mistral-7b-drug-prots_pair

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0454

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 4
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 64
  • total_eval_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 30
  • training_steps: 5000

Training results

Training Loss Epoch Step Validation Loss
0.1497 0.01 50 0.0787
0.0734 0.02 100 0.0740
0.0694 0.03 150 0.0688
0.0709 0.04 200 0.0679
0.0668 0.05 250 0.0662
0.0651 0.06 300 0.0643
0.0619 0.07 350 0.0627
0.0614 0.08 400 0.0627
0.0607 0.09 450 0.0612
0.0609 0.1 500 0.0601
0.0586 0.11 550 0.0596
0.0578 0.12 600 0.0590
0.0577 0.13 650 0.0584
0.0569 0.14 700 0.0580
0.0564 0.15 750 0.0575
0.0566 0.16 800 0.0569
0.0558 0.17 850 0.0562
0.0556 0.18 900 0.0556
0.0552 0.19 950 0.0551
0.0546 0.2 1000 0.0547
0.0544 0.21 1050 0.0543
0.0541 0.22 1100 0.0545
0.0532 0.23 1150 0.0536
0.0527 0.24 1200 0.0534
0.0526 0.25 1250 0.0525
0.0524 0.26 1300 0.0523
0.0518 0.27 1350 0.0517
0.0514 0.28 1400 0.0518
0.0507 0.29 1450 0.0516
0.0501 0.3 1500 0.0512
0.0493 0.31 1550 0.0508
0.0495 0.32 1600 0.0507
0.0484 0.33 1650 0.0506
0.0482 0.34 1700 0.0504
0.0477 0.35 1750 0.0501
0.0474 0.36 1800 0.0503
0.0467 0.37 1850 0.0501
0.0469 0.38 1900 0.0494
0.0462 0.39 1950 0.0492
0.0455 0.4 2000 0.0490
0.0456 0.41 2050 0.0486
0.0454 0.42 2100 0.0488
0.0452 0.43 2150 0.0487
0.0456 0.44 2200 0.0485
0.0444 0.45 2250 0.0484
0.0441 0.46 2300 0.0477
0.0439 0.47 2350 0.0476
0.0441 0.48 2400 0.0471
0.0437 0.49 2450 0.0475
0.0432 0.5 2500 0.0472
0.0428 0.51 2550 0.0471
0.0431 0.52 2600 0.0472
0.0425 0.53 2650 0.0467
0.0429 0.54 2700 0.0464
0.0421 0.55 2750 0.0464
0.0422 0.56 2800 0.0460
0.0417 0.57 2850 0.0464
0.0419 0.58 2900 0.0462
0.0407 0.59 2950 0.0461
0.0409 0.6 3000 0.0462
0.0408 0.61 3050 0.0459
0.04 0.62 3100 0.0458
0.0401 0.63 3150 0.0453
0.0398 0.64 3200 0.0454
0.0395 0.65 3250 0.0451
0.0395 0.66 3300 0.0452
0.0387 0.67 3350 0.0453
0.0384 0.68 3400 0.0454
0.0386 0.69 3450 0.0451
0.0385 0.7 3500 0.0451
0.0383 0.71 3550 0.0450
0.0385 0.72 3600 0.0448
0.0378 0.73 3650 0.0449
0.0378 0.74 3700 0.0448
0.0376 0.75 3750 0.0453
0.0376 0.76 3800 0.0453
0.0374 0.77 3850 0.0454
0.0376 0.78 3900 0.0454
0.0369 0.79 3950 0.0454
0.037 0.8 4000 0.0448
0.0367 0.81 4050 0.0451
0.0369 0.82 4100 0.0450
0.0366 0.83 4150 0.0449
0.0368 0.84 4200 0.0449
0.0361 0.85 4250 0.0453
0.037 0.86 4300 0.0454
0.0364 0.87 4350 0.0453
0.0356 0.88 4400 0.0453
0.0359 0.89 4450 0.0453
0.0357 0.9 4500 0.0455
0.0357 0.91 4550 0.0454
0.0351 0.92 4600 0.0464
0.0356 0.93 4650 0.0458
0.0353 0.94 4700 0.0460
0.0351 0.95 4750 0.0456
0.0348 0.96 4800 0.0456
0.0355 0.97 4850 0.0455
0.0354 0.98 4900 0.0454
0.0357 0.99 4950 0.0453
0.0356 1.0 5000 0.0454

Framework versions

  • Transformers 4.44.0.dev0
  • Pytorch 2.1.0+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1
Downloads last month
4
Safetensors
Model size
266k params
Tensor type
BF16
·
Inference API
Unable to determine this model's library. Check the docs .