss": 1.2046,
+      "step": 19523
+    },
+    {
+      "epoch": 0.8994333625097894,
+      "grad_norm": 1.4503700462631817,
+      "learning_rate": 5.258036626785578e-06,
+      "loss": 1.3264,
+      "step": 19524
+    },
+    {
+      "epoch": 0.8994794305984245,
+      "grad_norm": 0.8748296928595344,
+      "learning_rate": 5.253263090134842e-06,
+      "loss": 1.4082,
+      "step": 19525
+    },
+    {
+      "epoch": 0.8995254986870594,
+      "grad_norm": 0.9341534708307319,
+      "learning_rate": 5.248491662857713e-06,
+      "loss": 1.5077,
+      "step": 19526
+    },
+    {
+      "epoch": 0.8995715667756945,
+      "grad_norm": 1.4492768439938741,
+      "learning_rate": 5.243722345060487e-06,
+      "loss": 1.1572,
+      "step": 19527
+    },
+    {
+      "epoch": 0.8996176348643294,
+      "grad_norm": 0.8627529568685516,
+      "learning_rate": 5.238955136849289e-06,
+      "loss": 1.2318,
+      "step": 19528
+    },
+    {
+      "epoch": 0.8996637029529645,
+      "grad_norm": 0.9596523458460569,
+      "learning_rate": 5.2341900383303e-06,
+      "loss": 0.9791,
+      "step": 19529
+    },
+    {
+      "epoch": 0.8997097710415994,
+      "grad_norm": 1.285635948111411,
+      "learning_rate": 5.229427049609558e-06,
+      "loss": 1.4654,
+      "step": 19530
+    },
+    {
+      "epoch": 0.8997558391302345,
+      "grad_norm": 0.9199842089217661,
+      "learning_rate": 5.224666170793124e-06,
+      "loss": 1.3563,
+      "step": 19531
+    },
+    {
+      "epoch": 0.8998019072188694,
+      "grad_norm": 0.7502155830956965,
+      "learning_rate": 5.219907401987012e-06,
+      "loss": 1.2794,
+      "step": 19532
+    },
+    {
+      "epoch": 0.8998479753075045,
+      "grad_norm": 4.138639619352972,
+      "learning_rate": 5.215150743297137e-06,
+      "loss": 1.2817,
+      "step": 19533
+    },
+    {
+      "epoch": 0.8998940433961395,
+      "grad_norm": 1.014235778958493,
+      "learning_rate": 5.210396194829403e-06,
+      "loss": 1.2732,
+      "step": 19534
+    },
+    {
+      "epoch": 0.8999401114847745,
+      "grad_norm": 1.2029720561597308,
+      "learning_rate": 5.205643756689682e-06,
+      "loss": 1.4009,
+      "step": 19535
+    },
+    {
+      "epoch": 0.8999861795734095,
+      "grad_norm": 0.6729159012347463,
+      "learning_rate": 5.200893428983766e-06,
+      "loss": 1.2256,
+      "step": 19536
+    },
+    {
+      "epoch": 0.9000322476620445,
+      "grad_norm": 1.3747714604302717,
+      "learning_rate": 5.1961452118173825e-06,
+      "loss": 1.0871,
+      "step": 19537
+    },
+    {
+      "epoch": 0.9000783157506795,
+      "grad_norm": 2.0942970548696684,
+      "learning_rate": 5.191399105296313e-06,
+      "loss": 1.0224,
+      "step": 19538
+    },
+    {
+      "epoch": 0.9001243838393145,
+      "grad_norm": 1.355883941163037,
+      "learning_rate": 5.186655109526173e-06,
+      "loss": 1.2146,
+      "step": 19539
+    },
+    {
+      "epoch": 0.9001704519279495,
+      "grad_norm": 1.0748560895064536,
+      "learning_rate": 5.181913224612567e-06,
+      "loss": 1.2899,
+      "step": 19540
+    },
+    {
+      "epoch": 0.9002165200165845,
+      "grad_norm": 1.0968684731834109,
+      "learning_rate": 5.177173450661088e-06,
+      "loss": 1.374,
+      "step": 19541
+    },
+    {
+      "epoch": 0.9002625881052195,
+      "grad_norm": 1.0249362632352441,
+      "learning_rate": 5.172435787777263e-06,
+      "loss": 1.286,
+      "step": 19542
+    },
+    {
+      "epoch": 0.9003086561938545,
+      "grad_norm": 1.5069065955970227,
+      "learning_rate": 5.167700236066564e-06,
+      "loss": 1.0361,
+      "step": 19543
+    },
+    {
+      "epoch": 0.9003547242824895,
+      "grad_norm": 1.0406695302777271,
+      "learning_rate": 5.162966795634416e-06,
+      "loss": 0.9964,
+      "step": 19544
+    },
+    {
+      "epoch": 0.9004007923711245,
+      "grad_norm": 0.93084767383295,
+      "learning_rate": 5.158235466586192e-06,
+      "loss": 1.2033,
+      "step": 19545
+    },
+    {
+      "epoch": 0.9004468604597595,
+      "grad_norm": 0.7324319999004307,
+      "learning_rate": 5.15350624902724e-06,
+      "loss": 1.2419,
+      "step": 19546
+    },
+    {
+      "epoch": 0.9004929285483946,
+      "grad_norm": 1.0905090694421007,
+      "learning_rate": 5.148779143062854e-06,
+      "loss": 1.3896,
+      "step": 19547
+    },
+    {
+      "epoch": 0.9005389966370295,
+      "grad_norm": 1.3825527699755868,
+      "learning_rate": 5.1440541487982605e-06,
+      "loss": 1.0231,
+      "step": 19548
+    },
+    {
+      "epoch": 0.9005850647256646,
+      "grad_norm": 3.223703771347108,
+      "learning_rate": 5.139331266338654e-06,
+      "loss": 1.3021,
+      "step": 19549
+    },
+    {
+      "epoch": 0.9006311328142995,
+      "grad_norm": 0.8360655930736126,
+      "learning_rate": 5.134610495789194e-06,
+      "loss": 1.1579,
+      "step": 19550
+    },
+    {
+      "epoch": 0.9006772009029346,
+      "grad_norm": 1.1606503167007387,
+      "learning_rate": 5.129891837254963e-06,
+      "loss": 1.0723,
+      "step": 19551
+    },
+    {
+      "epoch": 0.9007232689915695,
+      "grad_norm": 0.9275783607870566,
+      "learning_rate": 5.125175290841022e-06,
+      "loss": 1.1797,
+      "step": 19552
+    },
+    {
+      "epoch": 0.9007693370802046,
+      "grad_norm": 0.6954938711323572,
+      "learning_rate": 5.120460856652376e-06,
+      "loss": 1.3194,
+      "step": 19553
+    },
+    {
+      "epoch": 0.9008154051688395,
+      "grad_norm": 2.2579010607036043,
+      "learning_rate": 5.115748534794007e-06,
+      "loss": 1.1794,
+      "step": 19554
+    },
+    {
+      "epoch": 0.9008614732574746,
+      "grad_norm": 4.459869256353044,
+      "learning_rate": 5.111038325370787e-06,
+      "loss": 1.2695,
+      "step": 19555
+    },
+    {
+      "epoch": 0.9009075413461095,
+      "grad_norm": 0.8754011283190151,
+      "learning_rate": 5.106330228487599e-06,
+      "loss": 0.9194,
+      "step": 19556
+    },
+    {
+      "epoch": 0.9009536094347446,
+      "grad_norm": 1.679623067049902,
+      "learning_rate": 5.10162424424927e-06,
+      "loss": 0.9863,
+      "step": 19557
+    },
+    {
+      "epoch": 0.9009996775233795,
+      "grad_norm": 1.2641235329547749,
+      "learning_rate": 5.09692037276055e-06,
+      "loss": 1.2022,
+      "step": 19558
+    },
+    {
+      "epoch": 0.9010457456120146,
+      "grad_norm": 0.8755881311032555,
+      "learning_rate": 5.092218614126177e-06,
+      "loss": 1.3741,
+      "step": 19559
+    },
+    {
+      "epoch": 0.9010918137006496,
+      "grad_norm": 0.8647972575395222,
+      "learning_rate": 5.087518968450822e-06,
+      "loss": 1.1112,
+      "step": 19560
+    },
+    {
+      "epoch": 0.9011378817892846,
+      "grad_norm": 1.424186157306916,
+      "learning_rate": 5.082821435839125e-06,
+      "loss": 1.2913,
+      "step": 19561
+    },
+    {
+      "epoch": 0.9011839498779196,
+      "grad_norm": 1.2624598694901459,
+      "learning_rate": 5.078126016395657e-06,
+      "loss": 0.864,
+      "step": 19562
+    },
+    {
+      "epoch": 0.9012300179665546,
+      "grad_norm": 1.1169182095103443,
+      "learning_rate": 5.0734327102249566e-06,
+      "loss": 1.3361,
+      "step": 19563
+    },
+    {
+      "epoch": 0.9012760860551896,
+      "grad_norm": 1.0508764496225067,
+      "learning_rate": 5.068741517431508e-06,
+      "loss": 1.3462,
+      "step": 19564
+    },
+    {
+      "epoch": 0.9013221541438245,
+      "grad_norm": 1.091391945028212,
+      "learning_rate": 5.0640524381197704e-06,
+      "loss": 1.3379,
+      "step": 19565
+    },
+    {
+      "epoch": 0.9013682222324596,
+      "grad_norm": 0.9358918183584433,
+      "learning_rate": 5.059365472394117e-06,
+      "loss": 1.3098,
+      "step": 19566
+    },
+    {
+      "epoch": 0.9014142903210945,
+      "grad_norm": 0.7628206959321403,
+      "learning_rate": 5.0546806203588985e-06,
+      "loss": 1.4046,
+      "step": 19567
+    },
+    {
+      "epoch": 0.9014603584097296,
+      "grad_norm": 0.733057669513978,
+      "learning_rate": 5.04999788211844e-06,
+      "loss": 1.065,
+      "step": 19568
+    },
+    {
+      "epoch": 0.9015064264983645,
+      "grad_norm": 0.7457497084527229,
+      "learning_rate": 5.0453172577769495e-06,
+      "loss": 1.1591,
+      "step": 19569
+    },
+    {
+      "epoch": 0.9015524945869996,
+      "grad_norm": 0.9414904300818759,
+      "learning_rate": 5.040638747438664e-06,
+      "loss": 1.0304,
+      "step": 19570
+    },
+    {
+      "epoch": 0.9015985626756345,
+      "grad_norm": 1.2778934512020645,
+      "learning_rate": 5.035962351207735e-06,
+      "loss": 1.4279,
+      "step": 19571
+    },
+    {
+      "epoch": 0.9016446307642696,
+      "grad_norm": 0.9457753968361435,
+      "learning_rate": 5.031288069188289e-06,
+      "loss": 1.2697,
+      "step": 19572
+    },
+    {
+      "epoch": 0.9016906988529046,
+      "grad_norm": 0.8403656655232197,
+      "learning_rate": 5.026615901484366e-06,
+      "loss": 1.18,
+      "step": 19573
+    },
+    {
+      "epoch": 0.9017367669415396,
+      "grad_norm": 0.966907627050994,
+      "learning_rate": 5.021945848199993e-06,
+      "loss": 1.3284,
+      "step": 19574
+    },
+    {
+      "epoch": 0.9017828350301746,
+      "grad_norm": 1.1234694917397865,
+      "learning_rate": 5.017277909439167e-06,
+      "loss": 1.2854,
+      "step": 19575
+    },
+    {
+      "epoch": 0.9018289031188096,
+      "grad_norm": 1.2536733499368484,
+      "learning_rate": 5.012612085305768e-06,
+      "loss": 1.2808,
+      "step": 19576
+    },
+    {
+      "epoch": 0.9018749712074446,
+      "grad_norm": 1.208996598790852,
+      "learning_rate": 5.0079483759036814e-06,
+      "loss": 1.3799,
+      "step": 19577
+    },
+    {
+      "epoch": 0.9019210392960796,
+      "grad_norm": 2.284683868234623,
+      "learning_rate": 5.003286781336758e-06,
+      "loss": 0.9859,
+      "step": 19578
+    },
+    {
+      "epoch": 0.9019671073847146,
+      "grad_norm": 0.8400733639187524,
+      "learning_rate": 4.998627301708792e-06,
+      "loss": 1.2292,
+      "step": 19579
+    },
+    {
+      "epoch": 0.9020131754733496,
+      "grad_norm": 0.777083794728647,
+      "learning_rate": 4.993969937123466e-06,
+      "loss": 1.2154,
+      "step": 19580
+    },
+    {
+      "epoch": 0.9020592435619846,
+      "grad_norm": 0.7696898354114914,
+      "learning_rate": 4.98931468768451e-06,
+      "loss": 1.12,
+      "step": 19581
+    },
+    {
+      "epoch": 0.9021053116506196,
+      "grad_norm": 3.114307948428883,
+      "learning_rate": 4.984661553495551e-06,
+      "loss": 1.2015,
+      "step": 19582
+    },
+    {
+      "epoch": 0.9021513797392546,
+      "grad_norm": 1.2823376589376343,
+      "learning_rate": 4.980010534660195e-06,
+      "loss": 1.1904,
+      "step": 19583
+    },
+    {
+      "epoch": 0.9021974478278896,
+      "grad_norm": 1.2796985547169941,
+      "learning_rate": 4.9753616312819714e-06,
+      "loss": 1.2589,
+      "step": 19584
+    },
+    {
+      "epoch": 0.9022435159165246,
+      "grad_norm": 0.8458794099053001,
+      "learning_rate": 4.970714843464397e-06,
+      "loss": 1.1121,
+      "step": 19585
+    },
+    {
+      "epoch": 0.9022895840051596,
+      "grad_norm": 0.9377481786117211,
+      "learning_rate": 4.966070171310921e-06,
+      "loss": 1.1715,
+      "step": 19586
+    },
+    {
+      "epoch": 0.9023356520937946,
+      "grad_norm": 0.939696926370881,
+      "learning_rate": 4.9614276149249295e-06,
+      "loss": 1.4801,
+      "step": 19587
+    },
+    {
+      "epoch": 0.9023817201824297,
+      "grad_norm": 1.1601902461457094,
+      "learning_rate": 4.956787174409794e-06,
+      "loss": 1.1699,
+      "step": 19588
+    },
+    {
+      "epoch": 0.9024277882710646,
+      "grad_norm": 0.9943836019080448,
+      "learning_rate": 4.952148849868831e-06,
+      "loss": 1.3075,
+      "step": 19589
+    },
+    {
+      "epoch": 0.9024738563596997,
+      "grad_norm": 1.0389885094292057,
+      "learning_rate": 4.947512641405316e-06,
+      "loss": 1.162,
+      "step": 19590
+    },
+    {
+      "epoch": 0.9025199244483346,
+      "grad_norm": 1.350176546153155,
+      "learning_rate": 4.9428785491224426e-06,
+      "loss": 0.809,
+      "step": 19591
+    },
+    {
+      "epoch": 0.9025659925369697,
+      "grad_norm": 1.68315837309256,
+      "learning_rate": 4.938246573123384e-06,
+      "loss": 1.2334,
+      "step": 19592
+    },
+    {
+      "epoch": 0.9026120606256046,
+      "grad_norm": 0.9293807438552075,
+      "learning_rate": 4.933616713511269e-06,
+      "loss": 1.0815,
+      "step": 19593
+    },
+    {
+      "epoch": 0.9026581287142397,
+      "grad_norm": 9.4321189874316,
+      "learning_rate": 4.928988970389192e-06,
+      "loss": 1.0812,
+      "step": 19594
+    },
+    {
+      "epoch": 0.9027041968028746,
+      "grad_norm": 4.208701809941647,
+      "learning_rate": 4.92436334386015e-06,
+      "loss": 1.348,
+      "step": 19595
+    },
+    {
+      "epoch": 0.9027502648915097,
+      "grad_norm": 1.571031182730826,
+      "learning_rate": 4.919739834027137e-06,
+      "loss": 1.2738,
+      "step": 19596
+    },
+    {
+      "epoch": 0.9027963329801446,
+      "grad_norm": 1.0929297167566916,
+      "learning_rate": 4.915118440993094e-06,
+      "loss": 1.4513,
+      "step": 19597
+    },
+    {
+      "epoch": 0.9028424010687797,
+      "grad_norm": 0.647427522818073,
+      "learning_rate": 4.9104991648609046e-06,
+      "loss": 1.3809,
+      "step": 19598
+    },
+    {
+      "epoch": 0.9028884691574146,
+      "grad_norm": 1.7395422045140894,
+      "learning_rate": 4.905882005733397e-06,
+      "loss": 1.3534,
+      "step": 19599
+    },
+    {
+      "epoch": 0.9029345372460497,
+      "grad_norm": 0.8468201104099372,
+      "learning_rate": 4.90126696371338e-06,
+      "loss": 1.11,
+      "step": 19600
+    },
+    {
+      "epoch": 0.9029806053346847,
+      "grad_norm": 1.2145701310750812,
+      "learning_rate": 4.896654038903603e-06,
+      "loss": 0.9625,
+      "step": 19601
+    },
+    {
+      "epoch": 0.9030266734233197,
+      "grad_norm": 0.9704524379979566,
+      "learning_rate": 4.892043231406751e-06,
+      "loss": 1.1866,
+      "step": 19602
+    },
+    {
+      "epoch": 0.9030727415119547,
+      "grad_norm": 0.8772793540966862,
+      "learning_rate": 4.887434541325475e-06,
+      "loss": 1.4684,
+      "step": 19603
+    },
+    {
+      "epoch": 0.9031188096005897,
+      "grad_norm": 1.0148193459330186,
+      "learning_rate": 4.882827968762405e-06,
+      "loss": 1.2362,
+      "step": 19604
+    },
+    {
+      "epoch": 0.9031648776892247,
+      "grad_norm": 1.3529923629791927,
+      "learning_rate": 4.878223513820068e-06,
+      "loss": 0.9407,
+      "step": 19605
+    },
+    {
+      "epoch": 0.9032109457778597,
+      "grad_norm": 0.8115737183523732,
+      "learning_rate": 4.873621176600984e-06,
+      "loss": 1.344,
+      "step": 19606
+    },
+    {
+      "epoch": 0.9032570138664947,
+      "grad_norm": 1.9512168730035075,
+      "learning_rate": 4.869020957207604e-06,
+      "loss": 1.0744,
+      "step": 19607
+    },
+    {
+      "epoch": 0.9033030819551296,
+      "grad_norm": 1.9451855834593437,
+      "learning_rate": 4.864422855742379e-06,
+      "loss": 1.2519,
+      "step": 19608
+    },
+    {
+      "epoch": 0.9033491500437647,
+      "grad_norm": 0.8102301200687408,
+      "learning_rate": 4.859826872307649e-06,
+      "loss": 1.2637,
+      "step": 19609
+    },
+    {
+      "epoch": 0.9033952181323996,
+      "grad_norm": 1.0113988275143366,
+      "learning_rate": 4.855233007005744e-06,
+      "loss": 1.4441,
+      "step": 19610
+    },
+    {
+      "epoch": 0.9034412862210347,
+      "grad_norm": 0.7681476591225336,
+      "learning_rate": 4.850641259938926e-06,
+      "loss": 1.4449,
+      "step": 19611
+    },
+    {
+      "epoch": 0.9034873543096696,
+      "grad_norm": 1.2291313693343344,
+      "learning_rate": 4.846051631209458e-06,
+      "loss": 1.0393,
+      "step": 19612
+    },
+    {
+      "epoch": 0.9035334223983047,
+      "grad_norm": 3.481346201425124,
+      "learning_rate": 4.841464120919492e-06,
+      "loss": 1.5051,
+      "step": 19613
+    },
+    {
+      "epoch": 0.9035794904869398,
+      "grad_norm": 0.9204984822021143,
+      "learning_rate": 4.836878729171146e-06,
+      "loss": 1.1791,
+      "step": 19614
+    },
+    {
+      "epoch": 0.9036255585755747,
+      "grad_norm": 1.1915038782026637,
+      "learning_rate": 4.832295456066538e-06,
+      "loss": 1.1433,
+      "step": 19615
+    },
+    {
+      "epoch": 0.9036716266642097,
+      "grad_norm": 1.0902560831198995,
+      "learning_rate": 4.827714301707686e-06,
+      "loss": 1.0605,
+      "step": 19616
+    },
+    {
+      "epoch": 0.9037176947528447,
+      "grad_norm": 1.2056615767613128,
+      "learning_rate": 4.82313526619661e-06,
+      "loss": 1.2112,
+      "step": 19617
+    },
+    {
+      "epoch": 0.9037637628414797,
+      "grad_norm": 1.3885634482829865,
+      "learning_rate": 4.818558349635205e-06,
+      "loss": 1.2221,
+      "step": 19618
+    },
+    {
+      "epoch": 0.9038098309301147,
+      "grad_norm": 1.6989692877306222,
+      "learning_rate": 4.813983552125423e-06,
+      "loss": 0.9179,
+      "step": 19619
+    },
+    {
+      "epoch": 0.9038558990187497,
+      "grad_norm": 1.2151317393382928,
+      "learning_rate": 4.809410873769071e-06,
+      "loss": 1.1142,
+      "step": 19620
+    },
+    {
+      "epoch": 0.9039019671073847,
+      "grad_norm": 1.8402797642099609,
+      "learning_rate": 4.804840314667969e-06,
+      "loss": 1.3867,
+      "step": 19621
+    },
+    {
+      "epoch": 0.9039480351960197,
+      "grad_norm": 0.9811444420111847,
+      "learning_rate": 4.800271874923889e-06,
+      "loss": 1.2712,
+      "step": 19622
+    },
+    {
+      "epoch": 0.9039941032846547,
+      "grad_norm": 1.6861388210485813,
+      "learning_rate": 4.795705554638519e-06,
+      "loss": 1.3965,
+      "step": 19623
+    },
+    {
+      "epoch": 0.9040401713732897,
+      "grad_norm": 0.8368589598488887,
+      "learning_rate": 4.79114135391352e-06,
+      "loss": 1.5263,
+      "step": 19624
+    },
+    {
+      "epoch": 0.9040862394619247,
+      "grad_norm": 4.216727286053156,
+      "learning_rate": 4.7865792728505e-06,
+      "loss": 1.1777,
+      "step": 19625
+    },
+    {
+      "epoch": 0.9041323075505597,
+      "grad_norm": 0.8303997331873212,
+      "learning_rate": 4.7820193115510555e-06,
+      "loss": 1.1894,
+      "step": 19626
+    },
+    {
+      "epoch": 0.9041783756391947,
+      "grad_norm": 1.237587854264464,
+      "learning_rate": 4.7774614701166735e-06,
+      "loss": 1.2057,
+      "step": 19627
+    },
+    {
+      "epoch": 0.9042244437278297,
+      "grad_norm": 1.1192665814457106,
+      "learning_rate": 4.772905748648848e-06,
+      "loss": 1.2751,
+      "step": 19628
+    },
+    {
+      "epoch": 0.9042705118164648,
+      "grad_norm": 0.701930626080491,
+      "learning_rate": 4.768352147248978e-06,
+      "loss": 1.2749,
+      "step": 19629
+    },
+    {
+      "epoch": 0.9043165799050997,
+      "grad_norm": 1.0026888041980557,
+      "learning_rate": 4.763800666018481e-06,
+      "loss": 1.105,
+      "step": 19630
+    },
+    {
+      "epoch": 0.9043626479937348,
+      "grad_norm": 1.0628833488666878,
+      "learning_rate": 4.759251305058676e-06,
+      "loss": 0.9809,
+      "step": 19631
+    },
+    {
+      "epoch": 0.9044087160823697,
+      "grad_norm": 1.0059396014416018,
+      "learning_rate": 4.754704064470816e-06,
+      "loss": 1.07,
+      "step": 19632
+    },
+    {
+      "epoch": 0.9044547841710048,
+      "grad_norm": 0.8079138131720065,
+      "learning_rate": 4.750158944356165e-06,
+      "loss": 0.9718,
+      "step": 19633
+    },
+    {
+      "epoch": 0.9045008522596397,
+      "grad_norm": 1.9517688060135505,
+      "learning_rate": 4.745615944815906e-06,
+      "loss": 1.5401,
+      "step": 19634
+    },
+    {
+      "epoch": 0.9045469203482748,
+      "grad_norm": 0.929405166957708,
+      "learning_rate": 4.741075065951195e-06,
+      "loss": 1.1448,
+      "step": 19635
+    },
+    {
+      "epoch": 0.9045929884369097,
+      "grad_norm": 1.5462733686743706,
+      "learning_rate": 4.736536307863082e-06,
+      "loss": 1.4265,
+      "step": 19636
+    },
+    {
+      "epoch": 0.9046390565255448,
+      "grad_norm": 1.4446130097810967,
+      "learning_rate": 4.731999670652676e-06,
+      "loss": 1.1164,
+      "step": 19637
+    },
+    {
+      "epoch": 0.9046851246141797,
+      "grad_norm": 4.976660523027023,
+      "learning_rate": 4.727465154420929e-06,
+      "loss": 1.0838,
+      "step": 19638
+    },
+    {
+      "epoch": 0.9047311927028148,
+      "grad_norm": 1.7561085558885918,
+      "learning_rate": 4.722932759268839e-06,
+      "loss": 1.1867,
+      "step": 19639
+    },
+    {
+      "epoch": 0.9047772607914497,
+      "grad_norm": 0.9536181614925523,
+      "learning_rate": 4.718402485297269e-06,
+      "loss": 1.3993,
+      "step": 19640
+    },
+    {
+      "epoch": 0.9048233288800848,
+      "grad_norm": 0.7422143753005401,
+      "learning_rate": 4.7138743326070824e-06,
+      "loss": 1.1965,
+      "step": 19641
+    },
+    {
+      "epoch": 0.9048693969687198,
+      "grad_norm": 1.257883619858369,
+      "learning_rate": 4.709348301299133e-06,
+      "loss": 0.9469,
+      "step": 19642
+    },
+    {
+      "epoch": 0.9049154650573548,
+      "grad_norm": 1.1685915645147942,
+      "learning_rate": 4.704824391474128e-06,
+      "loss": 1.0973,
+      "step": 19643
+    },
+    {
+      "epoch": 0.9049615331459898,
+      "grad_norm": 0.9407792930220861,
+      "learning_rate": 4.7003026032328205e-06,
+      "loss": 1.1148,
+      "step": 19644
+    },
+    {
+      "epoch": 0.9050076012346248,
+      "grad_norm": 0.8727317183637457,
+      "learning_rate": 4.695782936675864e-06,
+      "loss": 1.153,
+      "step": 19645
+    },
+    {
+      "epoch": 0.9050536693232598,
+      "grad_norm": 1.0318922440710177,
+      "learning_rate": 4.691265391903909e-06,
+      "loss": 1.3468,
+      "step": 19646
+    },
+    {
+      "epoch": 0.9050997374118948,
+      "grad_norm": 0.7756160446037362,
+      "learning_rate": 4.68674996901749e-06,
+      "loss": 1.2806,
+      "step": 19647
+    },
+    {
+      "epoch": 0.9051458055005298,
+      "grad_norm": 0.7535780975761017,
+      "learning_rate": 4.682236668117168e-06,
+      "loss": 1.3793,
+      "step": 19648
+    },
+    {
+      "epoch": 0.9051918735891648,
+      "grad_norm": 1.0470646710533928,
+      "learning_rate": 4.67772548930342e-06,
+      "loss": 1.3089,
+      "step": 19649
+    },
+    {
+      "epoch": 0.9052379416777998,
+      "grad_norm": 2.989104104090819,
+      "learning_rate": 4.673216432676652e-06,
+      "loss": 1.0581,
+      "step": 19650
+    },
+    {
+      "epoch": 0.9052840097664347,
+      "grad_norm": 2.25855537519059,
+      "learning_rate": 4.668709498337276e-06,
+      "loss": 1.3052,
+      "step": 19651
+    },
+    {
+      "epoch": 0.9053300778550698,
+      "grad_norm": 1.6188665493337488,
+      "learning_rate": 4.664204686385621e-06,
+      "loss": 0.9332,
+      "step": 19652
+    },
+    {
+      "epoch": 0.9053761459437047,
+      "grad_norm": 0.7964275162801708,
+      "learning_rate": 4.659701996922006e-06,
+      "loss": 1.2248,
+      "step": 19653
+    },
+    {
+      "epoch": 0.9054222140323398,
+      "grad_norm": 0.8048378962483782,
+      "learning_rate": 4.655201430046629e-06,
+      "loss": 0.9366,
+      "step": 19654
+    },
+    {
+      "epoch": 0.9054682821209749,
+      "grad_norm": 1.2058197888373208,
+      "learning_rate": 4.650702985859712e-06,
+      "loss": 1.1257,
+      "step": 19655
+    },
+    {
+      "epoch": 0.9055143502096098,
+      "grad_norm": 1.039310195384112,
+      "learning_rate": 4.646206664461405e-06,
+      "loss": 1.2711,
+      "step": 19656
+    },
+    {
+      "epoch": 0.9055604182982449,
+      "grad_norm": 1.3033392916661886,
+      "learning_rate": 4.64171246595182e-06,
+      "loss": 1.3039,
+      "step": 19657
+    },
+    {
+      "epoch": 0.9056064863868798,
+      "grad_norm": 0.9784705634917384,
+      "learning_rate": 4.6372203904309965e-06,
+      "loss": 1.0145,
+      "step": 19658
+    },
+    {
+      "epoch": 0.9056525544755148,
+      "grad_norm": 0.7607711037810065,
+      "learning_rate": 4.632730437998945e-06,
+      "loss": 1.214,
+      "step": 19659
+    },
+    {
+      "epoch": 0.9056986225641498,
+      "grad_norm": 1.1862692654585083,
+      "learning_rate": 4.628242608755639e-06,
+      "loss": 0.8859,
+      "step": 19660
+    },
+    {
+      "epoch": 0.9057446906527848,
+      "grad_norm": 0.7774880711240907,
+      "learning_rate": 4.62375690280098e-06,
+      "loss": 0.907,
+      "step": 19661
+    },
+    {
+      "epoch": 0.9057907587414198,
+      "grad_norm": 1.2617031707617252,
+      "learning_rate": 4.6192733202348295e-06,
+      "loss": 1.2861,
+      "step": 19662
+    },
+    {
+      "epoch": 0.9058368268300548,
+      "grad_norm": 0.8155915077241812,
+      "learning_rate": 4.61479186115702e-06,
+      "loss": 1.2897,
+      "step": 19663
+    },
+    {
+      "epoch": 0.9058828949186898,
+      "grad_norm": 0.9164798941918845,
+      "learning_rate": 4.610312525667338e-06,
+      "loss": 1.3074,
+      "step": 19664
+    },
+    {
+      "epoch": 0.9059289630073248,
+      "grad_norm": 4.730051593612944,
+      "learning_rate": 4.60583531386547e-06,
+      "loss": 0.9414,
+      "step": 19665
+    },
+    {
+      "epoch": 0.9059750310959598,
+      "grad_norm": 0.9625275423585608,
+      "learning_rate": 4.6013602258511145e-06,
+      "loss": 1.1978,
+      "step": 19666
+    },
+    {
+      "epoch": 0.9060210991845948,
+      "grad_norm": 1.0341815310338496,
+      "learning_rate": 4.596887261723903e-06,
+      "loss": 1.5066,
+      "step": 19667
+    },
+    {
+      "epoch": 0.9060671672732298,
+      "grad_norm": 0.8881991819857401,
+      "learning_rate": 4.592416421583412e-06,
+      "loss": 1.308,
+      "step": 19668
+    },
+    {
+      "epoch": 0.9061132353618648,
+      "grad_norm": 0.711891275923468,
+      "learning_rate": 4.587947705529183e-06,
+      "loss": 1.2998,
+      "step": 19669
+    },
+    {
+      "epoch": 0.9061593034504999,
+      "grad_norm": 1.8479316742261902,
+      "learning_rate": 4.583481113660692e-06,
+      "loss": 1.424,
+      "step": 19670
+    },
+    {
+      "epoch": 0.9062053715391348,
+      "grad_norm": 1.3975969363383467,
+      "learning_rate": 4.579016646077416e-06,
+      "loss": 1.0133,
+      "step": 19671
+    },
+    {
+      "epoch": 0.9062514396277699,
+      "grad_norm": 1.3102846606499976,
+      "learning_rate": 4.574554302878697e-06,
+      "loss": 1.2977,
+      "step": 19672
+    },
+    {
+      "epoch": 0.9062975077164048,
+      "grad_norm": 1.2390933798870571,
+      "learning_rate": 4.570094084163912e-06,
+      "loss": 1.3748,
+      "step": 19673
+    },
+    {
+      "epoch": 0.9063435758050399,
+      "grad_norm": 0.9528880517853313,
+      "learning_rate": 4.565635990032358e-06,
+      "loss": 1.3072,
+      "step": 19674
+    },
+    {
+      "epoch": 0.9063896438936748,
+      "grad_norm": 1.975388769842371,
+      "learning_rate": 4.5611800205833e-06,
+      "loss": 1.1755,
+      "step": 19675
+    },
+    {
+      "epoch": 0.9064357119823099,
+      "grad_norm": 0.679975990265317,
+      "learning_rate": 4.556726175915904e-06,
+      "loss": 1.3121,
+      "step": 19676
+    },
+    {
+      "epoch": 0.9064817800709448,
+      "grad_norm": 1.2094281970373055,
+      "learning_rate": 4.552274456129357e-06,
+      "loss": 0.993,
+      "step": 19677
+    },
+    {
+      "epoch": 0.9065278481595799,
+      "grad_norm": 1.1887818806422636,
+      "learning_rate": 4.547824861322769e-06,
+      "loss": 1.1953,
+      "step": 19678
+    },
+    {
+      "epoch": 0.9065739162482148,
+      "grad_norm": 1.0424068236291129,
+      "learning_rate": 4.543377391595182e-06,
+      "loss": 1.319,
+      "step": 19679
+    },
+    {
+      "epoch": 0.9066199843368499,
+      "grad_norm": 0.6749662283471642,
+      "learning_rate": 4.538932047045629e-06,
+      "loss": 1.2336,
+      "step": 19680
+    },
+    {
+      "epoch": 0.9066660524254848,
+      "grad_norm": 1.4334621052595562,
+      "learning_rate": 4.534488827773065e-06,
+      "loss": 1.0881,
+      "step": 19681
+    },
+    {
+      "epoch": 0.9067121205141199,
+      "grad_norm": 0.8119802117348098,
+      "learning_rate": 4.530047733876441e-06,
+      "loss": 1.108,
+      "step": 19682
+    },
+    {
+      "epoch": 0.9067581886027549,
+      "grad_norm": 1.3163552883547005,
+      "learning_rate": 4.525608765454592e-06,
+      "loss": 1.0633,
+      "step": 19683
+    },
+    {
+      "epoch": 0.9068042566913899,
+      "grad_norm": 0.9221243012578725,
+      "learning_rate": 4.52117192260636e-06,
+      "loss": 1.3912,
+      "step": 19684
+    },
+    {
+      "epoch": 0.9068503247800249,
+      "grad_norm": 1.0817155899017612,
+      "learning_rate": 4.516737205430532e-06,
+      "loss": 1.0629,
+      "step": 19685
+    },
+    {
+      "epoch": 0.9068963928686599,
+      "grad_norm": 1.119447502276221,
+      "learning_rate": 4.512304614025831e-06,
+      "loss": 0.95,
+      "step": 19686
+    },
+    {
+      "epoch": 0.9069424609572949,
+      "grad_norm": 1.6028830539228214,
+      "learning_rate": 4.507874148490931e-06,
+      "loss": 1.2895,
+      "step": 19687
+    },
+    {
+      "epoch": 0.9069885290459299,
+      "grad_norm": 1.3058600718794502,
+      "learning_rate": 4.503445808924478e-06,
+      "loss": 1.1236,
+      "step": 19688
+    },
+    {
+      "epoch": 0.9070345971345649,
+      "grad_norm": 1.0368047637738302,
+      "learning_rate": 4.499019595425081e-06,
+      "loss": 1.1393,
+      "step": 19689
+    },
+    {
+      "epoch": 0.9070806652231999,
+      "grad_norm": 0.6481751858174055,
+      "learning_rate": 4.494595508091249e-06,
+      "loss": 1.5221,
+      "step": 19690
+    },
+    {
+      "epoch": 0.9071267333118349,
+      "grad_norm": 1.2913716472744525,
+      "learning_rate": 4.490173547021492e-06,
+      "loss": 1.6083,
+      "step": 19691
+    },
+    {
+      "epoch": 0.9071728014004699,
+      "grad_norm": 1.3634404846745547,
+      "learning_rate": 4.485753712314255e-06,
+      "loss": 1.3275,
+      "step": 19692
+    },
+    {
+      "epoch": 0.9072188694891049,
+      "grad_norm": 0.9252139898919258,
+      "learning_rate": 4.4813360040679595e-06,
+      "loss": 1.0728,
+      "step": 19693
+    },
+    {
+      "epoch": 0.9072649375777398,
+      "grad_norm": 0.6754896787151593,
+      "learning_rate": 4.476920422380915e-06,
+      "loss": 1.3889,
+      "step": 19694
+    },
+    {
+      "epoch": 0.9073110056663749,
+      "grad_norm": 1.1387705075228087,
+      "learning_rate": 4.472506967351464e-06,
+      "loss": 1.0585,
+      "step": 19695
+    },
+    {
+      "epoch": 0.90735707375501,
+      "grad_norm": 0.7949818143754871,
+      "learning_rate": 4.468095639077852e-06,
+      "loss": 1.447,
+      "step": 19696
+    },
+    {
+      "epoch": 0.9074031418436449,
+      "grad_norm": 0.6990638606574388,
+      "learning_rate": 4.463686437658277e-06,
+      "loss": 1.392,
+      "step": 19697
+    },
+    {
+      "epoch": 0.90744920993228,
+      "grad_norm": 2.1596623536673656,
+      "learning_rate": 4.459279363190916e-06,
+      "loss": 1.1729,
+      "step": 19698
+    },
+    {
+      "epoch": 0.9074952780209149,
+      "grad_norm": 0.6956395237531072,
+      "learning_rate": 4.45487441577388e-06,
+      "loss": 1.0283,
+      "step": 19699
+    },
+    {
+      "epoch": 0.90754134610955,
+      "grad_norm": 1.7322359843721615,
+      "learning_rate": 4.450471595505256e-06,
+      "loss": 1.2098,
+      "step": 19700
+    },
+    {
+      "epoch": 0.9075874141981849,
+      "grad_norm": 1.027314480058277,
+      "learning_rate": 4.446070902483035e-06,
+      "loss": 1.1921,
+      "step": 19701
+    },
+    {
+      "epoch": 0.90763348228682,
+      "grad_norm": 1.0065641542935342,
+      "learning_rate": 4.441672336805214e-06,
+      "loss": 1.0753,
+      "step": 19702
+    },
+    {
+      "epoch": 0.9076795503754549,
+      "grad_norm": 1.1797796453362668,
+      "learning_rate": 4.437275898569704e-06,
+      "loss": 1.0658,
+      "step": 19703
+    },
+    {
+      "epoch": 0.90772561846409,
+      "grad_norm": 1.1870566704989363,
+      "learning_rate": 4.4328815878744046e-06,
+      "loss": 1.4685,
+      "step": 19704
+    },
+    {
+      "epoch": 0.9077716865527249,
+      "grad_norm": 0.8231770877572252,
+      "learning_rate": 4.428489404817127e-06,
+      "loss": 0.949,
+      "step": 19705
+    },
+    {
+      "epoch": 0.9078177546413599,
+      "grad_norm": 1.0574503197448801,
+      "learning_rate": 4.424099349495647e-06,
+      "loss": 1.5918,
+      "step": 19706
+    },
+    {
+      "epoch": 0.9078638227299949,
+      "grad_norm": 0.8360908896212248,
+      "learning_rate": 4.419711422007744e-06,
+      "loss": 1.2338,
+      "step": 19707
+    },
+    {
+      "epoch": 0.9079098908186299,
+      "grad_norm": 1.2673931532297285,
+      "learning_rate": 4.415325622451061e-06,
+      "loss": 1.2517,
+      "step": 19708
+    },
+    {
+      "epoch": 0.907955958907265,
+      "grad_norm": 0.9199185388570126,
+      "learning_rate": 4.410941950923264e-06,
+      "loss": 0.8029,
+      "step": 19709
+    },
+    {
+      "epoch": 0.9080020269958999,
+      "grad_norm": 0.8123704882364904,
+      "learning_rate": 4.406560407521942e-06,
+      "loss": 1.1913,
+      "step": 19710
+    },
+    {
+      "epoch": 0.908048095084535,
+      "grad_norm": 0.7055258371847954,
+      "learning_rate": 4.402180992344662e-06,
+      "loss": 1.3303,
+      "step": 19711
+    },
+    {
+      "epoch": 0.9080941631731699,
+      "grad_norm": 1.4922077600748018,
+      "learning_rate": 4.3978037054888895e-06,
+      "loss": 1.1468,
+      "step": 19712
+    },
+    {
+      "epoch": 0.908140231261805,
+      "grad_norm": 0.8621108914953917,
+      "learning_rate": 4.393428547052103e-06,
+      "loss": 1.5249,
+      "step": 19713
+    },
+    {
+      "epoch": 0.9081862993504399,
+      "grad_norm": 0.9687134059330246,
+      "learning_rate": 4.389055517131701e-06,
+      "loss": 1.3608,
+      "step": 19714
+    },
+    {
+      "epoch": 0.908232367439075,
+      "grad_norm": 0.7019057341190875,
+      "learning_rate": 4.384684615825041e-06,
+      "loss": 1.314,
+      "step": 19715
+    },
+    {
+      "epoch": 0.9082784355277099,
+      "grad_norm": 0.9902034965346928,
+      "learning_rate": 4.380315843229433e-06,
+      "loss": 1.4001,
+      "step": 19716
+    },
+    {
+      "epoch": 0.908324503616345,
+      "grad_norm": 0.8936214747838911,
+      "learning_rate": 4.37594919944212e-06,
+      "loss": 1.3154,
+      "step": 19717
+    },
+    {
+      "epoch": 0.9083705717049799,
+      "grad_norm": 0.7953307114793723,
+      "learning_rate": 4.3715846845603594e-06,
+      "loss": 1.2469,
+      "step": 19718
+    },
+    {
+      "epoch": 0.908416639793615,
+      "grad_norm": 0.6958461556922848,
+      "learning_rate": 4.367222298681295e-06,
+      "loss": 1.461,
+      "step": 19719
+    },
+    {
+      "epoch": 0.9084627078822499,
+      "grad_norm": 0.759903948768367,
+      "learning_rate": 4.3628620419020495e-06,
+      "loss": 1.2192,
+      "step": 19720
+    },
+    {
+      "epoch": 0.908508775970885,
+      "grad_norm": 0.9217904780365954,
+      "learning_rate": 4.358503914319689e-06,
+      "loss": 0.898,
+      "step": 19721
+    },
+    {
+      "epoch": 0.9085548440595199,
+      "grad_norm": 0.8427989583358221,
+      "learning_rate": 4.35414791603127e-06,
+      "loss": 1.2777,
+      "step": 19722
+    },
+    {
+      "epoch": 0.908600912148155,
+      "grad_norm": 0.9280744487211094,
+      "learning_rate": 4.3497940471337486e-06,
+      "loss": 1.1765,
+      "step": 19723
+    },
+    {
+      "epoch": 0.90864698023679,
+      "grad_norm": 1.553592779123735,
+      "learning_rate": 4.345442307724024e-06,
+      "loss": 1.2663,
+      "step": 19724
+    },
+    {
+      "epoch": 0.908693048325425,
+      "grad_norm": 1.1657377720113655,
+      "learning_rate": 4.341092697899052e-06,
+      "loss": 1.1815,
+      "step": 19725
+    },
+    {
+      "epoch": 0.90873911641406,
+      "grad_norm": 1.2949366357745227,
+      "learning_rate": 4.336745217755611e-06,
+      "loss": 1.2534,
+      "step": 19726
+    },
+    {
+      "epoch": 0.908785184502695,
+      "grad_norm": 1.1576915880949465,
+      "learning_rate": 4.332399867390524e-06,
+      "loss": 1.2826,
+      "step": 19727
+    },
+    {
+      "epoch": 0.90883125259133,
+      "grad_norm": 1.1123005295156987,
+      "learning_rate": 4.3280566469004915e-06,
+      "loss": 1.1198,
+      "step": 19728
+    },
+    {
+      "epoch": 0.908877320679965,
+      "grad_norm": 1.2452801289340183,
+      "learning_rate": 4.323715556382268e-06,
+      "loss": 1.0441,
+      "step": 19729
+    },
+    {
+      "epoch": 0.9089233887686,
+      "grad_norm": 0.8733429147436046,
+      "learning_rate": 4.319376595932445e-06,
+      "loss": 1.3991,
+      "step": 19730
+    },
+    {
+      "epoch": 0.908969456857235,
+      "grad_norm": 0.7479239649356865,
+      "learning_rate": 4.315039765647666e-06,
+      "loss": 1.4824,
+      "step": 19731
+    },
+    {
+      "epoch": 0.90901552494587,
+      "grad_norm": 1.9273961599556477,
+      "learning_rate": 4.310705065624454e-06,
+      "loss": 1.4184,
+      "step": 19732
+    },
+    {
+      "epoch": 0.909061593034505,
+      "grad_norm": 1.0781166145069951,
+      "learning_rate": 4.3063724959593095e-06,
+      "loss": 1.2627,
+      "step": 19733
+    },
+    {
+      "epoch": 0.90910766112314,
+      "grad_norm": 0.7993682374227729,
+      "learning_rate": 4.302042056748723e-06,
+      "loss": 1.077,
+      "step": 19734
+    },
+    {
+      "epoch": 0.909153729211775,
+      "grad_norm": 1.2598872113436455,
+      "learning_rate": 4.297713748089049e-06,
+      "loss": 0.9763,
+      "step": 19735
+    },
+    {
+      "epoch": 0.90919979730041,
+      "grad_norm": 4.526761286735517,
+      "learning_rate": 4.293387570076723e-06,
+      "loss": 1.247,
+      "step": 19736
+    },
+    {
+      "epoch": 0.9092458653890451,
+      "grad_norm": 1.0077066763791747,
+      "learning_rate": 4.289063522808001e-06,
+      "loss": 1.4648,
+      "step": 19737
+    },
+    {
+      "epoch": 0.90929193347768,
+      "grad_norm": 1.2194350347557472,
+      "learning_rate": 4.284741606379184e-06,
+      "loss": 1.2486,
+      "step": 19738
+    },
+    {
+      "epoch": 0.9093380015663151,
+      "grad_norm": 0.7858714705020649,
+      "learning_rate": 4.2804218208864484e-06,
+      "loss": 1.1033,
+      "step": 19739
+    },
+    {
+      "epoch": 0.90938406965495,
+      "grad_norm": 1.1035127179922692,
+      "learning_rate": 4.276104166426032e-06,
+      "loss": 1.0565,
+      "step": 19740
+    },
+    {
+      "epoch": 0.909430137743585,
+      "grad_norm": 0.768174672179032,
+      "learning_rate": 4.27178864309401e-06,
+      "loss": 1.1428,
+      "step": 19741
+    },
+    {
+      "epoch": 0.90947620583222,
+      "grad_norm": 2.3721071883721048,
+      "learning_rate": 4.267475250986475e-06,
+      "loss": 1.225,
+      "step": 19742
+    },
+    {
+      "epoch": 0.909522273920855,
+      "grad_norm": 1.6832890744772802,
+      "learning_rate": 4.2631639901994595e-06,
+      "loss": 1.2964,
+      "step": 19743
+    },
+    {
+      "epoch": 0.90956834200949,
+      "grad_norm": 1.33138628461376,
+      "learning_rate": 4.258854860828942e-06,
+      "loss": 1.1895,
+      "step": 19744
+    },
+    {
+      "epoch": 0.909614410098125,
+      "grad_norm": 1.2662337372836003,
+      "learning_rate": 4.25454786297087e-06,
+      "loss": 1.2208,
+      "step": 19745
+    },
+    {
+      "epoch": 0.90966047818676,
+      "grad_norm": 1.0102311602226695,
+      "learning_rate": 4.250242996721088e-06,
+      "loss": 1.65,
+      "step": 19746
+    },
+    {
+      "epoch": 0.909706546275395,
+      "grad_norm": 2.388358615521272,
+      "learning_rate": 4.245940262175507e-06,
+      "loss": 1.1551,
+      "step": 19747
+    },
+    {
+      "epoch": 0.90975261436403,
+      "grad_norm": 1.3259422333299793,
+      "learning_rate": 4.2416396594298745e-06,
+      "loss": 1.0549,
+      "step": 19748
+    },
+    {
+      "epoch": 0.909798682452665,
+      "grad_norm": 0.8616294890467217,
+      "learning_rate": 4.2373411885799465e-06,
+      "loss": 1.3582,
+      "step": 19749
+    },
+    {
+      "epoch": 0.9098447505413001,
+      "grad_norm": 1.1508313679809685,
+      "learning_rate": 4.233044849721412e-06,
+      "loss": 1.0714,
+      "step": 19750
+    },
+    {
+      "epoch": 0.909890818629935,
+      "grad_norm": 1.2116972839439466,
+      "learning_rate": 4.228750642949919e-06,
+      "loss": 1.0303,
+      "step": 19751
+    },
+    {
+      "epoch": 0.9099368867185701,
+      "grad_norm": 2.804304839884077,
+      "learning_rate": 4.224458568361101e-06,
+      "loss": 1.2846,
+      "step": 19752
+    },
+    {
+      "epoch": 0.909982954807205,
+      "grad_norm": 0.8121503241871205,
+      "learning_rate": 4.22016862605048e-06,
+      "loss": 1.3225,
+      "step": 19753
+    },
+    {
+      "epoch": 0.9100290228958401,
+      "grad_norm": 0.9172864183319763,
+      "learning_rate": 4.2158808161135825e-06,
+      "loss": 1.4908,
+      "step": 19754
+    },
+    {
+      "epoch": 0.910075090984475,
+      "grad_norm": 1.0599897196202916,
+      "learning_rate": 4.211595138645863e-06,
+      "loss": 1.3367,
+      "step": 19755
+    },
+    {
+      "epoch": 0.9101211590731101,
+      "grad_norm": 1.0844102592667786,
+      "learning_rate": 4.207311593742747e-06,
+      "loss": 1.1147,
+      "step": 19756
+    },
+    {
+      "epoch": 0.910167227161745,
+      "grad_norm": 0.9033349638611666,
+      "learning_rate": 4.203030181499568e-06,
+      "loss": 1.1685,
+      "step": 19757
+    },
+    {
+      "epoch": 0.9102132952503801,
+      "grad_norm": 3.4828987144474586,
+      "learning_rate": 4.1987509020116725e-06,
+      "loss": 1.0532,
+      "step": 19758
+    },
+    {
+      "epoch": 0.910259363339015,
+      "grad_norm": 0.866217698598193,
+      "learning_rate": 4.1944737553743394e-06,
+      "loss": 1.1455,
+      "step": 19759
+    },
+    {
+      "epoch": 0.9103054314276501,
+      "grad_norm": 1.5805673745936009,
+      "learning_rate": 4.190198741682761e-06,
+      "loss": 1.1846,
+      "step": 19760
+    },
+    {
+      "epoch": 0.910351499516285,
+      "grad_norm": 0.7484009211305157,
+      "learning_rate": 4.185925861032125e-06,
+      "loss": 1.4699,
+      "step": 19761
+    },
+    {
+      "epoch": 0.9103975676049201,
+      "grad_norm": 0.945596186512801,
+      "learning_rate": 4.181655113517569e-06,
+      "loss": 1.1289,
+      "step": 19762
+    },
+    {
+      "epoch": 0.910443635693555,
+      "grad_norm": 0.8134437608149166,
+      "learning_rate": 4.177386499234181e-06,
+      "loss": 1.4381,
+      "step": 19763
+    },
+    {
+      "epoch": 0.9104897037821901,
+      "grad_norm": 2.935484856147544,
+      "learning_rate": 4.173120018276966e-06,
+      "loss": 1.2018,
+      "step": 19764
+    },
+    {
+      "epoch": 0.9105357718708251,
+      "grad_norm": 0.9296310124871194,
+      "learning_rate": 4.168855670740934e-06,
+      "loss": 1.1394,
+      "step": 19765
+    },
+    {
+      "epoch": 0.9105818399594601,
+      "grad_norm": 0.8762072055456086,
+      "learning_rate": 4.164593456721011e-06,
+      "loss": 1.3668,
+      "step": 19766
+    },
+    {
+      "epoch": 0.9106279080480951,
+      "grad_norm": 0.7882466296035754,
+      "learning_rate": 4.1603333763120975e-06,
+      "loss": 1.2853,
+      "step": 19767
+    },
+    {
+      "epoch": 0.9106739761367301,
+      "grad_norm": 0.9104421235411495,
+      "learning_rate": 4.15607542960903e-06,
+      "loss": 1.265,
+      "step": 19768
+    },
+    {
+      "epoch": 0.9107200442253651,
+      "grad_norm": 1.089106763707707,
+      "learning_rate": 4.151819616706598e-06,
+      "loss": 1.021,
+      "step": 19769
+    },
+    {
+      "epoch": 0.9107661123140001,
+      "grad_norm": 2.092077831986323,
+      "learning_rate": 4.147565937699572e-06,
+      "loss": 1.0882,
+      "step": 19770
+    },
+    {
+      "epoch": 0.9108121804026351,
+      "grad_norm": 0.9151009172842781,
+      "learning_rate": 4.143314392682629e-06,
+      "loss": 1.1438,
+      "step": 19771
+    },
+    {
+      "epoch": 0.9108582484912701,
+      "grad_norm": 1.470912647386061,
+      "learning_rate": 4.139064981750429e-06,
+      "loss": 1.3031,
+      "step": 19772
+    },
+    {
+      "epoch": 0.9109043165799051,
+      "grad_norm": 0.9450757831237163,
+      "learning_rate": 4.134817704997584e-06,
+      "loss": 1.2681,
+      "step": 19773
+    },
+    {
+      "epoch": 0.9109503846685401,
+      "grad_norm": 1.388402116667581,
+      "learning_rate": 4.130572562518664e-06,
+      "loss": 1.0046,
+      "step": 19774
+    },
+    {
+      "epoch": 0.9109964527571751,
+      "grad_norm": 1.206232766598409,
+      "learning_rate": 4.126329554408148e-06,
+      "loss": 0.8653,
+      "step": 19775
+    },
+    {
+      "epoch": 0.91104252084581,
+      "grad_norm": 1.6515852057424532,
+      "learning_rate": 4.1220886807605165e-06,
+      "loss": 1.4133,
+      "step": 19776
+    },
+    {
+      "epoch": 0.9110885889344451,
+      "grad_norm": 0.9728030748518881,
+      "learning_rate": 4.117849941670204e-06,
+      "loss": 1.1275,
+      "step": 19777
+    },
+    {
+      "epoch": 0.9111346570230802,
+      "grad_norm": 0.7895371291768993,
+      "learning_rate": 4.113613337231537e-06,
+      "loss": 1.4974,
+      "step": 19778
+    },
+    {
+      "epoch": 0.9111807251117151,
+      "grad_norm": 0.8984169873687365,
+      "learning_rate": 4.109378867538871e-06,
+      "loss": 1.1243,
+      "step": 19779
+    },
+    {
+      "epoch": 0.9112267932003502,
+      "grad_norm": 2.941595033776481,
+      "learning_rate": 4.105146532686455e-06,
+      "loss": 1.0265,
+      "step": 19780
+    },
+    {
+      "epoch": 0.9112728612889851,
+      "grad_norm": 2.538397150153149,
+      "learning_rate": 4.1009163327685455e-06,
+      "loss": 1.3289,
+      "step": 19781
+    },
+    {
+      "epoch": 0.9113189293776202,
+      "grad_norm": 0.7310207777038371,
+      "learning_rate": 4.09668826787929e-06,
+      "loss": 1.2219,
+      "step": 19782
+    },
+    {
+      "epoch": 0.9113649974662551,
+      "grad_norm": 0.7933166358807026,
+      "learning_rate": 4.092462338112835e-06,
+      "loss": 1.2773,
+      "step": 19783
+    },
+    {
+      "epoch": 0.9114110655548902,
+      "grad_norm": 0.8172125064824098,
+      "learning_rate": 4.08823854356325e-06,
+      "loss": 1.5075,
+      "step": 19784
+    },
+    {
+      "epoch": 0.9114571336435251,
+      "grad_norm": 4.769953872922277,
+      "learning_rate": 4.084016884324604e-06,
+      "loss": 1.3612,
+      "step": 19785
+    },
+    {
+      "epoch": 0.9115032017321602,
+      "grad_norm": 0.9581019672761791,
+      "learning_rate": 4.079797360490834e-06,
+      "loss": 1.44,
+      "step": 19786
+    },
+    {
+      "epoch": 0.9115492698207951,
+      "grad_norm": 1.1905819747137079,
+      "learning_rate": 4.075579972155919e-06,
+      "loss": 1.0302,
+      "step": 19787
+    },
+    {
+      "epoch": 0.9115953379094301,
+      "grad_norm": 0.7813136466000326,
+      "learning_rate": 4.07136471941375e-06,
+      "loss": 1.2648,
+      "step": 19788
+    },
+    {
+      "epoch": 0.9116414059980651,
+      "grad_norm": 1.3322257204254686,
+      "learning_rate": 4.0671516023581545e-06,
+      "loss": 1.3821,
+      "step": 19789
+    },
+    {
+      "epoch": 0.9116874740867001,
+      "grad_norm": 0.9337619762746392,
+      "learning_rate": 4.062940621082933e-06,
+      "loss": 1.2433,
+      "step": 19790
+    },
+    {
+      "epoch": 0.9117335421753352,
+      "grad_norm": 1.2986001781122365,
+      "learning_rate": 4.058731775681846e-06,
+      "loss": 1.3049,
+      "step": 19791
+    },
+    {
+      "epoch": 0.9117796102639701,
+      "grad_norm": 1.249883503536425,
+      "learning_rate": 4.054525066248604e-06,
+      "loss": 1.2092,
+      "step": 19792
+    },
+    {
+      "epoch": 0.9118256783526052,
+      "grad_norm": 1.7675066177387184,
+      "learning_rate": 4.050320492876847e-06,
+      "loss": 1.2001,
+      "step": 19793
+    },
+    {
+      "epoch": 0.9118717464412401,
+      "grad_norm": 1.0969868656308526,
+      "learning_rate": 4.046118055660175e-06,
+      "loss": 1.1083,
+      "step": 19794
+    },
+    {
+      "epoch": 0.9119178145298752,
+      "grad_norm": 0.7327220130286513,
+      "learning_rate": 4.041917754692171e-06,
+      "loss": 1.1525,
+      "step": 19795
+    },
+    {
+      "epoch": 0.9119638826185101,
+      "grad_norm": 0.9512356580275226,
+      "learning_rate": 4.037719590066325e-06,
+      "loss": 1.2228,
+      "step": 19796
+    },
+    {
+      "epoch": 0.9120099507071452,
+      "grad_norm": 1.3345017990992738,
+      "learning_rate": 4.033523561876107e-06,
+      "loss": 1.0228,
