End of training

README.md

@@ -3,6 +3,8 @@ license: mit
 base_model: cointegrated/rubert-tiny
 tags:
 - generated_from_trainer
+metrics:
+- accuracy
 model-index:
 - name: finetuned-tiny-bert_second_attempt
   results: []
@@ -14,6 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 # finetuned-tiny-bert_second_attempt
 
 This model is a fine-tuned version of [cointegrated/rubert-tiny](https://huggingface.co/cointegrated/rubert-tiny) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.2433
+- Accuracy: 0.5666
 
 ## Model description
 

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 1.0,
-    "eval_accuracy": 0.5167136043030395,
-    "eval_loss": 2.5832159519195557,
-    "eval_runtime": 62.4114,
+    "epoch": 3.0,
+    "eval_accuracy": 0.5665694910957602,
+    "eval_loss": 2.2432701587677,
+    "eval_runtime": 62.5137,
     "eval_samples": 15263,
-    "eval_samples_per_second": 244.555,
-    "eval_steps_per_second": 7.643,
-    "perplexity": 13.23964783936109,
-    "train_loss": 3.04393067064311,
-    "train_runtime": 388.0084,
+    "eval_samples_per_second": 244.155,
+    "eval_steps_per_second": 7.63,
+    "perplexity": 9.424099252756871,
+    "train_loss": 2.6927158644757037,
+    "train_runtime": 1168.1664,
     "train_samples": 35613,
-    "train_samples_per_second": 91.784,
-    "train_steps_per_second": 2.868
+    "train_samples_per_second": 91.459,
+    "train_steps_per_second": 2.858
 }

eval_results.json

@@ -1,10 +1,10 @@
 {
-    "epoch": 1.0,
-    "eval_accuracy": 0.5167136043030395,
-    "eval_loss": 2.5832159519195557,
-    "eval_runtime": 62.4114,
+    "epoch": 3.0,
+    "eval_accuracy": 0.5665694910957602,
+    "eval_loss": 2.2432701587677,
+    "eval_runtime": 62.5137,
     "eval_samples": 15263,
-    "eval_samples_per_second": 244.555,
-    "eval_steps_per_second": 7.643,
-    "perplexity": 13.23964783936109
+    "eval_samples_per_second": 244.155,
+    "eval_steps_per_second": 7.63,
+    "perplexity": 9.424099252756871
 }

runs/Feb25_11-49-43_e77a61bc3de0/events.out.tfevents.1708863022.e77a61bc3de0.27155.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d433545a77706cfba0591526c7e36429313a4a024a34d87675ad3714c3bde36f
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.0,
-    "train_loss": 3.04393067064311,
-    "train_runtime": 388.0084,
+    "epoch": 3.0,
+    "train_loss": 2.6927158644757037,
+    "train_runtime": 1168.1664,
     "train_samples": 35613,
-    "train_samples_per_second": 91.784,
-    "train_steps_per_second": 2.868
+    "train_samples_per_second": 91.459,
+    "train_steps_per_second": 2.858
 }

trainer_state.json

@@ -1,43 +1,71 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.0,
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 1113,
+  "global_step": 3339,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.45,
-      "grad_norm": 5.323787212371826,
-      "learning_rate": 2.75381850853549e-05,
-      "loss": 3.2259,
+      "grad_norm": 5.29541015625,
+      "learning_rate": 4.251272836178497e-05,
+      "loss": 3.2051,
       "step": 500
     },
     {
       "epoch": 0.9,
-      "grad_norm": 5.575705528259277,
-      "learning_rate": 5.0763701707097935e-06,
-      "loss": 2.9056,
+      "grad_norm": 5.331775188446045,
+      "learning_rate": 3.502545672356993e-05,
+      "loss": 2.8264,
       "step": 1000
     },
     {
-      "epoch": 1.0,
-      "step": 1113,
-      "total_flos": 159583189651080.0,
-      "train_loss": 3.04393067064311,
-      "train_runtime": 388.0084,
-      "train_samples_per_second": 91.784,
-      "train_steps_per_second": 2.868
+      "epoch": 1.35,
+      "grad_norm": 6.496217250823975,
+      "learning_rate": 2.75381850853549e-05,
+      "loss": 2.6611,
+      "step": 1500
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.267590522766113,
+      "learning_rate": 2.0050913447139864e-05,
+      "loss": 2.5819,
+      "step": 2000
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.9951887130737305,
+      "learning_rate": 1.2563641808924828e-05,
+      "loss": 2.5289,
+      "step": 2500
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.234454154968262,
+      "learning_rate": 5.0763701707097935e-06,
+      "loss": 2.5066,
+      "step": 3000
+    },
+    {
+      "epoch": 3.0,
+      "step": 3339,
+      "total_flos": 480295528237080.0,
+      "train_loss": 2.6927158644757037,
+      "train_runtime": 1168.1664,
+      "train_samples_per_second": 91.459,
+      "train_steps_per_second": 2.858
     }
   ],
   "logging_steps": 500,
-  "max_steps": 1113,
+  "max_steps": 3339,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 1,
+  "num_train_epochs": 3,
   "save_steps": 500,
-  "total_flos": 159583189651080.0,
+  "total_flos": 480295528237080.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null