+      "step": 19797
+    },
+    {
+      "epoch": 0.9120560187957801,
+      "grad_norm": 0.8797494518567214,
+      "learning_rate": 4.029329670214943e-06,
+      "loss": 1.1109,
+      "step": 19798
+    },
+    {
+      "epoch": 0.9121020868844152,
+      "grad_norm": 0.7063612588823397,
+      "learning_rate": 4.0251379151762136e-06,
+      "loss": 1.1395,
+      "step": 19799
+    },
+    {
+      "epoch": 0.9121481549730501,
+      "grad_norm": 0.944461714520055,
+      "learning_rate": 4.020948296853222e-06,
+      "loss": 1.1391,
+      "step": 19800
+    },
+    {
+      "epoch": 0.9121942230616852,
+      "grad_norm": 0.8565707047084168,
+      "learning_rate": 4.016760815339238e-06,
+      "loss": 1.0479,
+      "step": 19801
+    },
+    {
+      "epoch": 0.9122402911503201,
+      "grad_norm": 2.9524773182779067,
+      "learning_rate": 4.012575470727498e-06,
+      "loss": 0.9373,
+      "step": 19802
+    },
+    {
+      "epoch": 0.9122863592389552,
+      "grad_norm": 2.08300799402221,
+      "learning_rate": 4.0083922631112044e-06,
+      "loss": 1.139,
+      "step": 19803
+    },
+    {
+      "epoch": 0.9123324273275901,
+      "grad_norm": 3.964242011726602,
+      "learning_rate": 4.004211192583451e-06,
+      "loss": 1.1067,
+      "step": 19804
+    },
+    {
+      "epoch": 0.9123784954162252,
+      "grad_norm": 0.7787038480308786,
+      "learning_rate": 4.000032259237329e-06,
+      "loss": 1.1933,
+      "step": 19805
+    },
+    {
+      "epoch": 0.9124245635048602,
+      "grad_norm": 0.6214783109703413,
+      "learning_rate": 3.995855463165898e-06,
+      "loss": 1.1545,
+      "step": 19806
+    },
+    {
+      "epoch": 0.9124706315934952,
+      "grad_norm": 0.7381884747269712,
+      "learning_rate": 3.991680804462128e-06,
+      "loss": 1.3327,
+      "step": 19807
+    },
+    {
+      "epoch": 0.9125166996821302,
+      "grad_norm": 1.118789971137547,
+      "learning_rate": 3.987508283218955e-06,
+      "loss": 0.9279,
+      "step": 19808
+    },
+    {
+      "epoch": 0.9125627677707652,
+      "grad_norm": 1.6392297898947568,
+      "learning_rate": 3.983337899529294e-06,
+      "loss": 1.0462,
+      "step": 19809
+    },
+    {
+      "epoch": 0.9126088358594002,
+      "grad_norm": 1.023552772378845,
+      "learning_rate": 3.979169653485993e-06,
+      "loss": 1.195,
+      "step": 19810
+    },
+    {
+      "epoch": 0.9126549039480352,
+      "grad_norm": 2.1512938574634366,
+      "learning_rate": 3.975003545181821e-06,
+      "loss": 1.5231,
+      "step": 19811
+    },
+    {
+      "epoch": 0.9127009720366702,
+      "grad_norm": 3.7233066135008253,
+      "learning_rate": 3.970839574709551e-06,
+      "loss": 1.1177,
+      "step": 19812
+    },
+    {
+      "epoch": 0.9127470401253052,
+      "grad_norm": 1.0574022444271707,
+      "learning_rate": 3.966677742161895e-06,
+      "loss": 1.1334,
+      "step": 19813
+    },
+    {
+      "epoch": 0.9127931082139402,
+      "grad_norm": 0.9238272220261623,
+      "learning_rate": 3.9625180476314915e-06,
+      "loss": 1.0627,
+      "step": 19814
+    },
+    {
+      "epoch": 0.9128391763025752,
+      "grad_norm": 1.0749939288538586,
+      "learning_rate": 3.958360491210955e-06,
+      "loss": 1.1468,
+      "step": 19815
+    },
+    {
+      "epoch": 0.9128852443912102,
+      "grad_norm": 2.1837597351831652,
+      "learning_rate": 3.9542050729928335e-06,
+      "loss": 1.3386,
+      "step": 19816
+    },
+    {
+      "epoch": 0.9129313124798452,
+      "grad_norm": 0.8450549889017903,
+      "learning_rate": 3.950051793069676e-06,
+      "loss": 1.2969,
+      "step": 19817
+    },
+    {
+      "epoch": 0.9129773805684802,
+      "grad_norm": 0.7700832198691918,
+      "learning_rate": 3.945900651533907e-06,
+      "loss": 1.0782,
+      "step": 19818
+    },
+    {
+      "epoch": 0.9130234486571153,
+      "grad_norm": 0.7807192055146807,
+      "learning_rate": 3.941751648477965e-06,
+      "loss": 1.1993,
+      "step": 19819
+    },
+    {
+      "epoch": 0.9130695167457502,
+      "grad_norm": 0.7768232497894919,
+      "learning_rate": 3.937604783994209e-06,
+      "loss": 1.4011,
+      "step": 19820
+    },
+    {
+      "epoch": 0.9131155848343853,
+      "grad_norm": 1.5422576726906516,
+      "learning_rate": 3.933460058174998e-06,
+      "loss": 1.2926,
+      "step": 19821
+    },
+    {
+      "epoch": 0.9131616529230202,
+      "grad_norm": 0.8476738501274754,
+      "learning_rate": 3.929317471112559e-06,
+      "loss": 0.9798,
+      "step": 19822
+    },
+    {
+      "epoch": 0.9132077210116553,
+      "grad_norm": 1.1360752399799596,
+      "learning_rate": 3.92517702289914e-06,
+      "loss": 1.3889,
+      "step": 19823
+    },
+    {
+      "epoch": 0.9132537891002902,
+      "grad_norm": 1.3469675905314145,
+      "learning_rate": 3.921038713626934e-06,
+      "loss": 1.0743,
+      "step": 19824
+    },
+    {
+      "epoch": 0.9132998571889253,
+      "grad_norm": 2.0889816978877356,
+      "learning_rate": 3.9169025433880564e-06,
+      "loss": 1.457,
+      "step": 19825
+    },
+    {
+      "epoch": 0.9133459252775602,
+      "grad_norm": 3.468519827556173,
+      "learning_rate": 3.912768512274601e-06,
+      "loss": 1.1243,
+      "step": 19826
+    },
+    {
+      "epoch": 0.9133919933661953,
+      "grad_norm": 1.2754718451771925,
+      "learning_rate": 3.908636620378581e-06,
+      "loss": 1.208,
+      "step": 19827
+    },
+    {
+      "epoch": 0.9134380614548302,
+      "grad_norm": 1.2374153309075224,
+      "learning_rate": 3.904506867792024e-06,
+      "loss": 1.437,
+      "step": 19828
+    },
+    {
+      "epoch": 0.9134841295434653,
+      "grad_norm": 1.2688140687346583,
+      "learning_rate": 3.900379254606845e-06,
+      "loss": 1.1939,
+      "step": 19829
+    },
+    {
+      "epoch": 0.9135301976321002,
+      "grad_norm": 1.1732419522617652,
+      "learning_rate": 3.89625378091496e-06,
+      "loss": 0.9624,
+      "step": 19830
+    },
+    {
+      "epoch": 0.9135762657207352,
+      "grad_norm": 0.7793858625438513,
+      "learning_rate": 3.892130446808173e-06,
+      "loss": 1.4012,
+      "step": 19831
+    },
+    {
+      "epoch": 0.9136223338093703,
+      "grad_norm": 1.2632188310631693,
+      "learning_rate": 3.888009252378333e-06,
+      "loss": 1.2482,
+      "step": 19832
+    },
+    {
+      "epoch": 0.9136684018980052,
+      "grad_norm": 1.4958955910219782,
+      "learning_rate": 3.883890197717177e-06,
+      "loss": 1.1409,
+      "step": 19833
+    },
+    {
+      "epoch": 0.9137144699866403,
+      "grad_norm": 0.7951384639573866,
+      "learning_rate": 3.879773282916377e-06,
+      "loss": 1.3156,
+      "step": 19834
+    },
+    {
+      "epoch": 0.9137605380752752,
+      "grad_norm": 0.7632074013936484,
+      "learning_rate": 3.875658508067636e-06,
+      "loss": 1.3346,
+      "step": 19835
+    },
+    {
+      "epoch": 0.9138066061639103,
+      "grad_norm": 2.009280277070377,
+      "learning_rate": 3.871545873262528e-06,
+      "loss": 1.1178,
+      "step": 19836
+    },
+    {
+      "epoch": 0.9138526742525452,
+      "grad_norm": 9.568368110788777,
+      "learning_rate": 3.867435378592643e-06,
+      "loss": 1.1313,
+      "step": 19837
+    },
+    {
+      "epoch": 0.9138987423411803,
+      "grad_norm": 1.0090059520396764,
+      "learning_rate": 3.8633270241494544e-06,
+      "loss": 1.0661,
+      "step": 19838
+    },
+    {
+      "epoch": 0.9139448104298152,
+      "grad_norm": 0.7541628093735648,
+      "learning_rate": 3.859220810024467e-06,
+      "loss": 1.2936,
+      "step": 19839
+    },
+    {
+      "epoch": 0.9139908785184503,
+      "grad_norm": 1.082050454961285,
+      "learning_rate": 3.8551167363090725e-06,
+      "loss": 1.2204,
+      "step": 19840
+    },
+    {
+      "epoch": 0.9140369466070852,
+      "grad_norm": 0.8692958645783974,
+      "learning_rate": 3.851014803094666e-06,
+      "loss": 1.2669,
+      "step": 19841
+    },
+    {
+      "epoch": 0.9140830146957203,
+      "grad_norm": 1.02739308664116,
+      "learning_rate": 3.846915010472541e-06,
+      "loss": 1.2287,
+      "step": 19842
+    },
+    {
+      "epoch": 0.9141290827843552,
+      "grad_norm": 1.9153625410911792,
+      "learning_rate": 3.84281735853399e-06,
+      "loss": 1.1812,
+      "step": 19843
+    },
+    {
+      "epoch": 0.9141751508729903,
+      "grad_norm": 1.3921814841672713,
+      "learning_rate": 3.838721847370241e-06,
+      "loss": 1.0098,
+      "step": 19844
+    },
+    {
+      "epoch": 0.9142212189616253,
+      "grad_norm": 6.550675295059426,
+      "learning_rate": 3.834628477072444e-06,
+      "loss": 1.3059,
+      "step": 19845
+    },
+    {
+      "epoch": 0.9142672870502603,
+      "grad_norm": 3.6406278438063437,
+      "learning_rate": 3.83053724773178e-06,
+      "loss": 1.0917,
+      "step": 19846
+    },
+    {
+      "epoch": 0.9143133551388953,
+      "grad_norm": 1.2660559030950107,
+      "learning_rate": 3.826448159439311e-06,
+      "loss": 1.0653,
+      "step": 19847
+    },
+    {
+      "epoch": 0.9143594232275303,
+      "grad_norm": 1.089185665686323,
+      "learning_rate": 3.822361212286063e-06,
+      "loss": 1.4893,
+      "step": 19848
+    },
+    {
+      "epoch": 0.9144054913161653,
+      "grad_norm": 3.3607012105562597,
+      "learning_rate": 3.818276406363019e-06,
+      "loss": 1.0893,
+      "step": 19849
+    },
+    {
+      "epoch": 0.9144515594048003,
+      "grad_norm": 0.6435994573138478,
+      "learning_rate": 3.8141937417611628e-06,
+      "loss": 1.4538,
+      "step": 19850
+    },
+    {
+      "epoch": 0.9144976274934353,
+      "grad_norm": 1.3129463695475507,
+      "learning_rate": 3.810113218571354e-06,
+      "loss": 1.1978,
+      "step": 19851
+    },
+    {
+      "epoch": 0.9145436955820703,
+      "grad_norm": 1.0638633183080406,
+      "learning_rate": 3.8060348368844314e-06,
+      "loss": 1.2325,
+      "step": 19852
+    },
+    {
+      "epoch": 0.9145897636707053,
+      "grad_norm": 0.8751139254889811,
+      "learning_rate": 3.801958596791211e-06,
+      "loss": 1.1599,
+      "step": 19853
+    },
+    {
+      "epoch": 0.9146358317593403,
+      "grad_norm": 0.7911862174576496,
+      "learning_rate": 3.7978844983824426e-06,
+      "loss": 1.4698,
+      "step": 19854
+    },
+    {
+      "epoch": 0.9146818998479753,
+      "grad_norm": 0.7608242776235148,
+      "learning_rate": 3.79381254174882e-06,
+      "loss": 1.2842,
+      "step": 19855
+    },
+    {
+      "epoch": 0.9147279679366103,
+      "grad_norm": 1.1493346375431692,
+      "learning_rate": 3.789742726981005e-06,
+      "loss": 1.2918,
+      "step": 19856
+    },
+    {
+      "epoch": 0.9147740360252453,
+      "grad_norm": 1.0277561305569072,
+      "learning_rate": 3.7856750541696018e-06,
+      "loss": 1.2352,
+      "step": 19857
+    },
+    {
+      "epoch": 0.9148201041138803,
+      "grad_norm": 0.885243617629935,
+      "learning_rate": 3.7816095234051607e-06,
+      "loss": 1.3602,
+      "step": 19858
+    },
+    {
+      "epoch": 0.9148661722025153,
+      "grad_norm": 1.3168100291006264,
+      "learning_rate": 3.7775461347782315e-06,
+      "loss": 0.9858,
+      "step": 19859
+    },
+    {
+      "epoch": 0.9149122402911504,
+      "grad_norm": 1.4106326651055685,
+      "learning_rate": 3.773484888379231e-06,
+      "loss": 1.0049,
+      "step": 19860
+    },
+    {
+      "epoch": 0.9149583083797853,
+      "grad_norm": 1.2336892303471751,
+      "learning_rate": 3.769425784298597e-06,
+      "loss": 1.1407,
+      "step": 19861
+    },
+    {
+      "epoch": 0.9150043764684204,
+      "grad_norm": 1.550232986453848,
+      "learning_rate": 3.7653688226267025e-06,
+      "loss": 1.1782,
+      "step": 19862
+    },
+    {
+      "epoch": 0.9150504445570553,
+      "grad_norm": 1.0328082841701995,
+      "learning_rate": 3.761314003453864e-06,
+      "loss": 1.3233,
+      "step": 19863
+    },
+    {
+      "epoch": 0.9150965126456904,
+      "grad_norm": 4.723382185896302,
+      "learning_rate": 3.7572613268703426e-06,
+      "loss": 1.1907,
+      "step": 19864
+    },
+    {
+      "epoch": 0.9151425807343253,
+      "grad_norm": 0.7905198841483012,
+      "learning_rate": 3.7532107929663885e-06,
+      "loss": 1.4135,
+      "step": 19865
+    },
+    {
+      "epoch": 0.9151886488229604,
+      "grad_norm": 1.3788553757114457,
+      "learning_rate": 3.7491624018321738e-06,
+      "loss": 0.8227,
+      "step": 19866
+    },
+    {
+      "epoch": 0.9152347169115953,
+      "grad_norm": 1.5331640058509626,
+      "learning_rate": 3.745116153557815e-06,
+      "loss": 1.2118,
+      "step": 19867
+    },
+    {
+      "epoch": 0.9152807850002304,
+      "grad_norm": 0.6300594132522352,
+      "learning_rate": 3.741072048233407e-06,
+      "loss": 1.5856,
+      "step": 19868
+    },
+    {
+      "epoch": 0.9153268530888653,
+      "grad_norm": 0.912213083146276,
+      "learning_rate": 3.737030085948989e-06,
+      "loss": 0.9646,
+      "step": 19869
+    },
+    {
+      "epoch": 0.9153729211775004,
+      "grad_norm": 0.9345013618580609,
+      "learning_rate": 3.7329902667945226e-06,
+      "loss": 1.1853,
+      "step": 19870
+    },
+    {
+      "epoch": 0.9154189892661353,
+      "grad_norm": 1.1832025697911768,
+      "learning_rate": 3.72895259085998e-06,
+      "loss": 1.126,
+      "step": 19871
+    },
+    {
+      "epoch": 0.9154650573547704,
+      "grad_norm": 1.030543740090033,
+      "learning_rate": 3.7249170582352223e-06,
+      "loss": 0.9565,
+      "step": 19872
+    },
+    {
+      "epoch": 0.9155111254434054,
+      "grad_norm": 0.7618789131785638,
+      "learning_rate": 3.7208836690101332e-06,
+      "loss": 1.2876,
+      "step": 19873
+    },
+    {
+      "epoch": 0.9155571935320403,
+      "grad_norm": 1.0694180879088444,
+      "learning_rate": 3.716852423274464e-06,
+      "loss": 1.1608,
+      "step": 19874
+    },
+    {
+      "epoch": 0.9156032616206754,
+      "grad_norm": 0.948234508014023,
+      "learning_rate": 3.7128233211179977e-06,
+      "loss": 1.2948,
+      "step": 19875
+    },
+    {
+      "epoch": 0.9156493297093103,
+      "grad_norm": 2.409050367212994,
+      "learning_rate": 3.708796362630418e-06,
+      "loss": 1.1856,
+      "step": 19876
+    },
+    {
+      "epoch": 0.9156953977979454,
+      "grad_norm": 0.8459059679321179,
+      "learning_rate": 3.7047715479013865e-06,
+      "loss": 1.0395,
+      "step": 19877
+    },
+    {
+      "epoch": 0.9157414658865803,
+      "grad_norm": 0.8284520457486426,
+      "learning_rate": 3.700748877020499e-06,
+      "loss": 1.384,
+      "step": 19878
+    },
+    {
+      "epoch": 0.9157875339752154,
+      "grad_norm": 1.0153148897685977,
+      "learning_rate": 3.6967283500773276e-06,
+      "loss": 1.3751,
+      "step": 19879
+    },
+    {
+      "epoch": 0.9158336020638503,
+      "grad_norm": 0.9353131109979336,
+      "learning_rate": 3.6927099671613785e-06,
+      "loss": 1.2068,
+      "step": 19880
+    },
+    {
+      "epoch": 0.9158796701524854,
+      "grad_norm": 1.08508182886095,
+      "learning_rate": 3.6886937283621024e-06,
+      "loss": 1.0396,
+      "step": 19881
+    },
+    {
+      "epoch": 0.9159257382411203,
+      "grad_norm": 0.781648718664658,
+      "learning_rate": 3.6846796337689172e-06,
+      "loss": 1.5194,
+      "step": 19882
+    },
+    {
+      "epoch": 0.9159718063297554,
+      "grad_norm": 1.1400099713383751,
+      "learning_rate": 3.680667683471206e-06,
+      "loss": 1.2791,
+      "step": 19883
+    },
+    {
+      "epoch": 0.9160178744183903,
+      "grad_norm": 2.0774989210746733,
+      "learning_rate": 3.676657877558276e-06,
+      "loss": 1.3318,
+      "step": 19884
+    },
+    {
+      "epoch": 0.9160639425070254,
+      "grad_norm": 1.6018424953757302,
+      "learning_rate": 3.6726502161193887e-06,
+      "loss": 1.251,
+      "step": 19885
+    },
+    {
+      "epoch": 0.9161100105956604,
+      "grad_norm": 0.868318711238014,
+      "learning_rate": 3.668644699243784e-06,
+      "loss": 1.2634,
+      "step": 19886
+    },
+    {
+      "epoch": 0.9161560786842954,
+      "grad_norm": 0.7656652589199243,
+      "learning_rate": 3.6646413270206457e-06,
+      "loss": 1.2528,
+      "step": 19887
+    },
+    {
+      "epoch": 0.9162021467729304,
+      "grad_norm": 0.8082543911570617,
+      "learning_rate": 3.66064009953907e-06,
+      "loss": 1.2724,
+      "step": 19888
+    },
+    {
+      "epoch": 0.9162482148615654,
+      "grad_norm": 1.1524357265357539,
+      "learning_rate": 3.6566410168881627e-06,
+      "loss": 1.1763,
+      "step": 19889
+    },
+    {
+      "epoch": 0.9162942829502004,
+      "grad_norm": 1.3037219442897021,
+      "learning_rate": 3.6526440791569415e-06,
+      "loss": 1.1108,
+      "step": 19890
+    },
+    {
+      "epoch": 0.9163403510388354,
+      "grad_norm": 1.8569305393675366,
+      "learning_rate": 3.6486492864344135e-06,
+      "loss": 1.1837,
+      "step": 19891
+    },
+    {
+      "epoch": 0.9163864191274704,
+      "grad_norm": 0.8313400318876536,
+      "learning_rate": 3.644656638809496e-06,
+      "loss": 1.3057,
+      "step": 19892
+    },
+    {
+      "epoch": 0.9164324872161054,
+      "grad_norm": 1.4322164085586946,
+      "learning_rate": 3.640666136371074e-06,
+      "loss": 0.9993,
+      "step": 19893
+    },
+    {
+      "epoch": 0.9164785553047404,
+      "grad_norm": 0.7303980577787762,
+      "learning_rate": 3.6366777792080087e-06,
+      "loss": 1.3359,
+      "step": 19894
+    },
+    {
+      "epoch": 0.9165246233933754,
+      "grad_norm": 1.1240209216448303,
+      "learning_rate": 3.6326915674090966e-06,
+      "loss": 1.401,
+      "step": 19895
+    },
+    {
+      "epoch": 0.9165706914820104,
+      "grad_norm": 1.9431067200310652,
+      "learning_rate": 3.6287075010630558e-06,
+      "loss": 1.142,
+      "step": 19896
+    },
+    {
+      "epoch": 0.9166167595706454,
+      "grad_norm": 1.2093424238903943,
+      "learning_rate": 3.6247255802586032e-06,
+      "loss": 1.2493,
+      "step": 19897
+    },
+    {
+      "epoch": 0.9166628276592804,
+      "grad_norm": 2.285676692639417,
+      "learning_rate": 3.620745805084402e-06,
+      "loss": 1.3426,
+      "step": 19898
+    },
+    {
+      "epoch": 0.9167088957479154,
+      "grad_norm": 0.9409483125668776,
+      "learning_rate": 3.6167681756290263e-06,
+      "loss": 1.1695,
+      "step": 19899
+    },
+    {
+      "epoch": 0.9167549638365504,
+      "grad_norm": 1.2430729537153944,
+      "learning_rate": 3.612792691981037e-06,
+      "loss": 1.3702,
+      "step": 19900
+    },
+    {
+      "epoch": 0.9168010319251855,
+      "grad_norm": 0.9364459131239243,
+      "learning_rate": 3.608819354228965e-06,
+      "loss": 1.0785,
+      "step": 19901
+    },
+    {
+      "epoch": 0.9168471000138204,
+      "grad_norm": 0.922598698099173,
+      "learning_rate": 3.6048481624612605e-06,
+      "loss": 1.1707,
+      "step": 19902
+    },
+    {
+      "epoch": 0.9168931681024555,
+      "grad_norm": 1.97403408505766,
+      "learning_rate": 3.600879116766309e-06,
+      "loss": 0.8951,
+      "step": 19903
+    },
+    {
+      "epoch": 0.9169392361910904,
+      "grad_norm": 0.7175036761664323,
+      "learning_rate": 3.5969122172325063e-06,
+      "loss": 1.174,
+      "step": 19904
+    },
+    {
+      "epoch": 0.9169853042797255,
+      "grad_norm": 1.2639727158507805,
+      "learning_rate": 3.5929474639481596e-06,
+      "loss": 1.4795,
+      "step": 19905
+    },
+    {
+      "epoch": 0.9170313723683604,
+      "grad_norm": 2.4406709679918497,
+      "learning_rate": 3.5889848570015317e-06,
+      "loss": 1.1158,
+      "step": 19906
+    },
+    {
+      "epoch": 0.9170774404569955,
+      "grad_norm": 1.2628682901887724,
+      "learning_rate": 3.5850243964808407e-06,
+      "loss": 1.1973,
+      "step": 19907
+    },
+    {
+      "epoch": 0.9171235085456304,
+      "grad_norm": 1.1584684479188987,
+      "learning_rate": 3.5810660824742714e-06,
+      "loss": 1.045,
+      "step": 19908
+    },
+    {
+      "epoch": 0.9171695766342655,
+      "grad_norm": 0.782644459123841,
+      "learning_rate": 3.577109915069954e-06,
+      "loss": 1.3904,
+      "step": 19909
+    },
+    {
+      "epoch": 0.9172156447229004,
+      "grad_norm": 1.0814115234052184,
+      "learning_rate": 3.5731558943559396e-06,
+      "loss": 0.9346,
+      "step": 19910
+    },
+    {
+      "epoch": 0.9172617128115355,
+      "grad_norm": 0.9585812037976197,
+      "learning_rate": 3.569204020420269e-06,
+      "loss": 0.9379,
+      "step": 19911
+    },
+    {
+      "epoch": 0.9173077809001704,
+      "grad_norm": 0.7551716019018937,
+      "learning_rate": 3.5652542933509394e-06,
+      "loss": 1.156,
+      "step": 19912
+    },
+    {
+      "epoch": 0.9173538489888055,
+      "grad_norm": 0.9872852427041423,
+      "learning_rate": 3.561306713235868e-06,
+      "loss": 1.3323,
+      "step": 19913
+    },
+    {
+      "epoch": 0.9173999170774405,
+      "grad_norm": 0.9082875846378188,
+      "learning_rate": 3.557361280162963e-06,
+      "loss": 1.0795,
+      "step": 19914
+    },
+    {
+      "epoch": 0.9174459851660755,
+      "grad_norm": 7.1844147589946346,
+      "learning_rate": 3.5534179942200096e-06,
+      "loss": 0.9265,
+      "step": 19915
+    },
+    {
+      "epoch": 0.9174920532547105,
+      "grad_norm": 0.741457727204681,
+      "learning_rate": 3.54947685549486e-06,
+      "loss": 1.3623,
+      "step": 19916
+    },
+    {
+      "epoch": 0.9175381213433454,
+      "grad_norm": 0.8750479096032687,
+      "learning_rate": 3.545537864075221e-06,
+      "loss": 1.5688,
+      "step": 19917
+    },
+    {
+      "epoch": 0.9175841894319805,
+      "grad_norm": 1.0910468003756786,
+      "learning_rate": 3.5416010200488014e-06,
+      "loss": 0.8453,
+      "step": 19918
+    },
+    {
+      "epoch": 0.9176302575206154,
+      "grad_norm": 1.1130728434899182,
+      "learning_rate": 3.5376663235032303e-06,
+      "loss": 1.0288,
+      "step": 19919
+    },
+    {
+      "epoch": 0.9176763256092505,
+      "grad_norm": 0.8296096304662918,
+      "learning_rate": 3.533733774526149e-06,
+      "loss": 1.1701,
+      "step": 19920
+    },
+    {
+      "epoch": 0.9177223936978854,
+      "grad_norm": 1.050814062224411,
+      "learning_rate": 3.529803373205065e-06,
+      "loss": 1.1369,
+      "step": 19921
+    },
+    {
+      "epoch": 0.9177684617865205,
+      "grad_norm": 0.9294683117564979,
+      "learning_rate": 3.5258751196274973e-06,
+      "loss": 1.3494,
+      "step": 19922
+    },
+    {
+      "epoch": 0.9178145298751554,
+      "grad_norm": 2.6135486645846484,
+      "learning_rate": 3.5219490138809095e-06,
+      "loss": 1.1069,
+      "step": 19923
+    },
+    {
+      "epoch": 0.9178605979637905,
+      "grad_norm": 0.6742843878618939,
+      "learning_rate": 3.5180250560526874e-06,
+      "loss": 1.4403,
+      "step": 19924
+    },
+    {
+      "epoch": 0.9179066660524254,
+      "grad_norm": 0.7292994433033204,
+      "learning_rate": 3.514103246230227e-06,
+      "loss": 1.2235,
+      "step": 19925
+    },
+    {
+      "epoch": 0.9179527341410605,
+      "grad_norm": 0.9841473365181073,
+      "learning_rate": 3.5101835845007926e-06,
+      "loss": 1.2664,
+      "step": 19926
+    },
+    {
+      "epoch": 0.9179988022296955,
+      "grad_norm": 1.2315688769911692,
+      "learning_rate": 3.5062660709517027e-06,
+      "loss": 1.2555,
+      "step": 19927
+    },
+    {
+      "epoch": 0.9180448703183305,
+      "grad_norm": 1.623839181146193,
+      "learning_rate": 3.502350705670132e-06,
+      "loss": 0.9507,
+      "step": 19928
+    },
+    {
+      "epoch": 0.9180909384069655,
+      "grad_norm": 0.697456899828365,
+      "learning_rate": 3.498437488743289e-06,
+      "loss": 1.073,
+      "step": 19929
+    },
+    {
+      "epoch": 0.9181370064956005,
+      "grad_norm": 2.0026787050188743,
+      "learning_rate": 3.4945264202582373e-06,
+      "loss": 1.2395,
+      "step": 19930
+    },
+    {
+      "epoch": 0.9181830745842355,
+      "grad_norm": 0.7483033530587245,
+      "learning_rate": 3.490617500302107e-06,
+      "loss": 1.5134,
+      "step": 19931
+    },
+    {
+      "epoch": 0.9182291426728705,
+      "grad_norm": 1.8183960813024598,
+      "learning_rate": 3.486710728961895e-06,
+      "loss": 1.2688,
+      "step": 19932
+    },
+    {
+      "epoch": 0.9182752107615055,
+      "grad_norm": 1.1545952855133665,
+      "learning_rate": 3.4828061063245655e-06,
+      "loss": 1.3233,
+      "step": 19933
+    },
+    {
+      "epoch": 0.9183212788501405,
+      "grad_norm": 0.9801637150496859,
+      "learning_rate": 3.478903632477093e-06,
+      "loss": 1.0943,
+      "step": 19934
+    },
+    {
+      "epoch": 0.9183673469387755,
+      "grad_norm": 1.4291865815430533,
+      "learning_rate": 3.475003307506308e-06,
+      "loss": 1.3937,
+      "step": 19935
+    },
+    {
+      "epoch": 0.9184134150274105,
+      "grad_norm": 1.9782495369486954,
+      "learning_rate": 3.4711051314990974e-06,
+      "loss": 1.1542,
+      "step": 19936
+    },
+    {
+      "epoch": 0.9184594831160455,
+      "grad_norm": 0.8223804683626152,
+      "learning_rate": 3.4672091045421905e-06,
+      "loss": 1.1426,
+      "step": 19937
+    },
+    {
+      "epoch": 0.9185055512046805,
+      "grad_norm": 1.0199759901582475,
+      "learning_rate": 3.4633152267223746e-06,
+      "loss": 1.1934,
+      "step": 19938
+    },
+    {
+      "epoch": 0.9185516192933155,
+      "grad_norm": 1.3026163772778754,
+      "learning_rate": 3.459423498126313e-06,
+      "loss": 1.3358,
+      "step": 19939
+    },
+    {
+      "epoch": 0.9185976873819506,
+      "grad_norm": 0.898595251952859,
+      "learning_rate": 3.455533918840659e-06,
+      "loss": 1.5726,
+      "step": 19940
+    },
+    {
+      "epoch": 0.9186437554705855,
+      "grad_norm": 0.7266491729618401,
+      "learning_rate": 3.4516464889519983e-06,
+      "loss": 1.3191,
+      "step": 19941
+    },
+    {
+      "epoch": 0.9186898235592206,
+      "grad_norm": 1.1024412516888278,
+      "learning_rate": 3.447761208546885e-06,
+      "loss": 1.0932,
+      "step": 19942
+    },
+    {
+      "epoch": 0.9187358916478555,
+      "grad_norm": 1.6987554593744296,
+      "learning_rate": 3.4438780777118263e-06,
+      "loss": 1.2586,
+      "step": 19943
+    },
+    {
+      "epoch": 0.9187819597364906,
+      "grad_norm": 0.7746221794079562,
+      "learning_rate": 3.439997096533243e-06,
+      "loss": 1.2238,
+      "step": 19944
+    },
+    {
+      "epoch": 0.9188280278251255,
+      "grad_norm": 1.1452000698873965,
+      "learning_rate": 3.4361182650975876e-06,
+      "loss": 1.1634,
+      "step": 19945
+    },
+    {
+      "epoch": 0.9188740959137606,
+      "grad_norm": 1.1758439211793386,
+      "learning_rate": 3.432241583491169e-06,
+      "loss": 1.3455,
+      "step": 19946
+    },
+    {
+      "epoch": 0.9189201640023955,
+      "grad_norm": 1.9523495519200582,
+      "learning_rate": 3.4283670518003187e-06,
+      "loss": 1.2023,
+      "step": 19947
+    },
+    {
+      "epoch": 0.9189662320910306,
+      "grad_norm": 0.8632058567360433,
+      "learning_rate": 3.424494670111278e-06,
+      "loss": 1.3451,
+      "step": 19948
+    },
+    {
+      "epoch": 0.9190123001796655,
+      "grad_norm": 1.3433218060088439,
+      "learning_rate": 3.4206244385102893e-06,
+      "loss": 1.0765,
+      "step": 19949
+    },
+    {
+      "epoch": 0.9190583682683006,
+      "grad_norm": 0.8925326073415308,
+      "learning_rate": 3.416756357083495e-06,
+      "loss": 1.1679,
+      "step": 19950
+    },
+    {
+      "epoch": 0.9191044363569355,
+      "grad_norm": 0.7798752283513025,
+      "learning_rate": 3.4128904259170257e-06,
+      "loss": 1.2562,
+      "step": 19951
+    },
+    {
+      "epoch": 0.9191505044455706,
+      "grad_norm": 1.0504244526271138,
+      "learning_rate": 3.4090266450969246e-06,
+      "loss": 1.3095,
+      "step": 19952
+    },
+    {
+      "epoch": 0.9191965725342055,
+      "grad_norm": 0.7982054460895823,
+      "learning_rate": 3.405165014709222e-06,
+      "loss": 1.0992,
+      "step": 19953
+    },
+    {
+      "epoch": 0.9192426406228406,
+      "grad_norm": 0.8476850843740792,
+      "learning_rate": 3.401305534839916e-06,
+      "loss": 1.3316,
+      "step": 19954
+    },
+    {
+      "epoch": 0.9192887087114756,
+      "grad_norm": 0.9406491310995618,
+      "learning_rate": 3.3974482055748937e-06,
+      "loss": 1.1318,
+      "step": 19955
+    },
+    {
+      "epoch": 0.9193347768001106,
+      "grad_norm": 0.9843082940822416,
+      "learning_rate": 3.3935930270000527e-06,
+      "loss": 1.1378,
+      "step": 19956
+    },
+    {
+      "epoch": 0.9193808448887456,
+      "grad_norm": 1.0386356356404303,
+      "learning_rate": 3.389739999201225e-06,
+      "loss": 1.1901,
+      "step": 19957
+    },
+    {
+      "epoch": 0.9194269129773806,
+      "grad_norm": 1.455794979025236,
+      "learning_rate": 3.385889122264185e-06,
+      "loss": 0.8662,
+      "step": 19958
+    },
+    {
+      "epoch": 0.9194729810660156,
+      "grad_norm": 1.423538441567622,
+      "learning_rate": 3.382040396274644e-06,
+      "loss": 1.3536,
+      "step": 19959
+    },
+    {
+      "epoch": 0.9195190491546505,
+      "grad_norm": 0.897008483370576,
+      "learning_rate": 3.378193821318343e-06,
+      "loss": 1.1091,
+      "step": 19960
+    },
+    {
+      "epoch": 0.9195651172432856,
+      "grad_norm": 0.8282050816170792,
+      "learning_rate": 3.3743493974808695e-06,
+      "loss": 1.1935,
+      "step": 19961
+    },
+    {
+      "epoch": 0.9196111853319205,
+      "grad_norm": 0.9594001133730264,
+      "learning_rate": 3.3705071248478327e-06,
+      "loss": 1.1123,
+      "step": 19962
+    },
+    {
+      "epoch": 0.9196572534205556,
+      "grad_norm": 1.4669519311737298,
+      "learning_rate": 3.3666670035047644e-06,
+      "loss": 1.1138,
+      "step": 19963
+    },
+    {
+      "epoch": 0.9197033215091905,
+      "grad_norm": 1.0206648266274312,
+      "learning_rate": 3.3628290335371624e-06,
+      "loss": 1.0597,
+      "step": 19964
+    },
+    {
+      "epoch": 0.9197493895978256,
+      "grad_norm": 1.6986923087986774,
+      "learning_rate": 3.3589932150304928e-06,
+      "loss": 1.5721,
+      "step": 19965
+    },
+    {
+      "epoch": 0.9197954576864605,
+      "grad_norm": 1.5628021938335446,
+      "learning_rate": 3.3551595480701194e-06,
+      "loss": 1.4676,
+      "step": 19966
+    },
+    {
+      "epoch": 0.9198415257750956,
+      "grad_norm": 1.6183635224232054,
+      "learning_rate": 3.3513280327414076e-06,
+      "loss": 0.9061,
+      "step": 19967
+    },
+    {
+      "epoch": 0.9198875938637306,
+      "grad_norm": 0.9346556042231412,
+      "learning_rate": 3.347498669129656e-06,
+      "loss": 1.4085,
+      "step": 19968
+    },
+    {
+      "epoch": 0.9199336619523656,
+      "grad_norm": 0.731379594445811,
+      "learning_rate": 3.343671457320141e-06,
+      "loss": 1.2339,
+      "step": 19969
+    },
+    {
+      "epoch": 0.9199797300410006,
+      "grad_norm": 0.8815593297941328,
+      "learning_rate": 3.339846397398039e-06,
+      "loss": 1.2808,
+      "step": 19970
+    },
+    {
+      "epoch": 0.9200257981296356,
+      "grad_norm": 0.8687886594478349,
+      "learning_rate": 3.3360234894485254e-06,
+      "loss": 0.9263,
+      "step": 19971
+    },
+    {
+      "epoch": 0.9200718662182706,
+      "grad_norm": 0.7827464470068947,
+      "learning_rate": 3.332202733556711e-06,
+      "loss": 1.2893,
+      "step": 19972
+    },
+    {
+      "epoch": 0.9201179343069056,
+      "grad_norm": 1.323882437457986,
+      "learning_rate": 3.328384129807638e-06,
+      "loss": 1.3218,
+      "step": 19973
+    },
+    {
+      "epoch": 0.9201640023955406,
+      "grad_norm": 1.6186059612831687,
+      "learning_rate": 3.324567678286339e-06,
+      "loss": 1.6152,
+      "step": 19974
+    },
+    {
+      "epoch": 0.9202100704841756,
+      "grad_norm": 0.9750498743724862,
+      "learning_rate": 3.320753379077779e-06,
+      "loss": 1.2718,
+      "step": 19975
+    },
+    {
+      "epoch": 0.9202561385728106,
+      "grad_norm": 0.948655249624706,
+      "learning_rate": 3.3169412322668904e-06,
+      "loss": 1.1021,
+      "step": 19976
+    },
+    {
+      "epoch": 0.9203022066614456,
+      "grad_norm": 1.6603518663131864,
+      "learning_rate": 3.313131237938516e-06,
+      "loss": 1.374,
+      "step": 19977
+    },
+    {
+      "epoch": 0.9203482747500806,
+      "grad_norm": 1.0473963057505007,
+      "learning_rate": 3.3093233961774885e-06,
+      "loss": 1.264,
+      "step": 19978
+    },
+    {
+      "epoch": 0.9203943428387156,
+      "grad_norm": 1.3432406977086364,
+      "learning_rate": 3.305517707068606e-06,
+      "loss": 1.2283,
+      "step": 19979
+    },
+    {
+      "epoch": 0.9204404109273506,
+      "grad_norm": 0.8235022193507915,
+      "learning_rate": 3.301714170696557e-06,
+      "loss": 1.0443,
+      "step": 19980
+    },
+    {
+      "epoch": 0.9204864790159857,
+      "grad_norm": 0.7951799360350604,
+      "learning_rate": 3.29791278714604e-06,
+      "loss": 1.308,
+      "step": 19981
+    },
+    {
+      "epoch": 0.9205325471046206,
+      "grad_norm": 0.9280929476777044,
+      "learning_rate": 3.294113556501688e-06,
+      "loss": 1.3387,
+      "step": 19982
+    },
+    {
+      "epoch": 0.9205786151932557,
+      "grad_norm": 0.8707003769420025,
+      "learning_rate": 3.290316478848099e-06,
+      "loss": 0.9207,
+      "step": 19983
+    },
+    {
+      "epoch": 0.9206246832818906,
+      "grad_norm": 7.461544683137335,
+      "learning_rate": 3.2865215542697726e-06,
+      "loss": 1.1856,
+      "step": 19984
+    },
+    {
+      "epoch": 0.9206707513705257,
+      "grad_norm": 0.8780673620270552,
+      "learning_rate": 3.282728782851219e-06,
+      "loss": 1.531,
+      "step": 19985
+    },
+    {
+      "epoch": 0.9207168194591606,
+      "grad_norm": 1.328953258165737,
+      "learning_rate": 3.2789381646768704e-06,
+      "loss": 1.0832,
+      "step": 19986
+    },
+    {
+      "epoch": 0.9207628875477957,
+      "grad_norm": 1.2346762921862762,
+      "learning_rate": 3.275149699831137e-06,
+      "loss": 0.9468,
+      "step": 19987
+    },
+    {
+      "epoch": 0.9208089556364306,
+      "grad_norm": 0.831317788232992,
+      "learning_rate": 3.271363388398341e-06,
+      "loss": 1.463,
+      "step": 19988
+    },
+    {
+      "epoch": 0.9208550237250657,
+      "grad_norm": 1.2884426022813216,
+      "learning_rate": 3.267579230462781e-06,
+      "loss": 1.0185,
+      "step": 19989
+    },
+    {
+      "epoch": 0.9209010918137006,
+      "grad_norm": 1.02760943889592,
+      "learning_rate": 3.2637972261087225e-06,
+      "loss": 0.9316,
+      "step": 19990
+    },
+    {
+      "epoch": 0.9209471599023357,
+      "grad_norm": 1.6515751506774854,
+      "learning_rate": 3.2600173754203432e-06,
+      "loss": 1.1467,
+      "step": 19991
+    },
+    {
+      "epoch": 0.9209932279909706,
+      "grad_norm": 0.7411621961456315,
+      "learning_rate": 3.2562396784817983e-06,
+      "loss": 1.0282,
+      "step": 19992
+    },
+    {
+      "epoch": 0.9210392960796057,
+      "grad_norm": 1.3953438183147293,
+      "learning_rate": 3.252464135377198e-06,
+      "loss": 1.2837,
+      "step": 19993
+    },
+    {
+      "epoch": 0.9210853641682406,
+      "grad_norm": 0.8628014989532476,
+      "learning_rate": 3.248690746190608e-06,
+      "loss": 1.5235,
+      "step": 19994
+    },
+    {
+      "epoch": 0.9211314322568757,
+      "grad_norm": 0.8696629283754932,
+      "learning_rate": 3.244919511006017e-06,
+      "loss": 0.9599,
+      "step": 19995
+    },
+    {
+      "epoch": 0.9211775003455107,
+      "grad_norm": 0.8796022267507912,
+      "learning_rate": 3.2411504299073913e-06,
+      "loss": 1.3409,
+      "step": 19996
+    },
+    {
+      "epoch": 0.9212235684341457,
+      "grad_norm": 1.00462838483939,
+      "learning_rate": 3.2373835029786528e-06,
+      "loss": 1.3282,
+      "step": 19997
+    },
+    {
+      "epoch": 0.9212696365227807,
+      "grad_norm": 0.8796682190112699,
+      "learning_rate": 3.2336187303036556e-06,
+      "loss": 1.1532,
+      "step": 19998
+    },
+    {
+      "epoch": 0.9213157046114157,
+      "grad_norm": 0.9778957232208438,
+      "learning_rate": 3.2298561119662117e-06,
+      "loss": 1.1155,
+      "step": 19999
+    },
+    {
+      "epoch": 0.9213617727000507,
+      "grad_norm": 1.227464909689238,
+      "learning_rate": 3.226095648050098e-06,
+      "loss": 1.2004,
+      "step": 20000
+    },
+    {
+      "epoch": 0.9214078407886857,
+      "grad_norm": 0.8778890842535816,
+      "learning_rate": 3.2223373386390476e-06,
+      "loss": 1.4061,
+      "step": 20001
+    },
+    {
+      "epoch": 0.9214539088773207,
+      "grad_norm": 1.2300169600133828,
+      "learning_rate": 3.2185811838167045e-06,
+      "loss": 1.0668,
+      "step": 20002
+    },
+    {
+      "epoch": 0.9214999769659556,
+      "grad_norm": 0.6209960033503867,
+      "learning_rate": 3.214827183666702e-06,
+      "loss": 1.5517,
+      "step": 20003
+    },
+    {
+      "epoch": 0.9215460450545907,
+      "grad_norm": 1.0637128455236564,
+      "learning_rate": 3.2110753382726287e-06,
+      "loss": 1.2501,
+      "step": 20004
+    },
+    {
+      "epoch": 0.9215921131432256,
+      "grad_norm": 1.3332163985802212,
+      "learning_rate": 3.207325647718018e-06,
+      "loss": 1.3189,
+      "step": 20005
+    },
+    {
+      "epoch": 0.9216381812318607,
+      "grad_norm": 0.7777878991025631,
+      "learning_rate": 3.203578112086314e-06,
+      "loss": 1.3938,
+      "step": 20006
+    },
+    {
+      "epoch": 0.9216842493204956,
+      "grad_norm": 1.2942790491042797,
+      "learning_rate": 3.1998327314609834e-06,
+      "loss": 1.3725,
+      "step": 20007
+    },
+    {
+      "epoch": 0.9217303174091307,
+      "grad_norm": 1.7768296821501104,
+      "learning_rate": 3.196089505925415e-06,
+      "loss": 1.1841,
+      "step": 20008
+    },
+    {
+      "epoch": 0.9217763854977657,
+      "grad_norm": 1.2888854666735658,
+      "learning_rate": 3.1923484355629084e-06,
+      "loss": 1.1348,
+      "step": 20009
+    },
+    {
+      "epoch": 0.9218224535864007,
+      "grad_norm": 1.6177384558759398,
+      "learning_rate": 3.1886095204567867e-06,
+      "loss": 0.914,
+      "step": 20010
+    },
+    {
+      "epoch": 0.9218685216750357,
+      "grad_norm": 0.9815907951268409,
+      "learning_rate": 3.1848727606902718e-06,
+      "loss": 1.2948,
+      "step": 20011
+    },
+    {
+      "epoch": 0.9219145897636707,
+      "grad_norm": 1.5820632891893998,
+      "learning_rate": 3.181138156346575e-06,
+      "loss": 1.1241,
+      "step": 20012
+    },
+    {
+      "epoch": 0.9219606578523057,
+      "grad_norm": 1.2752295742251503,
+      "learning_rate": 3.177405707508818e-06,
+      "loss": 1.0681,
+      "step": 20013
+    },
+    {
+      "epoch": 0.9220067259409407,
+      "grad_norm": 0.9029406307848568,
+      "learning_rate": 3.173675414260102e-06,
+      "loss": 1.1444,
+      "step": 20014
+    },
+    {
+      "epoch": 0.9220527940295757,
+      "grad_norm": 1.1965996436430022,
+      "learning_rate": 3.1699472766834936e-06,
+      "loss": 1.3701,
+      "step": 20015
+    },
+    {
+      "epoch": 0.9220988621182107,
+      "grad_norm": 0.8076744089703786,
+      "learning_rate": 3.1662212948619706e-06,
+      "loss": 1.0073,
+      "step": 20016
+    },
+    {
+      "epoch": 0.9221449302068457,
+      "grad_norm": 0.8381627944714313,
+      "learning_rate": 3.1624974688785e-06,
+      "loss": 1.2808,
+      "step": 20017
+    },
+    {
+      "epoch": 0.9221909982954807,
+      "grad_norm": 0.7234739077255173,
+      "learning_rate": 3.1587757988159826e-06,
+      "loss": 1.0364,
+      "step": 20018
+    },
+    {
+      "epoch": 0.9222370663841157,
+      "grad_norm": 2.649173399785697,
+      "learning_rate": 3.155056284757285e-06,
+      "loss": 1.1819,
+      "step": 20019
+    },
+    {
+      "epoch": 0.9222831344727507,
+      "grad_norm": 0.9988497933515472,
+      "learning_rate": 3.1513389267851968e-06,
+      "loss": 1.39,
+      "step": 20020
+    },
+    {
+      "epoch": 0.9223292025613857,
+      "grad_norm": 0.7428327019892679,
+      "learning_rate": 3.1476237249824848e-06,
+      "loss": 1.3674,
+      "step": 20021
+    },
+    {
+      "epoch": 0.9223752706500208,
+      "grad_norm": 0.9108153692550499,
+      "learning_rate": 3.1439106794318606e-06,
+      "loss": 1.3771,
+      "step": 20022
+    },
+    {
+      "epoch": 0.9224213387386557,
+      "grad_norm": 1.3169722011261635,
+      "learning_rate": 3.140199790216003e-06,
+      "loss": 1.1412,
+      "step": 20023
+    },
+    {
+      "epoch": 0.9224674068272908,
+      "grad_norm": 1.0399773158037802,
+      "learning_rate": 3.1364910574175233e-06,
+      "loss": 1.0361,
+      "step": 20024
+    },
+    {
+      "epoch": 0.9225134749159257,
+      "grad_norm": 0.9846279412000651,
+      "learning_rate": 3.1327844811189664e-06,
+      "loss": 1.3268,
+      "step": 20025
+    },
+    {
+      "epoch": 0.9225595430045608,
+      "grad_norm": 1.2898936165940298,
+      "learning_rate": 3.1290800614028893e-06,
+      "loss": 1.2523,
+      "step": 20026
+    },
+    {
+      "epoch": 0.9226056110931957,
+      "grad_norm": 2.1882210935606046,
+      "learning_rate": 3.1253777983517363e-06,
+      "loss": 1.3565,
+      "step": 20027
+    },
+    {
+      "epoch": 0.9226516791818308,
+      "grad_norm": 1.4883587857252023,
+      "learning_rate": 3.1216776920479527e-06,
+      "loss": 1.1263,
+      "step": 20028
+    },
+    {
+      "epoch": 0.9226977472704657,
+      "grad_norm": 1.1673139075040941,
+      "learning_rate": 3.1179797425738733e-06,
+      "loss": 1.3445,
+      "step": 20029
+    },
+    {
+      "epoch": 0.9227438153591008,
+      "grad_norm": 0.8379667381759561,
+      "learning_rate": 3.1142839500118983e-06,
+      "loss": 1.2368,
+      "step": 20030
+    },
+    {
+      "epoch": 0.9227898834477357,
+      "grad_norm": 1.570432064762237,
+      "learning_rate": 3.1105903144442393e-06,
+      "loss": 1.0555,
+      "step": 20031
+    },
+    {
+      "epoch": 0.9228359515363708,
+      "grad_norm": 1.4251612677037615,
+      "learning_rate": 3.106898835953165e-06,
+      "loss": 1.1279,
+      "step": 20032
+    },
+    {
+      "epoch": 0.9228820196250057,
+      "grad_norm": 1.0155426815753337,
+      "learning_rate": 3.103209514620864e-06,
+      "loss": 1.2548,
+      "step": 20033
+    },
+    {
+      "epoch": 0.9229280877136408,
+      "grad_norm": 1.4460741503491232,
+      "learning_rate": 3.099522350529449e-06,
+      "loss": 1.1674,
+      "step": 20034
+    },
+    {
+      "epoch": 0.9229741558022757,
+      "grad_norm": 1.0408833816172185,
+      "learning_rate": 3.095837343761032e-06,
+      "loss": 1.445,
+      "step": 20035
+    },
+    {
+      "epoch": 0.9230202238909108,
+      "grad_norm": 1.0448413242554686,
+      "learning_rate": 3.0921544943976143e-06,
+      "loss": 1.3419,
+      "step": 20036
+    },
+    {
+      "epoch": 0.9230662919795458,
+      "grad_norm": 0.7908311470784845,
+      "learning_rate": 3.088473802521241e-06,
+      "loss": 1.0672,
+      "step": 20037
+    },
+    {
+      "epoch": 0.9231123600681808,
+      "grad_norm": 2.459329032580726,
+      "learning_rate": 3.084795268213825e-06,
+      "loss": 1.0262,
+      "step": 20038
+    },
+    {
+      "epoch": 0.9231584281568158,
+      "grad_norm": 0.8736250169243258,
+      "learning_rate": 3.0811188915572774e-06,
+      "loss": 1.1552,
+      "step": 20039
+    },
+    {
+      "epoch": 0.9232044962454508,
+      "grad_norm": 1.2300575349255036,
+      "learning_rate": 3.0774446726334115e-06,
+      "loss": 1.3326,
+      "step": 20040
+    },
+    {
+      "epoch": 0.9232505643340858,
+      "grad_norm": 0.8043628163932955,
+      "learning_rate": 3.073772611524073e-06,
+      "loss": 1.2979,
+      "step": 20041
+    },
+    {
+      "epoch": 0.9232966324227208,
+      "grad_norm": 1.839564392203136,
+      "learning_rate": 3.070102708311007e-06,
+      "loss": 1.2073,
+      "step": 20042
+    },
+    {
+      "epoch": 0.9233427005113558,
+      "grad_norm": 0.6320583550633305,
+      "learning_rate": 3.0664349630758703e-06,
+      "loss": 1.433,
+      "step": 20043
+    },
+    {
+      "epoch": 0.9233887685999908,
+      "grad_norm": 1.378250413285059,
+      "learning_rate": 3.0627693759003762e-06,
+      "loss": 1.2632,
+      "step": 20044
+    },
+    {
+      "epoch": 0.9234348366886258,
+      "grad_norm": 0.969113061681989,
+      "learning_rate": 3.0591059468661033e-06,
+      "loss": 1.1245,
+      "step": 20045
+    },
+    {
+      "epoch": 0.9234809047772607,
+      "grad_norm": 1.4637922725355996,
+      "learning_rate": 3.0554446760546305e-06,
+      "loss": 1.316,
+      "step": 20046
+    },
+    {
+      "epoch": 0.9235269728658958,
+      "grad_norm": 1.0948623781508555,
+      "learning_rate": 3.051785563547438e-06,
+      "loss": 1.5058,
+      "step": 20047
+    },
+    {
+      "epoch": 0.9235730409545307,
+      "grad_norm": 0.9815872846017859,
+      "learning_rate": 3.0481286094260262e-06,
+      "loss": 1.2883,
+      "step": 20048
+    },
+    {
+      "epoch": 0.9236191090431658,
+      "grad_norm": 2.018248682599013,
+      "learning_rate": 3.0444738137717864e-06,
+      "loss": 1.0185,
+      "step": 20049
+    },
+    {
+      "epoch": 0.9236651771318009,
+      "grad_norm": 0.7994911288921134,
+      "learning_rate": 3.0408211766660977e-06,
+      "loss": 1.3446,
+      "step": 20050
+    },
+    {
+      "epoch": 0.9237112452204358,
+      "grad_norm": 0.9563727975072701,
+      "learning_rate": 3.0371706981902725e-06,
+      "loss": 1.3644,
+      "step": 20051
+    },
+    {
+      "epoch": 0.9237573133090708,
+      "grad_norm": 0.9095138996592937,
+      "learning_rate": 3.0335223784255793e-06,
+      "loss": 1.1392,
+      "step": 20052
+    },
+    {
+      "epoch": 0.9238033813977058,
+      "grad_norm": 0.995532812631559,
+      "learning_rate": 3.0298762174532647e-06,
+      "loss": 1.3649,
+      "step": 20053
+    },
+    {
+      "epoch": 0.9238494494863408,
+      "grad_norm": 1.0784799635540947,
+      "learning_rate": 3.0262322153544854e-06,
+      "loss": 1.1814,
+      "step": 20054
+    },
+    {
+      "epoch": 0.9238955175749758,
+      "grad_norm": 1.0214562433522119,
+      "learning_rate": 3.022590372210354e-06,
+      "loss": 1.1706,
+      "step": 20055
+    },
+    {
+      "epoch": 0.9239415856636108,
+      "grad_norm": 0.7857935400177093,
+      "learning_rate": 3.0189506881019845e-06,
+      "loss": 1.2593,
+      "step": 20056
+    },
+    {
+      "epoch": 0.9239876537522458,
+      "grad_norm": 1.3664920740270858,
+      "learning_rate": 3.0153131631103893e-06,
+      "loss": 1.2546,
+      "step": 20057
+    },
+    {
+      "epoch": 0.9240337218408808,
+      "grad_norm": 1.360278660306268,
+      "learning_rate": 3.0116777973165368e-06,
+      "loss": 1.2434,
+      "step": 20058
+    },
+    {
+      "epoch": 0.9240797899295158,
+      "grad_norm": 0.9451803098065084,
+      "learning_rate": 3.0080445908014067e-06,
+      "loss": 1.0465,
+      "step": 20059
+    },
+    {
+      "epoch": 0.9241258580181508,
+      "grad_norm": 0.8532924770994531,
+      "learning_rate": 3.0044135436458453e-06,
+      "loss": 1.1197,
+      "step": 20060
+    },
+    {
+      "epoch": 0.9241719261067858,
+      "grad_norm": 1.4780729957807868,
+      "learning_rate": 3.0007846559306995e-06,
+      "loss": 1.0575,
+      "step": 20061
+    },
+    {
+      "epoch": 0.9242179941954208,
+      "grad_norm": 1.0023630322520285,
+      "learning_rate": 2.997157927736771e-06,
+      "loss": 1.3718,
+      "step": 20062
+    },
+    {
+      "epoch": 0.9242640622840559,
+      "grad_norm": 0.892604290951937,
+      "learning_rate": 2.9935333591447844e-06,
+      "loss": 1.116,
+      "step": 20063
+    },
+    {
+      "epoch": 0.9243101303726908,
+      "grad_norm": 0.7362260823506159,
+      "learning_rate": 2.989910950235464e-06,
+      "loss": 1.3699,
+      "step": 20064
+    },
+    {
+      "epoch": 0.9243561984613259,
+      "grad_norm": 2.5743849492725577,
+      "learning_rate": 2.9862907010894337e-06,
+      "loss": 1.0324,
+      "step": 20065
+    },
+    {
+      "epoch": 0.9244022665499608,
+      "grad_norm": 7.555064879351826,
+      "learning_rate": 2.982672611787296e-06,
+      "loss": 1.3367,
+      "step": 20066
+    },
+    {
+      "epoch": 0.9244483346385959,
+      "grad_norm": 1.4137679078927445,
+      "learning_rate": 2.979056682409598e-06,
+      "loss": 0.88,
+      "step": 20067
+    },
+    {
+      "epoch": 0.9244944027272308,
+      "grad_norm": 0.8625476425737671,
+      "learning_rate": 2.9754429130368633e-06,
+      "loss": 1.2668,
+      "step": 20068
+    },
+    {
+      "epoch": 0.9245404708158659,
+      "grad_norm": 0.8366138685029437,
+      "learning_rate": 2.9718313037495172e-06,
+      "loss": 1.5302,
+      "step": 20069
+    },
+    {
+      "epoch": 0.9245865389045008,
+      "grad_norm": 0.9962616651441734,
+      "learning_rate": 2.968221854627984e-06,
+      "loss": 1.1316,
+      "step": 20070
+    },
+    {
+      "epoch": 0.9246326069931359,
+      "grad_norm": 0.9539969489804718,
+      "learning_rate": 2.9646145657526326e-06,
+      "loss": 1.2227,
+      "step": 20071
+    },
+    {
+      "epoch": 0.9246786750817708,
+      "grad_norm": 1.1366639914289365,
+      "learning_rate": 2.961009437203732e-06,
+      "loss": 0.8695,
+      "step": 20072
+    },
+    {
+      "epoch": 0.9247247431704059,
+      "grad_norm": 0.800071182999124,
+      "learning_rate": 2.9574064690615856e-06,
+      "loss": 1.1423,
+      "step": 20073
+    },
+    {
+      "epoch": 0.9247708112590408,
+      "grad_norm": 1.3827086028387976,
+      "learning_rate": 2.9538056614063836e-06,
+      "loss": 1.2092,
+      "step": 20074
+    },
+    {
+      "epoch": 0.9248168793476759,
+      "grad_norm": 3.093218814249684,
+      "learning_rate": 2.950207014318307e-06,
+      "loss": 1.4901,
+      "step": 20075
+    },
+    {
+      "epoch": 0.9248629474363109,
+      "grad_norm": 1.2355114894327697,
+      "learning_rate": 2.946610527877458e-06,
+      "loss": 1.0527,
+      "step": 20076
+    },
+    {
+      "epoch": 0.9249090155249459,
+      "grad_norm": 0.9187870177499198,
+      "learning_rate": 2.9430162021639175e-06,
+      "loss": 1.054,
+      "step": 20077
+    },
+    {
+      "epoch": 0.9249550836135809,
+      "grad_norm": 0.9049423534591426,
+      "learning_rate": 2.93942403725771e-06,
+      "loss": 1.4861,
+      "step": 20078
+    },
+    {
+      "epoch": 0.9250011517022159,
+      "grad_norm": 1.0050118326391446,
+      "learning_rate": 2.9358340332388044e-06,
+      "loss": 0.9536,
+      "step": 20079
+    },
+    {
+      "epoch": 0.9250472197908509,
+      "grad_norm": 1.4617236591718634,
+      "learning_rate": 2.9322461901871156e-06,
+      "loss": 1.4677,
+      "step": 20080
+    },
+    {
+      "epoch": 0.9250932878794859,
+      "grad_norm": 1.015406248960211,
+      "learning_rate": 2.9286605081825347e-06,
+      "loss": 1.2394,
+      "step": 20081
+    },
+    {
+      "epoch": 0.9251393559681209,
+      "grad_norm": 0.7824097855410922,
+      "learning_rate": 2.9250769873048868e-06,
+      "loss": 1.0617,
+      "step": 20082
+    },
+    {
+      "epoch": 0.9251854240567559,
+      "grad_norm": 1.556487445318175,
+      "learning_rate": 2.9214956276339523e-06,
+      "loss": 1.2456,
+      "step": 20083
+    },
+    {
+      "epoch": 0.9252314921453909,
+      "grad_norm": 0.853593218986566,
+      "learning_rate": 2.9179164292494565e-06,
+      "loss": 1.0397,
+      "step": 20084
+    },
+    {
+      "epoch": 0.9252775602340259,
+      "grad_norm": 1.2807579010749048,
+      "learning_rate": 2.9143393922311023e-06,
+      "loss": 1.0548,
+      "step": 20085
+    },
+    {
+      "epoch": 0.9253236283226609,
+      "grad_norm": 1.301473214634698,
+      "learning_rate": 2.910764516658515e-06,
+      "loss": 1.1523,
+      "step": 20086
+    },
+    {
+      "epoch": 0.9253696964112959,
+      "grad_norm": 1.105997850838418,
+      "learning_rate": 2.9071918026112753e-06,
+      "loss": 1.0936,
+      "step": 20087
+    },
+    {
+      "epoch": 0.9254157644999309,
+      "grad_norm": 1.0323221295686185,
+      "learning_rate": 2.903621250168931e-06,
+      "loss": 1.2514,
+      "step": 20088
+    },
+    {
+      "epoch": 0.9254618325885658,
+      "grad_norm": 2.057934474524191,
+      "learning_rate": 2.9000528594109844e-06,
+      "loss": 1.4357,
+      "step": 20089
+    },
+    {
+      "epoch": 0.9255079006772009,
+      "grad_norm": 2.0878203428891493,
+      "learning_rate": 2.8964866304168615e-06,
+      "loss": 1.2675,
+      "step": 20090
+    },
+    {
+      "epoch": 0.925553968765836,
+      "grad_norm": 1.8368569807532031,
+      "learning_rate": 2.892922563265965e-06,
+      "loss": 1.2765,
+      "step": 20091
+    },
+    {
+      "epoch": 0.9256000368544709,
+      "grad_norm": 1.2292376083442167,
+      "learning_rate": 2.889360658037654e-06,
+      "loss": 1.0854,
+      "step": 20092
+    },
+    {
+      "epoch": 0.925646104943106,
+      "grad_norm": 1.444994769048583,
+      "learning_rate": 2.885800914811221e-06,
+      "loss": 1.1838,
+      "step": 20093
+    },
+    {
+      "epoch": 0.9256921730317409,
+      "grad_norm": 0.7531743293424913,
+      "learning_rate": 2.882243333665902e-06,
+      "loss": 1.2471,
+      "step": 20094
+    },
+    {
+      "epoch": 0.925738241120376,
+      "grad_norm": 0.7922454348967033,
+      "learning_rate": 2.8786879146809222e-06,
+      "loss": 1.2072,
+      "step": 20095
+    },
+    {
+      "epoch": 0.9257843092090109,
+      "grad_norm": 0.785901152066121,
+      "learning_rate": 2.8751346579354188e-06,
+      "loss": 1.1257,
+      "step": 20096
+    },
+    {
+      "epoch": 0.925830377297646,
+      "grad_norm": 1.1694078811378763,
+      "learning_rate": 2.8715835635085285e-06,
+      "loss": 1.3477,
+      "step": 20097
+    },
+    {
+      "epoch": 0.9258764453862809,
+      "grad_norm": 1.1569728749286017,
+      "learning_rate": 2.8680346314792773e-06,
+      "loss": 1.2893,
+      "step": 20098
+    },
+    {
+      "epoch": 0.9259225134749159,
+      "grad_norm": 1.030616134280987,
+      "learning_rate": 2.8644878619266902e-06,
+      "loss": 1.1933,
+      "step": 20099
+    },
+    {
+      "epoch": 0.9259685815635509,
+      "grad_norm": 2.183208939818982,
+      "learning_rate": 2.8609432549297377e-06,
+      "loss": 1.5725,
+      "step": 20100
+    },
+    {
+      "epoch": 0.9260146496521859,
+      "grad_norm": 0.8804955688932363,
+      "learning_rate": 2.8574008105673123e-06,
+      "loss": 1.3145,
+      "step": 20101
+    },
+    {
+      "epoch": 0.9260607177408209,
+      "grad_norm": 1.21199458630209,
+      "learning_rate": 2.853860528918295e-06,
+      "loss": 1.6036,
+      "step": 20102
+    },
+    {
+      "epoch": 0.9261067858294559,
+      "grad_norm": 1.1636208689553313,
+      "learning_rate": 2.850322410061512e-06,
+      "loss": 1.1118,
+      "step": 20103
+    },
+    {
+      "epoch": 0.926152853918091,
+      "grad_norm": 1.7398516713725,
+      "learning_rate": 2.846786454075723e-06,
+      "loss": 1.2999,
+      "step": 20104
+    },
+    {
+      "epoch": 0.9261989220067259,
+      "grad_norm": 1.255942880917003,
+      "learning_rate": 2.8432526610396526e-06,
+      "loss": 0.9074,
+      "step": 20105
+    },
+    {
+      "epoch": 0.926244990095361,
+      "grad_norm": 0.8538098534632397,
+      "learning_rate": 2.8397210310319724e-06,
+      "loss": 1.2926,
+      "step": 20106
+    },
+    {
+      "epoch": 0.9262910581839959,
+      "grad_norm": 1.5817515566916944,
+      "learning_rate": 2.8361915641313074e-06,
+      "loss": 1.1017,
+      "step": 20107
+    },
+    {
+      "epoch": 0.926337126272631,
+      "grad_norm": 0.8527133706897982,
+      "learning_rate": 2.83266426041624e-06,
+      "loss": 1.0502,
+      "step": 20108
+    },
+    {
+      "epoch": 0.9263831943612659,
+      "grad_norm": 2.0451588974108925,
+      "learning_rate": 2.829139119965285e-06,
+      "loss": 1.4042,
+      "step": 20109
+    },
+    {
+      "epoch": 0.926429262449901,
+      "grad_norm": 1.066621636883474,
+      "learning_rate": 2.825616142856946e-06,
+      "loss": 1.3292,
+      "step": 20110
+    },
+    {
+      "epoch": 0.9264753305385359,
+      "grad_norm": 0.8787672452856983,
+      "learning_rate": 2.8220953291696496e-06,
+      "loss": 1.3426,
+      "step": 20111
+    },
+    {
+      "epoch": 0.926521398627171,
+      "grad_norm": 0.9050164567714716,
+      "learning_rate": 2.8185766789817657e-06,
+      "loss": 1.3793,
+      "step": 20112
+    },
+    {
+      "epoch": 0.9265674667158059,
+      "grad_norm": 1.6908747851671029,
+      "learning_rate": 2.8150601923716437e-06,
+      "loss": 1.1068,
+      "step": 20113
+    },
+    {
+      "epoch": 0.926613534804441,
+      "grad_norm": 0.8940175488645813,
+      "learning_rate": 2.8115458694175758e-06,
+      "loss": 1.2254,
+      "step": 20114
+    },
+    {
+      "epoch": 0.9266596028930759,
+      "grad_norm": 1.287389625539515,
+      "learning_rate": 2.8080337101977993e-06,
+      "loss": 1.2747,
+      "step": 20115
+    },
+    {
+      "epoch": 0.926705670981711,
+      "grad_norm": 1.7302462735260606,
+      "learning_rate": 2.8045237147904857e-06,
+      "loss": 0.9561,
+      "step": 20116
+    },
+    {
+      "epoch": 0.926751739070346,
+      "grad_norm": 0.9036061229180973,
+      "learning_rate": 2.8010158832738166e-06,
+      "loss": 1.3681,
+      "step": 20117
+    },
+    {
+      "epoch": 0.926797807158981,
+      "grad_norm": 0.7725593096998926,
+      "learning_rate": 2.7975102157258626e-06,
+      "loss": 1.3845,
+      "step": 20118
+    },
+    {
+      "epoch": 0.926843875247616,
+      "grad_norm": 1.9313874854181687,
+      "learning_rate": 2.7940067122246726e-06,
+      "loss": 1.1709,
+      "step": 20119
+    },
+    {
+      "epoch": 0.926889943336251,
+      "grad_norm": 1.3233464985981118,
+      "learning_rate": 2.790505372848251e-06,
+      "loss": 1.3205,
+      "step": 20120
+    },
+    {
+      "epoch": 0.926936011424886,
+      "grad_norm": 0.657389455329695,
+      "learning_rate": 2.7870061976745466e-06,
+      "loss": 1.4772,
+      "step": 20121
+    },
+    {
+      "epoch": 0.926982079513521,
+      "grad_norm": 1.3888548649419523,
+      "learning_rate": 2.7835091867814745e-06,
+      "loss": 0.9811,
+      "step": 20122
+    },
+    {
+      "epoch": 0.927028147602156,
+      "grad_norm": 0.896422330072768,
+      "learning_rate": 2.7800143402468724e-06,
+      "loss": 1.2738,
+      "step": 20123
+    },
+    {
+      "epoch": 0.927074215690791,
+      "grad_norm": 1.3721002584773934,
+      "learning_rate": 2.7765216581485565e-06,
+      "loss": 1.5169,
+      "step": 20124
+    },
+    {
+      "epoch": 0.927120283779426,
+      "grad_norm": 0.8837377517332169,
+      "learning_rate": 2.773031140564286e-06,
+      "loss": 1.3087,
+      "step": 20125
+    },
+    {
+      "epoch": 0.927166351868061,
+      "grad_norm": 0.7963877957307417,
+      "learning_rate": 2.769542787571766e-06,
+      "loss": 1.3118,
+      "step": 20126
+    },
+    {
+      "epoch": 0.927212419956696,
+      "grad_norm": 1.0816712465237184,
+      "learning_rate": 2.7660565992486676e-06,
+      "loss": 1.3377,
+      "step": 20127
+    },
+    {
+      "epoch": 0.927258488045331,
+      "grad_norm": 0.9854388351994596,
+      "learning_rate": 2.762572575672584e-06,
+      "loss": 1.2934,
+      "step": 20128
+    },
+    {
+      "epoch": 0.927304556133966,
+      "grad_norm": 0.7689302819723783,
+      "learning_rate": 2.7590907169211087e-06,
+      "loss": 1.4419,
+      "step": 20129
+    },
+    {
+      "epoch": 0.927350624222601,
+      "grad_norm": 1.3706458877039163,
+      "learning_rate": 2.755611023071736e-06,
+      "loss": 1.374,
+      "step": 20130
+    },
+    {
+      "epoch": 0.927396692311236,
+      "grad_norm": 1.363113890317845,
+      "learning_rate": 2.752133494201947e-06,
+      "loss": 1.2389,
+      "step": 20131
+    },
+    {
+      "epoch": 0.9274427603998711,
+      "grad_norm": 0.7947938697484348,
+      "learning_rate": 2.74865813038917e-06,
+      "loss": 1.1861,
+      "step": 20132
+    },
+    {
+      "epoch": 0.927488828488506,
+      "grad_norm": 1.2212193991674178,
+      "learning_rate": 2.745184931710765e-06,
+      "loss": 1.1512,
+      "step": 20133
+    },
+    {
+      "epoch": 0.927534896577141,
+      "grad_norm": 0.9775193664079249,
+      "learning_rate": 2.7417138982440693e-06,
+      "loss": 1.3634,
+      "step": 20134
+    },
+    {
+      "epoch": 0.927580964665776,
+      "grad_norm": 1.1798246101835015,
+      "learning_rate": 2.7382450300663332e-06,
+      "loss": 1.4862,
+      "step": 20135
+    },
+    {
+      "epoch": 0.927627032754411,
+      "grad_norm": 1.1446208342067254,
+      "learning_rate": 2.7347783272548166e-06,
+      "loss": 1.2871,
+      "step": 20136
+    },
+    {
+      "epoch": 0.927673100843046,
+      "grad_norm": 1.4934209135429564,
+      "learning_rate": 2.7313137898866804e-06,
+      "loss": 1.1187,
+      "step": 20137
+    },
+    {
+      "epoch": 0.927719168931681,
+      "grad_norm": 0.8851198063285156,
+      "learning_rate": 2.7278514180390735e-06,
+      "loss": 1.3574,
+      "step": 20138
+    },
+    {
+      "epoch": 0.927765237020316,
+      "grad_norm": 0.823689446511368,
+      "learning_rate": 2.7243912117890456e-06,
+      "loss": 1.3238,
+      "step": 20139
+    },
+    {
+      "epoch": 0.927811305108951,
+      "grad_norm": 0.74372816781041,
+      "learning_rate": 2.72093317121368e-06,
+      "loss": 1.1317,
+      "step": 20140
+    },
+    {
+      "epoch": 0.927857373197586,
+      "grad_norm": 1.2237139093235767,
+      "learning_rate": 2.7174772963899363e-06,
+      "loss": 1.2865,
+      "step": 20141
+    },
+    {
+      "epoch": 0.927903441286221,
+      "grad_norm": 0.7209720226539648,
+      "learning_rate": 2.7140235873947426e-06,
+      "loss": 1.3089,
+      "step": 20142
+    },
+    {
+      "epoch": 0.927949509374856,
+      "grad_norm": 2.8066307643694555,
+      "learning_rate": 2.7105720443050155e-06,
+      "loss": 1.2698,
+      "step": 20143
+    },
+    {
+      "epoch": 0.927995577463491,
+      "grad_norm": 1.29524391645714,
+      "learning_rate": 2.7071226671975815e-06,
+      "loss": 1.2563,
+      "step": 20144
+    },
+    {
+      "epoch": 0.9280416455521261,
+      "grad_norm": 1.8026965707484268,
+      "learning_rate": 2.7036754561492463e-06,
+      "loss": 1.1669,
+      "step": 20145
+    },
+    {
+      "epoch": 0.928087713640761,
+      "grad_norm": 0.8461383208900529,
+      "learning_rate": 2.700230411236737e-06,
+      "loss": 1.1613,
+      "step": 20146
+    },
+    {
+      "epoch": 0.9281337817293961,
+      "grad_norm": 1.8016546752075913,
+      "learning_rate": 2.696787532536782e-06,
+      "loss": 1.4807,
+      "step": 20147
+    },
+    {
+      "epoch": 0.928179849818031,
+      "grad_norm": 0.813743944496833,
+      "learning_rate": 2.693346820125997e-06,
+      "loss": 1.1712,
+      "step": 20148
+    },
+    {
+      "epoch": 0.9282259179066661,
+      "grad_norm": 3.3647989735690382,
+      "learning_rate": 2.68990827408101e-06,
+      "loss": 1.1113,
+      "step": 20149
+    },
+    {
+      "epoch": 0.928271985995301,
+      "grad_norm": 6.395439668539842,
+      "learning_rate": 2.6864718944783487e-06,
+      "loss": 1.1706,
+      "step": 20150
+    },
+    {
+      "epoch": 0.9283180540839361,
+      "grad_norm": 0.9291905240094169,
+      "learning_rate": 2.683037681394551e-06,
+      "loss": 1.0775,
+      "step": 20151
+    },
+    {
+      "epoch": 0.928364122172571,
+      "grad_norm": 0.9555328893067073,
+      "learning_rate": 2.679605634906057e-06,
+      "loss": 1.0016,
+      "step": 20152
+    },
+    {
+      "epoch": 0.9284101902612061,
+      "grad_norm": 0.8831319617126542,
+      "learning_rate": 2.676175755089272e-06,
+      "loss": 1.2407,
+      "step": 20153
+    },
+    {
+      "epoch": 0.928456258349841,
+      "grad_norm": 1.368715721317814,
+      "learning_rate": 2.6727480420205455e-06,
+      "loss": 1.3175,
+      "step": 20154
+    },
+    {
+      "epoch": 0.9285023264384761,
+      "grad_norm": 1.0361664707963223,
+      "learning_rate": 2.6693224957762163e-06,
+      "loss": 1.5943,
+      "step": 20155
+    },
+    {
+      "epoch": 0.928548394527111,
+      "grad_norm": 1.069569648605764,
+      "learning_rate": 2.665899116432535e-06,
+      "loss": 1.1603,
+      "step": 20156
+    },
+    {
+      "epoch": 0.9285944626157461,
+      "grad_norm": 0.9001479774995419,
+      "learning_rate": 2.6624779040657075e-06,
+      "loss": 1.0709,
+      "step": 20157
+    },
+    {
+      "epoch": 0.9286405307043811,
+      "grad_norm": 0.7118382324352117,
+      "learning_rate": 2.6590588587519282e-06,
+      "loss": 1.398,
+      "step": 20158
+    },
+    {
+      "epoch": 0.9286865987930161,
+      "grad_norm": 0.7669013051017345,
+      "learning_rate": 2.6556419805672917e-06,
+      "loss": 1.2524,
+      "step": 20159
+    },
+    {
+      "epoch": 0.9287326668816511,
+      "grad_norm": 0.7745464145126351,
+      "learning_rate": 2.6522272695878814e-06,
+      "loss": 1.3636,
+      "step": 20160
+    },
+    {
+      "epoch": 0.9287787349702861,
+      "grad_norm": 1.047107251661746,
+      "learning_rate": 2.6488147258897034e-06,
+      "loss": 1.1937,
+      "step": 20161
+    },
+    {
+      "epoch": 0.9288248030589211,
+      "grad_norm": 0.6566514498587994,
+      "learning_rate": 2.6454043495487525e-06,
+      "loss": 1.653,
+      "step": 20162
+    },
+    {
+      "epoch": 0.9288708711475561,
+      "grad_norm": 0.7706561430934347,
+      "learning_rate": 2.6419961406409454e-06,
+      "loss": 1.3732,
+      "step": 20163
+    },
+    {
+      "epoch": 0.9289169392361911,
+      "grad_norm": 2.2596711105688603,
+      "learning_rate": 2.638590099242155e-06,
+      "loss": 1.2512,
+      "step": 20164
+    },
+    {
+      "epoch": 0.9289630073248261,
+      "grad_norm": 2.344662315112977,
+      "learning_rate": 2.635186225428221e-06,
+      "loss": 0.9948,
+      "step": 20165
+    },
+    {
+      "epoch": 0.9290090754134611,
+      "grad_norm": 0.9610973766387957,
+      "learning_rate": 2.6317845192749158e-06,
+      "loss": 1.5078,
+      "step": 20166
+    },
+    {
+      "epoch": 0.9290551435020961,
+      "grad_norm": 0.8278726860127197,
+      "learning_rate": 2.6283849808579896e-06,
+      "loss": 1.0938,
+      "step": 20167
+    },
+    {
+      "epoch": 0.9291012115907311,
+      "grad_norm": 1.066540647850569,
+      "learning_rate": 2.6249876102531155e-06,
+      "loss": 1.4671,
+      "step": 20168
+    },
+    {
+      "epoch": 0.9291472796793661,
+      "grad_norm": 1.267483211521634,
+      "learning_rate": 2.621592407535922e-06,
+      "loss": 1.2934,
+      "step": 20169
+    },
+    {
+      "epoch": 0.9291933477680011,
+      "grad_norm": 1.312207906510994,
+      "learning_rate": 2.618199372782004e-06,
+      "loss": 1.049,
+      "step": 20170
+    },
+    {
+      "epoch": 0.929239415856636,
+      "grad_norm": 1.1748108608327277,
+      "learning_rate": 2.6148085060669127e-06,
+      "loss": 1.0175,
+      "step": 20171
+    },
+    {
+      "epoch": 0.9292854839452711,
+      "grad_norm": 0.799925769167349,
+      "learning_rate": 2.61141980746612e-06,
+      "loss": 1.3648,
+      "step": 20172
+    },
+    {
+      "epoch": 0.9293315520339062,
+      "grad_norm": 1.227248292725964,
+      "learning_rate": 2.6080332770550776e-06,
+      "loss": 1.0068,
+      "step": 20173
+    },
+    {
+      "epoch": 0.9293776201225411,
+      "grad_norm": 1.218211856770454,
+      "learning_rate": 2.6046489149091914e-06,
+      "loss": 0.9445,
+      "step": 20174
+    },
+    {
+      "epoch": 0.9294236882111762,
+      "grad_norm": 0.8945959422643096,
+      "learning_rate": 2.601266721103801e-06,
+      "loss": 1.0777,
+      "step": 20175
+    },
+    {
+      "epoch": 0.9294697562998111,
+      "grad_norm": 1.2506339031885536,
+      "learning_rate": 2.5978866957141913e-06,
+      "loss": 1.219,
+      "step": 20176
+    },
+    {
+      "epoch": 0.9295158243884462,
+      "grad_norm": 0.8636029880180144,
+      "learning_rate": 2.5945088388156347e-06,
+      "loss": 1.0663,
+      "step": 20177
+    },
+    {
+      "epoch": 0.9295618924770811,
+      "grad_norm": 0.8160193183529844,
+      "learning_rate": 2.591133150483327e-06,
+      "loss": 1.6163,
+      "step": 20178
+    },
+    {
+      "epoch": 0.9296079605657162,
+      "grad_norm": 1.5021404008854857,
+      "learning_rate": 2.587759630792419e-06,
+      "loss": 1.3154,
+      "step": 20179
+    },
+    {
+      "epoch": 0.9296540286543511,
+      "grad_norm": 0.9193233412671346,
+      "learning_rate": 2.5843882798180065e-06,
+      "loss": 1.0373,
+      "step": 20180
+    },
+    {
+      "epoch": 0.9297000967429861,
+      "grad_norm": 1.1804627754645964,
+      "learning_rate": 2.5810190976351736e-06,
+      "loss": 1.3351,
+      "step": 20181
+    },
+    {
+      "epoch": 0.9297461648316211,
+      "grad_norm": 1.3017366051994463,
+      "learning_rate": 2.5776520843189047e-06,
+      "loss": 1.2084,
+      "step": 20182
+    },
+    {
+      "epoch": 0.9297922329202561,
+      "grad_norm": 0.8847932661503664,
+      "learning_rate": 2.574287239944162e-06,
+      "loss": 1.2088,
+      "step": 20183
+    },
+    {
+      "epoch": 0.9298383010088911,
+      "grad_norm": 0.6977202599734658,
+      "learning_rate": 2.570924564585864e-06,
+      "loss": 1.3508,
+      "step": 20184
+    },
+    {
+      "epoch": 0.9298843690975261,
+      "grad_norm": 0.8241277571182559,
+      "learning_rate": 2.5675640583188944e-06,
+      "loss": 1.2396,
+      "step": 20185
+    },
+    {
+      "epoch": 0.9299304371861612,
+      "grad_norm": 1.0614758352292488,
+      "learning_rate": 2.564205721218027e-06,
+      "loss": 1.0573,
+      "step": 20186
+    },
+    {
+      "epoch": 0.9299765052747961,
+      "grad_norm": 1.3994909675003135,
+      "learning_rate": 2.560849553358058e-06,
+      "loss": 1.1168,
+      "step": 20187
+    },
+    {
+      "epoch": 0.9300225733634312,
+      "grad_norm": 1.2589525356678557,
+      "learning_rate": 2.5574955548137046e-06,
+      "loss": 1.2637,
+      "step": 20188
+    },
+    {
+      "epoch": 0.9300686414520661,
+      "grad_norm": 1.005752663504434,
+      "learning_rate": 2.5541437256596412e-06,
+      "loss": 1.2137,
+      "step": 20189
+    },
+    {
+      "epoch": 0.9301147095407012,
+      "grad_norm": 0.7863089561464581,
+      "learning_rate": 2.5507940659704856e-06,
+      "loss": 1.5873,
+      "step": 20190
+    },
+    {
+      "epoch": 0.9301607776293361,
+      "grad_norm": 1.1972933733962123,
+      "learning_rate": 2.5474465758208e-06,
+      "loss": 0.8506,
+      "step": 20191
+    },
+    {
+      "epoch": 0.9302068457179712,
+      "grad_norm": 0.9049306357364914,
+      "learning_rate": 2.544101255285125e-06,
+      "loss": 1.1812,
+      "step": 20192
+    },
+    {
+      "epoch": 0.9302529138066061,
+      "grad_norm": 0.8474744358648504,
+      "learning_rate": 2.5407581044379346e-06,
+      "loss": 1.555,
+      "step": 20193
+    },
+    {
+      "epoch": 0.9302989818952412,
+      "grad_norm": 0.8960115020797595,
+      "learning_rate": 2.5374171233536583e-06,
+      "loss": 1.2393,
+      "step": 20194
+    },
+    {
+      "epoch": 0.9303450499838761,
+      "grad_norm": 1.2118044538857486,
+      "learning_rate": 2.534078312106669e-06,
+      "loss": 1.1453,
+      "step": 20195
+    },
+    {
+      "epoch": 0.9303911180725112,
+      "grad_norm": 1.036907236366032,
+      "learning_rate": 2.5307416707713306e-06,
+      "loss": 1.259,
+      "step": 20196
+    },
+    {
+      "epoch": 0.9304371861611461,
+      "grad_norm": 1.2265559456042525,
+      "learning_rate": 2.5274071994218827e-06,
+      "loss": 1.0407,
+      "step": 20197
+    },
+    {
+      "epoch": 0.9304832542497812,
+      "grad_norm": 1.5164776261266344,
+      "learning_rate": 2.524074898132589e-06,
+      "loss": 1.0471,
+      "step": 20198
+    },
+    {
+      "epoch": 0.9305293223384162,
+      "grad_norm": 0.869621497821763,
+      "learning_rate": 2.5207447669776453e-06,
+      "loss": 1.5647,
+      "step": 20199
+    },
+    {
+      "epoch": 0.9305753904270512,
+      "grad_norm": 0.9492058019196594,
+      "learning_rate": 2.5174168060311585e-06,
+      "loss": 1.0995,
+      "step": 20200
+    },
+    {
+      "epoch": 0.9306214585156862,
+      "grad_norm": 0.7197355980773762,
+      "learning_rate": 2.514091015367248e-06,
+      "loss": 1.3556,
+      "step": 20201
+    },
+    {
+      "epoch": 0.9306675266043212,
+      "grad_norm": 0.8919463043887617,
+      "learning_rate": 2.510767395059954e-06,
+      "loss": 1.2881,
+      "step": 20202
+    },
+    {
+      "epoch": 0.9307135946929562,
+      "grad_norm": 0.867487289657351,
+      "learning_rate": 2.507445945183262e-06,
+      "loss": 1.3913,
+      "step": 20203
+    },
+    {
+      "epoch": 0.9307596627815912,
+      "grad_norm": 1.241330114065755,
+      "learning_rate": 2.504126665811124e-06,
+      "loss": 1.2317,
+      "step": 20204
+    },
+    {
+      "epoch": 0.9308057308702262,
+      "grad_norm": 0.9728838236140589,
+      "learning_rate": 2.500809557017425e-06,
+      "loss": 1.1906,
+      "step": 20205
+    },
+    {
+      "epoch": 0.9308517989588612,
+      "grad_norm": 0.6803917205171365,
+      "learning_rate": 2.4974946188760286e-06,
+      "loss": 1.5178,
+      "step": 20206
+    },
+    {
+      "epoch": 0.9308978670474962,
+      "grad_norm": 0.7851135825200124,
+      "learning_rate": 2.494181851460753e-06,
+      "loss": 1.3131,
+      "step": 20207
+    },
+    {
+      "epoch": 0.9309439351361312,
+      "grad_norm": 1.1948523081451918,
+      "learning_rate": 2.490871254845317e-06,
+      "loss": 1.0847,
+      "step": 20208
+    },
+    {
+      "epoch": 0.9309900032247662,
+      "grad_norm": 0.8520723748365955,
+      "learning_rate": 2.4875628291034404e-06,
+      "loss": 1.3238,
+      "step": 20209
+    },
+    {
+      "epoch": 0.9310360713134012,
+      "grad_norm": 1.1070065959211037,
+      "learning_rate": 2.4842565743087853e-06,
+      "loss": 1.1335,
+      "step": 20210
+    },
+    {
+      "epoch": 0.9310821394020362,
+      "grad_norm": 1.2053173943993627,
+      "learning_rate": 2.480952490534949e-06,
+      "loss": 1.0541,
+      "step": 20211
+    },
+    {
+      "epoch": 0.9311282074906713,
+      "grad_norm": 1.4084877126595083,
+      "learning_rate": 2.477650577855495e-06,
+      "loss": 0.9995,
+      "step": 20212
+    },
+    {
+      "epoch": 0.9311742755793062,
+      "grad_norm": 1.1153409068634532,
+      "learning_rate": 2.4743508363439305e-06,
+      "loss": 1.2059,
+      "step": 20213
+    },
+    {
+      "epoch": 0.9312203436679413,
+      "grad_norm": 1.0546554928645173,
+      "learning_rate": 2.471053266073742e-06,
+      "loss": 1.295,
+      "step": 20214
+    },
+    {
+      "epoch": 0.9312664117565762,
+      "grad_norm": 1.1853699757465275,
+      "learning_rate": 2.4677578671183035e-06,
+      "loss": 1.2708,
+      "step": 20215
+    },
+    {
+      "epoch": 0.9313124798452113,
+      "grad_norm": 0.7897924489397852,
+      "learning_rate": 2.464464639551012e-06,
+      "loss": 1.2122,
+      "step": 20216
+    },
+    {
+      "epoch": 0.9313585479338462,
+      "grad_norm": 0.863614994397496,
+      "learning_rate": 2.461173583445198e-06,
+      "loss": 1.025,
+      "step": 20217
+    },
+    {
+      "epoch": 0.9314046160224813,
+      "grad_norm": 0.7646982813385923,
+      "learning_rate": 2.457884698874091e-06,
+      "loss": 1.4293,
+      "step": 20218
+    },
+    {
+      "epoch": 0.9314506841111162,
+      "grad_norm": 0.7450580047316864,
+      "learning_rate": 2.4545979859109335e-06,
+      "loss": 1.2572,
+      "step": 20219
+    },
+    {
+      "epoch": 0.9314967521997513,
+      "grad_norm": 0.79781063858838,
+      "learning_rate": 2.4513134446288887e-06,
+      "loss": 1.1427,
+      "step": 20220
+    },
+    {
+      "epoch": 0.9315428202883862,
+      "grad_norm": 1.1898588741818064,
+      "learning_rate": 2.4480310751011093e-06,
+      "loss": 1.1558,
+      "step": 20221
+    },
+    {
+      "epoch": 0.9315888883770213,
+      "grad_norm": 0.8702814686517077,
+      "learning_rate": 2.4447508774006477e-06,
+      "loss": 1.3209,
+      "step": 20222
+    },
+    {
+      "epoch": 0.9316349564656562,
+      "grad_norm": 0.860914029601199,
+      "learning_rate": 2.4414728516005236e-06,
+      "loss": 1.2684,
+      "step": 20223
+    },
+    {
+      "epoch": 0.9316810245542912,
+      "grad_norm": 1.1660134452112005,
+      "learning_rate": 2.438196997773734e-06,
+      "loss": 1.0696,
+      "step": 20224
+    },
+    {
+      "epoch": 0.9317270926429262,
+      "grad_norm": 0.926928291644071,
+      "learning_rate": 2.4349233159932204e-06,
+      "loss": 1.0642,
+      "step": 20225
+    },
+    {
+      "epoch": 0.9317731607315612,
+      "grad_norm": 1.5243434504971958,
+      "learning_rate": 2.431651806331847e-06,
+      "loss": 1.1397,
+      "step": 20226
+    },
+    {
+      "epoch": 0.9318192288201963,
+      "grad_norm": 1.4485876854270596,
+      "learning_rate": 2.4283824688624335e-06,
+      "loss": 1.025,
+      "step": 20227
+    },
+    {
+      "epoch": 0.9318652969088312,
+      "grad_norm": 1.1193416258771822,
+      "learning_rate": 2.42511530365781e-06,
+      "loss": 1.0096,
+      "step": 20228
+    },
+    {
+      "epoch": 0.9319113649974663,
+      "grad_norm": 1.505075042317716,
+      "learning_rate": 2.4218503107906633e-06,
+      "loss": 1.0936,
+      "step": 20229
+    },
+    {
+      "epoch": 0.9319574330861012,
+      "grad_norm": 1.231603036519025,
+      "learning_rate": 2.4185874903337235e-06,
+      "loss": 0.8525,
+      "step": 20230
+    },
+    {
+      "epoch": 0.9320035011747363,
+      "grad_norm": 1.764263148060279,
+      "learning_rate": 2.415326842359611e-06,
+      "loss": 1.3249,
+      "step": 20231
+    },
+    {
+      "epoch": 0.9320495692633712,
+      "grad_norm": 1.357978593909601,
+      "learning_rate": 2.4120683669409337e-06,
+      "loss": 1.4477,
+      "step": 20232
+    },
+    {
+      "epoch": 0.9320956373520063,
+      "grad_norm": 1.032311127247488,
+      "learning_rate": 2.4088120641502233e-06,
+      "loss": 1.1169,
+      "step": 20233
+    },
+    {
+      "epoch": 0.9321417054406412,
+      "grad_norm": 2.6975575745889318,
+      "learning_rate": 2.4055579340599767e-06,
+      "loss": 1.0576,
+      "step": 20234
+    },
+    {
+      "epoch": 0.9321877735292763,
+      "grad_norm": 1.0745821517922942,
+      "learning_rate": 2.402305976742647e-06,
+      "loss": 1.3181,
+      "step": 20235
+    },
+    {
+      "epoch": 0.9322338416179112,
+      "grad_norm": 0.6695125481427887,
+      "learning_rate": 2.3990561922706325e-06,
+      "loss": 1.3599,
+      "step": 20236
+    },
+    {
+      "epoch": 0.9322799097065463,
+      "grad_norm": 0.9357769237087983,
+      "learning_rate": 2.3958085807162856e-06,
+      "loss": 1.2659,
+      "step": 20237
+    },
+    {
+      "epoch": 0.9323259777951812,
+      "grad_norm": 1.6938368927909122,
+      "learning_rate": 2.3925631421518824e-06,
+      "loss": 1.0419,
+      "step": 20238
+    },
+    {
+      "epoch": 0.9323720458838163,
+      "grad_norm": 0.9352907511898512,
+      "learning_rate": 2.3893198766497203e-06,
+      "loss": 1.2078,
+      "step": 20239
+    },
+    {
+      "epoch": 0.9324181139724513,
+      "grad_norm": 1.008855403325513,
+      "learning_rate": 2.3860787842819866e-06,
+      "loss": 1.2651,
+      "step": 20240
+    },
+    {
+      "epoch": 0.9324641820610863,
+      "grad_norm": 0.9905339153263586,
+      "learning_rate": 2.382839865120823e-06,
+      "loss": 1.2824,
+      "step": 20241
+    },
+    {
+      "epoch": 0.9325102501497213,
+      "grad_norm": 1.1668683498112304,
+      "learning_rate": 2.3796031192383606e-06,
+      "loss": 1.0985,
+      "step": 20242
+    },
+    {
+      "epoch": 0.9325563182383563,
+      "grad_norm": 0.8250604611875267,
+      "learning_rate": 2.376368546706653e-06,
+      "loss": 1.4456,
+      "step": 20243
+    },
+    {
+      "epoch": 0.9326023863269913,
+      "grad_norm": 1.4035982263000129,
+      "learning_rate": 2.373136147597721e-06,
+      "loss": 1.2913,
+      "step": 20244
+    },
+    {
+      "epoch": 0.9326484544156263,
+      "grad_norm": 1.4415589395209762,
+      "learning_rate": 2.369905921983495e-06,
+      "loss": 1.0935,
+      "step": 20245
+    },
+    {
+      "epoch": 0.9326945225042613,
+      "grad_norm": 0.9015936775921595,
+      "learning_rate": 2.3666778699359293e-06,
+      "loss": 1.4781,
+      "step": 20246
+    },
+    {
+      "epoch": 0.9327405905928963,
+      "grad_norm": 0.9531756941482625,
+      "learning_rate": 2.363451991526866e-06,
+      "loss": 1.3761,
+      "step": 20247
+    },
+    {
+      "epoch": 0.9327866586815313,
+      "grad_norm": 4.844241290724144,
+      "learning_rate": 2.3602282868281477e-06,
+      "loss": 1.0434,
+      "step": 20248
+    },
+    {
+      "epoch": 0.9328327267701663,
+      "grad_norm": 0.8108178196731595,
+      "learning_rate": 2.3570067559115062e-06,
+      "loss": 1.0869,
+      "step": 20249
+    },
+    {
+      "epoch": 0.9328787948588013,
+      "grad_norm": 1.0847874404743931,
+      "learning_rate": 2.353787398848717e-06,
+      "loss": 1.1849,
+      "step": 20250
+    },
+    {
+      "epoch": 0.9329248629474363,
+      "grad_norm": 0.8078650648018422,
+      "learning_rate": 2.3505702157114006e-06,
+      "loss": 1.1161,
+      "step": 20251
+    },
+    {
+      "epoch": 0.9329709310360713,
+      "grad_norm": 1.0125255905844117,
+      "learning_rate": 2.347355206571222e-06,
+      "loss": 1.1982,
+      "step": 20252
+    },
+    {
+      "epoch": 0.9330169991247064,
+      "grad_norm": 0.9012526605740228,
+      "learning_rate": 2.3441423714997357e-06,
+      "loss": 1.2963,
+      "step": 20253
+    },
+    {
+      "epoch": 0.9330630672133413,
+      "grad_norm": 1.030149442363325,
+      "learning_rate": 2.3409317105684726e-06,
+      "loss": 1.2619,
+      "step": 20254
+    },
+    {
+      "epoch": 0.9331091353019764,
+      "grad_norm": 1.6148378432390196,
+      "learning_rate": 2.33772322384892e-06,
+      "loss": 1.132,
+      "step": 20255
+    },
+    {
+      "epoch": 0.9331552033906113,
+      "grad_norm": 1.3270196229272437,
+      "learning_rate": 2.3345169114124876e-06,
+      "loss": 1.0197,
+      "step": 20256
+    },
+    {
+      "epoch": 0.9332012714792464,
+      "grad_norm": 0.9136481411284243,
+      "learning_rate": 2.331312773330596e-06,
+      "loss": 1.5099,
+      "step": 20257
+    },
+    {
+      "epoch": 0.9332473395678813,
+      "grad_norm": 3.336211616435383,
+      "learning_rate": 2.3281108096745554e-06,
+      "loss": 1.278,
+      "step": 20258
+    },
+    {
+      "epoch": 0.9332934076565164,
+      "grad_norm": 0.971885565867656,
+      "learning_rate": 2.324911020515663e-06,
+      "loss": 1.3176,
+      "step": 20259
+    },
+    {
+      "epoch": 0.9333394757451513,
+      "grad_norm": 1.176341846168286,
+      "learning_rate": 2.3217134059251298e-06,
+      "loss": 1.4871,
+      "step": 20260
+    },
+    {
+      "epoch": 0.9333855438337864,
+      "grad_norm": 1.0273825645041619,
+      "learning_rate": 2.318517965974176e-06,
+      "loss": 1.1484,
+      "step": 20261
+    },
+    {
+      "epoch": 0.9334316119224213,
+      "grad_norm": 1.1757366777594136,
+      "learning_rate": 2.315324700733945e-06,
+      "loss": 1.1434,
+      "step": 20262
+    },
+    {
+      "epoch": 0.9334776800110564,
+      "grad_norm": 2.5347609677844893,
+      "learning_rate": 2.312133610275502e-06,
+      "loss": 1.1125,
+      "step": 20263
+    },
+    {
+      "epoch": 0.9335237480996913,
+      "grad_norm": 0.7898988039688709,
+      "learning_rate": 2.308944694669901e-06,
+      "loss": 1.1205,
+      "step": 20264
+    },
+    {
+      "epoch": 0.9335698161883264,
+      "grad_norm": 0.9456595745115719,
+      "learning_rate": 2.3057579539881413e-06,
+      "loss": 1.4648,
+      "step": 20265
+    },
+    {
+      "epoch": 0.9336158842769613,
+      "grad_norm": 1.6419536079649246,
+      "learning_rate": 2.3025733883011878e-06,
+      "loss": 1.171,
+      "step": 20266
+    },
+    {
+      "epoch": 0.9336619523655963,
+      "grad_norm": 1.2405934714964648,
+      "learning_rate": 2.299390997679907e-06,
+      "loss": 1.0996,
+      "step": 20267
+    },
+    {
+      "epoch": 0.9337080204542314,
+      "grad_norm": 0.9263916155267254,
+      "learning_rate": 2.296210782195174e-06,
+      "loss": 0.8732,
+      "step": 20268
+    },
+    {
+      "epoch": 0.9337540885428663,
+      "grad_norm": 1.2330480490283466,
+      "learning_rate": 2.2930327419177775e-06,
+      "loss": 1.2285,
+      "step": 20269
+    },
+    {
+      "epoch": 0.9338001566315014,
+      "grad_norm": 0.9945674143103925,
+      "learning_rate": 2.2898568769184837e-06,
+      "loss": 1.34,
+      "step": 20270
+    },
+    {
+      "epoch": 0.9338462247201363,
+      "grad_norm": 1.1209845021362643,
+      "learning_rate": 2.2866831872679794e-06,
+      "loss": 1.32,
+      "step": 20271
+    },
+    {
+      "epoch": 0.9338922928087714,
+      "grad_norm": 0.9352746781912463,
+      "learning_rate": 2.283511673036931e-06,
+      "loss": 1.2454,
+      "step": 20272
+    },
+    {
+      "epoch": 0.9339383608974063,
+      "grad_norm": 0.8156954077115912,
+      "learning_rate": 2.2803423342959597e-06,
+      "loss": 1.6285,
+      "step": 20273
+    },
+    {
+      "epoch": 0.9339844289860414,
+      "grad_norm": 1.0121051683007032,
+      "learning_rate": 2.2771751711155974e-06,
+      "loss": 1.2186,
+      "step": 20274
+    },
+    {
+      "epoch": 0.9340304970746763,
+      "grad_norm": 0.9980720580331415,
+      "learning_rate": 2.2740101835663774e-06,
+      "loss": 1.1229,
+      "step": 20275
+    },
+    {
+      "epoch": 0.9340765651633114,
+      "grad_norm": 1.0876316811196773,
+      "learning_rate": 2.270847371718754e-06,
+      "loss": 1.3396,
+      "step": 20276
+    },
+    {
+      "epoch": 0.9341226332519463,
+      "grad_norm": 0.949729075214821,
+      "learning_rate": 2.2676867356431595e-06,
+      "loss": 1.3075,
+      "step": 20277
+    },
+    {
+      "epoch": 0.9341687013405814,
+      "grad_norm": 1.1367262778012872,
+      "learning_rate": 2.2645282754099274e-06,
+      "loss": 1.5555,
+      "step": 20278
+    },
+    {
+      "epoch": 0.9342147694292163,
+      "grad_norm": 1.5507693958793143,
+      "learning_rate": 2.2613719910893893e-06,
+      "loss": 1.082,
+      "step": 20279
+    },
+    {
+      "epoch": 0.9342608375178514,
+      "grad_norm": 1.736310688714488,
+      "learning_rate": 2.258217882751834e-06,
+      "loss": 1.0152,
+      "step": 20280
+    },
+    {
+      "epoch": 0.9343069056064864,
+      "grad_norm": 0.751699994665109,
+      "learning_rate": 2.255065950467461e-06,
+      "loss": 1.3694,
+      "step": 20281
+    },
+    {
+      "epoch": 0.9343529736951214,
+      "grad_norm": 1.314228340338799,
+      "learning_rate": 2.2519161943064473e-06,
+      "loss": 1.3963,
+      "step": 20282
+    },
+    {
+      "epoch": 0.9343990417837564,
+      "grad_norm": 1.213559731492623,
+      "learning_rate": 2.2487686143389142e-06,
+      "loss": 1.3454,
+      "step": 20283
+    },
+    {
+      "epoch": 0.9344451098723914,
+      "grad_norm": 1.1309268563835115,
+      "learning_rate": 2.2456232106349506e-06,
+      "loss": 1.2111,
+      "step": 20284
+    },
+    {
+      "epoch": 0.9344911779610264,
+      "grad_norm": 1.8166583065655104,
+      "learning_rate": 2.242479983264556e-06,
+      "loss": 1.3326,
+      "step": 20285
+    },
+    {
+      "epoch": 0.9345372460496614,
+      "grad_norm": 2.0085641092779216,
+      "learning_rate": 2.239338932297741e-06,
+      "loss": 1.2405,
+      "step": 20286
+    },
+    {
+      "epoch": 0.9345833141382964,
+      "grad_norm": 4.659555730427509,
+      "learning_rate": 2.236200057804405e-06,
+      "loss": 1.4788,
+      "step": 20287
+    },
+    {
+      "epoch": 0.9346293822269314,
+      "grad_norm": 0.9919551947936963,
+      "learning_rate": 2.2330633598544703e-06,
+      "loss": 1.628,
+      "step": 20288
+    },
+    {
+      "epoch": 0.9346754503155664,
+      "grad_norm": 0.9754522440399673,
+      "learning_rate": 2.2299288385177254e-06,
+      "loss": 1.0581,
+      "step": 20289
+    },
+    {
+      "epoch": 0.9347215184042014,
+      "grad_norm": 0.7912145107585222,
+      "learning_rate": 2.2267964938639807e-06,
+      "loss": 1.2765,
+      "step": 20290
+    },
+    {
+      "epoch": 0.9347675864928364,
+      "grad_norm": 0.8981022179283983,
+      "learning_rate": 2.223666325962981e-06,
+      "loss": 1.1857,
+      "step": 20291
+    },
+    {
+      "epoch": 0.9348136545814714,
+      "grad_norm": 1.2313199363262421,
+      "learning_rate": 2.2205383348843813e-06,
+      "loss": 0.7996,
+      "step": 20292
+    },
+    {
+      "epoch": 0.9348597226701064,
+      "grad_norm": 0.7592903277272174,
+      "learning_rate": 2.217412520697848e-06,
+      "loss": 1.3338,
+      "step": 20293
+    },
+    {
+      "epoch": 0.9349057907587415,
+      "grad_norm": 1.0274171922460373,
+      "learning_rate": 2.214288883472948e-06,
+      "loss": 1.3242,
+      "step": 20294
+    },
+    {
+      "epoch": 0.9349518588473764,
+      "grad_norm": 0.9111674084507639,
+      "learning_rate": 2.21116742327927e-06,
+      "loss": 1.0312,
+      "step": 20295
+    },
+    {
+      "epoch": 0.9349979269360115,
+      "grad_norm": 0.8903098369193222,
+      "learning_rate": 2.2080481401862583e-06,
+      "loss": 1.2039,
+      "step": 20296
+    },
+    {
+      "epoch": 0.9350439950246464,
+      "grad_norm": 1.0491602044331605,
+      "learning_rate": 2.2049310342633688e-06,
+      "loss": 1.1932,
+      "step": 20297
+    },
+    {
+      "epoch": 0.9350900631132815,
+      "grad_norm": 0.8896709318272288,
+      "learning_rate": 2.2018161055800233e-06,
+      "loss": 1.0536,
+      "step": 20298
+    },
+    {
+      "epoch": 0.9351361312019164,
+      "grad_norm": 0.7656867116279746,
+      "learning_rate": 2.1987033542055557e-06,
+      "loss": 1.0859,
+      "step": 20299
+    },
+    {
+      "epoch": 0.9351821992905515,
+      "grad_norm": 0.9362070128750976,
+      "learning_rate": 2.1955927802092547e-06,
+      "loss": 1.1466,
+      "step": 20300
+    },
+    {
+      "epoch": 0.9352282673791864,
+      "grad_norm": 0.8949425904726352,
+      "learning_rate": 2.192484383660376e-06,
+      "loss": 0.9937,
+      "step": 20301
+    },
+    {
+      "epoch": 0.9352743354678215,
+      "grad_norm": 2.784692867373546,
+      "learning_rate": 2.189378164628142e-06,
+      "loss": 1.3945,
+      "step": 20302
+    },
+    {
+      "epoch": 0.9353204035564564,
+      "grad_norm": 1.279598475576696,
+      "learning_rate": 2.1862741231816974e-06,
+      "loss": 1.1681,
+      "step": 20303
+    },
+    {
+      "epoch": 0.9353664716450915,
+      "grad_norm": 4.2126796665365855,
+      "learning_rate": 2.1831722593901315e-06,
+      "loss": 1.0061,
+      "step": 20304
+    },
+    {
+      "epoch": 0.9354125397337264,
+      "grad_norm": 0.8683941716749378,
+      "learning_rate": 2.180072573322511e-06,
+      "loss": 1.2209,
+      "step": 20305
+    },
+    {
+      "epoch": 0.9354586078223615,
+      "grad_norm": 0.6989839074578909,
+      "learning_rate": 2.1769750650478702e-06,
+      "loss": 1.0983,
+      "step": 20306
+    },
+    {
+      "epoch": 0.9355046759109965,
+      "grad_norm": 1.2048259299689823,
+      "learning_rate": 2.1738797346351316e-06,
+      "loss": 1.5423,
+      "step": 20307
+    },
+    {
+      "epoch": 0.9355507439996315,
+      "grad_norm": 0.9435978037455138,
+      "learning_rate": 2.170786582153239e-06,
+      "loss": 1.1007,
+      "step": 20308
+    },
+    {
+      "epoch": 0.9355968120882665,
+      "grad_norm": 0.8756906512844852,
+      "learning_rate": 2.167695607671039e-06,
+      "loss": 1.1501,
+      "step": 20309
+    },
+    {
+      "epoch": 0.9356428801769014,
+      "grad_norm": 1.5247519400211937,
+      "learning_rate": 2.1646068112573416e-06,
+      "loss": 0.9376,
+      "step": 20310
+    },
+    {
+      "epoch": 0.9356889482655365,
+      "grad_norm": 0.7120246829855866,
+      "learning_rate": 2.161520192980926e-06,
+      "loss": 1.4401,
+      "step": 20311
+    },
+    {
+      "epoch": 0.9357350163541714,
+      "grad_norm": 1.1442980649813637,
+      "learning_rate": 2.1584357529105037e-06,
+      "loss": 0.8943,
+      "step": 20312
+    },
+    {
+      "epoch": 0.9357810844428065,
+      "grad_norm": 3.361417224871388,
+      "learning_rate": 2.155353491114753e-06,
+      "loss": 1.4019,
+      "step": 20313
+    },
+    {
+      "epoch": 0.9358271525314414,
+      "grad_norm": 2.156954281369056,
+      "learning_rate": 2.1522734076622865e-06,
+      "loss": 1.3007,
+      "step": 20314
+    },
+    {
+      "epoch": 0.9358732206200765,
+      "grad_norm": 1.1305226989927057,
+      "learning_rate": 2.14919550262167e-06,
+      "loss": 1.1123,
+      "step": 20315
+    },
+    {
+      "epoch": 0.9359192887087114,
+      "grad_norm": 1.3112205192290372,
+      "learning_rate": 2.1461197760614504e-06,
+      "loss": 1.0953,
+      "step": 20316
+    },
+    {
+      "epoch": 0.9359653567973465,
+      "grad_norm": 1.0452767718323404,
+      "learning_rate": 2.143046228050083e-06,
+      "loss": 1.2079,
+      "step": 20317
+    },
+    {
+      "epoch": 0.9360114248859814,
+      "grad_norm": 0.8179102985213297,
+      "learning_rate": 2.139974858656013e-06,
+      "loss": 1.216,
+      "step": 20318
+    },
+    {
+      "epoch": 0.9360574929746165,
+      "grad_norm": 1.1553218848735527,
+      "learning_rate": 2.1369056679475975e-06,
+      "loss": 0.922,
+      "step": 20319
+    },
+    {
+      "epoch": 0.9361035610632514,
+      "grad_norm": 1.0139679234759273,
+      "learning_rate": 2.1338386559931812e-06,
+      "loss": 1.2227,
+      "step": 20320
+    },
+    {
+      "epoch": 0.9361496291518865,
+      "grad_norm": 0.9145341355618335,
+      "learning_rate": 2.130773822861043e-06,
+      "loss": 1.4527,
+      "step": 20321
+    },
+    {
+      "epoch": 0.9361956972405215,
+      "grad_norm": 0.8724757939542235,
+      "learning_rate": 2.1277111686194172e-06,
+      "loss": 1.3745,
+      "step": 20322
+    },
+    {
+      "epoch": 0.9362417653291565,
+      "grad_norm": 0.8758652085367638,
+      "learning_rate": 2.124650693336483e-06,
+      "loss": 0.9584,
+      "step": 20323
+    },
+    {
+      "epoch": 0.9362878334177915,
+      "grad_norm": 12.55611052886484,
+      "learning_rate": 2.1215923970803965e-06,
+      "loss": 1.2891,
+      "step": 20324
+    },
+    {
+      "epoch": 0.9363339015064265,
+      "grad_norm": 1.8931487971779333,
+      "learning_rate": 2.1185362799192143e-06,
+      "loss": 1.0835,
+      "step": 20325
+    },
+    {
+      "epoch": 0.9363799695950615,
+      "grad_norm": 1.424996819211925,
+      "learning_rate": 2.1154823419209934e-06,
+      "loss": 1.0862,
+      "step": 20326
+    },
+    {
+      "epoch": 0.9364260376836965,
+      "grad_norm": 0.8943105066639474,
+      "learning_rate": 2.1124305831537237e-06,
+      "loss": 1.4796,
+      "step": 20327
+    },
+    {
+      "epoch": 0.9364721057723315,
+      "grad_norm": 1.016048174304135,
+      "learning_rate": 2.109381003685351e-06,
+      "loss": 1.1143,
+      "step": 20328
+    },
+    {
+      "epoch": 0.9365181738609665,
+      "grad_norm": 1.638169437747193,
+      "learning_rate": 2.1063336035837544e-06,
+      "loss": 1.0058,
+      "step": 20329
+    },
+    {
+      "epoch": 0.9365642419496015,
+      "grad_norm": 0.9188246820302968,
+      "learning_rate": 2.1032883829167906e-06,
+      "loss": 1.3628,
+      "step": 20330
+    },
+    {
+      "epoch": 0.9366103100382365,
+      "grad_norm": 1.4826898500994699,
+      "learning_rate": 2.1002453417522604e-06,
+      "loss": 1.1099,
+      "step": 20331
+    },
+    {
+      "epoch": 0.9366563781268715,
+      "grad_norm": 1.2004638467632809,
+      "learning_rate": 2.0972044801578994e-06,
+      "loss": 1.1737,
+      "step": 20332
+    },
+    {
+      "epoch": 0.9367024462155065,
+      "grad_norm": 3.2308967654789456,
+      "learning_rate": 2.0941657982014084e-06,
+      "loss": 1.4214,
+      "step": 20333
+    },
+    {
+      "epoch": 0.9367485143041415,
+      "grad_norm": 0.7152325202211307,
+      "learning_rate": 2.091129295950456e-06,
+      "loss": 1.4012,
+      "step": 20334
+    },
+    {
+      "epoch": 0.9367945823927766,
+      "grad_norm": 1.137055070694628,
+      "learning_rate": 2.088094973472632e-06,
+      "loss": 1.0764,
+      "step": 20335
+    },
+    {
+      "epoch": 0.9368406504814115,
+      "grad_norm": 0.9570465276694526,
+      "learning_rate": 2.0850628308354826e-06,
+      "loss": 1.2106,
+      "step": 20336
+    },
+    {
+      "epoch": 0.9368867185700466,
+      "grad_norm": 0.8189278408848664,
+      "learning_rate": 2.0820328681065203e-06,
+      "loss": 1.0388,
+      "step": 20337
+    },
+    {
+      "epoch": 0.9369327866586815,
+      "grad_norm": 1.113223008999633,
+      "learning_rate": 2.079005085353214e-06,
+      "loss": 1.2916,
+      "step": 20338
+    },
+    {
+      "epoch": 0.9369788547473166,
+      "grad_norm": 2.2049951712087057,
+      "learning_rate": 2.0759794826429425e-06,
+      "loss": 1.1962,
+      "step": 20339
+    },
+    {
+      "epoch": 0.9370249228359515,
+      "grad_norm": 3.8118307766292983,
+      "learning_rate": 2.0729560600430964e-06,
+      "loss": 1.2168,
+      "step": 20340
+    },
+    {
+      "epoch": 0.9370709909245866,
+      "grad_norm": 1.2761456133131261,
+      "learning_rate": 2.0699348176209665e-06,
+      "loss": 0.9206,
+      "step": 20341
+    },
+    {
+      "epoch": 0.9371170590132215,
+      "grad_norm": 1.0104908746359456,
+      "learning_rate": 2.0669157554438433e-06,
+      "loss": 1.2432,
+      "step": 20342
+    },
+    {
+      "epoch": 0.9371631271018566,
+      "grad_norm": 0.826890891507593,
+      "learning_rate": 2.0638988735789067e-06,
+      "loss": 1.344,
+      "step": 20343
+    },
+    {
+      "epoch": 0.9372091951904915,
+      "grad_norm": 0.7024890766534164,
+      "learning_rate": 2.060884172093336e-06,
+      "loss": 1.1503,
+      "step": 20344
+    },
+    {
+      "epoch": 0.9372552632791266,
+      "grad_norm": 0.9681711297027266,
+      "learning_rate": 2.057871651054255e-06,
+      "loss": 1.2476,
+      "step": 20345
+    },
+    {
+      "epoch": 0.9373013313677615,
+      "grad_norm": 1.1446919106095554,
+      "learning_rate": 2.054861310528722e-06,
+      "loss": 1.2402,
+      "step": 20346
+    },
+    {
+      "epoch": 0.9373473994563966,
+      "grad_norm": 0.8712595484042629,
+      "learning_rate": 2.051853150583771e-06,
+      "loss": 1.1465,
+      "step": 20347
+    },
+    {
+      "epoch": 0.9373934675450316,
+      "grad_norm": 0.8341077525651434,
+      "learning_rate": 2.04884717128635e-06,
+      "loss": 1.1091,
+      "step": 20348
+    },
+    {
+      "epoch": 0.9374395356336666,
+      "grad_norm": 1.085537317558762,
+      "learning_rate": 2.0458433727034043e-06,
+      "loss": 0.9022,
+      "step": 20349
+    },
+    {
+      "epoch": 0.9374856037223016,
+      "grad_norm": 1.199772162051471,
+      "learning_rate": 2.042841754901803e-06,
+      "loss": 1.1682,
+      "step": 20350
+    },
+    {
+      "epoch": 0.9375316718109366,
+      "grad_norm": 0.6812454206337657,
+      "learning_rate": 2.0398423179483707e-06,
+      "loss": 1.3466,
+      "step": 20351
+    },
+    {
+      "epoch": 0.9375777398995716,
+      "grad_norm": 0.964646580743165,
+      "learning_rate": 2.0368450619098756e-06,
+      "loss": 1.1391,
+      "step": 20352
+    },
+    {
+      "epoch": 0.9376238079882065,
+      "grad_norm": 2.029187231172797,
+      "learning_rate": 2.033849986853065e-06,
+      "loss": 0.9062,
+      "step": 20353
+    },
+    {
+      "epoch": 0.9376698760768416,
+      "grad_norm": 1.3147032497844453,
+      "learning_rate": 2.0308570928446067e-06,
+      "loss": 1.1237,
+      "step": 20354
+    },
+    {
+      "epoch": 0.9377159441654765,
+      "grad_norm": 1.123005923029507,
+      "learning_rate": 2.0278663799511153e-06,
+      "loss": 0.9895,
+      "step": 20355
+    },
+    {
+      "epoch": 0.9377620122541116,
+      "grad_norm": 1.4259887482145681,
+      "learning_rate": 2.024877848239215e-06,
+      "loss": 1.3086,
+      "step": 20356
+    },
+    {
+      "epoch": 0.9378080803427465,
+      "grad_norm": 2.4095812240322267,
+      "learning_rate": 2.021891497775408e-06,
+      "loss": 0.9969,
+      "step": 20357
+    },
+    {
+      "epoch": 0.9378541484313816,
+      "grad_norm": 7.912342924817242,
+      "learning_rate": 2.0189073286262074e-06,
+      "loss": 0.9808,
+      "step": 20358
+    },
+    {
+      "epoch": 0.9379002165200165,
+      "grad_norm": 0.8711705150840713,
+      "learning_rate": 2.0159253408580046e-06,
+      "loss": 1.3084,
+      "step": 20359
+    },
+    {
+      "epoch": 0.9379462846086516,
+      "grad_norm": 1.1603904510058236,
+      "learning_rate": 2.012945534537236e-06,
+      "loss": 1.1441,
+      "step": 20360
+    },
+    {
+      "epoch": 0.9379923526972865,
+      "grad_norm": 1.6420326267860892,
+      "learning_rate": 2.009967909730215e-06,
+      "loss": 1.314,
+      "step": 20361
+    },
+    {
+      "epoch": 0.9380384207859216,
+      "grad_norm": 0.7497409794373695,
+      "learning_rate": 2.006992466503255e-06,
+      "loss": 1.012,
+      "step": 20362
+    },
+    {
+      "epoch": 0.9380844888745566,
+      "grad_norm": 0.8414019991893018,
+      "learning_rate": 2.0040192049225802e-06,
+      "loss": 1.0482,
+      "step": 20363
+    },
+    {
+      "epoch": 0.9381305569631916,
+      "grad_norm": 1.1827000139701218,
+      "learning_rate": 2.0010481250543833e-06,
+      "loss": 1.5028,
+      "step": 20364
+    },
+    {
+      "epoch": 0.9381766250518266,
+      "grad_norm": 1.1597014931242824,
+      "learning_rate": 1.998079226964822e-06,
+      "loss": 1.4222,
+      "step": 20365
+    },
+    {
+      "epoch": 0.9382226931404616,
+      "grad_norm": 0.9663297264998161,
+      "learning_rate": 1.9951125107199874e-06,
+      "loss": 1.0313,
+      "step": 20366
+    },
+    {
+      "epoch": 0.9382687612290966,
+      "grad_norm": 2.403514305871862,
+      "learning_rate": 1.9921479763859276e-06,
+      "loss": 0.84,
+      "step": 20367
+    },
+    {
+      "epoch": 0.9383148293177316,
+      "grad_norm": 0.8726868304766711,
+      "learning_rate": 1.9891856240286554e-06,
+      "loss": 1.3857,
+      "step": 20368
+    },
+    {
+      "epoch": 0.9383608974063666,
+      "grad_norm": 0.864107183403343,
+      "learning_rate": 1.9862254537141077e-06,
+      "loss": 0.9739,
+      "step": 20369
+    },
+    {
+      "epoch": 0.9384069654950016,
+      "grad_norm": 1.0836014266070046,
+      "learning_rate": 1.9832674655081873e-06,
+      "loss": 0.923,
+      "step": 20370
+    },
+    {
+      "epoch": 0.9384530335836366,
+      "grad_norm": 1.5906664790642617,
+      "learning_rate": 1.9803116594767635e-06,
+      "loss": 1.0515,
+      "step": 20371
+    },
+    {
+      "epoch": 0.9384991016722716,
+      "grad_norm": 0.8307131412711517,
+      "learning_rate": 1.977358035685639e-06,
+      "loss": 1.2418,
+      "step": 20372
+    },
+    {
+      "epoch": 0.9385451697609066,
+      "grad_norm": 1.67513907783584,
+      "learning_rate": 1.974406594200551e-06,
+      "loss": 1.2083,
+      "step": 20373
+    },
+    {
+      "epoch": 0.9385912378495416,
+      "grad_norm": 0.797995813509824,
+      "learning_rate": 1.9714573350872232e-06,
+      "loss": 1.0929,
+      "step": 20374
+    },
+    {
+      "epoch": 0.9386373059381766,
+      "grad_norm": 0.8212709956703076,
+      "learning_rate": 1.968510258411316e-06,
+      "loss": 1.1709,
+      "step": 20375
+    },
+    {
+      "epoch": 0.9386833740268117,
+      "grad_norm": 1.220392961087697,
+      "learning_rate": 1.965565364238442e-06,
+      "loss": 0.9726,
+      "step": 20376
+    },
+    {
+      "epoch": 0.9387294421154466,
+      "grad_norm": 0.8528548718897325,
+      "learning_rate": 1.9626226526341605e-06,
+      "loss": 1.2487,
+      "step": 20377
+    },
+    {
+      "epoch": 0.9387755102040817,
+      "grad_norm": 1.3889488207620218,
+      "learning_rate": 1.9596821236639863e-06,
+      "loss": 1.4104,
+      "step": 20378
+    },
+    {
+      "epoch": 0.9388215782927166,
+      "grad_norm": 1.0407666569808927,
+      "learning_rate": 1.9567437773933885e-06,
+      "loss": 1.1546,
+      "step": 20379
+    },
+    {
+      "epoch": 0.9388676463813517,
+      "grad_norm": 0.8300980278587667,
+      "learning_rate": 1.9538076138877924e-06,
+      "loss": 1.0253,
+      "step": 20380
+    },
+    {
+      "epoch": 0.9389137144699866,
+      "grad_norm": 1.0259385039254005,
+      "learning_rate": 1.9508736332125466e-06,
+      "loss": 1.4035,
+      "step": 20381
+    },
+    {
+      "epoch": 0.9389597825586217,
+      "grad_norm": 0.7239246421695257,
+      "learning_rate": 1.947941835432976e-06,
+      "loss": 0.925,
+      "step": 20382
+    },
+    {
+      "epoch": 0.9390058506472566,
+      "grad_norm": 1.3749251733263697,
+      "learning_rate": 1.9450122206143728e-06,
+      "loss": 1.291,
+      "step": 20383
+    },
+    {
+      "epoch": 0.9390519187358917,
+      "grad_norm": 1.0213509278752508,
+      "learning_rate": 1.9420847888219297e-06,
+      "loss": 1.2299,
+      "step": 20384
+    },
+    {
+      "epoch": 0.9390979868245266,
+      "grad_norm": 1.0588426321732882,
+      "learning_rate": 1.9391595401208383e-06,
+      "loss": 1.3848,
+      "step": 20385
+    },
+    {
+      "epoch": 0.9391440549131617,
+      "grad_norm": 0.9099381050984627,
+      "learning_rate": 1.9362364745762252e-06,
+      "loss": 1.1625,
+      "step": 20386
+    },
+    {
+      "epoch": 0.9391901230017966,
+      "grad_norm": 4.451871394271935,
+      "learning_rate": 1.93331559225316e-06,
+      "loss": 1.3992,
+      "step": 20387
+    },
+    {
+      "epoch": 0.9392361910904317,
+      "grad_norm": 1.0033438062728897,
+      "learning_rate": 1.9303968932166793e-06,
+      "loss": 1.3298,
+      "step": 20388
+    },
+    {
+      "epoch": 0.9392822591790667,
+      "grad_norm": 1.6865658024024142,
+      "learning_rate": 1.927480377531754e-06,
+      "loss": 1.1143,
+      "step": 20389
+    },
+    {
+      "epoch": 0.9393283272677017,
+      "grad_norm": 1.231826140458332,
+      "learning_rate": 1.9245660452633317e-06,
+      "loss": 1.033,
+      "step": 20390
+    },
+    {
+      "epoch": 0.9393743953563367,
+      "grad_norm": 0.8759488286316828,
+      "learning_rate": 1.921653896476261e-06,
+      "loss": 1.092,
+      "step": 20391
+    },
+    {
+      "epoch": 0.9394204634449717,
+      "grad_norm": 0.8697122150547337,
+      "learning_rate": 1.918743931235412e-06,
+      "loss": 1.0463,
+      "step": 20392
+    },
+    {
+      "epoch": 0.9394665315336067,
+      "grad_norm": 0.806433964021036,
+      "learning_rate": 1.9158361496055545e-06,
+      "loss": 1.5162,
+      "step": 20393
+    },
+    {
+      "epoch": 0.9395125996222417,
+      "grad_norm": 2.493920133497834,
+      "learning_rate": 1.9129305516514263e-06,
+      "loss": 1.19,
+      "step": 20394
+    },
+    {
+      "epoch": 0.9395586677108767,
+      "grad_norm": 0.8575670659697745,
+      "learning_rate": 1.91002713743772e-06,
+      "loss": 1.4605,
+      "step": 20395
+    },
+    {
+      "epoch": 0.9396047357995116,
+      "grad_norm": 2.809458449562846,
+      "learning_rate": 1.9071259070290615e-06,
+      "loss": 0.994,
+      "step": 20396
+    },
+    {
+      "epoch": 0.9396508038881467,
+      "grad_norm": 0.854732636667296,
+      "learning_rate": 1.9042268604900549e-06,
+      "loss": 1.2343,
+      "step": 20397
+    },
+    {
+      "epoch": 0.9396968719767816,
+      "grad_norm": 1.062616460502494,
+      "learning_rate": 1.9013299978852483e-06,
+      "loss": 1.1231,
+      "step": 20398
+    },
+    {
+      "epoch": 0.9397429400654167,
+      "grad_norm": 1.3044047709348363,
+      "learning_rate": 1.8984353192791238e-06,
+      "loss": 1.059,
+      "step": 20399
+    },
+    {
+      "epoch": 0.9397890081540516,
+      "grad_norm": 0.7993532250735396,
+      "learning_rate": 1.8955428247361184e-06,
+      "loss": 0.9436,
+      "step": 20400
+    },
+    {
+      "epoch": 0.9398350762426867,
+      "grad_norm": 0.7756536955530795,
+      "learning_rate": 1.8926525143206475e-06,
+      "loss": 1.3425,
+      "step": 20401
+    },
+    {
+      "epoch": 0.9398811443313216,
+      "grad_norm": 1.3997860686005037,
+      "learning_rate": 1.8897643880970485e-06,
+      "loss": 1.4351,
+      "step": 20402
+    },
+    {
+      "epoch": 0.9399272124199567,
+      "grad_norm": 0.8473761425949341,
+      "learning_rate": 1.886878446129625e-06,
+      "loss": 1.249,
+      "step": 20403
+    },
+    {
+      "epoch": 0.9399732805085917,
+      "grad_norm": 1.5885315647292708,
+      "learning_rate": 1.8839946884826153e-06,
+      "loss": 1.2528,
+      "step": 20404
+    },
+    {
+      "epoch": 0.9400193485972267,
+      "grad_norm": 0.7662675342766418,
+      "learning_rate": 1.8811131152202454e-06,
+      "loss": 1.4041,
+      "step": 20405
+    },
+    {
+      "epoch": 0.9400654166858617,
+      "grad_norm": 1.8518324078275616,
+      "learning_rate": 1.878233726406642e-06,
+      "loss": 1.3778,
+      "step": 20406
+    },
+    {
+      "epoch": 0.9401114847744967,
+      "grad_norm": 0.8694847706022835,
+      "learning_rate": 1.87535652210592e-06,
+      "loss": 1.1002,
+      "step": 20407
+    },
+    {
+      "epoch": 0.9401575528631317,
+      "grad_norm": 0.7208736300625059,
+      "learning_rate": 1.8724815023821507e-06,
+      "loss": 1.4603,
+      "step": 20408
+    },
+    {
+      "epoch": 0.9402036209517667,
+      "grad_norm": 1.3135731726771118,
+      "learning_rate": 1.8696086672993273e-06,
+      "loss": 1.3518,
+      "step": 20409
+    },
+    {
+      "epoch": 0.9402496890404017,
+      "grad_norm": 1.8770705604493596,
+      "learning_rate": 1.8667380169213987e-06,
+      "loss": 1.2412,
+      "step": 20410
+    },
+    {
+      "epoch": 0.9402957571290367,
+      "grad_norm": 1.4910471985508207,
+      "learning_rate": 1.8638695513122805e-06,
+      "loss": 1.1473,
+      "step": 20411
+    },
+    {
+      "epoch": 0.9403418252176717,
+      "grad_norm": 1.2270813562690484,
+      "learning_rate": 1.861003270535855e-06,
+      "loss": 1.2865,
+      "step": 20412
+    },
+    {
+      "epoch": 0.9403878933063067,
+      "grad_norm": 0.8785428815002858,
+      "learning_rate": 1.858139174655915e-06,
+      "loss": 0.9005,
+      "step": 20413
+    },
+    {
+      "epoch": 0.9404339613949417,
+      "grad_norm": 1.3519870912542813,
+      "learning_rate": 1.8552772637362214e-06,
+      "loss": 1.1745,
+      "step": 20414
+    },
+    {
+      "epoch": 0.9404800294835767,
+      "grad_norm": 0.9464885693986318,
+      "learning_rate": 1.8524175378405007e-06,
+      "loss": 1.2721,
+      "step": 20415
+    },
+    {
+      "epoch": 0.9405260975722117,
+      "grad_norm": 0.987350600127834,
+      "learning_rate": 1.849559997032435e-06,
+      "loss": 1.4195,
+      "step": 20416
+    },
+    {
+      "epoch": 0.9405721656608468,
+      "grad_norm": 0.8864290494941032,
+      "learning_rate": 1.8467046413755963e-06,
+      "loss": 1.0594,
+      "step": 20417
+    },
+    {
+      "epoch": 0.9406182337494817,
+      "grad_norm": 1.171503088257269,
+      "learning_rate": 1.8438514709335997e-06,
+      "loss": 1.0444,
+      "step": 20418
+    },
+    {
+      "epoch": 0.9406643018381168,
+      "grad_norm": 1.0396691397356663,
+      "learning_rate": 1.8410004857699504e-06,
+      "loss": 0.9628,
+      "step": 20419
+    },
+    {
+      "epoch": 0.9407103699267517,
+      "grad_norm": 0.8971618289544389,
+      "learning_rate": 1.8381516859481085e-06,
+      "loss": 1.2858,
+      "step": 20420
+    },
+    {
+      "epoch": 0.9407564380153868,
+      "grad_norm": 0.8061446750852348,
+      "learning_rate": 1.8353050715315012e-06,
+      "loss": 1.1715,
+      "step": 20421
+    },
+    {
+      "epoch": 0.9408025061040217,
+      "grad_norm": 0.828738807833547,
+      "learning_rate": 1.8324606425835223e-06,
+      "loss": 1.436,
+      "step": 20422
+    },
+    {
+      "epoch": 0.9408485741926568,
+      "grad_norm": 0.881822862763121,
+      "learning_rate": 1.8296183991674876e-06,
+      "loss": 1.2244,
+      "step": 20423
+    },
+    {
+      "epoch": 0.9408946422812917,
+      "grad_norm": 0.8070362285179127,
+      "learning_rate": 1.826778341346669e-06,
+      "loss": 1.4752,
+      "step": 20424
+    },
+    {
+      "epoch": 0.9409407103699268,
+      "grad_norm": 1.0548523082015808,
+      "learning_rate": 1.8239404691842932e-06,
+      "loss": 1.1295,
+      "step": 20425
+    },
+    {
+      "epoch": 0.9409867784585617,
+      "grad_norm": 1.5479751572484153,
+      "learning_rate": 1.8211047827435545e-06,
+      "loss": 1.3553,
+      "step": 20426
+    },
+    {
+      "epoch": 0.9410328465471968,
+      "grad_norm": 0.853541771477387,
+      "learning_rate": 1.8182712820875803e-06,
+      "loss": 1.1632,
+      "step": 20427
+    },
+    {
+      "epoch": 0.9410789146358317,
+      "grad_norm": 0.797670341902138,
+      "learning_rate": 1.815439967279442e-06,
+      "loss": 1.428,
+      "step": 20428
+    },
+    {
+      "epoch": 0.9411249827244668,
+      "grad_norm": 1.2218386604227593,
+      "learning_rate": 1.8126108383821782e-06,
+      "loss": 1.4708,
+      "step": 20429
+    },
+    {
+      "epoch": 0.9411710508131018,
+      "grad_norm": 1.087354693481131,
+      "learning_rate": 1.8097838954587943e-06,
+      "loss": 1.1325,
+      "step": 20430
+    },
+    {
+      "epoch": 0.9412171189017368,
+      "grad_norm": 0.7693045996518184,
+      "learning_rate": 1.8069591385722061e-06,
+      "loss": 1.2094,
+      "step": 20431
+    },
+    {
+      "epoch": 0.9412631869903718,
+      "grad_norm": 1.1704038814038047,
+      "learning_rate": 1.8041365677852974e-06,
+      "loss": 1.2192,
+      "step": 20432
+    },
+    {
+      "epoch": 0.9413092550790068,
+      "grad_norm": 1.8480247201225808,
+      "learning_rate": 1.8013161831609282e-06,
+      "loss": 1.1371,
+      "step": 20433
+    },
+    {
+      "epoch": 0.9413553231676418,
+      "grad_norm": 0.888307585118976,
+      "learning_rate": 1.7984979847618821e-06,
+      "loss": 0.9585,
+      "step": 20434
+    },
+    {
+      "epoch": 0.9414013912562768,
+      "grad_norm": 0.8593667881533951,
+      "learning_rate": 1.7956819726508979e-06,
+      "loss": 1.3499,
+      "step": 20435
+    },
+    {
+      "epoch": 0.9414474593449118,
+      "grad_norm": 0.9470972846008952,
+      "learning_rate": 1.7928681468906582e-06,
+      "loss": 0.9814,
+      "step": 20436
+    },
+    {
+      "epoch": 0.9414935274335468,
+      "grad_norm": 1.555019840686538,
+      "learning_rate": 1.7900565075438357e-06,
+      "loss": 1.1568,
+      "step": 20437
+    },
+    {
+      "epoch": 0.9415395955221818,
+      "grad_norm": 1.0039630945549158,
+      "learning_rate": 1.7872470546730024e-06,
+      "loss": 0.9376,
+      "step": 20438
+    },
+    {
+      "epoch": 0.9415856636108167,
+      "grad_norm": 3.4381092075521265,
+      "learning_rate": 1.7844397883407194e-06,
+      "loss": 1.0265,
+      "step": 20439
+    },
+    {
+      "epoch": 0.9416317316994518,
+      "grad_norm": 7.782890249126845,
+      "learning_rate": 1.7816347086094698e-06,
+      "loss": 1.1974,
+      "step": 20440
+    },
+    {
+      "epoch": 0.9416777997880867,
+      "grad_norm": 2.06143341601615,
+      "learning_rate": 1.7788318155417372e-06,
+      "loss": 1.5004,
+      "step": 20441
+    },
+    {
+      "epoch": 0.9417238678767218,
+      "grad_norm": 1.1111323590562487,
+      "learning_rate": 1.7760311091998827e-06,
+      "loss": 1.3466,
+      "step": 20442
+    },
+    {
+      "epoch": 0.9417699359653569,
+      "grad_norm": 1.5809118084836005,
+      "learning_rate": 1.7732325896462787e-06,
+      "loss": 1.0196,
+      "step": 20443
+    },
+    {
+      "epoch": 0.9418160040539918,
+      "grad_norm": 1.0059512805631088,
+      "learning_rate": 1.770436256943231e-06,
+      "loss": 1.237,
+      "step": 20444
+    },
+    {
+      "epoch": 0.9418620721426268,
+      "grad_norm": 0.9583241547183415,
+      "learning_rate": 1.7676421111530006e-06,
+      "loss": 1.3121,
+      "step": 20445
+    },
+    {
+      "epoch": 0.9419081402312618,
+      "grad_norm": 2.1907433403712466,
+      "learning_rate": 1.7648501523377826e-06,
+      "loss": 1.0641,
+      "step": 20446
+    },
+    {
+      "epoch": 0.9419542083198968,
+      "grad_norm": 0.9412872675022649,
+      "learning_rate": 1.7620603805597269e-06,
+      "loss": 0.7748,
+      "step": 20447
+    },
+    {
+      "epoch": 0.9420002764085318,
+      "grad_norm": 0.833097876967602,
+      "learning_rate": 1.7592727958809618e-06,
+      "loss": 1.0999,
+      "step": 20448
+    },
+    {
+      "epoch": 0.9420463444971668,
+      "grad_norm": 3.0862559791875848,
+      "learning_rate": 1.7564873983635377e-06,
+      "loss": 1.15,
+      "step": 20449
+    },
+    {
+      "epoch": 0.9420924125858018,
+      "grad_norm": 1.0201423252163249,
+      "learning_rate": 1.7537041880694827e-06,
+      "loss": 0.9707,
+      "step": 20450
+    },
+    {
+      "epoch": 0.9421384806744368,
+      "grad_norm": 2.239163567967822,
+      "learning_rate": 1.750923165060736e-06,
+      "loss": 1.1396,
+      "step": 20451
+    },
+    {
+      "epoch": 0.9421845487630718,
+      "grad_norm": 1.4915953615602142,
+      "learning_rate": 1.7481443293992261e-06,
+      "loss": 1.1943,
+      "step": 20452
+    },
+    {
+      "epoch": 0.9422306168517068,
+      "grad_norm": 0.8160030441572498,
+      "learning_rate": 1.7453676811468256e-06,
+      "loss": 1.3964,
+      "step": 20453
+    },
+    {
+      "epoch": 0.9422766849403418,
+      "grad_norm": 0.6605288762267282,
+      "learning_rate": 1.7425932203653294e-06,
+      "loss": 1.1365,
+      "step": 20454
+    },
+    {
+      "epoch": 0.9423227530289768,
+      "grad_norm": 1.0096774272797857,
+      "learning_rate": 1.7398209471165328e-06,
+      "loss": 1.0676,
+      "step": 20455
+    },
+    {
+      "epoch": 0.9423688211176118,
+      "grad_norm": 0.9422773351020263,
+      "learning_rate": 1.7370508614621306e-06,
+      "loss": 1.366,
+      "step": 20456
+    },
+    {
+      "epoch": 0.9424148892062468,
+      "grad_norm": 1.8193810020335306,
+      "learning_rate": 1.734282963463818e-06,
+      "loss": 1.2374,
+      "step": 20457
+    },
+    {
+      "epoch": 0.9424609572948819,
+      "grad_norm": 2.7893522632773173,
+      "learning_rate": 1.7315172531831902e-06,
+      "loss": 1.0714,
+      "step": 20458
+    },
+    {
+      "epoch": 0.9425070253835168,
+      "grad_norm": 1.5346867637661517,
+      "learning_rate": 1.7287537306818536e-06,
+      "loss": 1.2299,
+      "step": 20459
+    },
+    {
+      "epoch": 0.9425530934721519,
+      "grad_norm": 0.693254868530791,
+      "learning_rate": 1.7259923960213031e-06,
+      "loss": 1.2897,
+      "step": 20460
+    },
+    {
+      "epoch": 0.9425991615607868,
+      "grad_norm": 1.5276996200016348,
+      "learning_rate": 1.7232332492630453e-06,
+      "loss": 1.0185,
+      "step": 20461
+    },
+    {
+      "epoch": 0.9426452296494219,
+      "grad_norm": 1.0012674785031854,
+      "learning_rate": 1.7204762904684645e-06,
+      "loss": 1.27,
+      "step": 20462
+    },
+    {
+      "epoch": 0.9426912977380568,
+      "grad_norm": 0.920885458447328,
+      "learning_rate": 1.7177215196989893e-06,
+      "loss": 1.3652,
+      "step": 20463
+    },
+    {
+      "epoch": 0.9427373658266919,
+      "grad_norm": 0.7873440877825351,
+      "learning_rate": 1.714968937015915e-06,
+      "loss": 1.3631,
+      "step": 20464
+    },
+    {
+      "epoch": 0.9427834339153268,
+      "grad_norm": 0.7308236265004985,
+      "learning_rate": 1.712218542480537e-06,
+      "loss": 1.1715,
+      "step": 20465
+    },
+    {
+      "epoch": 0.9428295020039619,
+      "grad_norm": 0.9579580576277477,
+      "learning_rate": 1.7094703361540843e-06,
+      "loss": 1.4686,
+      "step": 20466
+    },
+    {
+      "epoch": 0.9428755700925968,
+      "grad_norm": 3.8914116045072746,
+      "learning_rate": 1.706724318097741e-06,
+      "loss": 0.9808,
+      "step": 20467
+    },
+    {
+      "epoch": 0.9429216381812319,
+      "grad_norm": 1.7422277237584651,
+      "learning_rate": 1.7039804883726584e-06,
+      "loss": 1.191,
+      "step": 20468
+    },
+    {
+      "epoch": 0.9429677062698668,
+      "grad_norm": 1.533584736330122,
+      "learning_rate": 1.7012388470398877e-06,
+      "loss": 1.3143,
+      "step": 20469
+    },
+    {
+      "epoch": 0.9430137743585019,
+      "grad_norm": 1.3589914786537787,
+      "learning_rate": 1.698499394160502e-06,
+      "loss": 1.0789,
+      "step": 20470
+    },
+    {
+      "epoch": 0.9430598424471369,
+      "grad_norm": 1.8628227080778026,
+      "learning_rate": 1.695762129795464e-06,
+      "loss": 1.4084,
+      "step": 20471
+    },
+    {
+      "epoch": 0.9431059105357719,
+      "grad_norm": 1.0226815290847397,
+      "learning_rate": 1.693027054005747e-06,
+      "loss": 1.196,
+      "step": 20472
+    },
+    {
+      "epoch": 0.9431519786244069,
+      "grad_norm": 1.1628086119301602,
+      "learning_rate": 1.6902941668522021e-06,
+      "loss": 1.222,
+      "step": 20473
+    },
+    {
+      "epoch": 0.9431980467130419,
+      "grad_norm": 1.1604239804399479,
+      "learning_rate": 1.687563468395703e-06,
+      "loss": 1.0668,
+      "step": 20474
+    },
+    {
+      "epoch": 0.9432441148016769,
+      "grad_norm": 0.9490497330371161,
+      "learning_rate": 1.6848349586970348e-06,
+      "loss": 1.246,
+      "step": 20475
+    },
+    {
+      "epoch": 0.9432901828903119,
+      "grad_norm": 1.3902796922402478,
+      "learning_rate": 1.682108637816937e-06,
+      "loss": 1.1725,
+      "step": 20476
+    },
+    {
+      "epoch": 0.9433362509789469,
+      "grad_norm": 0.9515711096436777,
+      "learning_rate": 1.6793845058161063e-06,
+      "loss": 1.2374,
+      "step": 20477
+    },
+    {
+      "epoch": 0.9433823190675819,
+      "grad_norm": 1.3875909387643188,
+      "learning_rate": 1.6766625627551935e-06,
+      "loss": 0.947,
+      "step": 20478
+    },
+    {
+      "epoch": 0.9434283871562169,
+      "grad_norm": 2.308620225534535,
+      "learning_rate": 1.6739428086948171e-06,
+      "loss": 1.1058,
+      "step": 20479
+    },
+    {
+      "epoch": 0.9434744552448519,
+      "grad_norm": 0.9703958263771364,
+      "learning_rate": 1.6712252436954955e-06,
+      "loss": 1.2105,
+      "step": 20480
+    },
+    {
+      "epoch": 0.9435205233334869,
+      "grad_norm": 1.501958345907626,
+      "learning_rate": 1.668509867817758e-06,
+      "loss": 1.3072,
+      "step": 20481
+    },
+    {
+      "epoch": 0.9435665914221218,
+      "grad_norm": 0.9423682863170922,
+      "learning_rate": 1.6657966811220449e-06,
+      "loss": 1.3573,
+      "step": 20482
+    },
+    {
+      "epoch": 0.9436126595107569,
+      "grad_norm": 0.9837396409586399,
+      "learning_rate": 1.6630856836687525e-06,
+      "loss": 1.1378,
+      "step": 20483
+    },
+    {
+      "epoch": 0.943658727599392,
+      "grad_norm": 2.0833649061208876,
+      "learning_rate": 1.6603768755182547e-06,
+      "loss": 1.4014,
+      "step": 20484
+    },
+    {
+      "epoch": 0.9437047956880269,
+      "grad_norm": 0.9222002307544995,
+      "learning_rate": 1.6576702567308367e-06,
+      "loss": 1.1114,
+      "step": 20485
+    },
+    {
+      "epoch": 0.943750863776662,
+      "grad_norm": 0.9012211599114744,
+      "learning_rate": 1.6549658273667945e-06,
+      "loss": 1.096,
+      "step": 20486
+    },
+    {
+      "epoch": 0.9437969318652969,
+      "grad_norm": 0.7890961359167721,
+      "learning_rate": 1.6522635874862913e-06,
+      "loss": 1.1763,
+      "step": 20487
+    },
+    {
+      "epoch": 0.943842999953932,
+      "grad_norm": 1.4314767834051025,
+      "learning_rate": 1.6495635371495121e-06,
+      "loss": 1.3401,
+      "step": 20488
+    },
+    {
+      "epoch": 0.9438890680425669,
+      "grad_norm": 0.7824726519352831,
+      "learning_rate": 1.6468656764165647e-06,
+      "loss": 1.2613,
+      "step": 20489
+    },
+    {
+      "epoch": 0.943935136131202,
+      "grad_norm": 1.5045488833579117,
+      "learning_rate": 1.6441700053475228e-06,
+      "loss": 1.0289,
+      "step": 20490
+    },
+    {
+      "epoch": 0.9439812042198369,
+      "grad_norm": 1.0735794979640794,
+      "learning_rate": 1.6414765240023832e-06,
+      "loss": 1.0393,
+      "step": 20491
+    },
+    {
+      "epoch": 0.9440272723084719,
+      "grad_norm": 2.5229367142744556,
+      "learning_rate": 1.6387852324411202e-06,
+      "loss": 0.9266,
+      "step": 20492
+    },
+    {
+      "epoch": 0.9440733403971069,
+      "grad_norm": 0.705632782407567,
+      "learning_rate": 1.6360961307236633e-06,
+      "loss": 1.1927,
+      "step": 20493
+    },
+    {
+      "epoch": 0.9441194084857419,
+      "grad_norm": 1.7127210434167384,
+      "learning_rate": 1.6334092189098539e-06,
+      "loss": 1.3036,
+      "step": 20494
+    },
+    {
+      "epoch": 0.9441654765743769,
+      "grad_norm": 1.2615913584192455,
+      "learning_rate": 1.6307244970595214e-06,
+      "loss": 1.2134,
+      "step": 20495
+    },
+    {
+      "epoch": 0.9442115446630119,
+      "grad_norm": 1.5618584292279265,
+      "learning_rate": 1.6280419652324519e-06,
+      "loss": 1.1676,
+      "step": 20496
+    },
+    {
+      "epoch": 0.9442576127516469,
+      "grad_norm": 16.235247599081063,
+      "learning_rate": 1.625361623488364e-06,
+      "loss": 1.4731,
+      "step": 20497
+    },
+    {
+      "epoch": 0.9443036808402819,
+      "grad_norm": 0.9863445000677156,
+      "learning_rate": 1.62268347188691e-06,
+      "loss": 0.9371,
+      "step": 20498
+    },
+    {
+      "epoch": 0.944349748928917,
+      "grad_norm": 1.0763353844884265,
+      "learning_rate": 1.62000751048772e-06,
+      "loss": 1.1049,
+      "step": 20499
+    },
+    {
+      "epoch": 0.9443958170175519,
+      "grad_norm": 1.119864648045888,
+      "learning_rate": 1.6173337393504017e-06,
+      "loss": 1.5886,
+      "step": 20500
+    },
+    {
+      "epoch": 0.944441885106187,
+      "grad_norm": 0.7145081363663861,
+      "learning_rate": 1.6146621585344413e-06,
+      "loss": 1.1943,
+      "step": 20501
+    },
+    {
+      "epoch": 0.9444879531948219,
+      "grad_norm": 1.6427038041644533,
+      "learning_rate": 1.6119927680993241e-06,
+      "loss": 1.3551,
+      "step": 20502
+    },
+    {
+      "epoch": 0.944534021283457,
+      "grad_norm": 0.8312024870763273,
+      "learning_rate": 1.6093255681045027e-06,
+      "loss": 1.2521,
+      "step": 20503
+    },
+    {
+      "epoch": 0.9445800893720919,
+      "grad_norm": 0.9197139352586551,
+      "learning_rate": 1.606660558609352e-06,
+      "loss": 1.3494,
+      "step": 20504
+    },
+    {
+      "epoch": 0.944626157460727,
+      "grad_norm": 1.0006337882616934,
+      "learning_rate": 1.6039977396731798e-06,
+      "loss": 0.8542,
+      "step": 20505
+    },
+    {
+      "epoch": 0.9446722255493619,
+      "grad_norm": 0.9831256084424175,
+      "learning_rate": 1.6013371113552834e-06,
+      "loss": 0.8874,
+      "step": 20506
+    },
+    {
+      "epoch": 0.944718293637997,
+      "grad_norm": 0.9565164868081132,
+      "learning_rate": 1.598678673714904e-06,
+      "loss": 1.2884,
+      "step": 20507
+    },
+    {
+      "epoch": 0.9447643617266319,
+      "grad_norm": 1.3945842608092787,
+      "learning_rate": 1.5960224268112279e-06,
+      "loss": 1.1761,
+      "step": 20508
+    },
+    {
+      "epoch": 0.944810429815267,
+      "grad_norm": 0.8678462903996913,
+      "learning_rate": 1.5933683707033854e-06,
+      "loss": 1.1951,
+      "step": 20509
+    },
+    {
+      "epoch": 0.9448564979039019,
+      "grad_norm": 0.7986445079971175,
+      "learning_rate": 1.5907165054504514e-06,
+      "loss": 1.1148,
+      "step": 20510
+    },
+    {
+      "epoch": 0.944902565992537,
+      "grad_norm": 1.278066696931918,
+      "learning_rate": 1.588066831111501e-06,
+      "loss": 1.124,
+      "step": 20511
+    },
+    {
+      "epoch": 0.944948634081172,
+      "grad_norm": 1.1796910913748138,
+      "learning_rate": 1.5854193477454871e-06,
+      "loss": 1.0803,
+      "step": 20512
+    },
+    {
+      "epoch": 0.944994702169807,
+      "grad_norm": 1.053343924255424,
+      "learning_rate": 1.5827740554113623e-06,
+      "loss": 1.3938,
+      "step": 20513
+    },
+    {
+      "epoch": 0.945040770258442,
+      "grad_norm": 1.4842049032161908,
+      "learning_rate": 1.5801309541680242e-06,
+      "loss": 1.1819,
+      "step": 20514
+    },
+    {
+      "epoch": 0.945086838347077,
+      "grad_norm": 0.8945158107696549,
+      "learning_rate": 1.5774900440743257e-06,
+      "loss": 1.2875,
+      "step": 20515
+    },
+    {
+      "epoch": 0.945132906435712,
+      "grad_norm": 3.9277125705994664,
+      "learning_rate": 1.574851325189053e-06,
+      "loss": 0.7916,
+      "step": 20516
+    },
+    {
+      "epoch": 0.945178974524347,
+      "grad_norm": 1.2457862910945068,
+      "learning_rate": 1.5722147975709478e-06,
+      "loss": 1.157,
+      "step": 20517
+    },
+    {
+      "epoch": 0.945225042612982,
+      "grad_norm": 0.8931479882068385,
+      "learning_rate": 1.569580461278719e-06,
+      "loss": 1.3461,
+      "step": 20518
+    },
+    {
+      "epoch": 0.945271110701617,
+      "grad_norm": 1.2603383794561884,
+      "learning_rate": 1.5669483163709974e-06,
+      "loss": 1.0577,
+      "step": 20519
+    },
+    {
+      "epoch": 0.945317178790252,
+      "grad_norm": 1.0487581120402056,
+      "learning_rate": 1.5643183629064029e-06,
+      "loss": 1.2455,
+      "step": 20520
+    },
+    {
+      "epoch": 0.945363246878887,
+      "grad_norm": 2.1650889337567936,
+      "learning_rate": 1.5616906009434774e-06,
+      "loss": 1.2821,
+      "step": 20521
+    },
+    {
+      "epoch": 0.945409314967522,
+      "grad_norm": 0.9597448952775227,
+      "learning_rate": 1.5590650305407295e-06,
+      "loss": 1.1389,
+      "step": 20522
+    },
+    {
+      "epoch": 0.945455383056157,
+      "grad_norm": 6.330792686609787,
+      "learning_rate": 1.5564416517566016e-06,
+      "loss": 1.0809,
+      "step": 20523
+    },
+    {
+      "epoch": 0.945501451144792,
+      "grad_norm": 0.6760877579122487,
+      "learning_rate": 1.5538204646495135e-06,
+      "loss": 1.14,
+      "step": 20524
+    },
+    {
+      "epoch": 0.9455475192334271,
+      "grad_norm": 1.043721304404405,
+      "learning_rate": 1.5512014692778075e-06,
+      "loss": 1.1343,
+      "step": 20525
+    },
+    {
+      "epoch": 0.945593587322062,
+      "grad_norm": 0.919162799005652,
+      "learning_rate": 1.5485846656998038e-06,
+      "loss": 1.1791,
+      "step": 20526
+    },
+    {
+      "epoch": 0.945639655410697,
+      "grad_norm": 0.7242044333029021,
+      "learning_rate": 1.5459700539737442e-06,
+      "loss": 1.4587,
+      "step": 20527
+    },
+    {
+      "epoch": 0.945685723499332,
+      "grad_norm": 1.2877652800724264,
+      "learning_rate": 1.5433576341578492e-06,
+      "loss": 1.1496,
+      "step": 20528
+    },
+    {
+      "epoch": 0.945731791587967,
+      "grad_norm": 1.6048013172263136,
+      "learning_rate": 1.5407474063102944e-06,
+      "loss": 1.2437,
+      "step": 20529
+    },
+    {
+      "epoch": 0.945777859676602,
+      "grad_norm": 1.0942542885278923,
+      "learning_rate": 1.5381393704891666e-06,
+      "loss": 0.9391,
+      "step": 20530
+    },
+    {
+      "epoch": 0.945823927765237,
+      "grad_norm": 0.7736364651945447,
+      "learning_rate": 1.5355335267525418e-06,
+      "loss": 1.1114,
+      "step": 20531
+    },
+    {
+      "epoch": 0.945869995853872,
+      "grad_norm": 0.9112663602258995,
+      "learning_rate": 1.5329298751584286e-06,
+      "loss": 1.0913,
+      "step": 20532
+    },
+    {
+      "epoch": 0.945916063942507,
+      "grad_norm": 1.1902500799167348,
+      "learning_rate": 1.5303284157648035e-06,
+      "loss": 1.4713,
+      "step": 20533
+    },
+    {
+      "epoch": 0.945962132031142,
+      "grad_norm": 1.1004417444455028,
+      "learning_rate": 1.5277291486295753e-06,
+      "loss": 0.962,
+      "step": 20534
+    },
+    {
+      "epoch": 0.946008200119777,
+      "grad_norm": 1.0102373801522084,
+      "learning_rate": 1.525132073810609e-06,
+      "loss": 1.1405,
+      "step": 20535
+    },
+    {
+      "epoch": 0.946054268208412,
+      "grad_norm": 0.8678483577553684,
+      "learning_rate": 1.522537191365736e-06,
+      "loss": 1.3388,
+      "step": 20536
+    },
+    {
+      "epoch": 0.946100336297047,
+      "grad_norm": 0.7099962114259651,
+      "learning_rate": 1.5199445013527102e-06,
+      "loss": 1.4212,
+      "step": 20537
+    },
+    {
+      "epoch": 0.946146404385682,
+      "grad_norm": 0.8775149920592453,
+      "learning_rate": 1.5173540038292745e-06,
+      "loss": 1.4462,
+      "step": 20538
+    },
+    {
+      "epoch": 0.946192472474317,
+      "grad_norm": 1.3209264395651834,
+      "learning_rate": 1.5147656988530822e-06,
+      "loss": 0.9668,
+      "step": 20539
+    },
+    {
+      "epoch": 0.9462385405629521,
+      "grad_norm": 0.9520660036477817,
+      "learning_rate": 1.5121795864817656e-06,
+      "loss": 1.1642,
+      "step": 20540
+    },
+    {
+      "epoch": 0.946284608651587,
+      "grad_norm": 1.0113188225688516,
+      "learning_rate": 1.5095956667729005e-06,
+      "loss": 1.0766,
+      "step": 20541
+    },
+    {
+      "epoch": 0.9463306767402221,
+      "grad_norm": 0.8120968427507876,
+      "learning_rate": 1.5070139397840188e-06,
+      "loss": 1.2624,
+      "step": 20542
+    },
+    {
+      "epoch": 0.946376744828857,
+      "grad_norm": 0.8978672635524882,
+      "learning_rate": 1.5044344055725857e-06,
+      "loss": 1.1044,
+      "step": 20543
+    },
+    {
+      "epoch": 0.9464228129174921,
+      "grad_norm": 0.9005400857123758,
+      "learning_rate": 1.501857064196044e-06,
+      "loss": 1.3184,
+      "step": 20544
+    },
+    {
+      "epoch": 0.946468881006127,
+      "grad_norm": 0.772397469696935,
+      "learning_rate": 1.49928191571177e-06,
+      "loss": 1.5124,
+      "step": 20545
+    },
+    {
+      "epoch": 0.9465149490947621,
+      "grad_norm": 1.1061077606034468,
+      "learning_rate": 1.4967089601770734e-06,
+      "loss": 1.3268,
+      "step": 20546
+    },
+    {
+      "epoch": 0.946561017183397,
+      "grad_norm": 0.8819417748833245,
+      "learning_rate": 1.4941381976492753e-06,
+      "loss": 0.943,
+      "step": 20547
+    },
+    {
+      "epoch": 0.9466070852720321,
+      "grad_norm": 1.4132950594188367,
+      "learning_rate": 1.491569628185574e-06,
+      "loss": 1.5201,
+      "step": 20548
+    },
+    {
+      "epoch": 0.946653153360667,
+      "grad_norm": 1.0044571533358677,
+      "learning_rate": 1.4890032518431795e-06,
+      "loss": 1.208,
+      "step": 20549
+    },
+    {
+      "epoch": 0.9466992214493021,
+      "grad_norm": 0.8059594550606786,
+      "learning_rate": 1.4864390686792018e-06,
+      "loss": 1.0677,
+      "step": 20550
+    },
+    {
+      "epoch": 0.946745289537937,
+      "grad_norm": 1.632095303389771,
+      "learning_rate": 1.4838770787507617e-06,
+      "loss": 1.2586,
+      "step": 20551
+    },
+    {
+      "epoch": 0.9467913576265721,
+      "grad_norm": 0.75501163271617,
+      "learning_rate": 1.481317282114869e-06,
+      "loss": 1.3705,
+      "step": 20552
+    },
+    {
+      "epoch": 0.9468374257152071,
+      "grad_norm": 1.2120011588403587,
+      "learning_rate": 1.478759678828523e-06,
+      "loss": 1.0689,
+      "step": 20553
+    },
+    {
+      "epoch": 0.9468834938038421,
+      "grad_norm": 1.132430029932999,
+      "learning_rate": 1.4762042689486665e-06,
+      "loss": 1.3593,
+      "step": 20554
+    },
+    {
+      "epoch": 0.9469295618924771,
+      "grad_norm": 1.2036061160923204,
+      "learning_rate": 1.473651052532199e-06,
+      "loss": 1.1945,
+      "step": 20555
+    },
+    {
+      "epoch": 0.9469756299811121,
+      "grad_norm": 0.7890633288543502,
+      "learning_rate": 1.4711000296359523e-06,
+      "loss": 1.0909,
+      "step": 20556
+    },
+    {
+      "epoch": 0.9470216980697471,
+      "grad_norm": 0.7318270745576653,
+      "learning_rate": 1.4685512003167145e-06,
+      "loss": 1.3148,
+      "step": 20557
+    },
+    {
+      "epoch": 0.9470677661583821,
+      "grad_norm": 0.7892931793526599,
+      "learning_rate": 1.4660045646312404e-06,
+      "loss": 1.2338,
+      "step": 20558
+    },
+    {
+      "epoch": 0.9471138342470171,
+      "grad_norm": 1.2598130921056103,
+      "learning_rate": 1.4634601226362287e-06,
+      "loss": 0.9716,
+      "step": 20559
+    },
+    {
+      "epoch": 0.9471599023356521,
+      "grad_norm": 2.3957846751545304,
+      "learning_rate": 1.4609178743883234e-06,
+      "loss": 1.3655,
+      "step": 20560
+    },
+    {
+      "epoch": 0.9472059704242871,
+      "grad_norm": 0.7123238133412108,
+      "learning_rate": 1.4583778199441122e-06,
+      "loss": 1.3243,
+      "step": 20561
+    },
+    {
+      "epoch": 0.9472520385129221,
+      "grad_norm": 0.8300694547272935,
+      "learning_rate": 1.4558399593601723e-06,
+      "loss": 1.065,
+      "step": 20562
+    },
+    {
+      "epoch": 0.9472981066015571,
+      "grad_norm": 1.2910930813101023,
+      "learning_rate": 1.4533042926929808e-06,
+      "loss": 1.0201,
+      "step": 20563
+    },
+    {
+      "epoch": 0.947344174690192,
+      "grad_norm": 0.8655524607847129,
+      "learning_rate": 1.4507708199990032e-06,
+      "loss": 1.2071,
+      "step": 20564
+    },
+    {
+      "epoch": 0.9473902427788271,
+      "grad_norm": 0.8194226728872478,
+      "learning_rate": 1.4482395413346396e-06,
+      "loss": 1.3966,
+      "step": 20565
+    },
+    {
+      "epoch": 0.9474363108674622,
+      "grad_norm": 0.8279402043833377,
+      "learning_rate": 1.445710456756233e-06,
+      "loss": 1.4057,
+      "step": 20566
+    },
+    {
+      "epoch": 0.9474823789560971,
+      "grad_norm": 3.4481798712343865,
+      "learning_rate": 1.4431835663201055e-06,
+      "loss": 1.0739,
+      "step": 20567
+    },
+    {
+      "epoch": 0.9475284470447322,
+      "grad_norm": 1.7641540176054258,
+      "learning_rate": 1.4406588700824897e-06,
+      "loss": 0.9088,
+      "step": 20568
+    },
+    {
+      "epoch": 0.9475745151333671,
+      "grad_norm": 0.8273726033194757,
+      "learning_rate": 1.4381363680996408e-06,
+      "loss": 1.1946,
+      "step": 20569
+    },
+    {
+      "epoch": 0.9476205832220022,
+      "grad_norm": 0.7765058312994679,
+      "learning_rate": 1.4356160604276691e-06,
+      "loss": 1.3613,
+      "step": 20570
+    },
+    {
+      "epoch": 0.9476666513106371,
+      "grad_norm": 0.8580789300789393,
+      "learning_rate": 1.433097947122708e-06,
+      "loss": 1.1251,
+      "step": 20571
+    },
+    {
+      "epoch": 0.9477127193992722,
+      "grad_norm": 0.8142466786559801,
+      "learning_rate": 1.4305820282408122e-06,
+      "loss": 1.2681,
+      "step": 20572
+    },
+    {
+      "epoch": 0.9477587874879071,
+      "grad_norm": 0.8320595200286587,
+      "learning_rate": 1.4280683038380038e-06,
+      "loss": 1.0778,
+      "step": 20573
+    },
+    {
+      "epoch": 0.9478048555765421,
+      "grad_norm": 1.003439224924453,
+      "learning_rate": 1.425556773970238e-06,
+      "loss": 1.2893,
+      "step": 20574
+    },
+    {
+      "epoch": 0.9478509236651771,
+      "grad_norm": 1.0154100119022926,
+      "learning_rate": 1.4230474386934255e-06,
+      "loss": 1.0698,
+      "step": 20575
+    },
+    {
+      "epoch": 0.9478969917538121,
+      "grad_norm": 0.8922380676425489,
+      "learning_rate": 1.420540298063444e-06,
+      "loss": 1.1766,
+      "step": 20576
+    },
+    {
+      "epoch": 0.9479430598424471,
+      "grad_norm": 2.4639766969886057,
+      "learning_rate": 1.4180353521361046e-06,
+      "loss": 1.484,
+      "step": 20577
+    },
+    {
+      "epoch": 0.9479891279310821,
+      "grad_norm": 1.413679036258904,
+      "learning_rate": 1.4155326009671843e-06,
+      "loss": 1.0695,
+      "step": 20578
+    },
+    {
+      "epoch": 0.9480351960197172,
+      "grad_norm": 1.0668169143026769,
+      "learning_rate": 1.4130320446123725e-06,
+      "loss": 1.1303,
+      "step": 20579
+    },
+    {
+      "epoch": 0.9480812641083521,
+      "grad_norm": 1.3203511242248007,
+      "learning_rate": 1.4105336831273796e-06,
+      "loss": 1.0783,
+      "step": 20580
+    },
+    {
+      "epoch": 0.9481273321969872,
+      "grad_norm": 1.325314884095572,
+      "learning_rate": 1.408037516567806e-06,
+      "loss": 1.2468,
+      "step": 20581
+    },
+    {
+      "epoch": 0.9481734002856221,
+      "grad_norm": 1.156619745028682,
+      "learning_rate": 1.4055435449892296e-06,
+      "loss": 0.9631,
+      "step": 20582
+    },
+    {
+      "epoch": 0.9482194683742572,
+      "grad_norm": 0.9212186968437794,
+      "learning_rate": 1.4030517684471723e-06,
+      "loss": 1.1803,
+      "step": 20583
+    },
+    {
+      "epoch": 0.9482655364628921,
+      "grad_norm": 1.003138252949685,
+      "learning_rate": 1.4005621869971007e-06,
+      "loss": 1.1348,
+      "step": 20584
+    },
+    {
+      "epoch": 0.9483116045515272,
+      "grad_norm": 0.7873106206219657,
+      "learning_rate": 1.398074800694471e-06,
+      "loss": 1.2607,
+      "step": 20585
+    },
+    {
+      "epoch": 0.9483576726401621,
+      "grad_norm": 0.8979554600241059,
+      "learning_rate": 1.3955896095946274e-06,
+      "loss": 1.137,
+      "step": 20586
+    },
+    {
+      "epoch": 0.9484037407287972,
+      "grad_norm": 1.244192864388837,
+      "learning_rate": 1.3931066137529037e-06,
+      "loss": 0.9329,
+      "step": 20587
+    },
+    {
+      "epoch": 0.9484498088174321,
+      "grad_norm": 0.8142044256357743,
+      "learning_rate": 1.3906258132245887e-06,
+      "loss": 1.0502,
+      "step": 20588
+    },
+    {
+      "epoch": 0.9484958769060672,
+      "grad_norm": 1.2273175987849494,
+      "learning_rate": 1.3881472080649272e-06,
+      "loss": 0.9022,
+      "step": 20589
+    },
+    {
+      "epoch": 0.9485419449947021,
+      "grad_norm": 1.5556689999263082,
+      "learning_rate": 1.385670798329064e-06,
+      "loss": 1.4361,
+      "step": 20590
+    },
+    {
+      "epoch": 0.9485880130833372,
+      "grad_norm": 3.335228776975115,
+      "learning_rate": 1.3831965840721662e-06,
+      "loss": 1.1132,
+      "step": 20591
+    },
+    {
+      "epoch": 0.9486340811719721,
+      "grad_norm": 0.937805148601518,
+      "learning_rate": 1.3807245653493005e-06,
+      "loss": 0.8051,
+      "step": 20592
+    },
+    {
+      "epoch": 0.9486801492606072,
+      "grad_norm": 0.8856459651966813,
+      "learning_rate": 1.3782547422155122e-06,
+      "loss": 1.4505,
+      "step": 20593
+    },
+    {
+      "epoch": 0.9487262173492422,
+      "grad_norm": 0.9405246834134052,
+      "learning_rate": 1.3757871147257684e-06,
+      "loss": 1.3291,
+      "step": 20594
+    },
+    {
+      "epoch": 0.9487722854378772,
+      "grad_norm": 0.7212422497275891,
+      "learning_rate": 1.373321682935036e-06,
+      "loss": 1.265,
+      "step": 20595
+    },
+    {
+      "epoch": 0.9488183535265122,
+      "grad_norm": 0.9740472609393681,
+      "learning_rate": 1.3708584468981822e-06,
+      "loss": 1.2143,
+      "step": 20596
+    },
+    {
+      "epoch": 0.9488644216151472,
+      "grad_norm": 0.8943897189908375,
+      "learning_rate": 1.3683974066700522e-06,
+      "loss": 1.2081,
+      "step": 20597
+    },
+    {
+      "epoch": 0.9489104897037822,
+      "grad_norm": 0.9527166399326926,
+      "learning_rate": 1.3659385623054355e-06,
+      "loss": 1.3676,
+      "step": 20598
+    },
+    {
+      "epoch": 0.9489565577924172,
+      "grad_norm": 0.9743378355271802,
+      "learning_rate": 1.363481913859066e-06,
+      "loss": 1.2851,
+      "step": 20599
+    },
+    {
+      "epoch": 0.9490026258810522,
+      "grad_norm": 0.7638717744764676,
+      "learning_rate": 1.3610274613856665e-06,
+      "loss": 1.5405,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9490486939696872,
+      "grad_norm": 0.8937195507347065,
+      "learning_rate": 1.3585752049398492e-06,
+      "loss": 1.1338,
+      "step": 20601
+    },
+    {
+      "epoch": 0.9490947620583222,
+      "grad_norm": 2.072279753872068,
+      "learning_rate": 1.3561251445762147e-06,
+      "loss": 1.1194,
+      "step": 20602
+    },
+    {
+      "epoch": 0.9491408301469572,
+      "grad_norm": 1.0821528345451734,
+      "learning_rate": 1.3536772803493304e-06,
+      "loss": 1.3417,
+      "step": 20603
+    },
+    {
+      "epoch": 0.9491868982355922,
+      "grad_norm": 0.8931329138077915,
+      "learning_rate": 1.3512316123136747e-06,
+      "loss": 1.1986,
+      "step": 20604
+    },
+    {
+      "epoch": 0.9492329663242272,
+      "grad_norm": 1.1438737512412673,
+      "learning_rate": 1.3487881405236934e-06,
+      "loss": 1.1798,
+      "step": 20605
+    },
+    {
+      "epoch": 0.9492790344128622,
+      "grad_norm": 0.7992160539577798,
+      "learning_rate": 1.3463468650337985e-06,
+      "loss": 1.1797,
+      "step": 20606
+    },
+    {
+      "epoch": 0.9493251025014973,
+      "grad_norm": 1.0200112248298503,
+      "learning_rate": 1.3439077858983461e-06,
+      "loss": 1.3624,
+      "step": 20607
+    },
+    {
+      "epoch": 0.9493711705901322,
+      "grad_norm": 2.2304180105964044,
+      "learning_rate": 1.3414709031716154e-06,
+      "loss": 1.5438,
+      "step": 20608
+    },
+    {
+      "epoch": 0.9494172386787673,
+      "grad_norm": 1.0303932148356194,
+      "learning_rate": 1.3390362169078741e-06,
+      "loss": 1.2549,
+      "step": 20609
+    },
+    {
+      "epoch": 0.9494633067674022,
+      "grad_norm": 1.2944760180424586,
+      "learning_rate": 1.336603727161323e-06,
+      "loss": 1.1279,
+      "step": 20610
+    },
+    {
+      "epoch": 0.9495093748560373,
+      "grad_norm": 1.8560050759788889,
+      "learning_rate": 1.3341734339861189e-06,
+      "loss": 1.5692,
+      "step": 20611
+    },
+    {
+      "epoch": 0.9495554429446722,
+      "grad_norm": 0.7321952899853399,
+      "learning_rate": 1.3317453374363741e-06,
+      "loss": 1.1881,
+      "step": 20612
+    },
+    {
+      "epoch": 0.9496015110333073,
+      "grad_norm": 1.0445291801141379,
+      "learning_rate": 1.3293194375661344e-06,
+      "loss": 1.0424,
+      "step": 20613
+    },
+    {
+      "epoch": 0.9496475791219422,
+      "grad_norm": 1.101238664116351,
+      "learning_rate": 1.326895734429412e-06,
+      "loss": 1.2327,
+      "step": 20614
+    },
+    {
+      "epoch": 0.9496936472105773,
+      "grad_norm": 5.716066352788483,
+      "learning_rate": 1.3244742280801747e-06,
+      "loss": 1.1547,
+      "step": 20615
+    },
+    {
+      "epoch": 0.9497397152992122,
+      "grad_norm": 1.0965758754428718,
+      "learning_rate": 1.322054918572313e-06,
+      "loss": 1.1006,
+      "step": 20616
+    },
+    {
+      "epoch": 0.9497857833878472,
+      "grad_norm": 0.7802993065992252,
+      "learning_rate": 1.319637805959706e-06,
+      "loss": 1.0179,
+      "step": 20617
+    },
+    {
+      "epoch": 0.9498318514764822,
+      "grad_norm": 1.2003101649660581,
+      "learning_rate": 1.317222890296177e-06,
+      "loss": 1.2647,
+      "step": 20618
+    },
+    {
+      "epoch": 0.9498779195651172,
+      "grad_norm": 0.9723936524842899,
+      "learning_rate": 1.314810171635461e-06,
+      "loss": 1.1645,
+      "step": 20619
+    },
+    {
+      "epoch": 0.9499239876537523,
+      "grad_norm": 1.095777358370352,
+      "learning_rate": 1.312399650031293e-06,
+      "loss": 1.1449,
+      "step": 20620
+    },
+    {
+      "epoch": 0.9499700557423872,
+      "grad_norm": 1.2078628268505427,
+      "learning_rate": 1.309991325537341e-06,
+      "loss": 1.2306,
+      "step": 20621
+    },
+    {
+      "epoch": 0.9500161238310223,
+      "grad_norm": 1.0721312958306066,
+      "learning_rate": 1.3075851982072064e-06,
+      "loss": 1.2522,
+      "step": 20622
+    },
+    {
+      "epoch": 0.9500621919196572,
+      "grad_norm": 0.7486736590011519,
+      "learning_rate": 1.3051812680944576e-06,
+      "loss": 1.3966,
+      "step": 20623
+    },
+    {
+      "epoch": 0.9501082600082923,
+      "grad_norm": 1.3562009229756888,
+      "learning_rate": 1.3027795352526405e-06,
+      "loss": 1.4429,
+      "step": 20624
+    },
+    {
+      "epoch": 0.9501543280969272,
+      "grad_norm": 1.142308472284249,
+      "learning_rate": 1.3003799997352018e-06,
+      "loss": 1.0986,
+      "step": 20625
+    },
+    {
+      "epoch": 0.9502003961855623,
+      "grad_norm": 0.7911253073365474,
+      "learning_rate": 1.2979826615955647e-06,
+      "loss": 1.5784,
+      "step": 20626
+    },
+    {
+      "epoch": 0.9502464642741972,
+      "grad_norm": 1.006490203366234,
+      "learning_rate": 1.2955875208870982e-06,
+      "loss": 1.1925,
+      "step": 20627
+    },
+    {
+      "epoch": 0.9502925323628323,
+      "grad_norm": 0.778338575899609,
+      "learning_rate": 1.2931945776631593e-06,
+      "loss": 1.1239,
+      "step": 20628
+    },
+    {
+      "epoch": 0.9503386004514672,
+      "grad_norm": 1.04204428729211,
+      "learning_rate": 1.2908038319769723e-06,
+      "loss": 1.3218,
+      "step": 20629
+    },
+    {
+      "epoch": 0.9503846685401023,
+      "grad_norm": 0.9379180943530011,
+      "learning_rate": 1.2884152838817942e-06,
+      "loss": 1.1191,
+      "step": 20630
+    },
+    {
+      "epoch": 0.9504307366287372,
+      "grad_norm": 0.835887923284965,
+      "learning_rate": 1.286028933430794e-06,
+      "loss": 1.3064,
+      "step": 20631
+    },
+    {
+      "epoch": 0.9504768047173723,
+      "grad_norm": 1.3851878741160282,
+      "learning_rate": 1.2836447806771178e-06,
+      "loss": 1.3436,
+      "step": 20632
+    },
+    {
+      "epoch": 0.9505228728060072,
+      "grad_norm": 0.9302040034387061,
+      "learning_rate": 1.2812628256738124e-06,
+      "loss": 0.9442,
+      "step": 20633
+    },
+    {
+      "epoch": 0.9505689408946423,
+      "grad_norm": 0.8711849031847536,
+      "learning_rate": 1.278883068473924e-06,
+      "loss": 1.3925,
+      "step": 20634
+    },
+    {
+      "epoch": 0.9506150089832773,
+      "grad_norm": 0.9016676354946576,
+      "learning_rate": 1.2765055091304434e-06,
+      "loss": 1.1293,
+      "step": 20635
+    },
+    {
+      "epoch": 0.9506610770719123,
+      "grad_norm": 0.9323233457436768,
+      "learning_rate": 1.2741301476962842e-06,
+      "loss": 1.2903,
+      "step": 20636
+    },
+    {
+      "epoch": 0.9507071451605473,
+      "grad_norm": 0.8980964263553394,
+      "learning_rate": 1.2717569842243482e-06,
+      "loss": 1.3929,
+      "step": 20637
+    },
+    {
+      "epoch": 0.9507532132491823,
+      "grad_norm": 0.8463910101144368,
+      "learning_rate": 1.2693860187674489e-06,
+      "loss": 1.3826,
+      "step": 20638
+    },
+    {
+      "epoch": 0.9507992813378173,
+      "grad_norm": 0.9220153518580276,
+      "learning_rate": 1.2670172513783884e-06,
+      "loss": 1.0543,
+      "step": 20639
+    },
+    {
+      "epoch": 0.9508453494264523,
+      "grad_norm": 1.4338345295984432,
+      "learning_rate": 1.2646506821099025e-06,
+      "loss": 0.9618,
+      "step": 20640
+    },
+    {
+      "epoch": 0.9508914175150873,
+      "grad_norm": 1.2571172672243092,
+      "learning_rate": 1.2622863110146598e-06,
+      "loss": 1.2315,
+      "step": 20641
+    },
+    {
+      "epoch": 0.9509374856037223,
+      "grad_norm": 0.642598167482563,
+      "learning_rate": 1.2599241381453298e-06,
+      "loss": 1.2497,
+      "step": 20642
+    },
+    {
+      "epoch": 0.9509835536923573,
+      "grad_norm": 1.2375462493904648,
+      "learning_rate": 1.257564163554481e-06,
+      "loss": 1.1754,
+      "step": 20643
+    },
+    {
+      "epoch": 0.9510296217809923,
+      "grad_norm": 0.9443461240093598,
+      "learning_rate": 1.2552063872946496e-06,
+      "loss": 1.3769,
+      "step": 20644
+    },
+    {
+      "epoch": 0.9510756898696273,
+      "grad_norm": 2.7430019585606757,
+      "learning_rate": 1.2528508094183488e-06,
+      "loss": 1.2608,
+      "step": 20645
+    },
+    {
+      "epoch": 0.9511217579582623,
+      "grad_norm": 2.0689166233127425,
+      "learning_rate": 1.2504974299780148e-06,
+      "loss": 0.9682,
+      "step": 20646
+    },
+    {
+      "epoch": 0.9511678260468973,
+      "grad_norm": 0.9005209248250362,
+      "learning_rate": 1.2481462490260165e-06,
+      "loss": 0.9172,
+      "step": 20647
+    },
+    {
+      "epoch": 0.9512138941355324,
+      "grad_norm": 0.9140332996924211,
+      "learning_rate": 1.2457972666147344e-06,
+      "loss": 1.4801,
+      "step": 20648
+    },
+    {
+      "epoch": 0.9512599622241673,
+      "grad_norm": 1.5379032566293007,
+      "learning_rate": 1.2434504827964377e-06,
+      "loss": 1.316,
+      "step": 20649
+    },
+    {
+      "epoch": 0.9513060303128024,
+      "grad_norm": 0.8894852565454164,
+      "learning_rate": 1.2411058976233959e-06,
+      "loss": 1.5448,
+      "step": 20650
+    },
+    {
+      "epoch": 0.9513520984014373,
+      "grad_norm": 1.4278995513888189,
+      "learning_rate": 1.238763511147789e-06,
+      "loss": 1.1396,
+      "step": 20651
+    },
+    {
+      "epoch": 0.9513981664900724,
+      "grad_norm": 1.2261652970661636,
+      "learning_rate": 1.236423323421776e-06,
+      "loss": 1.2582,
+      "step": 20652
+    },
+    {
+      "epoch": 0.9514442345787073,
+      "grad_norm": 1.0379511247250957,
+      "learning_rate": 1.234085334497459e-06,
+      "loss": 1.0512,
+      "step": 20653
+    },
+    {
+      "epoch": 0.9514903026673424,
+      "grad_norm": 0.7682227321364369,
+      "learning_rate": 1.2317495444268857e-06,
+      "loss": 1.1552,
+      "step": 20654
+    },
+    {
+      "epoch": 0.9515363707559773,
+      "grad_norm": 0.837228406897117,
+      "learning_rate": 1.2294159532620586e-06,
+      "loss": 1.5425,
+      "step": 20655
+    },
+    {
+      "epoch": 0.9515824388446124,
+      "grad_norm": 1.2247786535377914,
+      "learning_rate": 1.2270845610549143e-06,
+      "loss": 1.0499,
+      "step": 20656
+    },
+    {
+      "epoch": 0.9516285069332473,
+      "grad_norm": 0.815766217964573,
+      "learning_rate": 1.2247553678573776e-06,
+      "loss": 1.3951,
+      "step": 20657
+    },
+    {
+      "epoch": 0.9516745750218824,
+      "grad_norm": 1.3805394491110465,
+      "learning_rate": 1.222428373721307e-06,
+      "loss": 1.3549,
+      "step": 20658
+    },
+    {
+      "epoch": 0.9517206431105173,
+      "grad_norm": 1.9020304978482416,
+      "learning_rate": 1.220103578698506e-06,
+      "loss": 1.3332,
+      "step": 20659
+    },
+    {
+      "epoch": 0.9517667111991523,
+      "grad_norm": 1.0853855075523469,
+      "learning_rate": 1.2177809828407105e-06,
+      "loss": 1.1457,
+      "step": 20660
+    },
+    {
+      "epoch": 0.9518127792877874,
+      "grad_norm": 1.0644463089401008,
+      "learning_rate": 1.2154605861996571e-06,
+      "loss": 0.9779,
+      "step": 20661
+    },
+    {
+      "epoch": 0.9518588473764223,
+      "grad_norm": 1.0459773352538861,
+      "learning_rate": 1.2131423888269933e-06,
+      "loss": 1.2927,
+      "step": 20662
+    },
+    {
+      "epoch": 0.9519049154650574,
+      "grad_norm": 0.8465224661086407,
+      "learning_rate": 1.2108263907743333e-06,
+      "loss": 1.3938,
+      "step": 20663
+    },
+    {
+      "epoch": 0.9519509835536923,
+      "grad_norm": 1.4232237197042716,
+      "learning_rate": 1.208512592093236e-06,
+      "loss": 1.3776,
+      "step": 20664
+    },
+    {
+      "epoch": 0.9519970516423274,
+      "grad_norm": 0.9413806421168784,
+      "learning_rate": 1.2062009928352158e-06,
+      "loss": 0.9983,
+      "step": 20665
+    },
+    {
+      "epoch": 0.9520431197309623,
+      "grad_norm": 0.9892927097644785,
+      "learning_rate": 1.2038915930517314e-06,
+      "loss": 1.504,
+      "step": 20666
+    },
+    {
+      "epoch": 0.9520891878195974,
+      "grad_norm": 0.813163154834512,
+      "learning_rate": 1.2015843927941971e-06,
+      "loss": 0.9973,
+      "step": 20667
+    },
+    {
+      "epoch": 0.9521352559082323,
+      "grad_norm": 1.083272483380212,
+      "learning_rate": 1.1992793921139944e-06,
+      "loss": 1.5602,
+      "step": 20668
+    },
+    {
+      "epoch": 0.9521813239968674,
+      "grad_norm": 0.9490941532155919,
+      "learning_rate": 1.1969765910624264e-06,
+      "loss": 0.8569,
+      "step": 20669
+    },
+    {
+      "epoch": 0.9522273920855023,
+      "grad_norm": 0.7779019257333143,
+      "learning_rate": 1.1946759896907634e-06,
+      "loss": 1.1955,
+      "step": 20670
+    },
+    {
+      "epoch": 0.9522734601741374,
+      "grad_norm": 1.3406967367767395,
+      "learning_rate": 1.1923775880502085e-06,
+      "loss": 1.5515,
+      "step": 20671
+    },
+    {
+      "epoch": 0.9523195282627723,
+      "grad_norm": 0.920084841359822,
+      "learning_rate": 1.190081386191977e-06,
+      "loss": 1.34,
+      "step": 20672
+    },
+    {
+      "epoch": 0.9523655963514074,
+      "grad_norm": 0.8339256422588434,
+      "learning_rate": 1.1877873841671383e-06,
+      "loss": 1.1297,
+      "step": 20673
+    },
+    {
+      "epoch": 0.9524116644400424,
+      "grad_norm": 2.0941205997274968,
+      "learning_rate": 1.1854955820267965e-06,
+      "loss": 1.0391,
+      "step": 20674
+    },
+    {
+      "epoch": 0.9524577325286774,
+      "grad_norm": 0.7291491808467113,
+      "learning_rate": 1.1832059798219664e-06,
+      "loss": 1.5435,
+      "step": 20675
+    },
+    {
+      "epoch": 0.9525038006173124,
+      "grad_norm": 0.7548481268822865,
+      "learning_rate": 1.180918577603618e-06,
+      "loss": 1.2549,
+      "step": 20676
+    },
+    {
+      "epoch": 0.9525498687059474,
+      "grad_norm": 0.7820233367802026,
+      "learning_rate": 1.1786333754226776e-06,
+      "loss": 1.3455,
+      "step": 20677
+    },
+    {
+      "epoch": 0.9525959367945824,
+      "grad_norm": 1.4009885271016622,
+      "learning_rate": 1.1763503733300152e-06,
+      "loss": 1.2595,
+      "step": 20678
+    },
+    {
+      "epoch": 0.9526420048832174,
+      "grad_norm": 1.439845693852182,
+      "learning_rate": 1.1740695713764794e-06,
+      "loss": 1.4126,
+      "step": 20679
+    },
+    {
+      "epoch": 0.9526880729718524,
+      "grad_norm": 1.7861410702650935,
+      "learning_rate": 1.1717909696128293e-06,
+      "loss": 1.1464,
+      "step": 20680
+    },
+    {
+      "epoch": 0.9527341410604874,
+      "grad_norm": 0.8029992262841563,
+      "learning_rate": 1.1695145680898023e-06,
+      "loss": 1.3772,
+      "step": 20681
+    },
+    {
+      "epoch": 0.9527802091491224,
+      "grad_norm": 1.1681653182806253,
+      "learning_rate": 1.1672403668580689e-06,
+      "loss": 1.3032,
+      "step": 20682
+    },
+    {
+      "epoch": 0.9528262772377574,
+      "grad_norm": 1.0754288434186114,
+      "learning_rate": 1.1649683659682664e-06,
+      "loss": 1.0882,
+      "step": 20683
+    },
+    {
+      "epoch": 0.9528723453263924,
+      "grad_norm": 1.4985435646911274,
+      "learning_rate": 1.1626985654709877e-06,
+      "loss": 1.1518,
+      "step": 20684
+    },
+    {
+      "epoch": 0.9529184134150274,
+      "grad_norm": 0.8515615159006036,
+      "learning_rate": 1.1604309654167478e-06,
+      "loss": 1.1491,
+      "step": 20685
+    },
+    {
+      "epoch": 0.9529644815036624,
+      "grad_norm": 1.1748688078716687,
+      "learning_rate": 1.1581655658560398e-06,
+      "loss": 0.9879,
+      "step": 20686
+    },
+    {
+      "epoch": 0.9530105495922974,
+      "grad_norm": 0.8105388535931284,
+      "learning_rate": 1.155902366839301e-06,
+      "loss": 1.1713,
+      "step": 20687
+    },
+    {
+      "epoch": 0.9530566176809324,
+      "grad_norm": 0.921934466308288,
+      "learning_rate": 1.1536413684169134e-06,
+      "loss": 1.571,
+      "step": 20688
+    },
+    {
+      "epoch": 0.9531026857695675,
+      "grad_norm": 5.971258737149441,
+      "learning_rate": 1.1513825706392145e-06,
+      "loss": 1.0183,
+      "step": 20689
+    },
+    {
+      "epoch": 0.9531487538582024,
+      "grad_norm": 1.0003298877723934,
+      "learning_rate": 1.1491259735564863e-06,
+      "loss": 1.3939,
+      "step": 20690
+    },
+    {
+      "epoch": 0.9531948219468375,
+      "grad_norm": 1.4519652904206226,
+      "learning_rate": 1.1468715772189886e-06,
+      "loss": 1.2638,
+      "step": 20691
+    },
+    {
+      "epoch": 0.9532408900354724,
+      "grad_norm": 0.9314683745864617,
+      "learning_rate": 1.1446193816768923e-06,
+      "loss": 1.3055,
+      "step": 20692
+    },
+    {
+      "epoch": 0.9532869581241075,
+      "grad_norm": 0.6940016880993577,
+      "learning_rate": 1.1423693869803464e-06,
+      "loss": 1.3289,
+      "step": 20693
+    },
+    {
+      "epoch": 0.9533330262127424,
+      "grad_norm": 1.9253341287805679,
+      "learning_rate": 1.140121593179433e-06,
+      "loss": 1.0851,
+      "step": 20694
+    },
+    {
+      "epoch": 0.9533790943013775,
+      "grad_norm": 0.9701055220952275,
+      "learning_rate": 1.1378760003242227e-06,
+      "loss": 1.0566,
+      "step": 20695
+    },
+    {
+      "epoch": 0.9534251623900124,
+      "grad_norm": 0.8010419675439026,
+      "learning_rate": 1.1356326084646762e-06,
+      "loss": 1.2624,
+      "step": 20696
+    },
+    {
+      "epoch": 0.9534712304786475,
+      "grad_norm": 1.0392773408575717,
+      "learning_rate": 1.133391417650753e-06,
+      "loss": 1.201,
+      "step": 20697
+    },
+    {
+      "epoch": 0.9535172985672824,
+      "grad_norm": 0.8382323307029544,
+      "learning_rate": 1.1311524279323582e-06,
+      "loss": 1.0889,
+      "step": 20698
+    },
+    {
+      "epoch": 0.9535633666559175,
+      "grad_norm": 0.8302616995643859,
+      "learning_rate": 1.1289156393593291e-06,
+      "loss": 1.3382,
+      "step": 20699
+    },
+    {
+      "epoch": 0.9536094347445524,
+      "grad_norm": 2.8701014784099317,
+      "learning_rate": 1.1266810519814598e-06,
+      "loss": 1.4774,
+      "step": 20700
+    },
+    {
+      "epoch": 0.9536555028331875,
+      "grad_norm": 0.9352237342326892,
+      "learning_rate": 1.1244486658485098e-06,
+      "loss": 1.2126,
+      "step": 20701
+    },
+    {
+      "epoch": 0.9537015709218225,
+      "grad_norm": 0.7569872874179904,
+      "learning_rate": 1.1222184810101733e-06,
+      "loss": 0.893,
+      "step": 20702
+    },
+    {
+      "epoch": 0.9537476390104574,
+      "grad_norm": 1.245224164770059,
+      "learning_rate": 1.1199904975161101e-06,
+      "loss": 1.0502,
+      "step": 20703
+    },
+    {
+      "epoch": 0.9537937070990925,
+      "grad_norm": 3.2293827027267277,
+      "learning_rate": 1.1177647154159033e-06,
+      "loss": 1.3472,
+      "step": 20704
+    },
+    {
+      "epoch": 0.9538397751877274,
+      "grad_norm": 1.6112068073855215,
+      "learning_rate": 1.1155411347591237e-06,
+      "loss": 1.2334,
+      "step": 20705
+    },
+    {
+      "epoch": 0.9538858432763625,
+      "grad_norm": 0.8017736286341361,
+      "learning_rate": 1.1133197555952767e-06,
+      "loss": 1.5609,
+      "step": 20706
+    },
+    {
+      "epoch": 0.9539319113649974,
+      "grad_norm": 0.7377249351705725,
+      "learning_rate": 1.1111005779738004e-06,
+      "loss": 1.3189,
+      "step": 20707
+    },
+    {
+      "epoch": 0.9539779794536325,
+      "grad_norm": 1.1907745597865085,
+      "learning_rate": 1.1088836019441218e-06,
+      "loss": 1.3005,
+      "step": 20708
+    },
+    {
+      "epoch": 0.9540240475422674,
+      "grad_norm": 1.0371937482867486,
+      "learning_rate": 1.1066688275555793e-06,
+      "loss": 0.948,
+      "step": 20709
+    },
+    {
+      "epoch": 0.9540701156309025,
+      "grad_norm": 0.9868507977436374,
+      "learning_rate": 1.1044562548574998e-06,
+      "loss": 1.289,
+      "step": 20710
+    },
+    {
+      "epoch": 0.9541161837195374,
+      "grad_norm": 2.711715797769318,
+      "learning_rate": 1.1022458838991224e-06,
+      "loss": 1.0227,
+      "step": 20711
+    },
+    {
+      "epoch": 0.9541622518081725,
+      "grad_norm": 0.8296770971102785,
+      "learning_rate": 1.1000377147296736e-06,
+      "loss": 1.3679,
+      "step": 20712
+    },
+    {
+      "epoch": 0.9542083198968074,
+      "grad_norm": 1.5819701796558587,
+      "learning_rate": 1.0978317473983034e-06,
+      "loss": 1.0369,
+      "step": 20713
+    },
+    {
+      "epoch": 0.9542543879854425,
+      "grad_norm": 1.0869087773269421,
+      "learning_rate": 1.095627981954128e-06,
+      "loss": 1.1448,
+      "step": 20714
+    },
+    {
+      "epoch": 0.9543004560740775,
+      "grad_norm": 1.196792579575751,
+      "learning_rate": 1.0934264184462196e-06,
+      "loss": 0.9276,
+      "step": 20715
+    },
+    {
+      "epoch": 0.9543465241627125,
+      "grad_norm": 1.4996321400590278,
+      "learning_rate": 1.0912270569235716e-06,
+      "loss": 0.9689,
+      "step": 20716
+    },
+    {
+      "epoch": 0.9543925922513475,
+      "grad_norm": 0.8120063782852578,
+      "learning_rate": 1.0890298974351676e-06,
+      "loss": 1.2056,
+      "step": 20717
+    },
+    {
+      "epoch": 0.9544386603399825,
+      "grad_norm": 0.8050545673114602,
+      "learning_rate": 1.0868349400299128e-06,
+      "loss": 1.2291,
+      "step": 20718
+    },
+    {
+      "epoch": 0.9544847284286175,
+      "grad_norm": 1.4183291612452407,
+      "learning_rate": 1.084642184756679e-06,
+      "loss": 1.0888,
+      "step": 20719
+    },
+    {
+      "epoch": 0.9545307965172525,
+      "grad_norm": 1.0233736372417743,
+      "learning_rate": 1.082451631664294e-06,
+      "loss": 1.0372,
+      "step": 20720
+    },
+    {
+      "epoch": 0.9545768646058875,
+      "grad_norm": 1.6093985198738443,
+      "learning_rate": 1.080263280801508e-06,
+      "loss": 1.0155,
+      "step": 20721
+    },
+    {
+      "epoch": 0.9546229326945225,
+      "grad_norm": 0.8424833546326361,
+      "learning_rate": 1.078077132217048e-06,
+      "loss": 1.1858,
+      "step": 20722
+    },
+    {
+      "epoch": 0.9546690007831575,
+      "grad_norm": 0.7618592663448961,
+      "learning_rate": 1.0758931859595866e-06,
+      "loss": 1.2365,
+      "step": 20723
+    },
+    {
+      "epoch": 0.9547150688717925,
+      "grad_norm": 1.525575465433601,
+      "learning_rate": 1.0737114420777517e-06,
+      "loss": 1.2614,
+      "step": 20724
+    },
+    {
+      "epoch": 0.9547611369604275,
+      "grad_norm": 3.3166104306551003,
+      "learning_rate": 1.071531900620093e-06,
+      "loss": 1.0207,
+      "step": 20725
+    },
+    {
+      "epoch": 0.9548072050490625,
+      "grad_norm": 1.0937890183821892,
+      "learning_rate": 1.069354561635161e-06,
+      "loss": 1.0425,
+      "step": 20726
+    },
+    {
+      "epoch": 0.9548532731376975,
+      "grad_norm": 0.8490515817407707,
+      "learning_rate": 1.0671794251714163e-06,
+      "loss": 0.8262,
+      "step": 20727
+    },
+    {
+      "epoch": 0.9548993412263325,
+      "grad_norm": 0.9584950861440101,
+      "learning_rate": 1.0650064912772984e-06,
+      "loss": 1.2132,
+      "step": 20728
+    },
+    {
+      "epoch": 0.9549454093149675,
+      "grad_norm": 1.084671444726603,
+      "learning_rate": 1.0628357600011684e-06,
+      "loss": 1.0414,
+      "step": 20729
+    },
+    {
+      "epoch": 0.9549914774036026,
+      "grad_norm": 1.4144958210509324,
+      "learning_rate": 1.0606672313913545e-06,
+      "loss": 1.1117,
+      "step": 20730
+    },
+    {
+      "epoch": 0.9550375454922375,
+      "grad_norm": 1.3701129601704762,
+      "learning_rate": 1.0585009054961514e-06,
+      "loss": 0.9683,
+      "step": 20731
+    },
+    {
+      "epoch": 0.9550836135808726,
+      "grad_norm": 0.8936189211508259,
+      "learning_rate": 1.0563367823637648e-06,
+      "loss": 1.0215,
+      "step": 20732
+    },
+    {
+      "epoch": 0.9551296816695075,
+      "grad_norm": 1.6450529063369808,
+      "learning_rate": 1.0541748620423896e-06,
+      "loss": 1.2623,
+      "step": 20733
+    },
+    {
+      "epoch": 0.9551757497581426,
+      "grad_norm": 1.782822562462504,
+      "learning_rate": 1.0520151445801652e-06,
+      "loss": 1.355,
+      "step": 20734
+    },
+    {
+      "epoch": 0.9552218178467775,
+      "grad_norm": 0.910109886475648,
+      "learning_rate": 1.0498576300251528e-06,
+      "loss": 1.296,
+      "step": 20735
+    },
+    {
+      "epoch": 0.9552678859354126,
+      "grad_norm": 1.1400506115406033,
+      "learning_rate": 1.0477023184254031e-06,
+      "loss": 1.0924,
+      "step": 20736
+    },
+    {
+      "epoch": 0.9553139540240475,
+      "grad_norm": 0.9036199140602988,
+      "learning_rate": 1.0455492098288887e-06,
+      "loss": 1.1632,
+      "step": 20737
+    },
+    {
+      "epoch": 0.9553600221126826,
+      "grad_norm": 0.7931598384325144,
+      "learning_rate": 1.0433983042835604e-06,
+      "loss": 1.0003,
+      "step": 20738
+    },
+    {
+      "epoch": 0.9554060902013175,
+      "grad_norm": 1.049344991777815,
+      "learning_rate": 1.0412496018372907e-06,
+      "loss": 1.6062,
+      "step": 20739
+    },
+    {
+      "epoch": 0.9554521582899526,
+      "grad_norm": 1.6150234235615621,
+      "learning_rate": 1.0391031025379195e-06,
+      "loss": 1.2944,
+      "step": 20740
+    },
+    {
+      "epoch": 0.9554982263785875,
+      "grad_norm": 1.073473569519569,
+      "learning_rate": 1.0369588064332303e-06,
+      "loss": 1.3796,
+      "step": 20741
+    },
+    {
+      "epoch": 0.9555442944672226,
+      "grad_norm": 2.206382443240987,
+      "learning_rate": 1.0348167135709851e-06,
+      "loss": 1.0553,
+      "step": 20742
+    },
+    {
+      "epoch": 0.9555903625558576,
+      "grad_norm": 1.1390716577625177,
+      "learning_rate": 1.032676823998835e-06,
+      "loss": 1.2949,
+      "step": 20743
+    },
+    {
+      "epoch": 0.9556364306444926,
+      "grad_norm": 0.9284109354954119,
+      "learning_rate": 1.0305391377644524e-06,
+      "loss": 1.3885,
+      "step": 20744
+    },
+    {
+      "epoch": 0.9556824987331276,
+      "grad_norm": 1.0003555080904656,
+      "learning_rate": 1.0284036549154218e-06,
+      "loss": 1.243,
+      "step": 20745
+    },
+    {
+      "epoch": 0.9557285668217625,
+      "grad_norm": 0.8862727586957172,
+      "learning_rate": 1.026270375499283e-06,
+      "loss": 0.9677,
+      "step": 20746
+    },
+    {
+      "epoch": 0.9557746349103976,
+      "grad_norm": 1.0229304420008054,
+      "learning_rate": 1.0241392995635425e-06,
+      "loss": 0.8832,
+      "step": 20747
+    },
+    {
+      "epoch": 0.9558207029990325,
+      "grad_norm": 0.8385521321146913,
+      "learning_rate": 1.0220104271556065e-06,
+      "loss": 1.3283,
+      "step": 20748
+    },
+    {
+      "epoch": 0.9558667710876676,
+      "grad_norm": 1.0061402637304253,
+      "learning_rate": 1.019883758322926e-06,
+      "loss": 1.2103,
+      "step": 20749
+    },
+    {
+      "epoch": 0.9559128391763025,
+      "grad_norm": 1.8088574176435122,
+      "learning_rate": 1.0177592931127967e-06,
+      "loss": 1.2435,
+      "step": 20750
+    },
+    {
+      "epoch": 0.9559589072649376,
+      "grad_norm": 0.7247468563013848,
+      "learning_rate": 1.015637031572547e-06,
+      "loss": 0.9595,
+      "step": 20751
+    },
+    {
+      "epoch": 0.9560049753535725,
+      "grad_norm": 0.7679913382501443,
+      "learning_rate": 1.0135169737494177e-06,
+      "loss": 1.0612,
+      "step": 20752
+    },
+    {
+      "epoch": 0.9560510434422076,
+      "grad_norm": 1.737736095425739,
+      "learning_rate": 1.0113991196906148e-06,
+      "loss": 0.8458,
+      "step": 20753
+    },
+    {
+      "epoch": 0.9560971115308425,
+      "grad_norm": 1.0667464034880318,
+      "learning_rate": 1.0092834694432672e-06,
+      "loss": 1.4071,
+      "step": 20754
+    },
+    {
+      "epoch": 0.9561431796194776,
+      "grad_norm": 0.7865812206508831,
+      "learning_rate": 1.0071700230544933e-06,
+      "loss": 1.2145,
+      "step": 20755
+    },
+    {
+      "epoch": 0.9561892477081126,
+      "grad_norm": 1.1368786351983435,
+      "learning_rate": 1.0050587805713552e-06,
+      "loss": 1.0105,
+      "step": 20756
+    },
+    {
+      "epoch": 0.9562353157967476,
+      "grad_norm": 0.7847545282160049,
+      "learning_rate": 1.0029497420408263e-06,
+      "loss": 1.2863,
+      "step": 20757
+    },
+    {
+      "epoch": 0.9562813838853826,
+      "grad_norm": 1.2600186346860542,
+      "learning_rate": 1.0008429075098913e-06,
+      "loss": 1.2248,
+      "step": 20758
+    },
+    {
+      "epoch": 0.9563274519740176,
+      "grad_norm": 1.1674888646393564,
+      "learning_rate": 9.98738277025435e-07,
+      "loss": 1.2142,
+      "step": 20759
+    },
+    {
+      "epoch": 0.9563735200626526,
+      "grad_norm": 0.8695029505527038,
+      "learning_rate": 9.966358506343198e-07,
+      "loss": 1.2555,
+      "step": 20760
+    },
+    {
+      "epoch": 0.9564195881512876,
+      "grad_norm": 1.2740333282840133,
+      "learning_rate": 9.94535628383353e-07,
+      "loss": 1.3653,
+      "step": 20761
+    },
+    {
+      "epoch": 0.9564656562399226,
+      "grad_norm": 1.1020616116076234,
+      "learning_rate": 9.924376103193078e-07,
+      "loss": 1.094,
+      "step": 20762
+    },
+    {
+      "epoch": 0.9565117243285576,
+      "grad_norm": 1.513547577182085,
+      "learning_rate": 9.903417964888473e-07,
+      "loss": 1.2041,
+      "step": 20763
+    },
+    {
+      "epoch": 0.9565577924171926,
+      "grad_norm": 1.978739035639797,
+      "learning_rate": 9.882481869386895e-07,
+      "loss": 1.0842,
+      "step": 20764
+    },
+    {
+      "epoch": 0.9566038605058276,
+      "grad_norm": 1.397497289609534,
+      "learning_rate": 9.861567817154083e-07,
+      "loss": 0.8811,
+      "step": 20765
+    },
+    {
+      "epoch": 0.9566499285944626,
+      "grad_norm": 0.8563302725625089,
+      "learning_rate": 9.840675808655664e-07,
+      "loss": 1.2072,
+      "step": 20766
+    },
+    {
+      "epoch": 0.9566959966830976,
+      "grad_norm": 0.9720360682950826,
+      "learning_rate": 9.819805844356934e-07,
+      "loss": 0.9766,
+      "step": 20767
+    },
+    {
+      "epoch": 0.9567420647717326,
+      "grad_norm": 0.884840082081356,
+      "learning_rate": 9.798957924722407e-07,
+      "loss": 1.198,
+      "step": 20768
+    },
+    {
+      "epoch": 0.9567881328603676,
+      "grad_norm": 0.9412397008833282,
+      "learning_rate": 9.77813205021627e-07,
+      "loss": 1.0771,
+      "step": 20769
+    },
+    {
+      "epoch": 0.9568342009490026,
+      "grad_norm": 0.6341766782087197,
+      "learning_rate": 9.757328221302043e-07,
+      "loss": 1.5125,
+      "step": 20770
+    },
+    {
+      "epoch": 0.9568802690376377,
+      "grad_norm": 1.1413085627952144,
+      "learning_rate": 9.73654643844313e-07,
+      "loss": 0.9887,
+      "step": 20771
+    },
+    {
+      "epoch": 0.9569263371262726,
+      "grad_norm": 1.218037843131313,
+      "learning_rate": 9.715786702101936e-07,
+      "loss": 1.2325,
+      "step": 20772
+    },
+    {
+      "epoch": 0.9569724052149077,
+      "grad_norm": 1.6263064803433411,
+      "learning_rate": 9.695049012740987e-07,
+      "loss": 0.9172,
+      "step": 20773
+    },
+    {
+      "epoch": 0.9570184733035426,
+      "grad_norm": 2.525693169858531,
+      "learning_rate": 9.674333370821575e-07,
+      "loss": 1.2834,
+      "step": 20774
+    },
+    {
+      "epoch": 0.9570645413921777,
+      "grad_norm": 1.7425838804565656,
+      "learning_rate": 9.653639776805113e-07,
+      "loss": 0.9058,
+      "step": 20775
+    },
+    {
+      "epoch": 0.9571106094808126,
+      "grad_norm": 0.7366495277066631,
+      "learning_rate": 9.63296823115234e-07,
+      "loss": 1.1275,
+      "step": 20776
+    },
+    {
+      "epoch": 0.9571566775694477,
+      "grad_norm": 1.3564035697048427,
+      "learning_rate": 9.612318734323333e-07,
+      "loss": 1.2715,
+      "step": 20777
+    },
+    {
+      "epoch": 0.9572027456580826,
+      "grad_norm": 2.0238063382997686,
+      "learning_rate": 9.59169128677806e-07,
+      "loss": 1.005,
+      "step": 20778
+    },
+    {
+      "epoch": 0.9572488137467177,
+      "grad_norm": 0.8822891954391842,
+      "learning_rate": 9.571085888975484e-07,
+      "loss": 1.0975,
+      "step": 20779
+    },
+    {
+      "epoch": 0.9572948818353526,
+      "grad_norm": 1.1218584552174033,
+      "learning_rate": 9.550502541374463e-07,
+      "loss": 1.2202,
+      "step": 20780
+    },
+    {
+      "epoch": 0.9573409499239877,
+      "grad_norm": 0.743402483530794,
+      "learning_rate": 9.529941244433183e-07,
+      "loss": 1.4342,
+      "step": 20781
+    },
+    {
+      "epoch": 0.9573870180126226,
+      "grad_norm": 0.9363480237578687,
+      "learning_rate": 9.509401998609612e-07,
+      "loss": 1.3583,
+      "step": 20782
+    },
+    {
+      "epoch": 0.9574330861012577,
+      "grad_norm": 0.6615698589915914,
+      "learning_rate": 9.488884804360831e-07,
+      "loss": 1.129,
+      "step": 20783
+    },
+    {
+      "epoch": 0.9574791541898927,
+      "grad_norm": 0.8860390385068696,
+      "learning_rate": 9.468389662143584e-07,
+      "loss": 1.1732,
+      "step": 20784
+    },
+    {
+      "epoch": 0.9575252222785277,
+      "grad_norm": 0.8104833859626558,
+      "learning_rate": 9.447916572414284e-07,
+      "loss": 1.3249,
+      "step": 20785
+    },
+    {
+      "epoch": 0.9575712903671627,
+      "grad_norm": 1.126361642194431,
+      "learning_rate": 9.427465535628677e-07,
+      "loss": 1.0442,
+      "step": 20786
+    },
+    {
+      "epoch": 0.9576173584557977,
+      "grad_norm": 1.79580335931088,
+      "learning_rate": 9.407036552242176e-07,
+      "loss": 1.1904,
+      "step": 20787
+    },
+    {
+      "epoch": 0.9576634265444327,
+      "grad_norm": 0.8416933422218337,
+      "learning_rate": 9.386629622709419e-07,
+      "loss": 1.1602,
+      "step": 20788
+    },
+    {
+      "epoch": 0.9577094946330676,
+      "grad_norm": 1.64261125658198,
+      "learning_rate": 9.36624474748482e-07,
+      "loss": 1.3065,
+      "step": 20789
+    },
+    {
+      "epoch": 0.9577555627217027,
+      "grad_norm": 0.9311213980842125,
+      "learning_rate": 9.345881927022237e-07,
+      "loss": 0.9614,
+      "step": 20790
+    },
+    {
+      "epoch": 0.9578016308103376,
+      "grad_norm": 0.8708065695347974,
+      "learning_rate": 9.325541161774975e-07,
+      "loss": 1.3659,
+      "step": 20791
+    },
+    {
+      "epoch": 0.9578476988989727,
+      "grad_norm": 0.9579110377446058,
+      "learning_rate": 9.305222452195894e-07,
+      "loss": 1.351,
+      "step": 20792
+    },
+    {
+      "epoch": 0.9578937669876076,
+      "grad_norm": 0.8217612057462963,
+      "learning_rate": 9.284925798737409e-07,
+      "loss": 1.1437,
+      "step": 20793
+    },
+    {
+      "epoch": 0.9579398350762427,
+      "grad_norm": 1.8795303195994697,
+      "learning_rate": 9.264651201851271e-07,
+      "loss": 1.3932,
+      "step": 20794
+    },
+    {
+      "epoch": 0.9579859031648776,
+      "grad_norm": 1.4959983894243525,
+      "learning_rate": 9.244398661988896e-07,
+      "loss": 1.2799,
+      "step": 20795
+    },
+    {
+      "epoch": 0.9580319712535127,
+      "grad_norm": 0.7785656839943459,
+      "learning_rate": 9.224168179601256e-07,
+      "loss": 1.3777,
+      "step": 20796
+    },
+    {
+      "epoch": 0.9580780393421477,
+      "grad_norm": 1.031341361433983,
+      "learning_rate": 9.203959755138658e-07,
+      "loss": 1.3938,
+      "step": 20797
+    },
+    {
+      "epoch": 0.9581241074307827,
+      "grad_norm": 1.5288686479997478,
+      "learning_rate": 9.183773389051187e-07,
+      "loss": 1.3768,
+      "step": 20798
+    },
+    {
+      "epoch": 0.9581701755194177,
+      "grad_norm": 0.8236209874023557,
+      "learning_rate": 9.163609081787928e-07,
+      "loss": 1.3529,
+      "step": 20799
+    },
+    {
+      "epoch": 0.9582162436080527,
+      "grad_norm": 0.776730638905622,
+      "learning_rate": 9.143466833798076e-07,
+      "loss": 1.1305,
+      "step": 20800
+    },
+    {
+      "epoch": 0.9582623116966877,
+      "grad_norm": 0.9440966949642241,
+      "learning_rate": 9.12334664552994e-07,
+      "loss": 1.0873,
+      "step": 20801
+    },
+    {
+      "epoch": 0.9583083797853227,
+      "grad_norm": 1.3064025651294742,
+      "learning_rate": 9.103248517431606e-07,
+      "loss": 1.3986,
+      "step": 20802
+    },
+    {
+      "epoch": 0.9583544478739577,
+      "grad_norm": 1.1079318566516148,
+      "learning_rate": 9.083172449950383e-07,
+      "loss": 0.9679,
+      "step": 20803
+    },
+    {
+      "epoch": 0.9584005159625927,
+      "grad_norm": 0.6299159264156877,
+      "learning_rate": 9.063118443533136e-07,
+      "loss": 1.4837,
+      "step": 20804
+    },
+    {
+      "epoch": 0.9584465840512277,
+      "grad_norm": 0.9715892798897106,
+      "learning_rate": 9.043086498626619e-07,
+      "loss": 1.0172,
+      "step": 20805
+    },
+    {
+      "epoch": 0.9584926521398627,
+      "grad_norm": 0.9204115303293382,
+      "learning_rate": 9.023076615676585e-07,
+      "loss": 1.271,
+      "step": 20806
+    },
+    {
+      "epoch": 0.9585387202284977,
+      "grad_norm": 0.997494928962319,
+      "learning_rate": 9.003088795128567e-07,
+      "loss": 1.266,
+      "step": 20807
+    },
+    {
+      "epoch": 0.9585847883171327,
+      "grad_norm": 0.9313050274817283,
+      "learning_rate": 8.983123037427654e-07,
+      "loss": 1.1964,
+      "step": 20808
+    },
+    {
+      "epoch": 0.9586308564057677,
+      "grad_norm": 0.6175620915707115,
+      "learning_rate": 8.963179343018269e-07,
+      "loss": 1.005,
+      "step": 20809
+    },
+    {
+      "epoch": 0.9586769244944028,
+      "grad_norm": 1.4595966035883507,
+      "learning_rate": 8.943257712344389e-07,
+      "loss": 1.2849,
+      "step": 20810
+    },
+    {
+      "epoch": 0.9587229925830377,
+      "grad_norm": 0.7069266043671603,
+      "learning_rate": 8.923358145849547e-07,
+      "loss": 1.3223,
+      "step": 20811
+    },
+    {
+      "epoch": 0.9587690606716728,
+      "grad_norm": 0.7906553967803314,
+      "learning_rate": 8.903480643976947e-07,
+      "loss": 1.2106,
+      "step": 20812
+    },
+    {
+      "epoch": 0.9588151287603077,
+      "grad_norm": 1.085975057274499,
+      "learning_rate": 8.8836252071689e-07,
+      "loss": 0.8415,
+      "step": 20813
+    },
+    {
+      "epoch": 0.9588611968489428,
+      "grad_norm": 0.8437919661852362,
+      "learning_rate": 8.863791835867496e-07,
+      "loss": 1.2915,
+      "step": 20814
+    },
+    {
+      "epoch": 0.9589072649375777,
+      "grad_norm": 0.7718309896598464,
+      "learning_rate": 8.843980530514273e-07,
+      "loss": 1.0757,
+      "step": 20815
+    },
+    {
+      "epoch": 0.9589533330262128,
+      "grad_norm": 3.6434117916864412,
+      "learning_rate": 8.824191291550543e-07,
+      "loss": 1.3061,
+      "step": 20816
+    },
+    {
+      "epoch": 0.9589994011148477,
+      "grad_norm": 0.8808537013069161,
+      "learning_rate": 8.80442411941651e-07,
+      "loss": 1.0781,
+      "step": 20817
+    },
+    {
+      "epoch": 0.9590454692034828,
+      "grad_norm": 1.37574969467994,
+      "learning_rate": 8.784679014552599e-07,
+      "loss": 1.1707,
+      "step": 20818
+    },
+    {
+      "epoch": 0.9590915372921177,
+      "grad_norm": 1.6686041818316901,
+      "learning_rate": 8.764955977398126e-07,
+      "loss": 0.9445,
+      "step": 20819
+    },
+    {
+      "epoch": 0.9591376053807528,
+      "grad_norm": 0.7932564896726005,
+      "learning_rate": 8.745255008392406e-07,
+      "loss": 1.3204,
+      "step": 20820
+    },
+    {
+      "epoch": 0.9591836734693877,
+      "grad_norm": 1.1786034943567205,
+      "learning_rate": 8.725576107973865e-07,
+      "loss": 1.2733,
+      "step": 20821
+    },
+    {
+      "epoch": 0.9592297415580228,
+      "grad_norm": 2.4965805062546744,
+      "learning_rate": 8.705919276580598e-07,
+      "loss": 1.2836,
+      "step": 20822
+    },
+    {
+      "epoch": 0.9592758096466577,
+      "grad_norm": 0.9249840499846977,
+      "learning_rate": 8.686284514650589e-07,
+      "loss": 1.1755,
+      "step": 20823
+    },
+    {
+      "epoch": 0.9593218777352928,
+      "grad_norm": 2.2028506988634744,
+      "learning_rate": 8.666671822620487e-07,
+      "loss": 1.1291,
+      "step": 20824
+    },
+    {
+      "epoch": 0.9593679458239278,
+      "grad_norm": 0.9337488483991591,
+      "learning_rate": 8.647081200927276e-07,
+      "loss": 1.3754,
+      "step": 20825
+    },
+    {
+      "epoch": 0.9594140139125628,
+      "grad_norm": 0.932380440026809,
+      "learning_rate": 8.627512650006941e-07,
+      "loss": 1.3217,
+      "step": 20826
+    },
+    {
+      "epoch": 0.9594600820011978,
+      "grad_norm": 2.5023499658115953,
+      "learning_rate": 8.607966170295245e-07,
+      "loss": 1.6056,
+      "step": 20827
+    },
+    {
+      "epoch": 0.9595061500898328,
+      "grad_norm": 1.0499766479532446,
+      "learning_rate": 8.588441762227284e-07,
+      "loss": 1.2624,
+      "step": 20828
+    },
+    {
+      "epoch": 0.9595522181784678,
+      "grad_norm": 1.209477091349812,
+      "learning_rate": 8.568939426237821e-07,
+      "loss": 1.316,
+      "step": 20829
+    },
+    {
+      "epoch": 0.9595982862671028,
+      "grad_norm": 1.157741367948455,
+      "learning_rate": 8.549459162760953e-07,
+      "loss": 1.5213,
+      "step": 20830
+    },
+    {
+      "epoch": 0.9596443543557378,
+      "grad_norm": 0.9058537748291793,
+      "learning_rate": 8.530000972230335e-07,
+      "loss": 1.2486,
+      "step": 20831
+    },
+    {
+      "epoch": 0.9596904224443727,
+      "grad_norm": 0.9868156225407878,
+      "learning_rate": 8.510564855079284e-07,
+      "loss": 1.2775,
+      "step": 20832
+    },
+    {
+      "epoch": 0.9597364905330078,
+      "grad_norm": 1.7187443783280467,
+      "learning_rate": 8.491150811740456e-07,
+      "loss": 1.3006,
+      "step": 20833
+    },
+    {
+      "epoch": 0.9597825586216427,
+      "grad_norm": 0.9135332945630926,
+      "learning_rate": 8.471758842646171e-07,
+      "loss": 1.2586,
+      "step": 20834
+    },
+    {
+      "epoch": 0.9598286267102778,
+      "grad_norm": 0.8485587312090487,
+      "learning_rate": 8.452388948227974e-07,
+      "loss": 1.231,
+      "step": 20835
+    },
+    {
+      "epoch": 0.9598746947989127,
+      "grad_norm": 0.7369880630934058,
+      "learning_rate": 8.433041128917185e-07,
+      "loss": 1.1052,
+      "step": 20836
+    },
+    {
+      "epoch": 0.9599207628875478,
+      "grad_norm": 1.7788285738911658,
+      "learning_rate": 8.413715385144572e-07,
+      "loss": 1.2677,
+      "step": 20837
+    },
+    {
+      "epoch": 0.9599668309761828,
+      "grad_norm": 0.9991035622623041,
+      "learning_rate": 8.394411717340455e-07,
+      "loss": 1.2723,
+      "step": 20838
+    },
+    {
+      "epoch": 0.9600128990648178,
+      "grad_norm": 2.7539940948434545,
+      "learning_rate": 8.375130125934494e-07,
+      "loss": 1.3448,
+      "step": 20839
+    },
+    {
+      "epoch": 0.9600589671534528,
+      "grad_norm": 0.8502873426358717,
+      "learning_rate": 8.355870611355898e-07,
+      "loss": 1.1621,
+      "step": 20840
+    },
+    {
+      "epoch": 0.9601050352420878,
+      "grad_norm": 1.3428078467031066,
+      "learning_rate": 8.336633174033659e-07,
+      "loss": 0.9703,
+      "step": 20841
+    },
+    {
+      "epoch": 0.9601511033307228,
+      "grad_norm": 0.9440392403666185,
+      "learning_rate": 8.317417814395878e-07,
+      "loss": 1.1237,
+      "step": 20842
+    },
+    {
+      "epoch": 0.9601971714193578,
+      "grad_norm": 0.9572291792392288,
+      "learning_rate": 8.298224532870436e-07,
+      "loss": 1.0377,
+      "step": 20843
+    },
+    {
+      "epoch": 0.9602432395079928,
+      "grad_norm": 0.9637376352732039,
+      "learning_rate": 8.279053329884657e-07,
+      "loss": 1.2483,
+      "step": 20844
+    },
+    {
+      "epoch": 0.9602893075966278,
+      "grad_norm": 1.3742929106871027,
+      "learning_rate": 8.25990420586531e-07,
+      "loss": 0.9681,
+      "step": 20845
+    },
+    {
+      "epoch": 0.9603353756852628,
+      "grad_norm": 0.7347182448279094,
+      "learning_rate": 8.240777161238611e-07,
+      "loss": 1.1811,
+      "step": 20846
+    },
+    {
+      "epoch": 0.9603814437738978,
+      "grad_norm": 0.8872156409047146,
+      "learning_rate": 8.221672196430663e-07,
+      "loss": 1.1504,
+      "step": 20847
+    },
+    {
+      "epoch": 0.9604275118625328,
+      "grad_norm": 0.8071425350798876,
+      "learning_rate": 8.20258931186657e-07,
+      "loss": 1.1616,
+      "step": 20848
+    },
+    {
+      "epoch": 0.9604735799511678,
+      "grad_norm": 2.019383069516816,
+      "learning_rate": 8.183528507971217e-07,
+      "loss": 0.9884,
+      "step": 20849
+    },
+    {
+      "epoch": 0.9605196480398028,
+      "grad_norm": 1.4085131745422197,
+      "learning_rate": 8.164489785169038e-07,
+      "loss": 1.124,
+      "step": 20850
+    },
+    {
+      "epoch": 0.9605657161284379,
+      "grad_norm": 1.0544634326574327,
+      "learning_rate": 8.145473143883919e-07,
+      "loss": 1.013,
+      "step": 20851
+    },
+    {
+      "epoch": 0.9606117842170728,
+      "grad_norm": 0.8600754447068945,
+      "learning_rate": 8.126478584539077e-07,
+      "loss": 1.1569,
+      "step": 20852
+    },
+    {
+      "epoch": 0.9606578523057079,
+      "grad_norm": 1.2218498994503473,
+      "learning_rate": 8.107506107557617e-07,
+      "loss": 1.1325,
+      "step": 20853
+    },
+    {
+      "epoch": 0.9607039203943428,
+      "grad_norm": 1.791742387388766,
+      "learning_rate": 8.088555713361645e-07,
+      "loss": 0.8101,
+      "step": 20854
+    },
+    {
+      "epoch": 0.9607499884829779,
+      "grad_norm": 1.582191027034067,
+      "learning_rate": 8.06962740237327e-07,
+      "loss": 1.2958,
+      "step": 20855
+    },
+    {
+      "epoch": 0.9607960565716128,
+      "grad_norm": 1.2632311451181293,
+      "learning_rate": 8.050721175013931e-07,
+      "loss": 0.9628,
+      "step": 20856
+    },
+    {
+      "epoch": 0.9608421246602479,
+      "grad_norm": 1.2403695870680418,
+      "learning_rate": 8.031837031704515e-07,
+      "loss": 1.0676,
+      "step": 20857
+    },
+    {
+      "epoch": 0.9608881927488828,
+      "grad_norm": 0.9440886452881591,
+      "learning_rate": 8.012974972865128e-07,
+      "loss": 1.3201,
+      "step": 20858
+    },
+    {
+      "epoch": 0.9609342608375179,
+      "grad_norm": 1.1950606018400347,
+      "learning_rate": 7.994134998916214e-07,
+      "loss": 0.925,
+      "step": 20859
+    },
+    {
+      "epoch": 0.9609803289261528,
+      "grad_norm": 0.8983343084197585,
+      "learning_rate": 7.975317110276769e-07,
+      "loss": 1.1498,
+      "step": 20860
+    },
+    {
+      "epoch": 0.9610263970147879,
+      "grad_norm": 1.1605584064423036,
+      "learning_rate": 7.956521307366016e-07,
+      "loss": 1.584,
+      "step": 20861
+    },
+    {
+      "epoch": 0.9610724651034228,
+      "grad_norm": 0.9290879689332899,
+      "learning_rate": 7.937747590602284e-07,
+      "loss": 1.3102,
+      "step": 20862
+    },
+    {
+      "epoch": 0.9611185331920579,
+      "grad_norm": 0.9987063189653811,
+      "learning_rate": 7.918995960403685e-07,
+      "loss": 1.1407,
+      "step": 20863
+    },
+    {
+      "epoch": 0.9611646012806928,
+      "grad_norm": 2.203340037300371,
+      "learning_rate": 7.900266417187441e-07,
+      "loss": 1.3541,
+      "step": 20864
+    },
+    {
+      "epoch": 0.9612106693693279,
+      "grad_norm": 1.3511930512372516,
+      "learning_rate": 7.881558961370772e-07,
+      "loss": 1.2944,
+      "step": 20865
+    },
+    {
+      "epoch": 0.9612567374579629,
+      "grad_norm": 3.5765417447352874,
+      "learning_rate": 7.862873593370124e-07,
+      "loss": 1.3788,
+      "step": 20866
+    },
+    {
+      "epoch": 0.9613028055465979,
+      "grad_norm": 0.9990202308206093,
+      "learning_rate": 7.844210313601385e-07,
+      "loss": 1.0555,
+      "step": 20867
+    },
+    {
+      "epoch": 0.9613488736352329,
+      "grad_norm": 1.0008775134132648,
+      "learning_rate": 7.825569122480115e-07,
+      "loss": 1.25,
+      "step": 20868
+    },
+    {
+      "epoch": 0.9613949417238679,
+      "grad_norm": 0.9905931203347875,
+      "learning_rate": 7.806950020421311e-07,
+      "loss": 1.3309,
+      "step": 20869
+    },
+    {
+      "epoch": 0.9614410098125029,
+      "grad_norm": 1.3725028624266433,
+      "learning_rate": 7.788353007839532e-07,
+      "loss": 1.105,
+      "step": 20870
+    },
+    {
+      "epoch": 0.9614870779011379,
+      "grad_norm": 1.1716546025744023,
+      "learning_rate": 7.769778085148782e-07,
+      "loss": 1.2855,
+      "step": 20871
+    },
+    {
+      "epoch": 0.9615331459897729,
+      "grad_norm": 1.135131304122324,
+      "learning_rate": 7.751225252762728e-07,
+      "loss": 0.9873,
+      "step": 20872
+    },
+    {
+      "epoch": 0.9615792140784079,
+      "grad_norm": 1.0396134551458964,
+      "learning_rate": 7.732694511094041e-07,
+      "loss": 1.2445,
+      "step": 20873
+    },
+    {
+      "epoch": 0.9616252821670429,
+      "grad_norm": 1.1838046738269263,
+      "learning_rate": 7.714185860555834e-07,
+      "loss": 0.8777,
+      "step": 20874
+    },
+    {
+      "epoch": 0.9616713502556778,
+      "grad_norm": 1.0838532680951039,
+      "learning_rate": 7.695699301559778e-07,
+      "loss": 1.067,
+      "step": 20875
+    },
+    {
+      "epoch": 0.9617174183443129,
+      "grad_norm": 0.7554097859515387,
+      "learning_rate": 7.677234834517432e-07,
+      "loss": 0.9646,
+      "step": 20876
+    },
+    {
+      "epoch": 0.9617634864329478,
+      "grad_norm": 1.0418625585599994,
+      "learning_rate": 7.658792459840026e-07,
+      "loss": 1.0025,
+      "step": 20877
+    },
+    {
+      "epoch": 0.9618095545215829,
+      "grad_norm": 0.9183921939505625,
+      "learning_rate": 7.640372177938227e-07,
+      "loss": 1.2106,
+      "step": 20878
+    },
+    {
+      "epoch": 0.961855622610218,
+      "grad_norm": 0.7853207675832622,
+      "learning_rate": 7.621973989221931e-07,
+      "loss": 1.4862,
+      "step": 20879
+    },
+    {
+      "epoch": 0.9619016906988529,
+      "grad_norm": 3.8931400593875334,
+      "learning_rate": 7.6035978941007e-07,
+      "loss": 1.3974,
+      "step": 20880
+    },
+    {
+      "epoch": 0.961947758787488,
+      "grad_norm": 0.963883867380838,
+      "learning_rate": 7.585243892983985e-07,
+      "loss": 1.2274,
+      "step": 20881
+    },
+    {
+      "epoch": 0.9619938268761229,
+      "grad_norm": 0.8184481892328673,
+      "learning_rate": 7.566911986280012e-07,
+      "loss": 1.2935,
+      "step": 20882
+    },
+    {
+      "epoch": 0.962039894964758,
+      "grad_norm": 1.0704041829626596,
+      "learning_rate": 7.548602174397235e-07,
+      "loss": 1.3511,
+      "step": 20883
+    },
+    {
+      "epoch": 0.9620859630533929,
+      "grad_norm": 1.1740708275119265,
+      "learning_rate": 7.530314457742992e-07,
+      "loss": 1.1759,
+      "step": 20884
+    },
+    {
+      "epoch": 0.9621320311420279,
+      "grad_norm": 1.2475500014291936,
+      "learning_rate": 7.512048836724628e-07,
+      "loss": 1.0386,
+      "step": 20885
+    },
+    {
+      "epoch": 0.9621780992306629,
+      "grad_norm": 0.9193674141127809,
+      "learning_rate": 7.493805311748814e-07,
+      "loss": 1.181,
+      "step": 20886
+    },
+    {
+      "epoch": 0.9622241673192979,
+      "grad_norm": 0.8042605274263909,
+      "learning_rate": 7.475583883221559e-07,
+      "loss": 1.3285,
+      "step": 20887
+    },
+    {
+      "epoch": 0.9622702354079329,
+      "grad_norm": 0.7754123648618001,
+      "learning_rate": 7.457384551548652e-07,
+      "loss": 1.07,
+      "step": 20888
+    },
+    {
+      "epoch": 0.9623163034965679,
+      "grad_norm": 0.8805992230767343,
+      "learning_rate": 7.439207317135211e-07,
+      "loss": 1.3451,
+      "step": 20889
+    },
+    {
+      "epoch": 0.9623623715852029,
+      "grad_norm": 1.0892202593431046,
+      "learning_rate": 7.421052180386023e-07,
+      "loss": 0.9467,
+      "step": 20890
+    },
+    {
+      "epoch": 0.9624084396738379,
+      "grad_norm": 0.8011617215941126,
+      "learning_rate": 7.40291914170499e-07,
+      "loss": 1.4015,
+      "step": 20891
+    },
+    {
+      "epoch": 0.962454507762473,
+      "grad_norm": 1.7943418493510905,
+      "learning_rate": 7.38480820149623e-07,
+      "loss": 1.3286,
+      "step": 20892
+    },
+    {
+      "epoch": 0.9625005758511079,
+      "grad_norm": 0.8993583517718091,
+      "learning_rate": 7.366719360162754e-07,
+      "loss": 1.0919,
+      "step": 20893
+    },
+    {
+      "epoch": 0.962546643939743,
+      "grad_norm": 0.8869842991140605,
+      "learning_rate": 7.348652618107133e-07,
+      "loss": 1.0676,
+      "step": 20894
+    },
+    {
+      "epoch": 0.9625927120283779,
+      "grad_norm": 0.9126777405356904,
+      "learning_rate": 7.330607975731818e-07,
+      "loss": 1.0979,
+      "step": 20895
+    },
+    {
+      "epoch": 0.962638780117013,
+      "grad_norm": 1.1418367414983464,
+      "learning_rate": 7.312585433438379e-07,
+      "loss": 1.2887,
+      "step": 20896
+    },
+    {
+      "epoch": 0.9626848482056479,
+      "grad_norm": 1.0017435011265412,
+      "learning_rate": 7.294584991628272e-07,
+      "loss": 1.347,
+      "step": 20897
+    },
+    {
+      "epoch": 0.962730916294283,
+      "grad_norm": 0.8527137694686963,
+      "learning_rate": 7.276606650701956e-07,
+      "loss": 1.2129,
+      "step": 20898
+    },
+    {
+      "epoch": 0.9627769843829179,
+      "grad_norm": 0.9745488966993613,
+      "learning_rate": 7.258650411059886e-07,
+      "loss": 1.1165,
+      "step": 20899
+    },
+    {
+      "epoch": 0.962823052471553,
+      "grad_norm": 1.400411918350482,
+      "learning_rate": 7.240716273101856e-07,
+      "loss": 1.2008,
+      "step": 20900
+    },
+    {
+      "epoch": 0.9628691205601879,
+      "grad_norm": 1.3633841891265475,
+      "learning_rate": 7.222804237226988e-07,
+      "loss": 1.0153,
+      "step": 20901
+    },
+    {
+      "epoch": 0.962915188648823,
+      "grad_norm": 0.9809910184310477,
+      "learning_rate": 7.204914303834076e-07,
+      "loss": 1.0473,
+      "step": 20902
+    },
+    {
+      "epoch": 0.9629612567374579,
+      "grad_norm": 1.4931315270133985,
+      "learning_rate": 7.18704647332158e-07,
+      "loss": 1.1169,
+      "step": 20903
+    },
+    {
+      "epoch": 0.963007324826093,
+      "grad_norm": 1.8573580199051065,
+      "learning_rate": 7.169200746087179e-07,
+      "loss": 1.1698,
+      "step": 20904
+    },
+    {
+      "epoch": 0.9630533929147279,
+      "grad_norm": 1.1184442172631228,
+      "learning_rate": 7.151377122528113e-07,
+      "loss": 1.4602,
+      "step": 20905
+    },
+    {
+      "epoch": 0.963099461003363,
+      "grad_norm": 1.007345935984066,
+      "learning_rate": 7.133575603041287e-07,
+      "loss": 1.1631,
+      "step": 20906
+    },
+    {
+      "epoch": 0.963145529091998,
+      "grad_norm": 1.6254147542682382,
+      "learning_rate": 7.115796188023049e-07,
+      "loss": 0.7903,
+      "step": 20907
+    },
+    {
+      "epoch": 0.963191597180633,
+      "grad_norm": 0.7974164032571597,
+      "learning_rate": 7.098038877869084e-07,
+      "loss": 1.2492,
+      "step": 20908
+    },
+    {
+      "epoch": 0.963237665269268,
+      "grad_norm": 0.7355635358688276,
+      "learning_rate": 7.080303672974853e-07,
+      "loss": 1.4087,
+      "step": 20909
+    },
+    {
+      "epoch": 0.963283733357903,
+      "grad_norm": 0.6904839637504037,
+      "learning_rate": 7.062590573735261e-07,
+      "loss": 1.4105,
+      "step": 20910
+    },
+    {
+      "epoch": 0.963329801446538,
+      "grad_norm": 1.0288140503384016,
+      "learning_rate": 7.044899580544551e-07,
+      "loss": 1.4175,
+      "step": 20911
+    },
+    {
+      "epoch": 0.963375869535173,
+      "grad_norm": 0.7526498608449269,
+      "learning_rate": 7.027230693796516e-07,
+      "loss": 1.4261,
+      "step": 20912
+    },
+    {
+      "epoch": 0.963421937623808,
+      "grad_norm": 0.9023151391712568,
+      "learning_rate": 7.009583913884621e-07,
+      "loss": 1.2316,
+      "step": 20913
+    },
+    {
+      "epoch": 0.963468005712443,
+      "grad_norm": 4.493399797023746,
+      "learning_rate": 6.991959241201662e-07,
+      "loss": 1.2609,
+      "step": 20914
+    },
+    {
+      "epoch": 0.963514073801078,
+      "grad_norm": 0.8298969604065513,
+      "learning_rate": 6.974356676140215e-07,
+      "loss": 1.1894,
+      "step": 20915
+    },
+    {
+      "epoch": 0.963560141889713,
+      "grad_norm": 1.0497190731720112,
+      "learning_rate": 6.956776219091965e-07,
+      "loss": 1.1701,
+      "step": 20916
+    },
+    {
+      "epoch": 0.963606209978348,
+      "grad_norm": 0.7523854839557745,
+      "learning_rate": 6.939217870448489e-07,
+      "loss": 1.2697,
+      "step": 20917
+    },
+    {
+      "epoch": 0.963652278066983,
+      "grad_norm": 1.3181664679044423,
+      "learning_rate": 6.921681630600474e-07,
+      "loss": 1.4575,
+      "step": 20918
+    },
+    {
+      "epoch": 0.963698346155618,
+      "grad_norm": 0.8709572163832162,
+      "learning_rate": 6.904167499938497e-07,
+      "loss": 1.0853,
+      "step": 20919
+    },
+    {
+      "epoch": 0.9637444142442531,
+      "grad_norm": 1.9379509854691788,
+      "learning_rate": 6.886675478852467e-07,
+      "loss": 1.0016,
+      "step": 20920
+    },
+    {
+      "epoch": 0.963790482332888,
+      "grad_norm": 1.145087166036389,
+      "learning_rate": 6.86920556773174e-07,
+      "loss": 1.349,
+      "step": 20921
+    },
+    {
+      "epoch": 0.963836550421523,
+      "grad_norm": 1.6723559308437106,
+      "learning_rate": 6.851757766965339e-07,
+      "loss": 1.2217,
+      "step": 20922
+    },
+    {
+      "epoch": 0.963882618510158,
+      "grad_norm": 1.3977639883674349,
+      "learning_rate": 6.834332076941619e-07,
+      "loss": 1.1712,
+      "step": 20923
+    },
+    {
+      "epoch": 0.963928686598793,
+      "grad_norm": 1.8239130405516102,
+      "learning_rate": 6.816928498048602e-07,
+      "loss": 1.271,
+      "step": 20924
+    },
+    {
+      "epoch": 0.963974754687428,
+      "grad_norm": 0.7347652310374192,
+      "learning_rate": 6.799547030673647e-07,
+      "loss": 1.2753,
+      "step": 20925
+    },
+    {
+      "epoch": 0.964020822776063,
+      "grad_norm": 0.7733714746180789,
+      "learning_rate": 6.782187675203999e-07,
+      "loss": 1.4011,
+      "step": 20926
+    },
+    {
+      "epoch": 0.964066890864698,
+      "grad_norm": 0.9222058066417284,
+      "learning_rate": 6.764850432025793e-07,
+      "loss": 1.3431,
+      "step": 20927
+    },
+    {
+      "epoch": 0.964112958953333,
+      "grad_norm": 1.5060518306125885,
+      "learning_rate": 6.747535301525165e-07,
+      "loss": 1.2991,
+      "step": 20928
+    },
+    {
+      "epoch": 0.964159027041968,
+      "grad_norm": 1.7452123976692313,
+      "learning_rate": 6.730242284087585e-07,
+      "loss": 1.4022,
+      "step": 20929
+    },
+    {
+      "epoch": 0.964205095130603,
+      "grad_norm": 0.7592283601826941,
+      "learning_rate": 6.712971380098187e-07,
+      "loss": 1.1474,
+      "step": 20930
+    },
+    {
+      "epoch": 0.964251163219238,
+      "grad_norm": 1.4769163089222126,
+      "learning_rate": 6.695722589941222e-07,
+      "loss": 1.5499,
+      "step": 20931
+    },
+    {
+      "epoch": 0.964297231307873,
+      "grad_norm": 0.7642845143804418,
+      "learning_rate": 6.678495914000938e-07,
+      "loss": 1.2546,
+      "step": 20932
+    },
+    {
+      "epoch": 0.9643432993965081,
+      "grad_norm": 0.7155326392020925,
+      "learning_rate": 6.661291352660803e-07,
+      "loss": 1.275,
+      "step": 20933
+    },
+    {
+      "epoch": 0.964389367485143,
+      "grad_norm": 1.0277928613733185,
+      "learning_rate": 6.644108906303625e-07,
+      "loss": 1.4289,
+      "step": 20934
+    },
+    {
+      "epoch": 0.9644354355737781,
+      "grad_norm": 1.5867809289443844,
+      "learning_rate": 6.62694857531232e-07,
+      "loss": 1.3329,
+      "step": 20935
+    },
+    {
+      "epoch": 0.964481503662413,
+      "grad_norm": 0.8459951497149796,
+      "learning_rate": 6.60981036006858e-07,
+      "loss": 1.5823,
+      "step": 20936
+    },
+    {
+      "epoch": 0.9645275717510481,
+      "grad_norm": 1.7211831283724042,
+      "learning_rate": 6.592694260954213e-07,
+      "loss": 1.088,
+      "step": 20937
+    },
+    {
+      "epoch": 0.964573639839683,
+      "grad_norm": 0.8557486611916887,
+      "learning_rate": 6.575600278350136e-07,
+      "loss": 1.4782,
+      "step": 20938
+    },
+    {
+      "epoch": 0.9646197079283181,
+      "grad_norm": 0.9525218697185149,
+      "learning_rate": 6.558528412636933e-07,
+      "loss": 1.1689,
+      "step": 20939
+    },
+    {
+      "epoch": 0.964665776016953,
+      "grad_norm": 0.882087513054176,
+      "learning_rate": 6.541478664194744e-07,
+      "loss": 1.0889,
+      "step": 20940
+    },
+    {
+      "epoch": 0.9647118441055881,
+      "grad_norm": 1.632924737940814,
+      "learning_rate": 6.524451033403156e-07,
+      "loss": 1.1411,
+      "step": 20941
+    },
+    {
+      "epoch": 0.964757912194223,
+      "grad_norm": 1.013346669890821,
+      "learning_rate": 6.507445520641198e-07,
+      "loss": 1.114,
+      "step": 20942
+    },
+    {
+      "epoch": 0.9648039802828581,
+      "grad_norm": 1.0976483732165137,
+      "learning_rate": 6.490462126287455e-07,
+      "loss": 1.0401,
+      "step": 20943
+    },
+    {
+      "epoch": 0.964850048371493,
+      "grad_norm": 0.8004784805511983,
+      "learning_rate": 6.473500850720071e-07,
+      "loss": 1.1798,
+      "step": 20944
+    },
+    {
+      "epoch": 0.9648961164601281,
+      "grad_norm": 0.9656050561418924,
+      "learning_rate": 6.456561694316632e-07,
+      "loss": 1.4681,
+      "step": 20945
+    },
+    {
+      "epoch": 0.9649421845487631,
+      "grad_norm": 0.9134618739364343,
+      "learning_rate": 6.439644657454391e-07,
+      "loss": 1.3662,
+      "step": 20946
+    },
+    {
+      "epoch": 0.9649882526373981,
+      "grad_norm": 1.0373876216415556,
+      "learning_rate": 6.422749740509715e-07,
+      "loss": 1.4365,
+      "step": 20947
+    },
+    {
+      "epoch": 0.9650343207260331,
+      "grad_norm": 0.9505365699479872,
+      "learning_rate": 6.405876943859079e-07,
+      "loss": 1.5092,
+      "step": 20948
+    },
+    {
+      "epoch": 0.9650803888146681,
+      "grad_norm": 1.3682000023777576,
+      "learning_rate": 6.389026267877851e-07,
+      "loss": 1.5009,
+      "step": 20949
+    },
+    {
+      "epoch": 0.9651264569033031,
+      "grad_norm": 0.6235988281836881,
+      "learning_rate": 6.372197712941174e-07,
+      "loss": 1.2109,
+      "step": 20950
+    },
+    {
+      "epoch": 0.9651725249919381,
+      "grad_norm": 1.3133072041362244,
+      "learning_rate": 6.355391279423973e-07,
+      "loss": 1.2146,
+      "step": 20951
+    },
+    {
+      "epoch": 0.9652185930805731,
+      "grad_norm": 1.4275008146278045,
+      "learning_rate": 6.338606967700167e-07,
+      "loss": 1.1231,
+      "step": 20952
+    },
+    {
+      "epoch": 0.9652646611692081,
+      "grad_norm": 0.9775093856562029,
+      "learning_rate": 6.32184477814357e-07,
+      "loss": 1.0022,
+      "step": 20953
+    },
+    {
+      "epoch": 0.9653107292578431,
+      "grad_norm": 0.8060889824252419,
+      "learning_rate": 6.305104711127219e-07,
+      "loss": 1.0192,
+      "step": 20954
+    },
+    {
+      "epoch": 0.9653567973464781,
+      "grad_norm": 1.36638512797271,
+      "learning_rate": 6.288386767024035e-07,
+      "loss": 1.0119,
+      "step": 20955
+    },
+    {
+      "epoch": 0.9654028654351131,
+      "grad_norm": 1.9565452089288837,
+      "learning_rate": 6.271690946205944e-07,
+      "loss": 1.3919,
+      "step": 20956
+    },
+    {
+      "epoch": 0.965448933523748,
+      "grad_norm": 1.67405569025051,
+      "learning_rate": 6.255017249044759e-07,
+      "loss": 1.1706,
+      "step": 20957
+    },
+    {
+      "epoch": 0.9654950016123831,
+      "grad_norm": 0.9614900280416613,
+      "learning_rate": 6.23836567591185e-07,
+      "loss": 1.094,
+      "step": 20958
+    },
+    {
+      "epoch": 0.965541069701018,
+      "grad_norm": 1.1153687207839118,
+      "learning_rate": 6.221736227177699e-07,
+      "loss": 1.337,
+      "step": 20959
+    },
+    {
+      "epoch": 0.9655871377896531,
+      "grad_norm": 0.8879661857890148,
+      "learning_rate": 6.205128903212676e-07,
+      "loss": 1.0194,
+      "step": 20960
+    },
+    {
+      "epoch": 0.9656332058782882,
+      "grad_norm": 0.8583155736860142,
+      "learning_rate": 6.188543704386263e-07,
+      "loss": 1.2402,
+      "step": 20961
+    },
+    {
+      "epoch": 0.9656792739669231,
+      "grad_norm": 0.8582540758606859,
+      "learning_rate": 6.171980631068164e-07,
+      "loss": 1.3442,
+      "step": 20962
+    },
+    {
+      "epoch": 0.9657253420555582,
+      "grad_norm": 0.9163327808712429,
+      "learning_rate": 6.155439683626752e-07,
+      "loss": 1.1689,
+      "step": 20963
+    },
+    {
+      "epoch": 0.9657714101441931,
+      "grad_norm": 1.7710569466136146,
+      "learning_rate": 6.138920862430397e-07,
+      "loss": 1.0442,
+      "step": 20964
+    },
+    {
+      "epoch": 0.9658174782328282,
+      "grad_norm": 1.0592315341904877,
+      "learning_rate": 6.122424167846808e-07,
+      "loss": 1.2076,
+      "step": 20965
+    },
+    {
+      "epoch": 0.9658635463214631,
+      "grad_norm": 0.7667998476815199,
+      "learning_rate": 6.105949600243465e-07,
+      "loss": 1.1864,
+      "step": 20966
+    },
+    {
+      "epoch": 0.9659096144100981,
+      "grad_norm": 1.192116547553434,
+      "learning_rate": 6.089497159986857e-07,
+      "loss": 0.9588,
+      "step": 20967
+    },
+    {
+      "epoch": 0.9659556824987331,
+      "grad_norm": 0.8898793943351657,
+      "learning_rate": 6.073066847443465e-07,
+      "loss": 1.2796,
+      "step": 20968
+    },
+    {
+      "epoch": 0.9660017505873681,
+      "grad_norm": 0.9903142131354749,
+      "learning_rate": 6.056658662978998e-07,
+      "loss": 0.9856,
+      "step": 20969
+    },
+    {
+      "epoch": 0.9660478186760031,
+      "grad_norm": 1.143198283360183,
+      "learning_rate": 6.040272606958719e-07,
+      "loss": 1.1405,
+      "step": 20970
+    },
+    {
+      "epoch": 0.9660938867646381,
+      "grad_norm": 1.2331951717045282,
+      "learning_rate": 6.02390867974767e-07,
+      "loss": 1.1888,
+      "step": 20971
+    },
+    {
+      "epoch": 0.9661399548532731,
+      "grad_norm": 0.8856242669661023,
+      "learning_rate": 6.007566881709781e-07,
+      "loss": 1.3583,
+      "step": 20972
+    },
+    {
+      "epoch": 0.9661860229419081,
+      "grad_norm": 1.0845607684295069,
+      "learning_rate": 5.991247213209206e-07,
+      "loss": 1.0364,
+      "step": 20973
+    },
+    {
+      "epoch": 0.9662320910305432,
+      "grad_norm": 0.933821123867821,
+      "learning_rate": 5.974949674609099e-07,
+      "loss": 1.3551,
+      "step": 20974
+    },
+    {
+      "epoch": 0.9662781591191781,
+      "grad_norm": 4.26924260559336,
+      "learning_rate": 5.95867426627239e-07,
+      "loss": 1.0093,
+      "step": 20975
+    },
+    {
+      "epoch": 0.9663242272078132,
+      "grad_norm": 1.7754455842572723,
+      "learning_rate": 5.942420988561348e-07,
+      "loss": 1.4649,
+      "step": 20976
+    },
+    {
+      "epoch": 0.9663702952964481,
+      "grad_norm": 1.5145435923800554,
+      "learning_rate": 5.926189841837904e-07,
+      "loss": 1.2084,
+      "step": 20977
+    },
+    {
+      "epoch": 0.9664163633850832,
+      "grad_norm": 0.9780709159415781,
+      "learning_rate": 5.909980826463435e-07,
+      "loss": 1.3691,
+      "step": 20978
+    },
+    {
+      "epoch": 0.9664624314737181,
+      "grad_norm": 2.2550407502766765,
+      "learning_rate": 5.893793942798542e-07,
+      "loss": 1.4321,
+      "step": 20979
+    },
+    {
+      "epoch": 0.9665084995623532,
+      "grad_norm": 0.7383524549191389,
+      "learning_rate": 5.877629191203937e-07,
+      "loss": 1.3567,
+      "step": 20980
+    },
+    {
+      "epoch": 0.9665545676509881,
+      "grad_norm": 0.8350722708463323,
+      "learning_rate": 5.861486572039332e-07,
+      "loss": 1.4018,
+      "step": 20981
+    },
+    {
+      "epoch": 0.9666006357396232,
+      "grad_norm": 0.8297231524709585,
+      "learning_rate": 5.845366085664216e-07,
+      "loss": 1.3227,
+      "step": 20982
+    },
+    {
+      "epoch": 0.9666467038282581,
+      "grad_norm": 1.760238312918552,
+      "learning_rate": 5.829267732437304e-07,
+      "loss": 1.1522,
+      "step": 20983
+    },
+    {
+      "epoch": 0.9666927719168932,
+      "grad_norm": 0.8024847317887023,
+      "learning_rate": 5.813191512717198e-07,
+      "loss": 1.1819,
+      "step": 20984
+    },
+    {
+      "epoch": 0.9667388400055281,
+      "grad_norm": 1.218246661736331,
+      "learning_rate": 5.797137426861721e-07,
+      "loss": 1.4027,
+      "step": 20985
+    },
+    {
+      "epoch": 0.9667849080941632,
+      "grad_norm": 0.7332462778701105,
+      "learning_rate": 5.781105475228254e-07,
+      "loss": 1.3734,
+      "step": 20986
+    },
+    {
+      "epoch": 0.9668309761827982,
+      "grad_norm": 0.85735989503434,
+      "learning_rate": 5.765095658173735e-07,
+      "loss": 1.0993,
+      "step": 20987
+    },
+    {
+      "epoch": 0.9668770442714332,
+      "grad_norm": 1.3695567760884189,
+      "learning_rate": 5.749107976054657e-07,
+      "loss": 1.105,
+      "step": 20988
+    },
+    {
+      "epoch": 0.9669231123600682,
+      "grad_norm": 1.259521870424094,
+      "learning_rate": 5.733142429226845e-07,
+      "loss": 0.8537,
+      "step": 20989
+    },
+    {
+      "epoch": 0.9669691804487032,
+      "grad_norm": 0.6378692375971089,
+      "learning_rate": 5.717199018045794e-07,
+      "loss": 1.2607,
+      "step": 20990
+    },
+    {
+      "epoch": 0.9670152485373382,
+      "grad_norm": 0.9308245740932363,
+      "learning_rate": 5.701277742866551e-07,
+      "loss": 1.4575,
+      "step": 20991
+    },
+    {
+      "epoch": 0.9670613166259732,
+      "grad_norm": 0.7674112521780596,
+      "learning_rate": 5.685378604043501e-07,
+      "loss": 1.123,
+      "step": 20992
+    },
+    {
+      "epoch": 0.9671073847146082,
+      "grad_norm": 1.0625648631049087,
+      "learning_rate": 5.669501601930694e-07,
+      "loss": 1.0572,
+      "step": 20993
+    },
+    {
+      "epoch": 0.9671534528032432,
+      "grad_norm": 0.7943779417966186,
+      "learning_rate": 5.653646736881401e-07,
+      "loss": 1.1397,
+      "step": 20994
+    },
+    {
+      "epoch": 0.9671995208918782,
+      "grad_norm": 0.8761390229354855,
+      "learning_rate": 5.637814009248676e-07,
+      "loss": 1.2518,
+      "step": 20995
+    },
+    {
+      "epoch": 0.9672455889805132,
+      "grad_norm": 0.8118983774940502,
+      "learning_rate": 5.622003419385236e-07,
+      "loss": 1.2553,
+      "step": 20996
+    },
+    {
+      "epoch": 0.9672916570691482,
+      "grad_norm": 5.871959456427442,
+      "learning_rate": 5.606214967642798e-07,
+      "loss": 1.0841,
+      "step": 20997
+    },
+    {
+      "epoch": 0.9673377251577832,
+      "grad_norm": 0.8112602105865697,
+      "learning_rate": 5.59044865437297e-07,
+      "loss": 1.161,
+      "step": 20998
+    },
+    {
+      "epoch": 0.9673837932464182,
+      "grad_norm": 1.0310824583079146,
+      "learning_rate": 5.574704479926696e-07,
+      "loss": 1.3061,
+      "step": 20999
+    },
+    {
+      "epoch": 0.9674298613350532,
+      "grad_norm": 0.7380208238331403,
+      "learning_rate": 5.558982444654581e-07,
+      "loss": 1.3859,
+      "step": 21000
+    },
+    {
+      "epoch": 0.9674759294236882,
+      "grad_norm": 0.8056045817240463,
+      "learning_rate": 5.543282548906681e-07,
+      "loss": 0.8087,
+      "step": 21001
+    },
+    {
+      "epoch": 0.9675219975123233,
+      "grad_norm": 1.1379563481135726,
+      "learning_rate": 5.527604793032381e-07,
+      "loss": 1.1884,
+      "step": 21002
+    },
+    {
+      "epoch": 0.9675680656009582,
+      "grad_norm": 1.0759245801277193,
+      "learning_rate": 5.511949177380959e-07,
+      "loss": 0.9378,
+      "step": 21003
+    },
+    {
+      "epoch": 0.9676141336895933,
+      "grad_norm": 1.0957261401939917,
+      "learning_rate": 5.496315702300691e-07,
+      "loss": 1.1182,
+      "step": 21004
+    },
+    {
+      "epoch": 0.9676602017782282,
+      "grad_norm": 2.211554814222876,
+      "learning_rate": 5.480704368139744e-07,
+      "loss": 1.1162,
+      "step": 21005
+    },
+    {
+      "epoch": 0.9677062698668633,
+      "grad_norm": 0.7305791564369061,
+      "learning_rate": 5.465115175245616e-07,
+      "loss": 1.0422,
+      "step": 21006
+    },
+    {
+      "epoch": 0.9677523379554982,
+      "grad_norm": 3.3166680173713528,
+      "learning_rate": 5.449548123965586e-07,
+      "loss": 1.033,
+      "step": 21007
+    },
+    {
+      "epoch": 0.9677984060441333,
+      "grad_norm": 1.5232153090565588,
+      "learning_rate": 5.434003214646044e-07,
+      "loss": 1.3116,
+      "step": 21008
+    },
+    {
+      "epoch": 0.9678444741327682,
+      "grad_norm": 1.2769682470143349,
+      "learning_rate": 5.418480447633045e-07,
+      "loss": 1.264,
+      "step": 21009
+    },
+    {
+      "epoch": 0.9678905422214032,
+      "grad_norm": 5.253702117197599,
+      "learning_rate": 5.402979823272203e-07,
+      "loss": 1.0525,
+      "step": 21010
+    },
+    {
+      "epoch": 0.9679366103100382,
+      "grad_norm": 0.9299585491282656,
+      "learning_rate": 5.387501341908796e-07,
+      "loss": 1.198,
+      "step": 21011
+    },
+    {
+      "epoch": 0.9679826783986732,
+      "grad_norm": 1.9094630942100055,
+      "learning_rate": 5.372045003887106e-07,
+      "loss": 1.1265,
+      "step": 21012
+    },
+    {
+      "epoch": 0.9680287464873082,
+      "grad_norm": 1.0494607323556644,
+      "learning_rate": 5.356610809551521e-07,
+      "loss": 1.35,
+      "step": 21013
+    },
+    {
+      "epoch": 0.9680748145759432,
+      "grad_norm": 1.363132960068221,
+      "learning_rate": 5.341198759245547e-07,
+      "loss": 1.1245,
+      "step": 21014
+    },
+    {
+      "epoch": 0.9681208826645783,
+      "grad_norm": 0.9684782205679023,
+      "learning_rate": 5.325808853312242e-07,
+      "loss": 1.1586,
+      "step": 21015
+    },
+    {
+      "epoch": 0.9681669507532132,
+      "grad_norm": 1.1437822228816186,
+      "learning_rate": 5.310441092094331e-07,
+      "loss": 1.0792,
+      "step": 21016
+    },
+    {
+      "epoch": 0.9682130188418483,
+      "grad_norm": 1.531246042732847,
+      "learning_rate": 5.295095475933875e-07,
+      "loss": 1.2557,
+      "step": 21017
+    },
+    {
+      "epoch": 0.9682590869304832,
+      "grad_norm": 1.02918463321729,
+      "learning_rate": 5.279772005172712e-07,
+      "loss": 1.3713,
+      "step": 21018
+    },
+    {
+      "epoch": 0.9683051550191183,
+      "grad_norm": 0.9588118432259297,
+      "learning_rate": 5.26447068015179e-07,
+      "loss": 1.0243,
+      "step": 21019
+    },
+    {
+      "epoch": 0.9683512231077532,
+      "grad_norm": 0.8097808871468455,
+      "learning_rate": 5.249191501211726e-07,
+      "loss": 1.4809,
+      "step": 21020
+    },
+    {
+      "epoch": 0.9683972911963883,
+      "grad_norm": 1.0085293181965658,
+      "learning_rate": 5.233934468692914e-07,
+      "loss": 1.4278,
+      "step": 21021
+    },
+    {
+      "epoch": 0.9684433592850232,
+      "grad_norm": 1.0787680417786778,
+      "learning_rate": 5.21869958293475e-07,
+      "loss": 1.2427,
+      "step": 21022
+    },
+    {
+      "epoch": 0.9684894273736583,
+      "grad_norm": 1.16899877145216,
+      "learning_rate": 5.203486844276628e-07,
+      "loss": 1.3257,
+      "step": 21023
+    },
+    {
+      "epoch": 0.9685354954622932,
+      "grad_norm": 0.8637360852837145,
+      "learning_rate": 5.188296253057168e-07,
+      "loss": 1.3386,
+      "step": 21024
+    },
+    {
+      "epoch": 0.9685815635509283,
+      "grad_norm": 1.1137686389962302,
+      "learning_rate": 5.173127809614542e-07,
+      "loss": 1.4321,
+      "step": 21025
+    },
+    {
+      "epoch": 0.9686276316395632,
+      "grad_norm": 1.397941814858866,
+      "learning_rate": 5.157981514286481e-07,
+      "loss": 1.1672,
+      "step": 21026
+    },
+    {
+      "epoch": 0.9686736997281983,
+      "grad_norm": 1.0914654158082449,
+      "learning_rate": 5.142857367410159e-07,
+      "loss": 1.4238,
+      "step": 21027
+    },
+    {
+      "epoch": 0.9687197678168333,
+      "grad_norm": 1.0375406372147638,
+      "learning_rate": 5.127755369322307e-07,
+      "loss": 1.1595,
+      "step": 21028
+    },
+    {
+      "epoch": 0.9687658359054683,
+      "grad_norm": 1.2689506076059365,
+      "learning_rate": 5.112675520359211e-07,
+      "loss": 1.4444,
+      "step": 21029
+    },
+    {
+      "epoch": 0.9688119039941033,
+      "grad_norm": 7.044781810019725,
+      "learning_rate": 5.097617820856381e-07,
+      "loss": 1.2473,
+      "step": 21030
+    },
+    {
+      "epoch": 0.9688579720827383,
+      "grad_norm": 2.1339115076243518,
+      "learning_rate": 5.082582271149328e-07,
+      "loss": 1.381,
+      "step": 21031
+    },
+    {
+      "epoch": 0.9689040401713733,
+      "grad_norm": 0.8254510860182543,
+      "learning_rate": 5.06756887157267e-07,
+      "loss": 1.3811,
+      "step": 21032
+    },
+    {
+      "epoch": 0.9689501082600083,
+      "grad_norm": 2.6540411369975456,
+      "learning_rate": 5.052577622460586e-07,
+      "loss": 1.371,
+      "step": 21033
+    },
+    {
+      "epoch": 0.9689961763486433,
+      "grad_norm": 0.9829194570991752,
+      "learning_rate": 5.037608524146809e-07,
+      "loss": 1.4852,
+      "step": 21034
+    },
+    {
+      "epoch": 0.9690422444372783,
+      "grad_norm": 2.5510359299322474,
+      "learning_rate": 5.022661576964738e-07,
+      "loss": 1.2197,
+      "step": 21035
+    },
+    {
+      "epoch": 0.9690883125259133,
+      "grad_norm": 0.8777547953568571,
+      "learning_rate": 5.007736781247218e-07,
+      "loss": 0.8286,
+      "step": 21036
+    },
+    {
+      "epoch": 0.9691343806145483,
+      "grad_norm": 0.8371087770125958,
+      "learning_rate": 4.992834137326208e-07,
+      "loss": 1.3205,
+      "step": 21037
+    },
+    {
+      "epoch": 0.9691804487031833,
+      "grad_norm": 0.813009351483683,
+      "learning_rate": 4.977953645533773e-07,
+      "loss": 1.1901,
+      "step": 21038
+    },
+    {
+      "epoch": 0.9692265167918183,
+      "grad_norm": 2.504982796929434,
+      "learning_rate": 4.963095306201093e-07,
+      "loss": 0.9762,
+      "step": 21039
+    },
+    {
+      "epoch": 0.9692725848804533,
+      "grad_norm": 1.0050306973937917,
+      "learning_rate": 4.948259119659015e-07,
+      "loss": 1.4393,
+      "step": 21040
+    },
+    {
+      "epoch": 0.9693186529690884,
+      "grad_norm": 0.8587048358357702,
+      "learning_rate": 4.93344508623772e-07,
+      "loss": 1.0962,
+      "step": 21041
+    },
+    {
+      "epoch": 0.9693647210577233,
+      "grad_norm": 0.8202342870434315,
+      "learning_rate": 4.918653206267165e-07,
+      "loss": 0.9868,
+      "step": 21042
+    },
+    {
+      "epoch": 0.9694107891463584,
+      "grad_norm": 0.9719705734075861,
+      "learning_rate": 4.903883480076532e-07,
+      "loss": 0.9829,
+      "step": 21043
+    },
+    {
+      "epoch": 0.9694568572349933,
+      "grad_norm": 1.2549276885362803,
+      "learning_rate": 4.889135907994779e-07,
+      "loss": 1.248,
+      "step": 21044
+    },
+    {
+      "epoch": 0.9695029253236284,
+      "grad_norm": 0.8233273135362625,
+      "learning_rate": 4.87441049035009e-07,
+      "loss": 1.2328,
+      "step": 21045
+    },
+    {
+      "epoch": 0.9695489934122633,
+      "grad_norm": 1.6764179997353268,
+      "learning_rate": 4.859707227470423e-07,
+      "loss": 1.2649,
+      "step": 21046
+    },
+    {
+      "epoch": 0.9695950615008984,
+      "grad_norm": 0.8249611948045253,
+      "learning_rate": 4.845026119683071e-07,
+      "loss": 1.34,
+      "step": 21047
+    },
+    {
+      "epoch": 0.9696411295895333,
+      "grad_norm": 0.9501616300052824,
+      "learning_rate": 4.830367167314886e-07,
+      "loss": 1.2156,
+      "step": 21048
+    },
+    {
+      "epoch": 0.9696871976781684,
+      "grad_norm": 1.803270915228719,
+      "learning_rate": 4.815730370692273e-07,
+      "loss": 1.1393,
+      "step": 21049
+    },
+    {
+      "epoch": 0.9697332657668033,
+      "grad_norm": 0.9959961125292535,
+      "learning_rate": 4.801115730140971e-07,
+      "loss": 1.243,
+      "step": 21050
+    },
+    {
+      "epoch": 0.9697793338554384,
+      "grad_norm": 0.8480126915767322,
+      "learning_rate": 4.786523245986496e-07,
+      "loss": 1.4596,
+      "step": 21051
+    },
+    {
+      "epoch": 0.9698254019440733,
+      "grad_norm": 1.281475298900665,
+      "learning_rate": 4.771952918553591e-07,
+      "loss": 1.3801,
+      "step": 21052
+    },
+    {
+      "epoch": 0.9698714700327083,
+      "grad_norm": 1.1310236233649271,
+      "learning_rate": 4.7574047481667716e-07,
+      "loss": 1.2519,
+      "step": 21053
+    },
+    {
+      "epoch": 0.9699175381213433,
+      "grad_norm": 0.6528933708358319,
+      "learning_rate": 4.7428787351497804e-07,
+      "loss": 1.2788,
+      "step": 21054
+    },
+    {
+      "epoch": 0.9699636062099783,
+      "grad_norm": 0.8953376865715638,
+      "learning_rate": 4.7283748798261363e-07,
+      "loss": 1.3709,
+      "step": 21055
+    },
+    {
+      "epoch": 0.9700096742986134,
+      "grad_norm": 1.430199854718348,
+      "learning_rate": 4.7138931825186917e-07,
+      "loss": 1.0174,
+      "step": 21056
+    },
+    {
+      "epoch": 0.9700557423872483,
+      "grad_norm": 0.6077083899506508,
+      "learning_rate": 4.6994336435498554e-07,
+      "loss": 1.1756,
+      "step": 21057
+    },
+    {
+      "epoch": 0.9701018104758834,
+      "grad_norm": 0.9082724223797863,
+      "learning_rate": 4.684996263241592e-07,
+      "loss": 1.2714,
+      "step": 21058
+    },
+    {
+      "epoch": 0.9701478785645183,
+      "grad_norm": 1.3747932095724855,
+      "learning_rate": 4.670581041915312e-07,
+      "loss": 1.0192,
+      "step": 21059
+    },
+    {
+      "epoch": 0.9701939466531534,
+      "grad_norm": 1.0702397668303312,
+      "learning_rate": 4.656187979891868e-07,
+      "loss": 1.3433,
+      "step": 21060
+    },
+    {
+      "epoch": 0.9702400147417883,
+      "grad_norm": 0.9492304664788129,
+      "learning_rate": 4.6418170774917837e-07,
+      "loss": 1.1749,
+      "step": 21061
+    },
+    {
+      "epoch": 0.9702860828304234,
+      "grad_norm": 1.156419940817085,
+      "learning_rate": 4.6274683350349124e-07,
+      "loss": 1.1748,
+      "step": 21062
+    },
+    {
+      "epoch": 0.9703321509190583,
+      "grad_norm": 1.3227118606223451,
+      "learning_rate": 4.613141752840777e-07,
+      "loss": 1.4469,
+      "step": 21063
+    },
+    {
+      "epoch": 0.9703782190076934,
+      "grad_norm": 0.8852780633382087,
+      "learning_rate": 4.598837331228345e-07,
+      "loss": 1.2902,
+      "step": 21064
+    },
+    {
+      "epoch": 0.9704242870963283,
+      "grad_norm": 0.9956446739406158,
+      "learning_rate": 4.584555070516028e-07,
+      "loss": 1.529,
+      "step": 21065
+    },
+    {
+      "epoch": 0.9704703551849634,
+      "grad_norm": 1.1186682664054834,
+      "learning_rate": 4.5702949710217937e-07,
+      "loss": 1.5301,
+      "step": 21066
+    },
+    {
+      "epoch": 0.9705164232735983,
+      "grad_norm": 1.8527633573765314,
+      "learning_rate": 4.5560570330631657e-07,
+      "loss": 1.4652,
+      "step": 21067
+    },
+    {
+      "epoch": 0.9705624913622334,
+      "grad_norm": 1.5142967283325117,
+      "learning_rate": 4.5418412569572246e-07,
+      "loss": 1.0338,
+      "step": 21068
+    },
+    {
+      "epoch": 0.9706085594508684,
+      "grad_norm": 0.8492301541046421,
+      "learning_rate": 4.5276476430201606e-07,
+      "loss": 1.2963,
+      "step": 21069
+    },
+    {
+      "epoch": 0.9706546275395034,
+      "grad_norm": 0.9623172588294288,
+      "learning_rate": 4.5134761915682775e-07,
+      "loss": 0.9813,
+      "step": 21070
+    },
+    {
+      "epoch": 0.9707006956281384,
+      "grad_norm": 0.8126243630726285,
+      "learning_rate": 4.4993269029168784e-07,
+      "loss": 1.1012,
+      "step": 21071
+    },
+    {
+      "epoch": 0.9707467637167734,
+      "grad_norm": 1.1570960354287858,
+      "learning_rate": 4.485199777381155e-07,
+      "loss": 1.2303,
+      "step": 21072
+    },
+    {
+      "epoch": 0.9707928318054084,
+      "grad_norm": 0.8642638127307134,
+      "learning_rate": 4.471094815275412e-07,
+      "loss": 1.289,
+      "step": 21073
+    },
+    {
+      "epoch": 0.9708388998940434,
+      "grad_norm": 2.0554163067996196,
+      "learning_rate": 4.4570120169137306e-07,
+      "loss": 1.1584,
+      "step": 21074
+    },
+    {
+      "epoch": 0.9708849679826784,
+      "grad_norm": 0.8768950585961919,
+      "learning_rate": 4.44295138260975e-07,
+      "loss": 1.1568,
+      "step": 21075
+    },
+    {
+      "epoch": 0.9709310360713134,
+      "grad_norm": 1.1630846294973618,
+      "learning_rate": 4.4289129126765525e-07,
+      "loss": 1.1844,
+      "step": 21076
+    },
+    {
+      "epoch": 0.9709771041599484,
+      "grad_norm": 0.8437138256122254,
+      "learning_rate": 4.414896607426444e-07,
+      "loss": 1.2902,
+      "step": 21077
+    },
+    {
+      "epoch": 0.9710231722485834,
+      "grad_norm": 1.4007308130701868,
+      "learning_rate": 4.4009024671716195e-07,
+      "loss": 1.1949,
+      "step": 21078
+    },
+    {
+      "epoch": 0.9710692403372184,
+      "grad_norm": 0.7899688446866241,
+      "learning_rate": 4.3869304922236067e-07,
+      "loss": 1.1608,
+      "step": 21079
+    },
+    {
+      "epoch": 0.9711153084258534,
+      "grad_norm": 0.6771073470402948,
+      "learning_rate": 4.3729806828934906e-07,
+      "loss": 1.0549,
+      "step": 21080
+    },
+    {
+      "epoch": 0.9711613765144884,
+      "grad_norm": 1.1851279661036158,
+      "learning_rate": 4.3590530394919115e-07,
+      "loss": 1.489,
+      "step": 21081
+    },
+    {
+      "epoch": 0.9712074446031235,
+      "grad_norm": 1.1508784144653696,
+      "learning_rate": 4.345147562328733e-07,
+      "loss": 1.196,
+      "step": 21082
+    },
+    {
+      "epoch": 0.9712535126917584,
+      "grad_norm": 1.0687989718552908,
+      "learning_rate": 4.331264251713707e-07,
+      "loss": 1.0532,
+      "step": 21083
+    },
+    {
+      "epoch": 0.9712995807803935,
+      "grad_norm": 1.2900140013952708,
+      "learning_rate": 4.31740310795592e-07,
+      "loss": 1.2792,
+      "step": 21084
+    },
+    {
+      "epoch": 0.9713456488690284,
+      "grad_norm": 0.8709422471908733,
+      "learning_rate": 4.303564131363902e-07,
+      "loss": 1.0941,
+      "step": 21085
+    },
+    {
+      "epoch": 0.9713917169576635,
+      "grad_norm": 0.9850964259677648,
+      "learning_rate": 4.289747322245741e-07,
+      "loss": 1.4186,
+      "step": 21086
+    },
+    {
+      "epoch": 0.9714377850462984,
+      "grad_norm": 3.2263483034956484,
+      "learning_rate": 4.275952680909079e-07,
+      "loss": 1.3932,
+      "step": 21087
+    },
+    {
+      "epoch": 0.9714838531349335,
+      "grad_norm": 1.1156704177592924,
+      "learning_rate": 4.262180207661004e-07,
+      "loss": 1.4338,
+      "step": 21088
+    },
+    {
+      "epoch": 0.9715299212235684,
+      "grad_norm": 1.1708751390550294,
+      "learning_rate": 4.2484299028081596e-07,
+      "loss": 1.0522,
+      "step": 21089
+    },
+    {
+      "epoch": 0.9715759893122035,
+      "grad_norm": 1.57696479716273,
+      "learning_rate": 4.234701766656746e-07,
+      "loss": 1.0733,
+      "step": 21090
+    },
+    {
+      "epoch": 0.9716220574008384,
+      "grad_norm": 1.3757655603017798,
+      "learning_rate": 4.220995799512295e-07,
+      "loss": 0.8824,
+      "step": 21091
+    },
+    {
+      "epoch": 0.9716681254894735,
+      "grad_norm": 1.0816463916457746,
+      "learning_rate": 4.2073120016800083e-07,
+      "loss": 1.5185,
+      "step": 21092
+    },
+    {
+      "epoch": 0.9717141935781084,
+      "grad_norm": 0.8662112106278788,
+      "learning_rate": 4.1936503734644195e-07,
+      "loss": 1.2231,
+      "step": 21093
+    },
+    {
+      "epoch": 0.9717602616667435,
+      "grad_norm": 0.9003016646386848,
+      "learning_rate": 4.1800109151698407e-07,
+      "loss": 1.307,
+      "step": 21094
+    },
+    {
+      "epoch": 0.9718063297553784,
+      "grad_norm": 0.6999629445728986,
+      "learning_rate": 4.166393627099918e-07,
+      "loss": 1.1402,
+      "step": 21095
+    },
+    {
+      "epoch": 0.9718523978440134,
+      "grad_norm": 0.8720528290236851,
+      "learning_rate": 4.1527985095576316e-07,
+      "loss": 0.9398,
+      "step": 21096
+    },
+    {
+      "epoch": 0.9718984659326485,
+      "grad_norm": 0.9067174465211906,
+      "learning_rate": 4.139225562845739e-07,
+      "loss": 1.1519,
+      "step": 21097
+    },
+    {
+      "epoch": 0.9719445340212834,
+      "grad_norm": 1.0886410650674174,
+      "learning_rate": 4.1256747872665533e-07,
+      "loss": 1.5562,
+      "step": 21098
+    },
+    {
+      "epoch": 0.9719906021099185,
+      "grad_norm": 1.1153356655023978,
+      "learning_rate": 4.1121461831216124e-07,
+      "loss": 1.1573,
+      "step": 21099
+    },
+    {
+      "epoch": 0.9720366701985534,
+      "grad_norm": 1.8645152367222237,
+      "learning_rate": 4.0986397507121187e-07,
+      "loss": 1.0837,
+      "step": 21100
+    },
+    {
+      "epoch": 0.9720827382871885,
+      "grad_norm": 1.7126990618057119,
+      "learning_rate": 4.0851554903389435e-07,
+      "loss": 1.2981,
+      "step": 21101
+    },
+    {
+      "epoch": 0.9721288063758234,
+      "grad_norm": 1.6408318444930818,
+      "learning_rate": 4.0716934023020683e-07,
+      "loss": 1.1298,
+      "step": 21102
+    },
+    {
+      "epoch": 0.9721748744644585,
+      "grad_norm": 1.4494199673928299,
+      "learning_rate": 4.058253486901253e-07,
+      "loss": 0.9678,
+      "step": 21103
+    },
+    {
+      "epoch": 0.9722209425530934,
+      "grad_norm": 0.9721631538554378,
+      "learning_rate": 4.0448357444358155e-07,
+      "loss": 0.9453,
+      "step": 21104
+    },
+    {
+      "epoch": 0.9722670106417285,
+      "grad_norm": 0.942634949290194,
+      "learning_rate": 4.031440175204404e-07,
+      "loss": 1.1903,
+      "step": 21105
+    },
+    {
+      "epoch": 0.9723130787303634,
+      "grad_norm": 1.0659814178052691,
+      "learning_rate": 4.018066779505225e-07,
+      "loss": 1.1358,
+      "step": 21106
+    },
+    {
+      "epoch": 0.9723591468189985,
+      "grad_norm": 0.9586478707429877,
+      "learning_rate": 4.004715557636041e-07,
+      "loss": 1.0069,
+      "step": 21107
+    },
+    {
+      "epoch": 0.9724052149076334,
+      "grad_norm": 0.8975335349407545,
+      "learning_rate": 3.9913865098941684e-07,
+      "loss": 1.215,
+      "step": 21108
+    },
+    {
+      "epoch": 0.9724512829962685,
+      "grad_norm": 0.9777034778663388,
+      "learning_rate": 3.9780796365761484e-07,
+      "loss": 1.2128,
+      "step": 21109
+    },
+    {
+      "epoch": 0.9724973510849035,
+      "grad_norm": 3.184963484451482,
+      "learning_rate": 3.9647949379785223e-07,
+      "loss": 1.0507,
+      "step": 21110
+    },
+    {
+      "epoch": 0.9725434191735385,
+      "grad_norm": 4.56980500530989,
+      "learning_rate": 3.9515324143968304e-07,
+      "loss": 0.9522,
+      "step": 21111
+    },
+    {
+      "epoch": 0.9725894872621735,
+      "grad_norm": 0.9963030274585695,
+      "learning_rate": 3.9382920661263924e-07,
+      "loss": 1.4731,
+      "step": 21112
+    },
+    {
+      "epoch": 0.9726355553508085,
+      "grad_norm": 0.9362453815515587,
+      "learning_rate": 3.925073893462083e-07,
+      "loss": 1.4197,
+      "step": 21113
+    },
+    {
+      "epoch": 0.9726816234394435,
+      "grad_norm": 0.8020464495401222,
+      "learning_rate": 3.9118778966980017e-07,
+      "loss": 1.3084,
+      "step": 21114
+    },
+    {
+      "epoch": 0.9727276915280785,
+      "grad_norm": 1.1309201760649787,
+      "learning_rate": 3.8987040761281344e-07,
+      "loss": 1.3708,
+      "step": 21115
+    },
+    {
+      "epoch": 0.9727737596167135,
+      "grad_norm": 0.9574137459668906,
+      "learning_rate": 3.8855524320455807e-07,
+      "loss": 1.3633,
+      "step": 21116
+    },
+    {
+      "epoch": 0.9728198277053485,
+      "grad_norm": 1.0361253682126652,
+      "learning_rate": 3.8724229647433277e-07,
+      "loss": 1.3588,
+      "step": 21117
+    },
+    {
+      "epoch": 0.9728658957939835,
+      "grad_norm": 0.9559500373106524,
+      "learning_rate": 3.859315674513475e-07,
+      "loss": 1.3103,
+      "step": 21118
+    },
+    {
+      "epoch": 0.9729119638826185,
+      "grad_norm": 0.860299446027978,
+      "learning_rate": 3.846230561648012e-07,
+      "loss": 1.4521,
+      "step": 21119
+    },
+    {
+      "epoch": 0.9729580319712535,
+      "grad_norm": 0.7590878284933502,
+      "learning_rate": 3.83316762643815e-07,
+      "loss": 1.353,
+      "step": 21120
+    },
+    {
+      "epoch": 0.9730041000598885,
+      "grad_norm": 0.7024458182446184,
+      "learning_rate": 3.820126869174878e-07,
+      "loss": 1.2717,
+      "step": 21121
+    },
+    {
+      "epoch": 0.9730501681485235,
+      "grad_norm": 1.3775467446707343,
+      "learning_rate": 3.807108290148298e-07,
+      "loss": 1.3658,
+      "step": 21122
+    },
+    {
+      "epoch": 0.9730962362371586,
+      "grad_norm": 0.6926593265068647,
+      "learning_rate": 3.7941118896482887e-07,
+      "loss": 1.5881,
+      "step": 21123
+    },
+    {
+      "epoch": 0.9731423043257935,
+      "grad_norm": 1.2339874083142768,
+      "learning_rate": 3.781137667964285e-07,
+      "loss": 1.2186,
+      "step": 21124
+    },
+    {
+      "epoch": 0.9731883724144286,
+      "grad_norm": 1.0480829690279012,
+      "learning_rate": 3.768185625385168e-07,
+      "loss": 1.3272,
+      "step": 21125
+    },
+    {
+      "epoch": 0.9732344405030635,
+      "grad_norm": 0.8324052911307875,
+      "learning_rate": 3.75525576219915e-07,
+      "loss": 1.2474,
+      "step": 21126
+    },
+    {
+      "epoch": 0.9732805085916986,
+      "grad_norm": 2.1776287191171204,
+      "learning_rate": 3.7423480786942247e-07,
+      "loss": 1.026,
+      "step": 21127
+    },
+    {
+      "epoch": 0.9733265766803335,
+      "grad_norm": 0.9330085060458039,
+      "learning_rate": 3.7294625751577164e-07,
+      "loss": 1.1378,
+      "step": 21128
+    },
+    {
+      "epoch": 0.9733726447689686,
+      "grad_norm": 0.895108399425919,
+      "learning_rate": 3.716599251876396e-07,
+      "loss": 1.3131,
+      "step": 21129
+    },
+    {
+      "epoch": 0.9734187128576035,
+      "grad_norm": 1.5313102392820102,
+      "learning_rate": 3.703758109136812e-07,
+      "loss": 0.8308,
+      "step": 21130
+    },
+    {
+      "epoch": 0.9734647809462386,
+      "grad_norm": 1.4645606167617635,
+      "learning_rate": 3.6909391472247367e-07,
+      "loss": 1.2599,
+      "step": 21131
+    },
+    {
+      "epoch": 0.9735108490348735,
+      "grad_norm": 1.262803823680034,
+      "learning_rate": 3.6781423664256075e-07,
+      "loss": 1.2382,
+      "step": 21132
+    },
+    {
+      "epoch": 0.9735569171235086,
+      "grad_norm": 0.9490317296592755,
+      "learning_rate": 3.665367767024308e-07,
+      "loss": 1.1855,
+      "step": 21133
+    },
+    {
+      "epoch": 0.9736029852121435,
+      "grad_norm": 0.8272788606891935,
+      "learning_rate": 3.652615349305277e-07,
+      "loss": 1.2381,
+      "step": 21134
+    },
+    {
+      "epoch": 0.9736490533007786,
+      "grad_norm": 0.799737395631192,
+      "learning_rate": 3.639885113552399e-07,
+      "loss": 1.0419,
+      "step": 21135
+    },
+    {
+      "epoch": 0.9736951213894135,
+      "grad_norm": 1.3260239762037056,
+      "learning_rate": 3.627177060049114e-07,
+      "loss": 1.1014,
+      "step": 21136
+    },
+    {
+      "epoch": 0.9737411894780486,
+      "grad_norm": 0.8602047160033433,
+      "learning_rate": 3.614491189078195e-07,
+      "loss": 1.0885,
+      "step": 21137
+    },
+    {
+      "epoch": 0.9737872575666836,
+      "grad_norm": 1.912233544336441,
+      "learning_rate": 3.6018275009223057e-07,
+      "loss": 0.9648,
+      "step": 21138
+    },
+    {
+      "epoch": 0.9738333256553185,
+      "grad_norm": 0.9652585762436086,
+      "learning_rate": 3.58918599586322e-07,
+      "loss": 1.3304,
+      "step": 21139
+    },
+    {
+      "epoch": 0.9738793937439536,
+      "grad_norm": 1.3640689790321605,
+      "learning_rate": 3.576566674182491e-07,
+      "loss": 1.0773,
+      "step": 21140
+    },
+    {
+      "epoch": 0.9739254618325885,
+      "grad_norm": 1.2265841128265338,
+      "learning_rate": 3.5639695361608936e-07,
+      "loss": 0.9005,
+      "step": 21141
+    },
+    {
+      "epoch": 0.9739715299212236,
+      "grad_norm": 1.4711010906707531,
+      "learning_rate": 3.551394582079093e-07,
+      "loss": 1.1376,
+      "step": 21142
+    },
+    {
+      "epoch": 0.9740175980098585,
+      "grad_norm": 0.9901826209027569,
+      "learning_rate": 3.5388418122168643e-07,
+      "loss": 1.352,
+      "step": 21143
+    },
+    {
+      "epoch": 0.9740636660984936,
+      "grad_norm": 0.995803645849434,
+      "learning_rate": 3.5263112268538735e-07,
+      "loss": 0.9947,
+      "step": 21144
+    },
+    {
+      "epoch": 0.9741097341871285,
+      "grad_norm": 0.9725896044857113,
+      "learning_rate": 3.513802826268897e-07,
+      "loss": 1.2995,
+      "step": 21145
+    },
+    {
+      "epoch": 0.9741558022757636,
+      "grad_norm": 2.3190195802537454,
+      "learning_rate": 3.5013166107404907e-07,
+      "loss": 0.9028,
+      "step": 21146
+    },
+    {
+      "epoch": 0.9742018703643985,
+      "grad_norm": 0.8837636852980699,
+      "learning_rate": 3.4888525805466534e-07,
+      "loss": 1.4158,
+      "step": 21147
+    },
+    {
+      "epoch": 0.9742479384530336,
+      "grad_norm": 1.5514354262191792,
+      "learning_rate": 3.47641073596483e-07,
+      "loss": 1.1808,
+      "step": 21148
+    },
+    {
+      "epoch": 0.9742940065416685,
+      "grad_norm": 4.318532092441203,
+      "learning_rate": 3.463991077272133e-07,
+      "loss": 1.3682,
+      "step": 21149
+    },
+    {
+      "epoch": 0.9743400746303036,
+      "grad_norm": 1.6163169092069671,
+      "learning_rate": 3.451593604744896e-07,
+      "loss": 1.314,
+      "step": 21150
+    },
+    {
+      "epoch": 0.9743861427189386,
+      "grad_norm": 1.3698021620889438,
+      "learning_rate": 3.439218318659232e-07,
+      "loss": 0.8009,
+      "step": 21151
+    },
+    {
+      "epoch": 0.9744322108075736,
+      "grad_norm": 2.234016462690623,
+      "learning_rate": 3.426865219290698e-07,
+      "loss": 1.2827,
+      "step": 21152
+    },
+    {
+      "epoch": 0.9744782788962086,
+      "grad_norm": 0.7962020919083023,
+      "learning_rate": 3.4145343069142964e-07,
+      "loss": 1.2053,
+      "step": 21153
+    },
+    {
+      "epoch": 0.9745243469848436,
+      "grad_norm": 1.298438909474489,
+      "learning_rate": 3.4022255818043636e-07,
+      "loss": 1.2293,
+      "step": 21154
+    },
+    {
+      "epoch": 0.9745704150734786,
+      "grad_norm": 2.0449596491054547,
+      "learning_rate": 3.389939044235124e-07,
+      "loss": 1.4993,
+      "step": 21155
+    },
+    {
+      "epoch": 0.9746164831621136,
+      "grad_norm": 1.0062432150209468,
+      "learning_rate": 3.3776746944801374e-07,
+      "loss": 1.0509,
+      "step": 21156
+    },
+    {
+      "epoch": 0.9746625512507486,
+      "grad_norm": 1.246338518304605,
+      "learning_rate": 3.365432532812296e-07,
+      "loss": 1.4292,
+      "step": 21157
+    },
+    {
+      "epoch": 0.9747086193393836,
+      "grad_norm": 9.524883751019233,
+      "learning_rate": 3.3532125595043816e-07,
+      "loss": 1.09,
+      "step": 21158
+    },
+    {
+      "epoch": 0.9747546874280186,
+      "grad_norm": 0.9767695754713954,
+      "learning_rate": 3.3410147748280663e-07,
+      "loss": 1.2878,
+      "step": 21159
+    },
+    {
+      "epoch": 0.9748007555166536,
+      "grad_norm": 0.8776089476986962,
+      "learning_rate": 3.328839179055354e-07,
+      "loss": 1.0169,
+      "step": 21160
+    },
+    {
+      "epoch": 0.9748468236052886,
+      "grad_norm": 0.9546669046470443,
+      "learning_rate": 3.316685772457029e-07,
+      "loss": 1.1661,
+      "step": 21161
+    },
+    {
+      "epoch": 0.9748928916939236,
+      "grad_norm": 1.0817428938045022,
+      "learning_rate": 3.304554555303763e-07,
+      "loss": 1.1538,
+      "step": 21162
+    },
+    {
+      "epoch": 0.9749389597825586,
+      "grad_norm": 0.7529264347780198,
+      "learning_rate": 3.292445527865673e-07,
+      "loss": 1.1362,
+      "step": 21163
+    },
+    {
+      "epoch": 0.9749850278711937,
+      "grad_norm": 1.0205344710405015,
+      "learning_rate": 3.280358690412211e-07,
+      "loss": 1.3483,
+      "step": 21164
+    },
+    {
+      "epoch": 0.9750310959598286,
+      "grad_norm": 0.7301848018266032,
+      "learning_rate": 3.268294043212605e-07,
+      "loss": 1.1181,
+      "step": 21165
+    },
+    {
+      "epoch": 0.9750771640484637,
+      "grad_norm": 0.717951421837857,
+      "learning_rate": 3.256251586535419e-07,
+      "loss": 1.3692,
+      "step": 21166
+    },
+    {
+      "epoch": 0.9751232321370986,
+      "grad_norm": 0.7519492439707648,
+      "learning_rate": 3.2442313206486607e-07,
+      "loss": 0.9334,
+      "step": 21167
+    },
+    {
+      "epoch": 0.9751693002257337,
+      "grad_norm": 1.4034252417548745,
+      "learning_rate": 3.232233245820115e-07,
+      "loss": 1.3059,
+      "step": 21168
+    },
+    {
+      "epoch": 0.9752153683143686,
+      "grad_norm": 1.3660495112274935,
+      "learning_rate": 3.220257362316792e-07,
+      "loss": 1.2662,
+      "step": 21169
+    },
+    {
+      "epoch": 0.9752614364030037,
+      "grad_norm": 0.9144597190556295,
+      "learning_rate": 3.208303670405255e-07,
+      "loss": 1.1707,
+      "step": 21170
+    },
+    {
+      "epoch": 0.9753075044916386,
+      "grad_norm": 0.8918258528864261,
+      "learning_rate": 3.196372170351847e-07,
+      "loss": 1.2019,
+      "step": 21171
+    },
+    {
+      "epoch": 0.9753535725802737,
+      "grad_norm": 0.9887440887393723,
+      "learning_rate": 3.184462862421911e-07,
+      "loss": 1.3696,
+      "step": 21172
+    },
+    {
+      "epoch": 0.9753996406689086,
+      "grad_norm": 2.6281410577591573,
+      "learning_rate": 3.17257574688079e-07,
+      "loss": 1.0994,
+      "step": 21173
+    },
+    {
+      "epoch": 0.9754457087575437,
+      "grad_norm": 0.9609554115838606,
+      "learning_rate": 3.1607108239930515e-07,
+      "loss": 1.3233,
+      "step": 21174
+    },
+    {
+      "epoch": 0.9754917768461786,
+      "grad_norm": 0.8801339178260763,
+      "learning_rate": 3.1488680940229274e-07,
+      "loss": 1.1041,
+      "step": 21175
+    },
+    {
+      "epoch": 0.9755378449348137,
+      "grad_norm": 1.1422856377446962,
+      "learning_rate": 3.137047557234096e-07,
+      "loss": 1.2231,
+      "step": 21176
+    },
+    {
+      "epoch": 0.9755839130234487,
+      "grad_norm": 1.9834653321948619,
+      "learning_rate": 3.1252492138894584e-07,
+      "loss": 1.2626,
+      "step": 21177
+    },
+    {
+      "epoch": 0.9756299811120837,
+      "grad_norm": 1.0371717517873278,
+      "learning_rate": 3.113473064252026e-07,
+      "loss": 1.4046,
+      "step": 21178
+    },
+    {
+      "epoch": 0.9756760492007187,
+      "grad_norm": 0.8265475756026227,
+      "learning_rate": 3.1017191085838115e-07,
+      "loss": 1.0223,
+      "step": 21179
+    },
+    {
+      "epoch": 0.9757221172893537,
+      "grad_norm": 1.5616914544909952,
+      "learning_rate": 3.0899873471464947e-07,
+      "loss": 1.1799,
+      "step": 21180
+    },
+    {
+      "epoch": 0.9757681853779887,
+      "grad_norm": 0.7128538775859697,
+      "learning_rate": 3.078277780201089e-07,
+      "loss": 1.1234,
+      "step": 21181
+    },
+    {
+      "epoch": 0.9758142534666236,
+      "grad_norm": 1.0365942714712062,
+      "learning_rate": 3.066590408008718e-07,
+      "loss": 0.7107,
+      "step": 21182
+    },
+    {
+      "epoch": 0.9758603215552587,
+      "grad_norm": 3.052239254511449,
+      "learning_rate": 3.0549252308291755e-07,
+      "loss": 1.407,
+      "step": 21183
+    },
+    {
+      "epoch": 0.9759063896438936,
+      "grad_norm": 1.3339387601301735,
+      "learning_rate": 3.0432822489224745e-07,
+      "loss": 1.28,
+      "step": 21184
+    },
+    {
+      "epoch": 0.9759524577325287,
+      "grad_norm": 1.2815110596446246,
+      "learning_rate": 3.031661462547519e-07,
+      "loss": 1.0846,
+      "step": 21185
+    },
+    {
+      "epoch": 0.9759985258211636,
+      "grad_norm": 1.2017259929548751,
+      "learning_rate": 3.020062871963325e-07,
+      "loss": 1.0021,
+      "step": 21186
+    },
+    {
+      "epoch": 0.9760445939097987,
+      "grad_norm": 1.0993589204272938,
+      "learning_rate": 3.0084864774280185e-07,
+      "loss": 1.2892,
+      "step": 21187
+    },
+    {
+      "epoch": 0.9760906619984336,
+      "grad_norm": 1.0462732498288707,
+      "learning_rate": 2.9969322791990606e-07,
+      "loss": 1.2299,
+      "step": 21188
+    },
+    {
+      "epoch": 0.9761367300870687,
+      "grad_norm": 1.0918425447584696,
+      "learning_rate": 2.985400277534134e-07,
+      "loss": 1.2712,
+      "step": 21189
+    },
+    {
+      "epoch": 0.9761827981757036,
+      "grad_norm": 0.9258318292480436,
+      "learning_rate": 2.973890472689589e-07,
+      "loss": 1.2308,
+      "step": 21190
+    },
+    {
+      "epoch": 0.9762288662643387,
+      "grad_norm": 0.9659501880515466,
+      "learning_rate": 2.962402864921998e-07,
+      "loss": 1.2839,
+      "step": 21191
+    },
+    {
+      "epoch": 0.9762749343529737,
+      "grad_norm": 1.3529326062723643,
+      "learning_rate": 2.950937454486824e-07,
+      "loss": 0.8212,
+      "step": 21192
+    },
+    {
+      "epoch": 0.9763210024416087,
+      "grad_norm": 0.8813861625946684,
+      "learning_rate": 2.939494241639529e-07,
+      "loss": 1.2156,
+      "step": 21193
+    },
+    {
+      "epoch": 0.9763670705302437,
+      "grad_norm": 0.7981310447875732,
+      "learning_rate": 2.9280732266347976e-07,
+      "loss": 1.3363,
+      "step": 21194
+    },
+    {
+      "epoch": 0.9764131386188787,
+      "grad_norm": 1.3121779136684535,
+      "learning_rate": 2.9166744097268716e-07,
+      "loss": 1.3199,
+      "step": 21195
+    },
+    {
+      "epoch": 0.9764592067075137,
+      "grad_norm": 0.9129194535482179,
+      "learning_rate": 2.905297791169548e-07,
+      "loss": 1.2246,
+      "step": 21196
+    },
+    {
+      "epoch": 0.9765052747961487,
+      "grad_norm": 0.6444238949310115,
+      "learning_rate": 2.8939433712161787e-07,
+      "loss": 1.1474,
+      "step": 21197
+    },
+    {
+      "epoch": 0.9765513428847837,
+      "grad_norm": 1.1936108003917931,
+      "learning_rate": 2.882611150119452e-07,
+      "loss": 1.1073,
+      "step": 21198
+    },
+    {
+      "epoch": 0.9765974109734187,
+      "grad_norm": 1.1958131751985013,
+      "learning_rate": 2.8713011281316093e-07,
+      "loss": 0.967,
+      "step": 21199
+    },
+    {
+      "epoch": 0.9766434790620537,
+      "grad_norm": 1.5473239034210489,
+      "learning_rate": 2.860013305504672e-07,
+      "loss": 1.2018,
+      "step": 21200
+    },
+    {
+      "epoch": 0.9766895471506887,
+      "grad_norm": 0.8450164610348756,
+      "learning_rate": 2.848747682489772e-07,
+      "loss": 1.3699,
+      "step": 21201
+    },
+    {
+      "epoch": 0.9767356152393237,
+      "grad_norm": 0.9502254005035047,
+      "learning_rate": 2.83750425933782e-07,
+      "loss": 1.1577,
+      "step": 21202
+    },
+    {
+      "epoch": 0.9767816833279587,
+      "grad_norm": 0.9538596569411414,
+      "learning_rate": 2.826283036298949e-07,
+      "loss": 1.2203,
+      "step": 21203
+    },
+    {
+      "epoch": 0.9768277514165937,
+      "grad_norm": 1.0545447053204628,
+      "learning_rate": 2.8150840136231814e-07,
+      "loss": 1.4029,
+      "step": 21204
+    },
+    {
+      "epoch": 0.9768738195052288,
+      "grad_norm": 2.3655824702963564,
+      "learning_rate": 2.8039071915597625e-07,
+      "loss": 1.1448,
+      "step": 21205
+    },
+    {
+      "epoch": 0.9769198875938637,
+      "grad_norm": 0.8806414629060967,
+      "learning_rate": 2.792752570357493e-07,
+      "loss": 1.1393,
+      "step": 21206
+    },
+    {
+      "epoch": 0.9769659556824988,
+      "grad_norm": 0.9579875771812763,
+      "learning_rate": 2.78162015026473e-07,
+      "loss": 1.3581,
+      "step": 21207
+    },
+    {
+      "epoch": 0.9770120237711337,
+      "grad_norm": 1.3622136763914405,
+      "learning_rate": 2.7705099315292747e-07,
+      "loss": 1.3187,
+      "step": 21208
+    },
+    {
+      "epoch": 0.9770580918597688,
+      "grad_norm": 2.2124202143246725,
+      "learning_rate": 2.7594219143987075e-07,
+      "loss": 1.0763,
+      "step": 21209
+    },
+    {
+      "epoch": 0.9771041599484037,
+      "grad_norm": 0.8729723975389349,
+      "learning_rate": 2.7483560991194976e-07,
+      "loss": 1.1075,
+      "step": 21210
+    },
+    {
+      "epoch": 0.9771502280370388,
+      "grad_norm": 1.0239059123873018,
+      "learning_rate": 2.737312485938337e-07,
+      "loss": 1.1576,
+      "step": 21211
+    },
+    {
+      "epoch": 0.9771962961256737,
+      "grad_norm": 0.9165979617406633,
+      "learning_rate": 2.7262910751010284e-07,
+      "loss": 1.194,
+      "step": 21212
+    },
+    {
+      "epoch": 0.9772423642143088,
+      "grad_norm": 0.9104723618714932,
+      "learning_rate": 2.7152918668528206e-07,
+      "loss": 1.1754,
+      "step": 21213
+    },
+    {
+      "epoch": 0.9772884323029437,
+      "grad_norm": 0.7736278430373944,
+      "learning_rate": 2.7043148614386284e-07,
+      "loss": 1.3318,
+      "step": 21214
+    },
+    {
+      "epoch": 0.9773345003915788,
+      "grad_norm": 1.521184070930073,
+      "learning_rate": 2.693360059102923e-07,
+      "loss": 1.4039,
+      "step": 21215
+    },
+    {
+      "epoch": 0.9773805684802137,
+      "grad_norm": 1.8795403718011847,
+      "learning_rate": 2.6824274600896207e-07,
+      "loss": 1.4383,
+      "step": 21216
+    },
+    {
+      "epoch": 0.9774266365688488,
+      "grad_norm": 1.0846179715951452,
+      "learning_rate": 2.6715170646418597e-07,
+      "loss": 1.0283,
+      "step": 21217
+    },
+    {
+      "epoch": 0.9774727046574838,
+      "grad_norm": 1.8899889642097711,
+      "learning_rate": 2.6606288730028904e-07,
+      "loss": 1.3099,
+      "step": 21218
+    },
+    {
+      "epoch": 0.9775187727461188,
+      "grad_norm": 0.8310349118272449,
+      "learning_rate": 2.6497628854148526e-07,
+      "loss": 1.2266,
+      "step": 21219
+    },
+    {
+      "epoch": 0.9775648408347538,
+      "grad_norm": 1.0790832976580582,
+      "learning_rate": 2.6389191021197747e-07,
+      "loss": 0.9552,
+      "step": 21220
+    },
+    {
+      "epoch": 0.9776109089233888,
+      "grad_norm": 0.9194194976387122,
+      "learning_rate": 2.628097523359019e-07,
+      "loss": 1.3354,
+      "step": 21221
+    },
+    {
+      "epoch": 0.9776569770120238,
+      "grad_norm": 2.0479610677578157,
+      "learning_rate": 2.617298149373615e-07,
+      "loss": 0.9733,
+      "step": 21222
+    },
+    {
+      "epoch": 0.9777030451006588,
+      "grad_norm": 1.169404092601652,
+      "learning_rate": 2.606520980403926e-07,
+      "loss": 1.2506,
+      "step": 21223
+    },
+    {
+      "epoch": 0.9777491131892938,
+      "grad_norm": 1.1624370053158473,
+      "learning_rate": 2.5957660166897604e-07,
+      "loss": 1.1368,
+      "step": 21224
+    },
+    {
+      "epoch": 0.9777951812779287,
+      "grad_norm": 0.7591320310501957,
+      "learning_rate": 2.585033258470815e-07,
+      "loss": 1.0662,
+      "step": 21225
+    },
+    {
+      "epoch": 0.9778412493665638,
+      "grad_norm": 0.6445484923450219,
+      "learning_rate": 2.574322705985788e-07,
+      "loss": 1.1904,
+      "step": 21226
+    },
+    {
+      "epoch": 0.9778873174551987,
+      "grad_norm": 2.732883865638248,
+      "learning_rate": 2.563634359473377e-07,
+      "loss": 1.2176,
+      "step": 21227
+    },
+    {
+      "epoch": 0.9779333855438338,
+      "grad_norm": 0.6858753469753139,
+      "learning_rate": 2.5529682191713924e-07,
+      "loss": 1.1363,
+      "step": 21228
+    },
+    {
+      "epoch": 0.9779794536324687,
+      "grad_norm": 0.8475263779905189,
+      "learning_rate": 2.542324285317199e-07,
+      "loss": 1.2362,
+      "step": 21229
+    },
+    {
+      "epoch": 0.9780255217211038,
+      "grad_norm": 1.0759885832660872,
+      "learning_rate": 2.5317025581479414e-07,
+      "loss": 1.0516,
+      "step": 21230
+    },
+    {
+      "epoch": 0.9780715898097387,
+      "grad_norm": 0.7933084574009732,
+      "learning_rate": 2.5211030379000967e-07,
+      "loss": 0.9362,
+      "step": 21231
+    },
+    {
+      "epoch": 0.9781176578983738,
+      "grad_norm": 0.6704706864475085,
+      "learning_rate": 2.5105257248095873e-07,
+      "loss": 1.4371,
+      "step": 21232
+    },
+    {
+      "epoch": 0.9781637259870088,
+      "grad_norm": 1.2850577400081955,
+      "learning_rate": 2.49997061911178e-07,
+      "loss": 1.1785,
+      "step": 21233
+    },
+    {
+      "epoch": 0.9782097940756438,
+      "grad_norm": 1.52322099371598,
+      "learning_rate": 2.489437721041932e-07,
+      "loss": 1.1047,
+      "step": 21234
+    },
+    {
+      "epoch": 0.9782558621642788,
+      "grad_norm": 2.4381377444695067,
+      "learning_rate": 2.4789270308342993e-07,
+      "loss": 1.0182,
+      "step": 21235
+    },
+    {
+      "epoch": 0.9783019302529138,
+      "grad_norm": 1.3547768206251152,
+      "learning_rate": 2.468438548722918e-07,
+      "loss": 1.3044,
+      "step": 21236
+    },
+    {
+      "epoch": 0.9783479983415488,
+      "grad_norm": 1.0506968635281781,
+      "learning_rate": 2.457972274941378e-07,
+      "loss": 0.9746,
+      "step": 21237
+    },
+    {
+      "epoch": 0.9783940664301838,
+      "grad_norm": 1.6755922988501415,
+      "learning_rate": 2.447528209722716e-07,
+      "loss": 1.3711,
+      "step": 21238
+    },
+    {
+      "epoch": 0.9784401345188188,
+      "grad_norm": 1.6101009593788775,
+      "learning_rate": 2.4371063532994123e-07,
+      "loss": 1.3698,
+      "step": 21239
+    },
+    {
+      "epoch": 0.9784862026074538,
+      "grad_norm": 2.6286481091434024,
+      "learning_rate": 2.426706705903503e-07,
+      "loss": 1.0483,
+      "step": 21240
+    },
+    {
+      "epoch": 0.9785322706960888,
+      "grad_norm": 1.1534326349215405,
+      "learning_rate": 2.4163292677664705e-07,
+      "loss": 1.0934,
+      "step": 21241
+    },
+    {
+      "epoch": 0.9785783387847238,
+      "grad_norm": 1.10848192814595,
+      "learning_rate": 2.405974039119352e-07,
+      "loss": 1.2851,
+      "step": 21242
+    },
+    {
+      "epoch": 0.9786244068733588,
+      "grad_norm": 1.4351352187080155,
+      "learning_rate": 2.395641020192629e-07,
+      "loss": 1.1258,
+      "step": 21243
+    },
+    {
+      "epoch": 0.9786704749619938,
+      "grad_norm": 3.533553349379314,
+      "learning_rate": 2.3853302112165633e-07,
+      "loss": 1.3492,
+      "step": 21244
+    },
+    {
+      "epoch": 0.9787165430506288,
+      "grad_norm": 1.1025155701676128,
+      "learning_rate": 2.3750416124205254e-07,
+      "loss": 1.0425,
+      "step": 21245
+    },
+    {
+      "epoch": 0.9787626111392639,
+      "grad_norm": 1.2755445114939579,
+      "learning_rate": 2.3647752240336661e-07,
+      "loss": 1.1243,
+      "step": 21246
+    },
+    {
+      "epoch": 0.9788086792278988,
+      "grad_norm": 1.3197806749324141,
+      "learning_rate": 2.3545310462844695e-07,
+      "loss": 1.0491,
+      "step": 21247
+    },
+    {
+      "epoch": 0.9788547473165339,
+      "grad_norm": 0.8510919253869964,
+      "learning_rate": 2.3443090794010858e-07,
+      "loss": 1.2026,
+      "step": 21248
+    },
+    {
+      "epoch": 0.9789008154051688,
+      "grad_norm": 1.359069284100844,
+      "learning_rate": 2.334109323611e-07,
+      "loss": 1.0918,
+      "step": 21249
+    },
+    {
+      "epoch": 0.9789468834938039,
+      "grad_norm": 1.1666322081965894,
+      "learning_rate": 2.323931779141364e-07,
+      "loss": 1.4452,
+      "step": 21250
+    },
+    {
+      "epoch": 0.9789929515824388,
+      "grad_norm": 1.0568399738450323,
+      "learning_rate": 2.3137764462187738e-07,
+      "loss": 0.9107,
+      "step": 21251
+    },
+    {
+      "epoch": 0.9790390196710739,
+      "grad_norm": 1.1528612787268109,
+      "learning_rate": 2.3036433250692714e-07,
+      "loss": 1.1997,
+      "step": 21252
+    },
+    {
+      "epoch": 0.9790850877597088,
+      "grad_norm": 1.5598502226475104,
+      "learning_rate": 2.2935324159184535e-07,
+      "loss": 1.1272,
+      "step": 21253
+    },
+    {
+      "epoch": 0.9791311558483439,
+      "grad_norm": 0.7271105929175444,
+      "learning_rate": 2.2834437189913627e-07,
+      "loss": 1.0543,
+      "step": 21254
+    },
+    {
+      "epoch": 0.9791772239369788,
+      "grad_norm": 0.7699599796354021,
+      "learning_rate": 2.273377234512819e-07,
+      "loss": 1.2328,
+      "step": 21255
+    },
+    {
+      "epoch": 0.9792232920256139,
+      "grad_norm": 0.9042840321781481,
+      "learning_rate": 2.2633329627067545e-07,
+      "loss": 1.3182,
+      "step": 21256
+    },
+    {
+      "epoch": 0.9792693601142488,
+      "grad_norm": 1.434997294971927,
+      "learning_rate": 2.253310903796768e-07,
+      "loss": 1.0047,
+      "step": 21257
+    },
+    {
+      "epoch": 0.9793154282028839,
+      "grad_norm": 0.6553884702530819,
+      "learning_rate": 2.2433110580060147e-07,
+      "loss": 1.2259,
+      "step": 21258
+    },
+    {
+      "epoch": 0.9793614962915189,
+      "grad_norm": 0.7853876217436601,
+      "learning_rate": 2.2333334255573157e-07,
+      "loss": 1.2477,
+      "step": 21259
+    },
+    {
+      "epoch": 0.9794075643801539,
+      "grad_norm": 1.2808177018415887,
+      "learning_rate": 2.2233780066724941e-07,
+      "loss": 1.2366,
+      "step": 21260
+    },
+    {
+      "epoch": 0.9794536324687889,
+      "grad_norm": 1.0275977436694401,
+      "learning_rate": 2.213444801573372e-07,
+      "loss": 0.9564,
+      "step": 21261
+    },
+    {
+      "epoch": 0.9794997005574239,
+      "grad_norm": 1.6363514439969666,
+      "learning_rate": 2.2035338104811064e-07,
+      "loss": 1.4406,
+      "step": 21262
+    },
+    {
+      "epoch": 0.9795457686460589,
+      "grad_norm": 1.547687489932154,
+      "learning_rate": 2.1936450336162984e-07,
+      "loss": 1.188,
+      "step": 21263
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 1.0176015457115688,
+      "learning_rate": 2.1837784711991048e-07,
+      "loss": 1.285,
+      "step": 21264
+    },
+    {
+      "epoch": 0.9796379048233289,
+      "grad_norm": 1.0590997922822296,
+      "learning_rate": 2.1739341234491283e-07,
+      "loss": 1.3419,
+      "step": 21265
+    },
+    {
+      "epoch": 0.9796839729119639,
+      "grad_norm": 1.0177045762844241,
+      "learning_rate": 2.1641119905857487e-07,
+      "loss": 1.3365,
+      "step": 21266
+    },
+    {
+      "epoch": 0.9797300410005989,
+      "grad_norm": 1.53745688946991,
+      "learning_rate": 2.154312072827347e-07,
+      "loss": 1.2437,
+      "step": 21267
+    },
+    {
+      "epoch": 0.9797761090892338,
+      "grad_norm": 0.6800635937015436,
+      "learning_rate": 2.1445343703924147e-07,
+      "loss": 1.2187,
+      "step": 21268
+    },
+    {
+      "epoch": 0.9798221771778689,
+      "grad_norm": 0.8916874819343877,
+      "learning_rate": 2.134778883498334e-07,
+      "loss": 1.1632,
+      "step": 21269
+    },
+    {
+      "epoch": 0.9798682452665038,
+      "grad_norm": 1.3889090848841414,
+      "learning_rate": 2.125045612362486e-07,
+      "loss": 1.357,
+      "step": 21270
+    },
+    {
+      "epoch": 0.9799143133551389,
+      "grad_norm": 1.3167724599313209,
+      "learning_rate": 2.115334557201476e-07,
+      "loss": 1.4333,
+      "step": 21271
+    },
+    {
+      "epoch": 0.9799603814437738,
+      "grad_norm": 1.3797323190302817,
+      "learning_rate": 2.1056457182314637e-07,
+      "loss": 1.2384,
+      "step": 21272
+    },
+    {
+      "epoch": 0.9800064495324089,
+      "grad_norm": 1.3552553327779142,
+      "learning_rate": 2.0959790956683877e-07,
+      "loss": 1.1106,
+      "step": 21273
+    },
+    {
+      "epoch": 0.980052517621044,
+      "grad_norm": 0.8568059788281315,
+      "learning_rate": 2.0863346897271873e-07,
+      "loss": 1.2942,
+      "step": 21274
+    },
+    {
+      "epoch": 0.9800985857096789,
+      "grad_norm": 0.8388115359002825,
+      "learning_rate": 2.0767125006226906e-07,
+      "loss": 1.2643,
+      "step": 21275
+    },
+    {
+      "epoch": 0.980144653798314,
+      "grad_norm": 1.6536024808194598,
+      "learning_rate": 2.0671125285691707e-07,
+      "loss": 1.4981,
+      "step": 21276
+    },
+    {
+      "epoch": 0.9801907218869489,
+      "grad_norm": 1.7666493599080697,
+      "learning_rate": 2.0575347737803452e-07,
+      "loss": 0.9661,
+      "step": 21277
+    },
+    {
+      "epoch": 0.9802367899755839,
+      "grad_norm": 2.2622104983143445,
+      "learning_rate": 2.0479792364692662e-07,
+      "loss": 0.7331,
+      "step": 21278
+    },
+    {
+      "epoch": 0.9802828580642189,
+      "grad_norm": 5.449387332879009,
+      "learning_rate": 2.0384459168489857e-07,
+      "loss": 1.0827,
+      "step": 21279
+    },
+    {
+      "epoch": 0.9803289261528539,
+      "grad_norm": 1.1616550029619188,
+      "learning_rate": 2.0289348151314446e-07,
+      "loss": 1.0824,
+      "step": 21280
+    },
+    {
+      "epoch": 0.9803749942414889,
+      "grad_norm": 0.8232229235066113,
+      "learning_rate": 2.0194459315284742e-07,
+      "loss": 1.3167,
+      "step": 21281
+    },
+    {
+      "epoch": 0.9804210623301239,
+      "grad_norm": 0.8531655425277541,
+      "learning_rate": 2.0099792662514606e-07,
+      "loss": 1.0485,
+      "step": 21282
+    },
+    {
+      "epoch": 0.9804671304187589,
+      "grad_norm": 1.120431266946184,
+      "learning_rate": 2.000534819511124e-07,
+      "loss": 1.2652,
+      "step": 21283
+    },
+    {
+      "epoch": 0.9805131985073939,
+      "grad_norm": 0.7399286792952937,
+      "learning_rate": 1.9911125915175188e-07,
+      "loss": 1.0261,
+      "step": 21284
+    },
+    {
+      "epoch": 0.9805592665960289,
+      "grad_norm": 1.06667808287782,
+      "learning_rate": 1.9817125824806992e-07,
+      "loss": 1.1101,
+      "step": 21285
+    },
+    {
+      "epoch": 0.9806053346846639,
+      "grad_norm": 1.582554783626487,
+      "learning_rate": 1.97233479260972e-07,
+      "loss": 1.0815,
+      "step": 21286
+    },
+    {
+      "epoch": 0.980651402773299,
+      "grad_norm": 0.8950843294979548,
+      "learning_rate": 1.9629792221135257e-07,
+      "loss": 1.4313,
+      "step": 21287
+    },
+    {
+      "epoch": 0.9806974708619339,
+      "grad_norm": 0.7457745230305606,
+      "learning_rate": 1.9536458712003935e-07,
+      "loss": 1.3049,
+      "step": 21288
+    },
+    {
+      "epoch": 0.980743538950569,
+      "grad_norm": 1.047180439611744,
+      "learning_rate": 1.9443347400779355e-07,
+      "loss": 1.1784,
+      "step": 21289
+    },
+    {
+      "epoch": 0.9807896070392039,
+      "grad_norm": 1.5726362116605064,
+      "learning_rate": 1.935045828953652e-07,
+      "loss": 1.3332,
+      "step": 21290
+    },
+    {
+      "epoch": 0.980835675127839,
+      "grad_norm": 1.1301664528599065,
+      "learning_rate": 1.9257791380341562e-07,
+      "loss": 1.3508,
+      "step": 21291
+    },
+    {
+      "epoch": 0.9808817432164739,
+      "grad_norm": 0.6828789028768992,
+      "learning_rate": 1.916534667525949e-07,
+      "loss": 1.5248,
+      "step": 21292
+    },
+    {
+      "epoch": 0.980927811305109,
+      "grad_norm": 1.3629769854189455,
+      "learning_rate": 1.9073124176346435e-07,
+      "loss": 1.3695,
+      "step": 21293
+    },
+    {
+      "epoch": 0.9809738793937439,
+      "grad_norm": 1.0120142649122763,
+      "learning_rate": 1.8981123885658536e-07,
+      "loss": 1.2475,
+      "step": 21294
+    },
+    {
+      "epoch": 0.981019947482379,
+      "grad_norm": 3.048849525342148,
+      "learning_rate": 1.888934580523971e-07,
+      "loss": 1.2323,
+      "step": 21295
+    },
+    {
+      "epoch": 0.9810660155710139,
+      "grad_norm": 1.6673080048760671,
+      "learning_rate": 1.8797789937137211e-07,
+      "loss": 1.4501,
+      "step": 21296
+    },
+    {
+      "epoch": 0.981112083659649,
+      "grad_norm": 0.7358594636156428,
+      "learning_rate": 1.8706456283388296e-07,
+      "loss": 1.4818,
+      "step": 21297
+    },
+    {
+      "epoch": 0.9811581517482839,
+      "grad_norm": 0.8294842266245888,
+      "learning_rate": 1.8615344846024673e-07,
+      "loss": 1.3808,
+      "step": 21298
+    },
+    {
+      "epoch": 0.981204219836919,
+      "grad_norm": 13.726352018753618,
+      "learning_rate": 1.8524455627076943e-07,
+      "loss": 1.0944,
+      "step": 21299
+    },
+    {
+      "epoch": 0.981250287925554,
+      "grad_norm": 0.6586477026213925,
+      "learning_rate": 1.8433788628566816e-07,
+      "loss": 1.0888,
+      "step": 21300
+    },
+    {
+      "epoch": 0.981296356014189,
+      "grad_norm": 1.418269723226657,
+      "learning_rate": 1.8343343852513795e-07,
+      "loss": 1.1161,
+      "step": 21301
+    },
+    {
+      "epoch": 0.981342424102824,
+      "grad_norm": 0.7904568689900827,
+      "learning_rate": 1.825312130093182e-07,
+      "loss": 1.3968,
+      "step": 21302
+    },
+    {
+      "epoch": 0.981388492191459,
+      "grad_norm": 0.8445648168031943,
+      "learning_rate": 1.8163120975829284e-07,
+      "loss": 1.2086,
+      "step": 21303
+    },
+    {
+      "epoch": 0.981434560280094,
+      "grad_norm": 1.0710284317322951,
+      "learning_rate": 1.807334287920903e-07,
+      "loss": 1.47,
+      "step": 21304
+    },
+    {
+      "epoch": 0.981480628368729,
+      "grad_norm": 1.0591385691513449,
+      "learning_rate": 1.7983787013070574e-07,
+      "loss": 0.9647,
+      "step": 21305
+    },
+    {
+      "epoch": 0.981526696457364,
+      "grad_norm": 0.7693197403139278,
+      "learning_rate": 1.7894453379407872e-07,
+      "loss": 1.3978,
+      "step": 21306
+    },
+    {
+      "epoch": 0.981572764545999,
+      "grad_norm": 0.9163781781551068,
+      "learning_rate": 1.7805341980209335e-07,
+      "loss": 1.0911,
+      "step": 21307
+    },
+    {
+      "epoch": 0.981618832634634,
+      "grad_norm": 0.8938438877336689,
+      "learning_rate": 1.7716452817458929e-07,
+      "loss": 1.189,
+      "step": 21308
+    },
+    {
+      "epoch": 0.981664900723269,
+      "grad_norm": 1.1959565879885472,
+      "learning_rate": 1.7627785893136183e-07,
+      "loss": 1.0719,
+      "step": 21309
+    },
+    {
+      "epoch": 0.981710968811904,
+      "grad_norm": 0.6089783648341386,
+      "learning_rate": 1.7539341209213966e-07,
+      "loss": 0.989,
+      "step": 21310
+    },
+    {
+      "epoch": 0.981757036900539,
+      "grad_norm": 1.451437872703676,
+      "learning_rate": 1.7451118767662922e-07,
+      "loss": 1.2274,
+      "step": 21311
+    },
+    {
+      "epoch": 0.981803104989174,
+      "grad_norm": 1.5885671656033449,
+      "learning_rate": 1.7363118570445923e-07,
+      "loss": 1.2614,
+      "step": 21312
+    },
+    {
+      "epoch": 0.9818491730778091,
+      "grad_norm": 1.3724148101653255,
+      "learning_rate": 1.727534061952252e-07,
+      "loss": 1.1564,
+      "step": 21313
+    },
+    {
+      "epoch": 0.981895241166444,
+      "grad_norm": 1.574495637932532,
+      "learning_rate": 1.7187784916846695e-07,
+      "loss": 1.183,
+      "step": 21314
+    },
+    {
+      "epoch": 0.981941309255079,
+      "grad_norm": 1.387844377601636,
+      "learning_rate": 1.7100451464368006e-07,
+      "loss": 1.0791,
+      "step": 21315
+    },
+    {
+      "epoch": 0.981987377343714,
+      "grad_norm": 9.271707644848593,
+      "learning_rate": 1.7013340264031562e-07,
+      "loss": 1.3555,
+      "step": 21316
+    },
+    {
+      "epoch": 0.982033445432349,
+      "grad_norm": 0.9402920325125699,
+      "learning_rate": 1.6926451317774707e-07,
+      "loss": 1.4838,
+      "step": 21317
+    },
+    {
+      "epoch": 0.982079513520984,
+      "grad_norm": 0.7370154240827043,
+      "learning_rate": 1.683978462753366e-07,
+      "loss": 1.2444,
+      "step": 21318
+    },
+    {
+      "epoch": 0.982125581609619,
+      "grad_norm": 1.218912892669844,
+      "learning_rate": 1.6753340195238e-07,
+      "loss": 1.3426,
+      "step": 21319
+    },
+    {
+      "epoch": 0.982171649698254,
+      "grad_norm": 0.8124956765840855,
+      "learning_rate": 1.6667118022811734e-07,
+      "loss": 1.3736,
+      "step": 21320
+    },
+    {
+      "epoch": 0.982217717786889,
+      "grad_norm": 1.9175329016162401,
+      "learning_rate": 1.6581118112173332e-07,
+      "loss": 1.2237,
+      "step": 21321
+    },
+    {
+      "epoch": 0.982263785875524,
+      "grad_norm": 0.8662250870777687,
+      "learning_rate": 1.649534046523904e-07,
+      "loss": 1.0293,
+      "step": 21322
+    },
+    {
+      "epoch": 0.982309853964159,
+      "grad_norm": 0.9540375308890195,
+      "learning_rate": 1.6409785083917328e-07,
+      "loss": 1.4144,
+      "step": 21323
+    },
+    {
+      "epoch": 0.982355922052794,
+      "grad_norm": 1.0603906779963974,
+      "learning_rate": 1.6324451970114452e-07,
+      "loss": 1.4106,
+      "step": 21324
+    },
+    {
+      "epoch": 0.982401990141429,
+      "grad_norm": 1.080854056561023,
+      "learning_rate": 1.623934112572889e-07,
+      "loss": 1.4804,
+      "step": 21325
+    },
+    {
+      "epoch": 0.982448058230064,
+      "grad_norm": 1.7994729845418893,
+      "learning_rate": 1.6154452552655796e-07,
+      "loss": 0.9674,
+      "step": 21326
+    },
+    {
+      "epoch": 0.982494126318699,
+      "grad_norm": 1.310570985392946,
+      "learning_rate": 1.606978625278477e-07,
+      "loss": 1.2264,
+      "step": 21327
+    },
+    {
+      "epoch": 0.9825401944073341,
+      "grad_norm": 2.6901893128697414,
+      "learning_rate": 1.5985342228002077e-07,
+      "loss": 1.0277,
+      "step": 21328
+    },
+    {
+      "epoch": 0.982586262495969,
+      "grad_norm": 0.9991933191018795,
+      "learning_rate": 1.5901120480186215e-07,
+      "loss": 1.1729,
+      "step": 21329
+    },
+    {
+      "epoch": 0.9826323305846041,
+      "grad_norm": 1.401241556683779,
+      "learning_rate": 1.5817121011213466e-07,
+      "loss": 1.1344,
+      "step": 21330
+    },
+    {
+      "epoch": 0.982678398673239,
+      "grad_norm": 1.1209434095770061,
+      "learning_rate": 1.573334382295233e-07,
+      "loss": 1.3499,
+      "step": 21331
+    },
+    {
+      "epoch": 0.9827244667618741,
+      "grad_norm": 0.9601731995208118,
+      "learning_rate": 1.5649788917270202e-07,
+      "loss": 1.5376,
+      "step": 21332
+    },
+    {
+      "epoch": 0.982770534850509,
+      "grad_norm": 1.4570087132838294,
+      "learning_rate": 1.55664562960256e-07,
+      "loss": 1.3034,
+      "step": 21333
+    },
+    {
+      "epoch": 0.9828166029391441,
+      "grad_norm": 0.704854704859564,
+      "learning_rate": 1.5483345961073703e-07,
+      "loss": 1.0897,
+      "step": 21334
+    },
+    {
+      "epoch": 0.982862671027779,
+      "grad_norm": 1.6800987565721905,
+      "learning_rate": 1.5400457914265254e-07,
+      "loss": 1.3719,
+      "step": 21335
+    },
+    {
+      "epoch": 0.9829087391164141,
+      "grad_norm": 1.3881224018249767,
+      "learning_rate": 1.5317792157446553e-07,
+      "loss": 1.2335,
+      "step": 21336
+    },
+    {
+      "epoch": 0.982954807205049,
+      "grad_norm": 0.8902752541595667,
+      "learning_rate": 1.523534869245613e-07,
+      "loss": 1.1536,
+      "step": 21337
+    },
+    {
+      "epoch": 0.9830008752936841,
+      "grad_norm": 3.42445142558305,
+      "learning_rate": 1.5153127521131405e-07,
+      "loss": 1.0874,
+      "step": 21338
+    },
+    {
+      "epoch": 0.983046943382319,
+      "grad_norm": 1.1873486161580764,
+      "learning_rate": 1.5071128645300914e-07,
+      "loss": 1.1522,
+      "step": 21339
+    },
+    {
+      "epoch": 0.9830930114709541,
+      "grad_norm": 1.7981417989696926,
+      "learning_rate": 1.4989352066792083e-07,
+      "loss": 1.1639,
+      "step": 21340
+    },
+    {
+      "epoch": 0.9831390795595891,
+      "grad_norm": 1.017943968088381,
+      "learning_rate": 1.490779778742457e-07,
+      "loss": 1.2563,
+      "step": 21341
+    },
+    {
+      "epoch": 0.9831851476482241,
+      "grad_norm": 1.2918458290080395,
+      "learning_rate": 1.482646580901359e-07,
+      "loss": 1.2849,
+      "step": 21342
+    },
+    {
+      "epoch": 0.9832312157368591,
+      "grad_norm": 1.050199252680953,
+      "learning_rate": 1.4745356133371024e-07,
+      "loss": 1.1539,
+      "step": 21343
+    },
+    {
+      "epoch": 0.9832772838254941,
+      "grad_norm": 1.2842042574570818,
+      "learning_rate": 1.4664468762302098e-07,
+      "loss": 1.3006,
+      "step": 21344
+    },
+    {
+      "epoch": 0.9833233519141291,
+      "grad_norm": 2.162931644788413,
+      "learning_rate": 1.4583803697607589e-07,
+      "loss": 1.4455,
+      "step": 21345
+    },
+    {
+      "epoch": 0.9833694200027641,
+      "grad_norm": 0.9488033100643292,
+      "learning_rate": 1.4503360941083844e-07,
+      "loss": 1.1909,
+      "step": 21346
+    },
+    {
+      "epoch": 0.9834154880913991,
+      "grad_norm": 1.0823653657889556,
+      "learning_rate": 1.4423140494520536e-07,
+      "loss": 0.9986,
+      "step": 21347
+    },
+    {
+      "epoch": 0.9834615561800341,
+      "grad_norm": 0.7813801828074728,
+      "learning_rate": 1.434314235970513e-07,
+      "loss": 1.3534,
+      "step": 21348
+    },
+    {
+      "epoch": 0.9835076242686691,
+      "grad_norm": 0.9290748606712516,
+      "learning_rate": 1.426336653841731e-07,
+      "loss": 1.1898,
+      "step": 21349
+    },
+    {
+      "epoch": 0.983553692357304,
+      "grad_norm": 1.3820653237914526,
+      "learning_rate": 1.4183813032433434e-07,
+      "loss": 1.1574,
+      "step": 21350
+    },
+    {
+      "epoch": 0.9835997604459391,
+      "grad_norm": 0.7962487488130506,
+      "learning_rate": 1.4104481843525418e-07,
+      "loss": 1.0835,
+      "step": 21351
+    },
+    {
+      "epoch": 0.983645828534574,
+      "grad_norm": 1.0296927290411328,
+      "learning_rate": 1.4025372973459626e-07,
+      "loss": 1.433,
+      "step": 21352
+    },
+    {
+      "epoch": 0.9836918966232091,
+      "grad_norm": 5.121948152100529,
+      "learning_rate": 1.3946486423995763e-07,
+      "loss": 1.1152,
+      "step": 21353
+    },
+    {
+      "epoch": 0.9837379647118442,
+      "grad_norm": 1.685711384433397,
+      "learning_rate": 1.386782219689131e-07,
+      "loss": 1.1102,
+      "step": 21354
+    },
+    {
+      "epoch": 0.9837840328004791,
+      "grad_norm": 1.1543053258644917,
+      "learning_rate": 1.378938029389709e-07,
+      "loss": 1.3176,
+      "step": 21355
+    },
+    {
+      "epoch": 0.9838301008891142,
+      "grad_norm": 0.9513687501479227,
+      "learning_rate": 1.3711160716759486e-07,
+      "loss": 1.1843,
+      "step": 21356
+    },
+    {
+      "epoch": 0.9838761689777491,
+      "grad_norm": 1.3746360025756372,
+      "learning_rate": 1.3633163467220434e-07,
+      "loss": 1.1183,
+      "step": 21357
+    },
+    {
+      "epoch": 0.9839222370663842,
+      "grad_norm": 1.6327577265090973,
+      "learning_rate": 1.3555388547015213e-07,
+      "loss": 1.1569,
+      "step": 21358
+    },
+    {
+      "epoch": 0.9839683051550191,
+      "grad_norm": 0.796394243462509,
+      "learning_rate": 1.3477835957876882e-07,
+      "loss": 1.2808,
+      "step": 21359
+    },
+    {
+      "epoch": 0.9840143732436542,
+      "grad_norm": 0.9455716779969628,
+      "learning_rate": 1.3400505701531842e-07,
+      "loss": 0.9396,
+      "step": 21360
+    },
+    {
+      "epoch": 0.9840604413322891,
+      "grad_norm": 2.1856826719768323,
+      "learning_rate": 1.332339777969982e-07,
+      "loss": 1.6685,
+      "step": 21361
+    },
+    {
+      "epoch": 0.9841065094209241,
+      "grad_norm": 1.1766901273290002,
+      "learning_rate": 1.3246512194100557e-07,
+      "loss": 1.1636,
+      "step": 21362
+    },
+    {
+      "epoch": 0.9841525775095591,
+      "grad_norm": 1.2342305996534382,
+      "learning_rate": 1.3169848946442686e-07,
+      "loss": 0.9446,
+      "step": 21363
+    },
+    {
+      "epoch": 0.9841986455981941,
+      "grad_norm": 1.272098755940301,
+      "learning_rate": 1.309340803843484e-07,
+      "loss": 1.0203,
+      "step": 21364
+    },
+    {
+      "epoch": 0.9842447136868291,
+      "grad_norm": 0.8872627085678602,
+      "learning_rate": 1.301718947177899e-07,
+      "loss": 1.1629,
+      "step": 21365
+    },
+    {
+      "epoch": 0.9842907817754641,
+      "grad_norm": 0.9969972605098085,
+      "learning_rate": 1.2941193248170447e-07,
+      "loss": 0.9431,
+      "step": 21366
+    },
+    {
+      "epoch": 0.9843368498640991,
+      "grad_norm": 0.9030868535380183,
+      "learning_rate": 1.2865419369303412e-07,
+      "loss": 1.4231,
+      "step": 21367
+    },
+    {
+      "epoch": 0.9843829179527341,
+      "grad_norm": 2.016924386516641,
+      "learning_rate": 1.278986783686209e-07,
+      "loss": 1.2281,
+      "step": 21368
+    },
+    {
+      "epoch": 0.9844289860413692,
+      "grad_norm": 0.7487134170601552,
+      "learning_rate": 1.2714538652530695e-07,
+      "loss": 1.283,
+      "step": 21369
+    },
+    {
+      "epoch": 0.9844750541300041,
+      "grad_norm": 0.6019220838711096,
+      "learning_rate": 1.263943181798566e-07,
+      "loss": 1.4771,
+      "step": 21370
+    },
+    {
+      "epoch": 0.9845211222186392,
+      "grad_norm": 0.8940708461699914,
+      "learning_rate": 1.2564547334898978e-07,
+      "loss": 1.1785,
+      "step": 21371
+    },
+    {
+      "epoch": 0.9845671903072741,
+      "grad_norm": 0.7124088143511166,
+      "learning_rate": 1.2489885204937101e-07,
+      "loss": 1.5826,
+      "step": 21372
+    },
+    {
+      "epoch": 0.9846132583959092,
+      "grad_norm": 0.7516025043474508,
+      "learning_rate": 1.241544542976425e-07,
+      "loss": 1.235,
+      "step": 21373
+    },
+    {
+      "epoch": 0.9846593264845441,
+      "grad_norm": 1.6594589124103318,
+      "learning_rate": 1.2341228011035765e-07,
+      "loss": 1.3431,
+      "step": 21374
+    },
+    {
+      "epoch": 0.9847053945731792,
+      "grad_norm": 1.005746278536623,
+      "learning_rate": 1.226723295040477e-07,
+      "loss": 1.4585,
+      "step": 21375
+    },
+    {
+      "epoch": 0.9847514626618141,
+      "grad_norm": 0.6502307136189948,
+      "learning_rate": 1.2193460249517729e-07,
+      "loss": 1.3764,
+      "step": 21376
+    },
+    {
+      "epoch": 0.9847975307504492,
+      "grad_norm": 0.8283818831365904,
+      "learning_rate": 1.211990991001888e-07,
+      "loss": 1.0753,
+      "step": 21377
+    },
+    {
+      "epoch": 0.9848435988390841,
+      "grad_norm": 1.3636010182159102,
+      "learning_rate": 1.2046581933544688e-07,
+      "loss": 1.2561,
+      "step": 21378
+    },
+    {
+      "epoch": 0.9848896669277192,
+      "grad_norm": 0.7173462296659684,
+      "learning_rate": 1.1973476321727183e-07,
+      "loss": 1.2686,
+      "step": 21379
+    },
+    {
+      "epoch": 0.9849357350163541,
+      "grad_norm": 3.3235505101245764,
+      "learning_rate": 1.1900593076195066e-07,
+      "loss": 1.2289,
+      "step": 21380
+    },
+    {
+      "epoch": 0.9849818031049892,
+      "grad_norm": 3.1570984135266045,
+      "learning_rate": 1.1827932198570369e-07,
+      "loss": 1.349,
+      "step": 21381
+    },
+    {
+      "epoch": 0.9850278711936242,
+      "grad_norm": 0.9498346314904313,
+      "learning_rate": 1.1755493690470687e-07,
+      "loss": 1.2841,
+      "step": 21382
+    },
+    {
+      "epoch": 0.9850739392822592,
+      "grad_norm": 0.8321344811850913,
+      "learning_rate": 1.1683277553508065e-07,
+      "loss": 1.3623,
+      "step": 21383
+    },
+    {
+      "epoch": 0.9851200073708942,
+      "grad_norm": 1.082173095323247,
+      "learning_rate": 1.1611283789291216e-07,
+      "loss": 1.199,
+      "step": 21384
+    },
+    {
+      "epoch": 0.9851660754595292,
+      "grad_norm": 0.9976638830866008,
+      "learning_rate": 1.1539512399422192e-07,
+      "loss": 1.2484,
+      "step": 21385
+    },
+    {
+      "epoch": 0.9852121435481642,
+      "grad_norm": 1.026454363258026,
+      "learning_rate": 1.1467963385500824e-07,
+      "loss": 0.8226,
+      "step": 21386
+    },
+    {
+      "epoch": 0.9852582116367992,
+      "grad_norm": 1.5278226261319752,
+      "learning_rate": 1.1396636749118061e-07,
+      "loss": 1.0256,
+      "step": 21387
+    },
+    {
+      "epoch": 0.9853042797254342,
+      "grad_norm": 0.8631344165581315,
+      "learning_rate": 1.1325532491861524e-07,
+      "loss": 1.371,
+      "step": 21388
+    },
+    {
+      "epoch": 0.9853503478140692,
+      "grad_norm": 0.8343346864905248,
+      "learning_rate": 1.1254650615316609e-07,
+      "loss": 1.1799,
+      "step": 21389
+    },
+    {
+      "epoch": 0.9853964159027042,
+      "grad_norm": 1.285770507794458,
+      "learning_rate": 1.1183991121058724e-07,
+      "loss": 1.2093,
+      "step": 21390
+    },
+    {
+      "epoch": 0.9854424839913392,
+      "grad_norm": 3.4313953852921277,
+      "learning_rate": 1.1113554010662164e-07,
+      "loss": 1.3369,
+      "step": 21391
+    },
+    {
+      "epoch": 0.9854885520799742,
+      "grad_norm": 0.9369289551486728,
+      "learning_rate": 1.1043339285694565e-07,
+      "loss": 1.2822,
+      "step": 21392
+    },
+    {
+      "epoch": 0.9855346201686092,
+      "grad_norm": 0.6384145361466804,
+      "learning_rate": 1.0973346947721341e-07,
+      "loss": 1.3141,
+      "step": 21393
+    },
+    {
+      "epoch": 0.9855806882572442,
+      "grad_norm": 0.8906097033570118,
+      "learning_rate": 1.0903576998297915e-07,
+      "loss": 1.0296,
+      "step": 21394
+    },
+    {
+      "epoch": 0.9856267563458793,
+      "grad_norm": 0.7966904590604532,
+      "learning_rate": 1.0834029438978599e-07,
+      "loss": 1.2759,
+      "step": 21395
+    },
+    {
+      "epoch": 0.9856728244345142,
+      "grad_norm": 0.9884482937094367,
+      "learning_rate": 1.0764704271312154e-07,
+      "loss": 0.9392,
+      "step": 21396
+    },
+    {
+      "epoch": 0.9857188925231493,
+      "grad_norm": 0.8240123223114187,
+      "learning_rate": 1.069560149684179e-07,
+      "loss": 1.0595,
+      "step": 21397
+    },
+    {
+      "epoch": 0.9857649606117842,
+      "grad_norm": 0.7403125654379836,
+      "learning_rate": 1.0626721117105165e-07,
+      "loss": 1.4394,
+      "step": 21398
+    },
+    {
+      "epoch": 0.9858110287004193,
+      "grad_norm": 0.8639876016632156,
+      "learning_rate": 1.0558063133637719e-07,
+      "loss": 1.184,
+      "step": 21399
+    },
+    {
+      "epoch": 0.9858570967890542,
+      "grad_norm": 0.8606646218781049,
+      "learning_rate": 1.0489627547966007e-07,
+      "loss": 1.3402,
+      "step": 21400
+    },
+    {
+      "epoch": 0.9859031648776893,
+      "grad_norm": 1.1990638584723046,
+      "learning_rate": 1.0421414361614368e-07,
+      "loss": 1.188,
+      "step": 21401
+    },
+    {
+      "epoch": 0.9859492329663242,
+      "grad_norm": 0.8070643350225084,
+      "learning_rate": 1.0353423576101584e-07,
+      "loss": 0.9334,
+      "step": 21402
+    },
+    {
+      "epoch": 0.9859953010549593,
+      "grad_norm": 1.3361887695392152,
+      "learning_rate": 1.0285655192942001e-07,
+      "loss": 1.2474,
+      "step": 21403
+    },
+    {
+      "epoch": 0.9860413691435942,
+      "grad_norm": 4.793793691552706,
+      "learning_rate": 1.021810921364219e-07,
+      "loss": 1.1535,
+      "step": 21404
+    },
+    {
+      "epoch": 0.9860874372322292,
+      "grad_norm": 0.8392216553295039,
+      "learning_rate": 1.0150785639708726e-07,
+      "loss": 1.1419,
+      "step": 21405
+    },
+    {
+      "epoch": 0.9861335053208642,
+      "grad_norm": 1.6053806473106258,
+      "learning_rate": 1.0083684472638189e-07,
+      "loss": 1.2196,
+      "step": 21406
+    },
+    {
+      "epoch": 0.9861795734094992,
+      "grad_norm": 0.8512584674008034,
+      "learning_rate": 1.0016805713926047e-07,
+      "loss": 1.2711,
+      "step": 21407
+    },
+    {
+      "epoch": 0.9862256414981343,
+      "grad_norm": 0.9157025012990848,
+      "learning_rate": 9.950149365060003e-08,
+      "loss": 1.1288,
+      "step": 21408
+    },
+    {
+      "epoch": 0.9862717095867692,
+      "grad_norm": 3.1612166369510843,
+      "learning_rate": 9.88371542752442e-08,
+      "loss": 1.228,
+      "step": 21409
+    },
+    {
+      "epoch": 0.9863177776754043,
+      "grad_norm": 1.3493136467449192,
+      "learning_rate": 9.81750390279923e-08,
+      "loss": 1.1831,
+      "step": 21410
+    },
+    {
+      "epoch": 0.9863638457640392,
+      "grad_norm": 0.8082535688498906,
+      "learning_rate": 9.751514792357696e-08,
+      "loss": 1.1498,
+      "step": 21411
+    },
+    {
+      "epoch": 0.9864099138526743,
+      "grad_norm": 1.5486653773904362,
+      "learning_rate": 9.685748097668645e-08,
+      "loss": 1.1892,
+      "step": 21412
+    },
+    {
+      "epoch": 0.9864559819413092,
+      "grad_norm": 1.4036987893903399,
+      "learning_rate": 9.62020382019757e-08,
+      "loss": 1.3368,
+      "step": 21413
+    },
+    {
+      "epoch": 0.9865020500299443,
+      "grad_norm": 1.285860402524864,
+      "learning_rate": 9.554881961402196e-08,
+      "loss": 1.2096,
+      "step": 21414
+    },
+    {
+      "epoch": 0.9865481181185792,
+      "grad_norm": 0.7881331821790941,
+      "learning_rate": 9.489782522736911e-08,
+      "loss": 1.0813,
+      "step": 21415
+    },
+    {
+      "epoch": 0.9865941862072143,
+      "grad_norm": 2.63902509416951,
+      "learning_rate": 9.424905505652782e-08,
+      "loss": 1.1751,
+      "step": 21416
+    },
+    {
+      "epoch": 0.9866402542958492,
+      "grad_norm": 1.3469856811606378,
+      "learning_rate": 9.360250911591983e-08,
+      "loss": 1.0162,
+      "step": 21417
+    },
+    {
+      "epoch": 0.9866863223844843,
+      "grad_norm": 3.1372572051755023,
+      "learning_rate": 9.295818741995587e-08,
+      "loss": 1.0314,
+      "step": 21418
+    },
+    {
+      "epoch": 0.9867323904731192,
+      "grad_norm": 1.687104536060591,
+      "learning_rate": 9.231608998296892e-08,
+      "loss": 1.2643,
+      "step": 21419
+    },
+    {
+      "epoch": 0.9867784585617543,
+      "grad_norm": 0.9082657896067061,
+      "learning_rate": 9.167621681925864e-08,
+      "loss": 1.1197,
+      "step": 21420
+    },
+    {
+      "epoch": 0.9868245266503892,
+      "grad_norm": 4.611550193961549,
+      "learning_rate": 9.10385679430803e-08,
+      "loss": 1.2872,
+      "step": 21421
+    },
+    {
+      "epoch": 0.9868705947390243,
+      "grad_norm": 7.05508091435411,
+      "learning_rate": 9.040314336861145e-08,
+      "loss": 1.2904,
+      "step": 21422
+    },
+    {
+      "epoch": 0.9869166628276593,
+      "grad_norm": 0.7860316442771421,
+      "learning_rate": 8.976994311000742e-08,
+      "loss": 1.3494,
+      "step": 21423
+    },
+    {
+      "epoch": 0.9869627309162943,
+      "grad_norm": 1.1000501533213995,
+      "learning_rate": 8.913896718137915e-08,
+      "loss": 1.2066,
+      "step": 21424
+    },
+    {
+      "epoch": 0.9870087990049293,
+      "grad_norm": 1.1718447650901416,
+      "learning_rate": 8.851021559674877e-08,
+      "loss": 1.1071,
+      "step": 21425
+    },
+    {
+      "epoch": 0.9870548670935643,
+      "grad_norm": 0.996199483885355,
+      "learning_rate": 8.788368837012729e-08,
+      "loss": 1.5427,
+      "step": 21426
+    },
+    {
+      "epoch": 0.9871009351821993,
+      "grad_norm": 0.9363017147672633,
+      "learning_rate": 8.725938551547019e-08,
+      "loss": 1.151,
+      "step": 21427
+    },
+    {
+      "epoch": 0.9871470032708343,
+      "grad_norm": 0.8615957026989614,
+      "learning_rate": 8.663730704666639e-08,
+      "loss": 1.3542,
+      "step": 21428
+    },
+    {
+      "epoch": 0.9871930713594693,
+      "grad_norm": 0.8607955697878606,
+      "learning_rate": 8.601745297757146e-08,
+      "loss": 1.2439,
+      "step": 21429
+    },
+    {
+      "epoch": 0.9872391394481043,
+      "grad_norm": 0.8536825086688999,
+      "learning_rate": 8.539982332197438e-08,
+      "loss": 1.4544,
+      "step": 21430
+    },
+    {
+      "epoch": 0.9872852075367393,
+      "grad_norm": 1.7921256370495078,
+      "learning_rate": 8.47844180936308e-08,
+      "loss": 1.3741,
+      "step": 21431
+    },
+    {
+      "epoch": 0.9873312756253743,
+      "grad_norm": 0.9561877581963368,
+      "learning_rate": 8.417123730625198e-08,
+      "loss": 1.1157,
+      "step": 21432
+    },
+    {
+      "epoch": 0.9873773437140093,
+      "grad_norm": 1.0009980909135063,
+      "learning_rate": 8.356028097348256e-08,
+      "loss": 1.1316,
+      "step": 21433
+    },
+    {
+      "epoch": 0.9874234118026443,
+      "grad_norm": 0.8223068731140694,
+      "learning_rate": 8.295154910892278e-08,
+      "loss": 1.292,
+      "step": 21434
+    },
+    {
+      "epoch": 0.9874694798912793,
+      "grad_norm": 0.8807563966461158,
+      "learning_rate": 8.234504172612845e-08,
+      "loss": 1.3071,
+      "step": 21435
+    },
+    {
+      "epoch": 0.9875155479799144,
+      "grad_norm": 1.218784875283841,
+      "learning_rate": 8.174075883858878e-08,
+      "loss": 0.918,
+      "step": 21436
+    },
+    {
+      "epoch": 0.9875616160685493,
+      "grad_norm": 0.9170543627492248,
+      "learning_rate": 8.113870045978189e-08,
+      "loss": 1.1829,
+      "step": 21437
+    },
+    {
+      "epoch": 0.9876076841571844,
+      "grad_norm": 1.9986162997516503,
+      "learning_rate": 8.053886660308596e-08,
+      "loss": 1.0619,
+      "step": 21438
+    },
+    {
+      "epoch": 0.9876537522458193,
+      "grad_norm": 0.6982747865334745,
+      "learning_rate": 7.994125728187918e-08,
+      "loss": 1.6372,
+      "step": 21439
+    },
+    {
+      "epoch": 0.9876998203344544,
+      "grad_norm": 0.8921409168660523,
+      "learning_rate": 7.934587250943981e-08,
+      "loss": 1.164,
+      "step": 21440
+    },
+    {
+      "epoch": 0.9877458884230893,
+      "grad_norm": 0.720123504156479,
+      "learning_rate": 7.875271229904612e-08,
+      "loss": 1.2695,
+      "step": 21441
+    },
+    {
+      "epoch": 0.9877919565117244,
+      "grad_norm": 1.289256819385449,
+      "learning_rate": 7.816177666389868e-08,
+      "loss": 1.0191,
+      "step": 21442
+    },
+    {
+      "epoch": 0.9878380246003593,
+      "grad_norm": 1.797452532430289,
+      "learning_rate": 7.75730656171425e-08,
+      "loss": 1.1698,
+      "step": 21443
+    },
+    {
+      "epoch": 0.9878840926889944,
+      "grad_norm": 0.9198959754406533,
+      "learning_rate": 7.698657917188934e-08,
+      "loss": 1.0341,
+      "step": 21444
+    },
+    {
+      "epoch": 0.9879301607776293,
+      "grad_norm": 1.637168246562619,
+      "learning_rate": 7.640231734120651e-08,
+      "loss": 1.3182,
+      "step": 21445
+    },
+    {
+      "epoch": 0.9879762288662643,
+      "grad_norm": 0.701563789934579,
+      "learning_rate": 7.582028013809473e-08,
+      "loss": 1.2389,
+      "step": 21446
+    },
+    {
+      "epoch": 0.9880222969548993,
+      "grad_norm": 0.7785631526286646,
+      "learning_rate": 7.524046757549919e-08,
+      "loss": 0.9576,
+      "step": 21447
+    },
+    {
+      "epoch": 0.9880683650435343,
+      "grad_norm": 1.1435161777705953,
+      "learning_rate": 7.466287966635399e-08,
+      "loss": 1.182,
+      "step": 21448
+    },
+    {
+      "epoch": 0.9881144331321694,
+      "grad_norm": 1.0339943285648168,
+      "learning_rate": 7.408751642349333e-08,
+      "loss": 1.272,
+      "step": 21449
+    },
+    {
+      "epoch": 0.9881605012208043,
+      "grad_norm": 1.361690492289987,
+      "learning_rate": 7.351437785975134e-08,
+      "loss": 1.0692,
+      "step": 21450
+    },
+    {
+      "epoch": 0.9882065693094394,
+      "grad_norm": 0.7813356468657566,
+      "learning_rate": 7.294346398786233e-08,
+      "loss": 1.0225,
+      "step": 21451
+    },
+    {
+      "epoch": 0.9882526373980743,
+      "grad_norm": 0.904532103919656,
+      "learning_rate": 7.237477482056055e-08,
+      "loss": 1.2308,
+      "step": 21452
+    },
+    {
+      "epoch": 0.9882987054867094,
+      "grad_norm": 0.8667210315566675,
+      "learning_rate": 7.18083103704803e-08,
+      "loss": 1.2413,
+      "step": 21453
+    },
+    {
+      "epoch": 0.9883447735753443,
+      "grad_norm": 0.66270546521172,
+      "learning_rate": 7.124407065025596e-08,
+      "loss": 1.4723,
+      "step": 21454
+    },
+    {
+      "epoch": 0.9883908416639794,
+      "grad_norm": 2.3403413457468347,
+      "learning_rate": 7.068205567244413e-08,
+      "loss": 0.9778,
+      "step": 21455
+    },
+    {
+      "epoch": 0.9884369097526143,
+      "grad_norm": 0.7189274425259843,
+      "learning_rate": 7.012226544955703e-08,
+      "loss": 1.1947,
+      "step": 21456
+    },
+    {
+      "epoch": 0.9884829778412494,
+      "grad_norm": 1.2138188692628564,
+      "learning_rate": 6.956469999404025e-08,
+      "loss": 1.269,
+      "step": 21457
+    },
+    {
+      "epoch": 0.9885290459298843,
+      "grad_norm": 5.23137566823524,
+      "learning_rate": 6.900935931833941e-08,
+      "loss": 1.0628,
+      "step": 21458
+    },
+    {
+      "epoch": 0.9885751140185194,
+      "grad_norm": 0.8153614883408526,
+      "learning_rate": 6.845624343478908e-08,
+      "loss": 1.1164,
+      "step": 21459
+    },
+    {
+      "epoch": 0.9886211821071543,
+      "grad_norm": 0.9083099464078995,
+      "learning_rate": 6.790535235571271e-08,
+      "loss": 1.0444,
+      "step": 21460
+    },
+    {
+      "epoch": 0.9886672501957894,
+      "grad_norm": 0.8974159820913327,
+      "learning_rate": 6.735668609337831e-08,
+      "loss": 1.1401,
+      "step": 21461
+    },
+    {
+      "epoch": 0.9887133182844243,
+      "grad_norm": 0.9016318073755829,
+      "learning_rate": 6.681024466000941e-08,
+      "loss": 1.5449,
+      "step": 21462
+    },
+    {
+      "epoch": 0.9887593863730594,
+      "grad_norm": 1.7340320722154967,
+      "learning_rate": 6.626602806775184e-08,
+      "loss": 1.1025,
+      "step": 21463
+    },
+    {
+      "epoch": 0.9888054544616944,
+      "grad_norm": 1.3651026679276772,
+      "learning_rate": 6.572403632872925e-08,
+      "loss": 1.248,
+      "step": 21464
+    },
+    {
+      "epoch": 0.9888515225503294,
+      "grad_norm": 0.8803154802457505,
+      "learning_rate": 6.518426945502088e-08,
+      "loss": 0.8673,
+      "step": 21465
+    },
+    {
+      "epoch": 0.9888975906389644,
+      "grad_norm": 0.8906927950134561,
+      "learning_rate": 6.464672745862821e-08,
+      "loss": 1.0579,
+      "step": 21466
+    },
+    {
+      "epoch": 0.9889436587275994,
+      "grad_norm": 1.079436243072993,
+      "learning_rate": 6.411141035151947e-08,
+      "loss": 0.846,
+      "step": 21467
+    },
+    {
+      "epoch": 0.9889897268162344,
+      "grad_norm": 1.304421526718231,
+      "learning_rate": 6.357831814561843e-08,
+      "loss": 1.4343,
+      "step": 21468
+    },
+    {
+      "epoch": 0.9890357949048694,
+      "grad_norm": 1.7049409979154149,
+      "learning_rate": 6.304745085279339e-08,
+      "loss": 1.0084,
+      "step": 21469
+    },
+    {
+      "epoch": 0.9890818629935044,
+      "grad_norm": 2.253629269304835,
+      "learning_rate": 6.251880848486825e-08,
+      "loss": 1.049,
+      "step": 21470
+    },
+    {
+      "epoch": 0.9891279310821394,
+      "grad_norm": 1.3207662227557078,
+      "learning_rate": 6.199239105360022e-08,
+      "loss": 0.8832,
+      "step": 21471
+    },
+    {
+      "epoch": 0.9891739991707744,
+      "grad_norm": 1.003632730278362,
+      "learning_rate": 6.146819857071329e-08,
+      "loss": 1.2874,
+      "step": 21472
+    },
+    {
+      "epoch": 0.9892200672594094,
+      "grad_norm": 1.2710488676136764,
+      "learning_rate": 6.094623104787589e-08,
+      "loss": 1.2029,
+      "step": 21473
+    },
+    {
+      "epoch": 0.9892661353480444,
+      "grad_norm": 1.4043738940378974,
+      "learning_rate": 6.042648849672316e-08,
+      "loss": 1.1653,
+      "step": 21474
+    },
+    {
+      "epoch": 0.9893122034366794,
+      "grad_norm": 0.8757231095742909,
+      "learning_rate": 5.990897092880144e-08,
+      "loss": 1.3349,
+      "step": 21475
+    },
+    {
+      "epoch": 0.9893582715253144,
+      "grad_norm": 2.8041986833179386,
+      "learning_rate": 5.939367835565701e-08,
+      "loss": 1.3222,
+      "step": 21476
+    },
+    {
+      "epoch": 0.9894043396139495,
+      "grad_norm": 1.1408125885651437,
+      "learning_rate": 5.8880610788747404e-08,
+      "loss": 1.3258,
+      "step": 21477
+    },
+    {
+      "epoch": 0.9894504077025844,
+      "grad_norm": 0.9563987276344745,
+      "learning_rate": 5.8369768239496804e-08,
+      "loss": 1.2984,
+      "step": 21478
+    },
+    {
+      "epoch": 0.9894964757912195,
+      "grad_norm": 0.7678392180507336,
+      "learning_rate": 5.786115071928499e-08,
+      "loss": 1.35,
+      "step": 21479
+    },
+    {
+      "epoch": 0.9895425438798544,
+      "grad_norm": 0.9195740556762552,
+      "learning_rate": 5.7354758239425154e-08,
+      "loss": 0.9235,
+      "step": 21480
+    },
+    {
+      "epoch": 0.9895886119684895,
+      "grad_norm": 0.7076911342791739,
+      "learning_rate": 5.6850590811197146e-08,
+      "loss": 1.4356,
+      "step": 21481
+    },
+    {
+      "epoch": 0.9896346800571244,
+      "grad_norm": 1.398295256338401,
+      "learning_rate": 5.634864844582533e-08,
+      "loss": 0.9753,
+      "step": 21482
+    },
+    {
+      "epoch": 0.9896807481457595,
+      "grad_norm": 0.84997161086932,
+      "learning_rate": 5.584893115448964e-08,
+      "loss": 1.2774,
+      "step": 21483
+    },
+    {
+      "epoch": 0.9897268162343944,
+      "grad_norm": 0.9801098170731017,
+      "learning_rate": 5.535143894830342e-08,
+      "loss": 1.1434,
+      "step": 21484
+    },
+    {
+      "epoch": 0.9897728843230295,
+      "grad_norm": 2.0768177846685196,
+      "learning_rate": 5.485617183834668e-08,
+      "loss": 1.3335,
+      "step": 21485
+    },
+    {
+      "epoch": 0.9898189524116644,
+      "grad_norm": 0.8309075965052446,
+      "learning_rate": 5.436312983565506e-08,
+      "loss": 1.4726,
+      "step": 21486
+    },
+    {
+      "epoch": 0.9898650205002995,
+      "grad_norm": 0.6767710230783759,
+      "learning_rate": 5.3872312951197546e-08,
+      "loss": 1.2539,
+      "step": 21487
+    },
+    {
+      "epoch": 0.9899110885889344,
+      "grad_norm": 0.938825637663928,
+      "learning_rate": 5.338372119589874e-08,
+      "loss": 1.2933,
+      "step": 21488
+    },
+    {
+      "epoch": 0.9899571566775694,
+      "grad_norm": 0.8263768344978846,
+      "learning_rate": 5.289735458063882e-08,
+      "loss": 1.3029,
+      "step": 21489
+    },
+    {
+      "epoch": 0.9900032247662045,
+      "grad_norm": 0.8471539278525955,
+      "learning_rate": 5.241321311625358e-08,
+      "loss": 1.2838,
+      "step": 21490
+    },
+    {
+      "epoch": 0.9900492928548394,
+      "grad_norm": 0.8063831431839324,
+      "learning_rate": 5.1931296813512166e-08,
+      "loss": 1.1777,
+      "step": 21491
+    },
+    {
+      "epoch": 0.9900953609434745,
+      "grad_norm": 0.8877826480220252,
+      "learning_rate": 5.145160568315044e-08,
+      "loss": 1.6106,
+      "step": 21492
+    },
+    {
+      "epoch": 0.9901414290321094,
+      "grad_norm": 1.174468354483745,
+      "learning_rate": 5.0974139735826546e-08,
+      "loss": 1.2456,
+      "step": 21493
+    },
+    {
+      "epoch": 0.9901874971207445,
+      "grad_norm": 0.8862306942344459,
+      "learning_rate": 5.049889898220972e-08,
+      "loss": 1.2246,
+      "step": 21494
+    },
+    {
+      "epoch": 0.9902335652093794,
+      "grad_norm": 1.0165630477250387,
+      "learning_rate": 5.0025883432847084e-08,
+      "loss": 1.2043,
+      "step": 21495
+    },
+    {
+      "epoch": 0.9902796332980145,
+      "grad_norm": 0.7647763887284175,
+      "learning_rate": 4.9555093098274664e-08,
+      "loss": 1.1709,
+      "step": 21496
+    },
+    {
+      "epoch": 0.9903257013866494,
+      "grad_norm": 0.8714196500178097,
+      "learning_rate": 4.908652798899516e-08,
+      "loss": 1.1688,
+      "step": 21497
+    },
+    {
+      "epoch": 0.9903717694752845,
+      "grad_norm": 0.764348045619764,
+      "learning_rate": 4.862018811541136e-08,
+      "loss": 1.2635,
+      "step": 21498
+    },
+    {
+      "epoch": 0.9904178375639194,
+      "grad_norm": 0.8686509621076581,
+      "learning_rate": 4.8156073487926055e-08,
+      "loss": 1.0961,
+      "step": 21499
+    },
+    {
+      "epoch": 0.9904639056525545,
+      "grad_norm": 1.059816461235981,
+      "learning_rate": 4.7694184116864324e-08,
+      "loss": 0.9751,
+      "step": 21500
+    },
+    {
+      "epoch": 0.9905099737411894,
+      "grad_norm": 0.942700758245986,
+      "learning_rate": 4.723452001250683e-08,
+      "loss": 1.4425,
+      "step": 21501
+    },
+    {
+      "epoch": 0.9905560418298245,
+      "grad_norm": 5.386742649487362,
+      "learning_rate": 4.677708118508983e-08,
+      "loss": 1.0698,
+      "step": 21502
+    },
+    {
+      "epoch": 0.9906021099184594,
+      "grad_norm": 0.6414032462502491,
+      "learning_rate": 4.632186764480517e-08,
+      "loss": 1.0348,
+      "step": 21503
+    },
+    {
+      "epoch": 0.9906481780070945,
+      "grad_norm": 0.8399374153966813,
+      "learning_rate": 4.586887940176699e-08,
+      "loss": 1.4007,
+      "step": 21504
+    },
+    {
+      "epoch": 0.9906942460957295,
+      "grad_norm": 1.1784298059039426,
+      "learning_rate": 4.54181164660894e-08,
+      "loss": 1.013,
+      "step": 21505
+    },
+    {
+      "epoch": 0.9907403141843645,
+      "grad_norm": 0.9445735664284673,
+      "learning_rate": 4.496957884778663e-08,
+      "loss": 1.4703,
+      "step": 21506
+    },
+    {
+      "epoch": 0.9907863822729995,
+      "grad_norm": 0.8032126350975317,
+      "learning_rate": 4.452326655685068e-08,
+      "loss": 1.357,
+      "step": 21507
+    },
+    {
+      "epoch": 0.9908324503616345,
+      "grad_norm": 1.1730182937411677,
+      "learning_rate": 4.4079179603218055e-08,
+      "loss": 1.0204,
+      "step": 21508
+    },
+    {
+      "epoch": 0.9908785184502695,
+      "grad_norm": 1.2582671967211982,
+      "learning_rate": 4.363731799676973e-08,
+      "loss": 1.3475,
+      "step": 21509
+    },
+    {
+      "epoch": 0.9909245865389045,
+      "grad_norm": 1.7294591582235157,
+      "learning_rate": 4.319768174735339e-08,
+      "loss": 1.0274,
+      "step": 21510
+    },
+    {
+      "epoch": 0.9909706546275395,
+      "grad_norm": 1.0469817450846388,
+      "learning_rate": 4.2760270864738994e-08,
+      "loss": 1.1504,
+      "step": 21511
+    },
+    {
+      "epoch": 0.9910167227161745,
+      "grad_norm": 0.8339323652457116,
+      "learning_rate": 4.2325085358696504e-08,
+      "loss": 1.1542,
+      "step": 21512
+    },
+    {
+      "epoch": 0.9910627908048095,
+      "grad_norm": 0.9148113247697388,
+      "learning_rate": 4.189212523888486e-08,
+      "loss": 1.4128,
+      "step": 21513
+    },
+    {
+      "epoch": 0.9911088588934445,
+      "grad_norm": 1.072223405558702,
+      "learning_rate": 4.14613905149519e-08,
+      "loss": 1.0067,
+      "step": 21514
+    },
+    {
+      "epoch": 0.9911549269820795,
+      "grad_norm": 0.7933852853971807,
+      "learning_rate": 4.103288119648996e-08,
+      "loss": 1.3823,
+      "step": 21515
+    },
+    {
+      "epoch": 0.9912009950707145,
+      "grad_norm": 0.7261390721292842,
+      "learning_rate": 4.060659729303584e-08,
+      "loss": 1.1281,
+      "step": 21516
+    },
+    {
+      "epoch": 0.9912470631593495,
+      "grad_norm": 0.9003791199373986,
+      "learning_rate": 4.018253881409306e-08,
+      "loss": 1.1811,
+      "step": 21517
+    },
+    {
+      "epoch": 0.9912931312479846,
+      "grad_norm": 1.1947955753146302,
+      "learning_rate": 3.976070576908742e-08,
+      "loss": 1.3176,
+      "step": 21518
+    },
+    {
+      "epoch": 0.9913391993366195,
+      "grad_norm": 6.846984030349887,
+      "learning_rate": 3.934109816741138e-08,
+      "loss": 1.2388,
+      "step": 21519
+    },
+    {
+      "epoch": 0.9913852674252546,
+      "grad_norm": 2.6987845022894623,
+      "learning_rate": 3.8923716018401943e-08,
+      "loss": 1.2566,
+      "step": 21520
+    },
+    {
+      "epoch": 0.9914313355138895,
+      "grad_norm": 0.7541139281044368,
+      "learning_rate": 3.850855933137387e-08,
+      "loss": 0.9042,
+      "step": 21521
+    },
+    {
+      "epoch": 0.9914774036025246,
+      "grad_norm": 3.311043853373493,
+      "learning_rate": 3.8095628115553114e-08,
+      "loss": 1.243,
+      "step": 21522
+    },
+    {
+      "epoch": 0.9915234716911595,
+      "grad_norm": 0.9586566637745106,
+      "learning_rate": 3.7684922380132324e-08,
+      "loss": 1.147,
+      "step": 21523
+    },
+    {
+      "epoch": 0.9915695397797946,
+      "grad_norm": 0.9359946330825151,
+      "learning_rate": 3.727644213424863e-08,
+      "loss": 1.0982,
+      "step": 21524
+    },
+    {
+      "epoch": 0.9916156078684295,
+      "grad_norm": 0.9731633495925633,
+      "learning_rate": 3.687018738701697e-08,
+      "loss": 1.2446,
+      "step": 21525
+    },
+    {
+      "epoch": 0.9916616759570646,
+      "grad_norm": 1.4342280541565777,
+      "learning_rate": 3.646615814747456e-08,
+      "loss": 1.4469,
+      "step": 21526
+    },
+    {
+      "epoch": 0.9917077440456995,
+      "grad_norm": 1.3033387849211402,
+      "learning_rate": 3.6064354424603096e-08,
+      "loss": 1.2553,
+      "step": 21527
+    },
+    {
+      "epoch": 0.9917538121343346,
+      "grad_norm": 0.7709586580426135,
+      "learning_rate": 3.566477622736208e-08,
+      "loss": 1.5003,
+      "step": 21528
+    },
+    {
+      "epoch": 0.9917998802229695,
+      "grad_norm": 1.0000894519468093,
+      "learning_rate": 3.526742356463331e-08,
+      "loss": 1.0965,
+      "step": 21529
+    },
+    {
+      "epoch": 0.9918459483116046,
+      "grad_norm": 1.1231825108623885,
+      "learning_rate": 3.487229644528744e-08,
+      "loss": 1.4456,
+      "step": 21530
+    },
+    {
+      "epoch": 0.9918920164002396,
+      "grad_norm": 1.392070634372661,
+      "learning_rate": 3.447939487809526e-08,
+      "loss": 1.2185,
+      "step": 21531
+    },
+    {
+      "epoch": 0.9919380844888745,
+      "grad_norm": 0.9414901630248458,
+      "learning_rate": 3.408871887182752e-08,
+      "loss": 1.0054,
+      "step": 21532
+    },
+    {
+      "epoch": 0.9919841525775096,
+      "grad_norm": 0.784432037613797,
+      "learning_rate": 3.370026843515506e-08,
+      "loss": 1.3494,
+      "step": 21533
+    },
+    {
+      "epoch": 0.9920302206661445,
+      "grad_norm": 1.1215894921327614,
+      "learning_rate": 3.3314043576748724e-08,
+      "loss": 1.4281,
+      "step": 21534
+    },
+    {
+      "epoch": 0.9920762887547796,
+      "grad_norm": 0.7832511915640359,
+      "learning_rate": 3.293004430520163e-08,
+      "loss": 1.481,
+      "step": 21535
+    },
+    {
+      "epoch": 0.9921223568434145,
+      "grad_norm": 0.8330566484962368,
+      "learning_rate": 3.2548270629051415e-08,
+      "loss": 1.0621,
+      "step": 21536
+    },
+    {
+      "epoch": 0.9921684249320496,
+      "grad_norm": 1.0721497039114856,
+      "learning_rate": 3.216872255681347e-08,
+      "loss": 1.1758,
+      "step": 21537
+    },
+    {
+      "epoch": 0.9922144930206845,
+      "grad_norm": 0.7260520718611978,
+      "learning_rate": 3.1791400096925495e-08,
+      "loss": 1.5477,
+      "step": 21538
+    },
+    {
+      "epoch": 0.9922605611093196,
+      "grad_norm": 1.1098146309429233,
+      "learning_rate": 3.141630325780298e-08,
+      "loss": 1.5049,
+      "step": 21539
+    },
+    {
+      "epoch": 0.9923066291979545,
+      "grad_norm": 1.8347651432396874,
+      "learning_rate": 3.10434320477726e-08,
+      "loss": 1.0742,
+      "step": 21540
+    },
+    {
+      "epoch": 0.9923526972865896,
+      "grad_norm": 2.2776015845200916,
+      "learning_rate": 3.067278647516103e-08,
+      "loss": 1.1936,
+      "step": 21541
+    },
+    {
+      "epoch": 0.9923987653752245,
+      "grad_norm": 1.0355190821448266,
+      "learning_rate": 3.030436654820612e-08,
+      "loss": 1.2198,
+      "step": 21542
+    },
+    {
+      "epoch": 0.9924448334638596,
+      "grad_norm": 2.4355829952567385,
+      "learning_rate": 2.993817227511242e-08,
+      "loss": 1.1147,
+      "step": 21543
+    },
+    {
+      "epoch": 0.9924909015524946,
+      "grad_norm": 0.984979819700457,
+      "learning_rate": 2.9574203664040068e-08,
+      "loss": 1.3694,
+      "step": 21544
+    },
+    {
+      "epoch": 0.9925369696411296,
+      "grad_norm": 1.60805779871656,
+      "learning_rate": 2.921246072307149e-08,
+      "loss": 1.4798,
+      "step": 21545
+    },
+    {
+      "epoch": 0.9925830377297646,
+      "grad_norm": 1.3650514569339907,
+      "learning_rate": 2.8852943460289107e-08,
+      "loss": 1.2715,
+      "step": 21546
+    },
+    {
+      "epoch": 0.9926291058183996,
+      "grad_norm": 1.2625718262332608,
+      "learning_rate": 2.8495651883675424e-08,
+      "loss": 1.3787,
+      "step": 21547
+    },
+    {
+      "epoch": 0.9926751739070346,
+      "grad_norm": 0.8021584725122108,
+      "learning_rate": 2.8140586001190737e-08,
+      "loss": 1.3049,
+      "step": 21548
+    },
+    {
+      "epoch": 0.9927212419956696,
+      "grad_norm": 0.8168092830830008,
+      "learning_rate": 2.7787745820739842e-08,
+      "loss": 1.1923,
+      "step": 21549
+    },
+    {
+      "epoch": 0.9927673100843046,
+      "grad_norm": 0.8427335930115224,
+      "learning_rate": 2.743713135018311e-08,
+      "loss": 1.1541,
+      "step": 21550
+    },
+    {
+      "epoch": 0.9928133781729396,
+      "grad_norm": 1.2501986070685778,
+      "learning_rate": 2.708874259732541e-08,
+      "loss": 1.1531,
+      "step": 21551
+    },
+    {
+      "epoch": 0.9928594462615746,
+      "grad_norm": 0.8944408353826961,
+      "learning_rate": 2.6742579569916104e-08,
+      "loss": 1.3751,
+      "step": 21552
+    },
+    {
+      "epoch": 0.9929055143502096,
+      "grad_norm": 0.8662694939770277,
+      "learning_rate": 2.6398642275660135e-08,
+      "loss": 1.3265,
+      "step": 21553
+    },
+    {
+      "epoch": 0.9929515824388446,
+      "grad_norm": 1.0111099240739703,
+      "learning_rate": 2.6056930722229145e-08,
+      "loss": 1.0567,
+      "step": 21554
+    },
+    {
+      "epoch": 0.9929976505274796,
+      "grad_norm": 0.9690077754479983,
+      "learning_rate": 2.571744491721706e-08,
+      "loss": 1.0349,
+      "step": 21555
+    },
+    {
+      "epoch": 0.9930437186161146,
+      "grad_norm": 0.6430214364296446,
+      "learning_rate": 2.53801848681956e-08,
+      "loss": 1.4695,
+      "step": 21556
+    },
+    {
+      "epoch": 0.9930897867047496,
+      "grad_norm": 0.7420680068160221,
+      "learning_rate": 2.5045150582658773e-08,
+      "loss": 1.4982,
+      "step": 21557
+    },
+    {
+      "epoch": 0.9931358547933846,
+      "grad_norm": 3.67220224157274,
+      "learning_rate": 2.4712342068067273e-08,
+      "loss": 0.8213,
+      "step": 21558
+    },
+    {
+      "epoch": 0.9931819228820197,
+      "grad_norm": 2.50986289612842,
+      "learning_rate": 2.4381759331837395e-08,
+      "loss": 1.2143,
+      "step": 21559
+    },
+    {
+      "epoch": 0.9932279909706546,
+      "grad_norm": 0.8257838615002613,
+      "learning_rate": 2.4053402381318814e-08,
+      "loss": 1.2333,
+      "step": 21560
+    },
+    {
+      "epoch": 0.9932740590592897,
+      "grad_norm": 1.205163641885224,
+      "learning_rate": 2.3727271223839e-08,
+      "loss": 1.1157,
+      "step": 21561
+    },
+    {
+      "epoch": 0.9933201271479246,
+      "grad_norm": 0.8336985402702446,
+      "learning_rate": 2.3403365866647707e-08,
+      "loss": 0.9462,
+      "step": 21562
+    },
+    {
+      "epoch": 0.9933661952365597,
+      "grad_norm": 1.4013151797733856,
+      "learning_rate": 2.3081686316961393e-08,
+      "loss": 1.3943,
+      "step": 21563
+    },
+    {
+      "epoch": 0.9934122633251946,
+      "grad_norm": 0.7985012764218158,
+      "learning_rate": 2.276223258192989e-08,
+      "loss": 1.3747,
+      "step": 21564
+    },
+    {
+      "epoch": 0.9934583314138297,
+      "grad_norm": 1.1690954634900848,
+      "learning_rate": 2.2445004668680824e-08,
+      "loss": 0.9557,
+      "step": 21565
+    },
+    {
+      "epoch": 0.9935043995024646,
+      "grad_norm": 0.9623389466769615,
+      "learning_rate": 2.213000258426412e-08,
+      "loss": 0.9685,
+      "step": 21566
+    },
+    {
+      "epoch": 0.9935504675910997,
+      "grad_norm": 2.1843226135301275,
+      "learning_rate": 2.1817226335696384e-08,
+      "loss": 0.9524,
+      "step": 21567
+    },
+    {
+      "epoch": 0.9935965356797346,
+      "grad_norm": 0.8918643218214802,
+      "learning_rate": 2.1506675929938713e-08,
+      "loss": 1.0139,
+      "step": 21568
+    },
+    {
+      "epoch": 0.9936426037683697,
+      "grad_norm": 0.789604746189584,
+      "learning_rate": 2.11983513739189e-08,
+      "loss": 1.5443,
+      "step": 21569
+    },
+    {
+      "epoch": 0.9936886718570046,
+      "grad_norm": 1.2709132143502189,
+      "learning_rate": 2.089225267447592e-08,
+      "loss": 1.1413,
+      "step": 21570
+    },
+    {
+      "epoch": 0.9937347399456397,
+      "grad_norm": 0.9530093706340396,
+      "learning_rate": 2.0588379838448745e-08,
+      "loss": 1.1142,
+      "step": 21571
+    },
+    {
+      "epoch": 0.9937808080342747,
+      "grad_norm": 0.8921788520759392,
+      "learning_rate": 2.0286732872598634e-08,
+      "loss": 1.4164,
+      "step": 21572
+    },
+    {
+      "epoch": 0.9938268761229097,
+      "grad_norm": 2.4443203418969848,
+      "learning_rate": 1.9987311783631335e-08,
+      "loss": 1.1639,
+      "step": 21573
+    },
+    {
+      "epoch": 0.9938729442115447,
+      "grad_norm": 1.1668353021113063,
+      "learning_rate": 1.9690116578219286e-08,
+      "loss": 1.3764,
+      "step": 21574
+    },
+    {
+      "epoch": 0.9939190123001796,
+      "grad_norm": 1.010623947500593,
+      "learning_rate": 1.9395147262968317e-08,
+      "loss": 1.2973,
+      "step": 21575
+    },
+    {
+      "epoch": 0.9939650803888147,
+      "grad_norm": 0.8898451427839096,
+      "learning_rate": 1.9102403844462047e-08,
+      "loss": 1.2691,
+      "step": 21576
+    },
+    {
+      "epoch": 0.9940111484774496,
+      "grad_norm": 0.7914659980565069,
+      "learning_rate": 1.8811886329217487e-08,
+      "loss": 1.4245,
+      "step": 21577
+    },
+    {
+      "epoch": 0.9940572165660847,
+      "grad_norm": 0.7832894858322704,
+      "learning_rate": 1.8523594723685032e-08,
+      "loss": 1.362,
+      "step": 21578
+    },
+    {
+      "epoch": 0.9941032846547196,
+      "grad_norm": 0.8955949346465102,
+      "learning_rate": 1.8237529034292876e-08,
+      "loss": 1.3057,
+      "step": 21579
+    },
+    {
+      "epoch": 0.9941493527433547,
+      "grad_norm": 0.6765891435245239,
+      "learning_rate": 1.7953689267413698e-08,
+      "loss": 1.0109,
+      "step": 21580
+    },
+    {
+      "epoch": 0.9941954208319896,
+      "grad_norm": 0.970128827793069,
+      "learning_rate": 1.7672075429364665e-08,
+      "loss": 1.2171,
+      "step": 21581
+    },
+    {
+      "epoch": 0.9942414889206247,
+      "grad_norm": 1.2721599430804214,
+      "learning_rate": 1.7392687526407437e-08,
+      "loss": 1.161,
+      "step": 21582
+    },
+    {
+      "epoch": 0.9942875570092596,
+      "grad_norm": 1.1817903579639597,
+      "learning_rate": 1.711552556477036e-08,
+      "loss": 1.2052,
+      "step": 21583
+    },
+    {
+      "epoch": 0.9943336250978947,
+      "grad_norm": 1.059992984071674,
+      "learning_rate": 1.6840589550626284e-08,
+      "loss": 1.065,
+      "step": 21584
+    },
+    {
+      "epoch": 0.9943796931865297,
+      "grad_norm": 0.8051901491068411,
+      "learning_rate": 1.656787949009253e-08,
+      "loss": 1.4083,
+      "step": 21585
+    },
+    {
+      "epoch": 0.9944257612751647,
+      "grad_norm": 0.9453133488951609,
+      "learning_rate": 1.629739538924202e-08,
+      "loss": 1.3522,
+      "step": 21586
+    },
+    {
+      "epoch": 0.9944718293637997,
+      "grad_norm": 1.5231809399877834,
+      "learning_rate": 1.602913725408106e-08,
+      "loss": 1.3357,
+      "step": 21587
+    },
+    {
+      "epoch": 0.9945178974524347,
+      "grad_norm": 0.7759721620339927,
+      "learning_rate": 1.576310509060486e-08,
+      "loss": 1.191,
+      "step": 21588
+    },
+    {
+      "epoch": 0.9945639655410697,
+      "grad_norm": 0.9574306333903998,
+      "learning_rate": 1.5499298904730897e-08,
+      "loss": 1.2079,
+      "step": 21589
+    },
+    {
+      "epoch": 0.9946100336297047,
+      "grad_norm": 0.8984157351262388,
+      "learning_rate": 1.5237718702321157e-08,
+      "loss": 1.3609,
+      "step": 21590
+    },
+    {
+      "epoch": 0.9946561017183397,
+      "grad_norm": 1.2733647977353184,
+      "learning_rate": 1.4978364489215413e-08,
+      "loss": 1.3661,
+      "step": 21591
+    },
+    {
+      "epoch": 0.9947021698069747,
+      "grad_norm": 1.5219904926183827,
+      "learning_rate": 1.472123627117572e-08,
+      "loss": 1.2004,
+      "step": 21592
+    },
+    {
+      "epoch": 0.9947482378956097,
+      "grad_norm": 0.9349084111819203,
+      "learning_rate": 1.446633405393083e-08,
+      "loss": 1.3272,
+      "step": 21593
+    },
+    {
+      "epoch": 0.9947943059842447,
+      "grad_norm": 2.5778124045810222,
+      "learning_rate": 1.421365784314288e-08,
+      "loss": 1.0451,
+      "step": 21594
+    },
+    {
+      "epoch": 0.9948403740728797,
+      "grad_norm": 0.9365064782027099,
+      "learning_rate": 1.3963207644462906e-08,
+      "loss": 1.0895,
+      "step": 21595
+    },
+    {
+      "epoch": 0.9948864421615147,
+      "grad_norm": 0.8330735209837793,
+      "learning_rate": 1.3714983463442022e-08,
+      "loss": 1.1102,
+      "step": 21596
+    },
+    {
+      "epoch": 0.9949325102501497,
+      "grad_norm": 0.9561616455105378,
+      "learning_rate": 1.346898530562024e-08,
+      "loss": 1.2836,
+      "step": 21597
+    },
+    {
+      "epoch": 0.9949785783387847,
+      "grad_norm": 0.7071857362610257,
+      "learning_rate": 1.3225213176482065e-08,
+      "loss": 1.2363,
+      "step": 21598
+    },
+    {
+      "epoch": 0.9950246464274197,
+      "grad_norm": 0.8004012362110456,
+      "learning_rate": 1.298366708143428e-08,
+      "loss": 1.2631,
+      "step": 21599
+    },
+    {
+      "epoch": 0.9950707145160548,
+      "grad_norm": 0.8141940190741423,
+      "learning_rate": 1.2744347025861469e-08,
+      "loss": 1.5518,
+      "step": 21600
+    },
+    {
+      "epoch": 0.9951167826046897,
+      "grad_norm": 1.0142219264331416,
+      "learning_rate": 1.2507253015103804e-08,
+      "loss": 1.3571,
+      "step": 21601
+    },
+    {
+      "epoch": 0.9951628506933248,
+      "grad_norm": 0.9337790912361478,
+      "learning_rate": 1.227238505442374e-08,
+      "loss": 1.1516,
+      "step": 21602
+    },
+    {
+      "epoch": 0.9952089187819597,
+      "grad_norm": 1.0019159278288796,
+      "learning_rate": 1.203974314905043e-08,
+      "loss": 1.5332,
+      "step": 21603
+    },
+    {
+      "epoch": 0.9952549868705948,
+      "grad_norm": 1.3255552639470847,
+      "learning_rate": 1.1809327304179717e-08,
+      "loss": 0.9523,
+      "step": 21604
+    },
+    {
+      "epoch": 0.9953010549592297,
+      "grad_norm": 2.3569328810589836,
+      "learning_rate": 1.158113752492973e-08,
+      "loss": 1.0178,
+      "step": 21605
+    },
+    {
+      "epoch": 0.9953471230478648,
+      "grad_norm": 0.9469794474905348,
+      "learning_rate": 1.1355173816385289e-08,
+      "loss": 1.0026,
+      "step": 21606
+    },
+    {
+      "epoch": 0.9953931911364997,
+      "grad_norm": 1.0444323780434388,
+      "learning_rate": 1.1131436183564603e-08,
+      "loss": 1.3561,
+      "step": 21607
+    },
+    {
+      "epoch": 0.9954392592251348,
+      "grad_norm": 0.8986112639676435,
+      "learning_rate": 1.0909924631463675e-08,
+      "loss": 1.0371,
+      "step": 21608
+    },
+    {
+      "epoch": 0.9954853273137697,
+      "grad_norm": 2.373261525594255,
+      "learning_rate": 1.0690639165011895e-08,
+      "loss": 1.2115,
+      "step": 21609
+    },
+    {
+      "epoch": 0.9955313954024048,
+      "grad_norm": 1.122171472399203,
+      "learning_rate": 1.0473579789083143e-08,
+      "loss": 1.3309,
+      "step": 21610
+    },
+    {
+      "epoch": 0.9955774634910397,
+      "grad_norm": 1.8809967982797045,
+      "learning_rate": 1.0258746508517991e-08,
+      "loss": 1.1064,
+      "step": 21611
+    },
+    {
+      "epoch": 0.9956235315796748,
+      "grad_norm": 0.6738030569641423,
+      "learning_rate": 1.00461393281015e-08,
+      "loss": 1.3553,
+      "step": 21612
+    },
+    {
+      "epoch": 0.9956695996683098,
+      "grad_norm": 1.2808290035083636,
+      "learning_rate": 9.835758252552118e-09,
+      "loss": 1.3224,
+      "step": 21613
+    },
+    {
+      "epoch": 0.9957156677569448,
+      "grad_norm": 1.053570231478655,
+      "learning_rate": 9.627603286566088e-09,
+      "loss": 1.1095,
+      "step": 21614
+    },
+    {
+      "epoch": 0.9957617358455798,
+      "grad_norm": 1.163503084500989,
+      "learning_rate": 9.421674434773042e-09,
+      "loss": 1.3688,
+      "step": 21615
+    },
+    {
+      "epoch": 0.9958078039342148,
+      "grad_norm": 2.5903259835818186,
+      "learning_rate": 9.2179717017693e-09,
+      "loss": 1.1505,
+      "step": 21616
+    },
+    {
+      "epoch": 0.9958538720228498,
+      "grad_norm": 1.3895324590981457,
+      "learning_rate": 9.016495092073474e-09,
+      "loss": 1.011,
+      "step": 21617
+    },
+    {
+      "epoch": 0.9958999401114847,
+      "grad_norm": 1.8823726980006361,
+      "learning_rate": 8.817244610170861e-09,
+      "loss": 1.1866,
+      "step": 21618
+    },
+    {
+      "epoch": 0.9959460082001198,
+      "grad_norm": 0.8503392238673622,
+      "learning_rate": 8.620220260513456e-09,
+      "loss": 1.299,
+      "step": 21619
+    },
+    {
+      "epoch": 0.9959920762887547,
+      "grad_norm": 0.8624416934110696,
+      "learning_rate": 8.42542204747554e-09,
+      "loss": 1.3522,
+      "step": 21620
+    },
+    {
+      "epoch": 0.9960381443773898,
+      "grad_norm": 1.0867326324653512,
+      "learning_rate": 8.232849975398082e-09,
+      "loss": 1.1821,
+      "step": 21621
+    },
+    {
+      "epoch": 0.9960842124660247,
+      "grad_norm": 0.8147434696121018,
+      "learning_rate": 8.042504048555444e-09,
+      "loss": 1.4199,
+      "step": 21622
+    },
+    {
+      "epoch": 0.9961302805546598,
+      "grad_norm": 0.8654331092126935,
+      "learning_rate": 7.85438427121088e-09,
+      "loss": 1.3494,
+      "step": 21623
+    },
+    {
+      "epoch": 0.9961763486432947,
+      "grad_norm": 1.6205454650407503,
+      "learning_rate": 7.668490647538828e-09,
+      "loss": 1.1523,
+      "step": 21624
+    },
+    {
+      "epoch": 0.9962224167319298,
+      "grad_norm": 0.9633457313784664,
+      "learning_rate": 7.48482318166932e-09,
+      "loss": 1.2058,
+      "step": 21625
+    },
+    {
+      "epoch": 0.9962684848205648,
+      "grad_norm": 0.805306176189691,
+      "learning_rate": 7.303381877699078e-09,
+      "loss": 1.0053,
+      "step": 21626
+    },
+    {
+      "epoch": 0.9963145529091998,
+      "grad_norm": 2.0945837494826427,
+      "learning_rate": 7.124166739680416e-09,
+      "loss": 1.1672,
+      "step": 21627
+    },
+    {
+      "epoch": 0.9963606209978348,
+      "grad_norm": 1.1790408885617623,
+      "learning_rate": 6.94717777157683e-09,
+      "loss": 1.0993,
+      "step": 21628
+    },
+    {
+      "epoch": 0.9964066890864698,
+      "grad_norm": 3.2801472917143237,
+      "learning_rate": 6.772414977351816e-09,
+      "loss": 1.3596,
+      "step": 21629
+    },
+    {
+      "epoch": 0.9964527571751048,
+      "grad_norm": 4.62041686026284,
+      "learning_rate": 6.599878360891154e-09,
+      "loss": 1.1421,
+      "step": 21630
+    },
+    {
+      "epoch": 0.9964988252637398,
+      "grad_norm": 1.0338053298478405,
+      "learning_rate": 6.429567926025115e-09,
+      "loss": 1.0906,
+      "step": 21631
+    },
+    {
+      "epoch": 0.9965448933523748,
+      "grad_norm": 0.9214427909361785,
+      "learning_rate": 6.2614836765506614e-09,
+      "loss": 1.2614,
+      "step": 21632
+    },
+    {
+      "epoch": 0.9965909614410098,
+      "grad_norm": 0.9468374577747518,
+      "learning_rate": 6.095625616220346e-09,
+      "loss": 1.2062,
+      "step": 21633
+    },
+    {
+      "epoch": 0.9966370295296448,
+      "grad_norm": 1.3576864307028402,
+      "learning_rate": 5.931993748709008e-09,
+      "loss": 1.0961,
+      "step": 21634
+    },
+    {
+      "epoch": 0.9966830976182798,
+      "grad_norm": 1.2507886070479686,
+      "learning_rate": 5.770588077669281e-09,
+      "loss": 1.263,
+      "step": 21635
+    },
+    {
+      "epoch": 0.9967291657069148,
+      "grad_norm": 0.6878309986692126,
+      "learning_rate": 5.611408606698287e-09,
+      "loss": 1.2194,
+      "step": 21636
+    },
+    {
+      "epoch": 0.9967752337955498,
+      "grad_norm": 0.8163139930922514,
+      "learning_rate": 5.4544553393376386e-09,
+      "loss": 1.436,
+      "step": 21637
+    },
+    {
+      "epoch": 0.9968213018841848,
+      "grad_norm": 0.7109340908969963,
+      "learning_rate": 5.2997282790734345e-09,
+      "loss": 1.2167,
+      "step": 21638
+    },
+    {
+      "epoch": 0.9968673699728198,
+      "grad_norm": 1.6286097660220318,
+      "learning_rate": 5.147227429369572e-09,
+      "loss": 0.9125,
+      "step": 21639
+    },
+    {
+      "epoch": 0.9969134380614548,
+      "grad_norm": 2.8495076490148104,
+      "learning_rate": 4.996952793590026e-09,
+      "loss": 0.8081,
+      "step": 21640
+    },
+    {
+      "epoch": 0.9969595061500899,
+      "grad_norm": 1.8684533521782878,
+      "learning_rate": 4.848904375109875e-09,
+      "loss": 1.0316,
+      "step": 21641
+    },
+    {
+      "epoch": 0.9970055742387248,
+      "grad_norm": 0.9122003848331615,
+      "learning_rate": 4.703082177215378e-09,
+      "loss": 0.7153,
+      "step": 21642
+    },
+    {
+      "epoch": 0.9970516423273599,
+      "grad_norm": 0.8297542801866371,
+      "learning_rate": 4.559486203148388e-09,
+      "loss": 1.1442,
+      "step": 21643
+    },
+    {
+      "epoch": 0.9970977104159948,
+      "grad_norm": 1.099243554454167,
+      "learning_rate": 4.418116456106347e-09,
+      "loss": 1.066,
+      "step": 21644
+    },
+    {
+      "epoch": 0.9971437785046299,
+      "grad_norm": 0.7118212662014688,
+      "learning_rate": 4.278972939242287e-09,
+      "loss": 1.5408,
+      "step": 21645
+    },
+    {
+      "epoch": 0.9971898465932648,
+      "grad_norm": 1.4362812302876498,
+      "learning_rate": 4.142055655653732e-09,
+      "loss": 1.3404,
+      "step": 21646
+    },
+    {
+      "epoch": 0.9972359146818999,
+      "grad_norm": 1.3629430661994493,
+      "learning_rate": 4.0073646083715884e-09,
+      "loss": 1.0888,
+      "step": 21647
+    },
+    {
+      "epoch": 0.9972819827705348,
+      "grad_norm": 2.2100855393400276,
+      "learning_rate": 3.8748998004267676e-09,
+      "loss": 1.112,
+      "step": 21648
+    },
+    {
+      "epoch": 0.9973280508591699,
+      "grad_norm": 0.7158178246776601,
+      "learning_rate": 3.744661234739155e-09,
+      "loss": 1.3644,
+      "step": 21649
+    },
+    {
+      "epoch": 0.9973741189478048,
+      "grad_norm": 0.942036119220107,
+      "learning_rate": 3.6166489142286377e-09,
+      "loss": 1.3107,
+      "step": 21650
+    },
+    {
+      "epoch": 0.9974201870364399,
+      "grad_norm": 1.5908115889363665,
+      "learning_rate": 3.4908628417373857e-09,
+      "loss": 1.2474,
+      "step": 21651
+    },
+    {
+      "epoch": 0.9974662551250748,
+      "grad_norm": 0.8861920310940888,
+      "learning_rate": 3.3673030200520594e-09,
+      "loss": 0.9509,
+      "step": 21652
+    },
+    {
+      "epoch": 0.9975123232137099,
+      "grad_norm": 1.1194164105994997,
+      "learning_rate": 3.2459694519482162e-09,
+      "loss": 1.1007,
+      "step": 21653
+    },
+    {
+      "epoch": 0.9975583913023449,
+      "grad_norm": 1.7755772555508103,
+      "learning_rate": 3.1268621401125964e-09,
+      "loss": 1.079,
+      "step": 21654
+    },
+    {
+      "epoch": 0.9976044593909799,
+      "grad_norm": 0.8351181880154243,
+      "learning_rate": 3.009981087198632e-09,
+      "loss": 1.0855,
+      "step": 21655
+    },
+    {
+      "epoch": 0.9976505274796149,
+      "grad_norm": 0.8626554454964646,
+      "learning_rate": 2.895326295804246e-09,
+      "loss": 1.2608,
+      "step": 21656
+    },
+    {
+      "epoch": 0.9976965955682499,
+      "grad_norm": 0.9514880970684986,
+      "learning_rate": 2.7828977684940525e-09,
+      "loss": 1.0801,
+      "step": 21657
+    },
+    {
+      "epoch": 0.9977426636568849,
+      "grad_norm": 0.7842554312013447,
+      "learning_rate": 2.672695507754952e-09,
+      "loss": 1.1484,
+      "step": 21658
+    },
+    {
+      "epoch": 0.9977887317455199,
+      "grad_norm": 0.9639220640203608,
+      "learning_rate": 2.564719516062741e-09,
+      "loss": 1.0309,
+      "step": 21659
+    },
+    {
+      "epoch": 0.9978347998341549,
+      "grad_norm": 5.893335510855347,
+      "learning_rate": 2.4589697958044e-09,
+      "loss": 1.0896,
+      "step": 21660
+    },
+    {
+      "epoch": 0.9978808679227898,
+      "grad_norm": 1.237578097349438,
+      "learning_rate": 2.355446349333601e-09,
+      "loss": 1.0671,
+      "step": 21661
+    },
+    {
+      "epoch": 0.9979269360114249,
+      "grad_norm": 3.309447811014566,
+      "learning_rate": 2.254149178959608e-09,
+      "loss": 1.2633,
+      "step": 21662
+    },
+    {
+      "epoch": 0.9979730041000598,
+      "grad_norm": 1.1766115766294134,
+      "learning_rate": 2.155078286936174e-09,
+      "loss": 1.5265,
+      "step": 21663
+    },
+    {
+      "epoch": 0.9980190721886949,
+      "grad_norm": 1.261377135630768,
+      "learning_rate": 2.058233675472643e-09,
+      "loss": 1.3161,
+      "step": 21664
+    },
+    {
+      "epoch": 0.9980651402773298,
+      "grad_norm": 2.9265974640791073,
+      "learning_rate": 1.963615346733949e-09,
+      "loss": 0.9197,
+      "step": 21665
+    },
+    {
+      "epoch": 0.9981112083659649,
+      "grad_norm": 1.124163723768451,
+      "learning_rate": 1.87122330279621e-09,
+      "loss": 1.2767,
+      "step": 21666
+    },
+    {
+      "epoch": 0.9981572764546,
+      "grad_norm": 0.8206219469866542,
+      "learning_rate": 1.781057545746645e-09,
+      "loss": 0.9884,
+      "step": 21667
+    },
+    {
+      "epoch": 0.9982033445432349,
+      "grad_norm": 0.6893408698138835,
+      "learning_rate": 1.6931180775836553e-09,
+      "loss": 1.4566,
+      "step": 21668
+    },
+    {
+      "epoch": 0.99824941263187,
+      "grad_norm": 0.7513770563975293,
+      "learning_rate": 1.6074049002612335e-09,
+      "loss": 1.4497,
+      "step": 21669
+    },
+    {
+      "epoch": 0.9982954807205049,
+      "grad_norm": 1.0794222136615717,
+      "learning_rate": 1.5239180156889633e-09,
+      "loss": 1.4434,
+      "step": 21670
+    },
+    {
+      "epoch": 0.99834154880914,
+      "grad_norm": 1.2735045093264037,
+      "learning_rate": 1.442657425720917e-09,
+      "loss": 1.3901,
+      "step": 21671
+    },
+    {
+      "epoch": 0.9983876168977749,
+      "grad_norm": 1.083357360371921,
+      "learning_rate": 1.3636231321778602e-09,
+      "loss": 1.0627,
+      "step": 21672
+    },
+    {
+      "epoch": 0.9984336849864099,
+      "grad_norm": 1.5561575605915832,
+      "learning_rate": 1.2868151368028435e-09,
+      "loss": 1.1931,
+      "step": 21673
+    },
+    {
+      "epoch": 0.9984797530750449,
+      "grad_norm": 0.8498417639523637,
+      "learning_rate": 1.2122334413278146e-09,
+      "loss": 1.0351,
+      "step": 21674
+    },
+    {
+      "epoch": 0.9985258211636799,
+      "grad_norm": 1.0869334696996453,
+      "learning_rate": 1.1398780473959037e-09,
+      "loss": 1.2696,
+      "step": 21675
+    },
+    {
+      "epoch": 0.9985718892523149,
+      "grad_norm": 0.9256138759220954,
+      "learning_rate": 1.0697489566280362e-09,
+      "loss": 1.1768,
+      "step": 21676
+    },
+    {
+      "epoch": 0.9986179573409499,
+      "grad_norm": 1.1068591098386078,
+      "learning_rate": 1.001846170567422e-09,
+      "loss": 1.3349,
+      "step": 21677
+    },
+    {
+      "epoch": 0.9986640254295849,
+      "grad_norm": 0.7613604227444113,
+      "learning_rate": 9.361696907572714e-10,
+      "loss": 0.9784,
+      "step": 21678
+    },
+    {
+      "epoch": 0.9987100935182199,
+      "grad_norm": 1.3884342450751241,
+      "learning_rate": 8.727195186297721e-10,
+      "loss": 1.1655,
+      "step": 21679
+    },
+    {
+      "epoch": 0.998756161606855,
+      "grad_norm": 0.7689778929243138,
+      "learning_rate": 8.114956556171116e-10,
+      "loss": 1.3665,
+      "step": 21680
+    },
+    {
+      "epoch": 0.9988022296954899,
+      "grad_norm": 1.259485983943609,
+      "learning_rate": 7.524981030626598e-10,
+      "loss": 1.1323,
+      "step": 21681
+    },
+    {
+      "epoch": 0.998848297784125,
+      "grad_norm": 0.7771101060171148,
+      "learning_rate": 6.957268622986846e-10,
+      "loss": 1.1608,
+      "step": 21682
+    },
+    {
+      "epoch": 0.9988943658727599,
+      "grad_norm": 1.288212435458082,
+      "learning_rate": 6.411819345908399e-10,
+      "loss": 1.1237,
+      "step": 21683
+    },
+    {
+      "epoch": 0.998940433961395,
+      "grad_norm": 1.0930381944300036,
+      "learning_rate": 5.888633211381667e-10,
+      "loss": 1.2058,
+      "step": 21684
+    },
+    {
+      "epoch": 0.9989865020500299,
+      "grad_norm": 1.1138745850551712,
+      "learning_rate": 5.387710231063991e-10,
+      "loss": 1.19,
+      "step": 21685
+    },
+    {
+      "epoch": 0.999032570138665,
+      "grad_norm": 0.9071826069395514,
+      "learning_rate": 4.909050416168626e-10,
+      "loss": 1.2528,
+      "step": 21686
+    },
+    {
+      "epoch": 0.9990786382272999,
+      "grad_norm": 2.0569238529909164,
+      "learning_rate": 4.4526537774647324e-10,
+      "loss": 1.18,
+      "step": 21687
+    },
+    {
+      "epoch": 0.999124706315935,
+      "grad_norm": 0.7839265892155309,
+      "learning_rate": 4.018520324944319e-10,
+      "loss": 1.2705,
+      "step": 21688
+    },
+    {
+      "epoch": 0.9991707744045699,
+      "grad_norm": 1.0155075719562054,
+      "learning_rate": 3.6066500682663264e-10,
+      "loss": 1.3008,
+      "step": 21689
+    },
+    {
+      "epoch": 0.999216842493205,
+      "grad_norm": 0.6882484767145587,
+      "learning_rate": 3.217043016756627e-10,
+      "loss": 1.0982,
+      "step": 21690
+    },
+    {
+      "epoch": 0.9992629105818399,
+      "grad_norm": 1.2088299639494535,
+      "learning_rate": 2.8496991788529157e-10,
+      "loss": 1.1173,
+      "step": 21691
+    },
+    {
+      "epoch": 0.999308978670475,
+      "grad_norm": 2.247488672969709,
+      "learning_rate": 2.5046185629928885e-10,
+      "loss": 0.9761,
+      "step": 21692
+    },
+    {
+      "epoch": 0.9993550467591099,
+      "grad_norm": 0.7724605019725375,
+      "learning_rate": 2.1818011767260615e-10,
+      "loss": 1.1151,
+      "step": 21693
+    },
+    {
+      "epoch": 0.999401114847745,
+      "grad_norm": 1.916832753280351,
+      "learning_rate": 1.881247027157862e-10,
+      "loss": 1.2292,
+      "step": 21694
+    },
+    {
+      "epoch": 0.99944718293638,
+      "grad_norm": 0.9861343770803095,
+      "learning_rate": 1.6029561211716727e-10,
+      "loss": 1.1573,
+      "step": 21695
+    },
+    {
+      "epoch": 0.999493251025015,
+      "grad_norm": 1.0597993957715799,
+      "learning_rate": 1.34692846487372e-10,
+      "loss": 1.2659,
+      "step": 21696
+    },
+    {
+      "epoch": 0.99953931911365,
+      "grad_norm": 1.2057802830555464,
+      "learning_rate": 1.113164063926142e-10,
+      "loss": 1.1392,
+      "step": 21697
+    },
+    {
+      "epoch": 0.999585387202285,
+      "grad_norm": 0.7118516881250079,
+      "learning_rate": 9.016629235469864e-11,
+      "loss": 1.2013,
+      "step": 21698
+    },
+    {
+      "epoch": 0.99963145529092,
+      "grad_norm": 0.9711977682864302,
+      "learning_rate": 7.124250485102124e-11,
+      "loss": 1.2241,
+      "step": 21699
+    },
+    {
+      "epoch": 0.999677523379555,
+      "grad_norm": 1.1853315913804563,
+      "learning_rate": 5.454504429236451e-11,
+      "loss": 1.2042,
+      "step": 21700
+    },
+    {
+      "epoch": 0.99972359146819,
+      "grad_norm": 0.7261232294024941,
+      "learning_rate": 4.0073911056204285e-11,
+      "loss": 1.1514,
+      "step": 21701
+    },
+    {
+      "epoch": 0.999769659556825,
+      "grad_norm": 3.9712089223972984,
+      "learning_rate": 2.7829105475607465e-11,
+      "loss": 1.5327,
+      "step": 21702
+    },
+    {
+      "epoch": 0.99981572764546,
+      "grad_norm": 0.910668734726166,
+      "learning_rate": 1.781062779482312e-11,
+      "loss": 0.9263,
+      "step": 21703
+    },
+    {
+      "epoch": 0.999861795734095,
+      "grad_norm": 0.9994139658018105,
+      "learning_rate": 1.0018478269202547e-11,
+      "loss": 1.3145,
+      "step": 21704
+    },
+    {
+      "epoch": 0.99990786382273,
+      "grad_norm": 0.8901729947063939,
+      "learning_rate": 4.452657054176968e-12,
+      "loss": 1.2288,
+      "step": 21705
+    },
+    {
+      "epoch": 0.999953931911365,
+      "grad_norm": 0.8098067441407663,
+      "learning_rate": 1.1131642718709146e-12,
+      "loss": 1.1766,
+      "step": 21706
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0981499096860188,
+      "learning_rate": 0.0,
+      "loss": 0.9875,
+      "step": 21707
+    },
+    {
+      "epoch": 1.0,
+      "step": 21707,
+      "total_flos": 6571206790545408.0,
+      "train_loss": 0.011552139646460999,
+      "train_runtime": 2160.6123,
+      "train_samples_per_second": 321.488,
+      "train_steps_per_second": 10.047
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 21707,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6571206790545408.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}