End of training

README.md

@@ -2,9 +2,10 @@
 library_name: transformers
 base_model: IVN-RIN/bioBIT
 tags:
+- token-classification
 - generated_from_trainer
 datasets:
-- drugtemist-it-fasttext-75-ner
+- Rodrigo1771/drugtemist-it-fasttext-75-ner
 metrics:
 - precision
 - recall
@@ -17,8 +18,8 @@ model-index:
       name: Token Classification
       type: token-classification
     dataset:
-      name: drugtemist-it-fasttext-75-ner
-      type: drugtemist-it-fasttext-75-ner
+      name: Rodrigo1771/drugtemist-it-fasttext-75-ner
+      type: Rodrigo1771/drugtemist-it-fasttext-75-ner
       config: DrugTEMIST Italian NER
       split: validation
       args: DrugTEMIST Italian NER
@@ -42,7 +43,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # output
 
-This model is a fine-tuned version of [IVN-RIN/bioBIT](https://huggingface.co/IVN-RIN/bioBIT) on the drugtemist-it-fasttext-75-ner dataset.
+This model is a fine-tuned version of [IVN-RIN/bioBIT](https://huggingface.co/IVN-RIN/bioBIT) on the Rodrigo1771/drugtemist-it-fasttext-75-ner dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.0094
 - Precision: 0.9169

all_results.json

@@ -1,26 +1,26 @@
 {
-    "epoch": 9.988518943742825,
-    "eval_accuracy": 0.9986674294644602,
-    "eval_f1": 0.9284386617100372,
-    "eval_loss": 0.0068497806787490845,
-    "eval_precision": 0.9258572752548656,
-    "eval_recall": 0.9310344827586207,
-    "eval_runtime": 15.1137,
-    "eval_samples": 6946,
-    "eval_samples_per_second": 459.582,
-    "eval_steps_per_second": 57.497,
-    "predict_accuracy": 0.9987433820748771,
-    "predict_f1": 0.9237312163311596,
-    "predict_loss": 0.00740663381293416,
-    "predict_precision": 0.9024930747922437,
-    "predict_recall": 0.945993031358885,
-    "predict_runtime": 28.4889,
-    "predict_samples_per_second": 516.517,
-    "predict_steps_per_second": 64.587,
-    "total_flos": 1.1084127968547612e+16,
-    "train_loss": 0.0030981004957495064,
-    "train_runtime": 1801.8371,
-    "train_samples": 27841,
-    "train_samples_per_second": 154.515,
-    "train_steps_per_second": 2.414
+    "epoch": 9.989701338825952,
+    "eval_accuracy": 0.9986302259153467,
+    "eval_f1": 0.9230769230769231,
+    "eval_loss": 0.009366312995553017,
+    "eval_precision": 0.9169054441260746,
+    "eval_recall": 0.9293320425943853,
+    "eval_runtime": 17.51,
+    "eval_samples": 6798,
+    "eval_samples_per_second": 388.236,
+    "eval_steps_per_second": 48.544,
+    "predict_accuracy": 0.998192886032094,
+    "predict_f1": 0.8964950711938664,
+    "predict_loss": 0.012696487829089165,
+    "predict_precision": 0.8772775991425509,
+    "predict_recall": 0.9165733482642777,
+    "predict_runtime": 33.039,
+    "predict_samples_per_second": 442.053,
+    "predict_steps_per_second": 55.268,
+    "total_flos": 1.7736053837017554e+16,
+    "train_loss": 0.002068036902580679,
+    "train_runtime": 2701.3395,
+    "train_samples": 31053,
+    "train_samples_per_second": 114.954,
+    "train_steps_per_second": 1.795
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 9.988518943742825,
-    "eval_accuracy": 0.9986674294644602,
-    "eval_f1": 0.9284386617100372,
-    "eval_loss": 0.0068497806787490845,
-    "eval_precision": 0.9258572752548656,
-    "eval_recall": 0.9310344827586207,
-    "eval_runtime": 15.1137,
-    "eval_samples": 6946,
-    "eval_samples_per_second": 459.582,
-    "eval_steps_per_second": 57.497
+    "epoch": 9.989701338825952,
+    "eval_accuracy": 0.9986302259153467,
+    "eval_f1": 0.9230769230769231,
+    "eval_loss": 0.009366312995553017,
+    "eval_precision": 0.9169054441260746,
+    "eval_recall": 0.9293320425943853,
+    "eval_runtime": 17.51,
+    "eval_samples": 6798,
+    "eval_samples_per_second": 388.236,
+    "eval_steps_per_second": 48.544
 }

predict_results.json

@@ -1,10 +1,10 @@
 {
-    "predict_accuracy": 0.9987433820748771,
-    "predict_f1": 0.9237312163311596,
-    "predict_loss": 0.00740663381293416,
-    "predict_precision": 0.9024930747922437,
-    "predict_recall": 0.945993031358885,
-    "predict_runtime": 28.4889,
-    "predict_samples_per_second": 516.517,
-    "predict_steps_per_second": 64.587
+    "predict_accuracy": 0.998192886032094,
+    "predict_f1": 0.8964950711938664,
+    "predict_loss": 0.012696487829089165,
+    "predict_precision": 0.8772775991425509,
+    "predict_recall": 0.9165733482642777,
+    "predict_runtime": 33.039,
+    "predict_samples_per_second": 442.053,
+    "predict_steps_per_second": 55.268
 }

tb/events.out.tfevents.1725898709.0a1c9bec2a53.65267.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64471db1189ffc9cdb3c758e8bad5d7fe51fb1f89b23f34814e661acc0324845
+size 560

train.log

@@ -1588,3 +1588,51 @@ Training completed. Do not forget to share your model on huggingface.co/models =
 {'eval_loss': 0.009366312995553017, 'eval_precision': 0.9169054441260746, 'eval_recall': 0.9293320425943853, 'eval_f1': 0.9230769230769231, 'eval_accuracy': 0.9986302259153467, 'eval_runtime': 17.4762, 'eval_samples_per_second': 388.987, 'eval_steps_per_second': 48.638, 'epoch': 9.99}
 {'train_runtime': 2701.3395, 'train_samples_per_second': 114.954, 'train_steps_per_second': 1.795, 'train_loss': 0.002068036902580679, 'epoch': 9.99}
 
+***** train metrics *****
+  epoch                    =     9.9897
+  total_flos               = 16517987GF
+  train_loss               =     0.0021
+  train_runtime            = 0:45:01.33
+  train_samples            =      31053
+  train_samples_per_second =    114.954
+  train_steps_per_second   =      1.795
+09/09/2024 16:18:12 - INFO - __main__ -   *** Evaluate ***
+[INFO|trainer.py:811] 2024-09-09 16:18:12,179 >> The following columns in the evaluation set don't have a corresponding argument in `BertForTokenClassification.forward` and have been ignored: id, tokens, ner_tags. If id, tokens, ner_tags are not expected by `BertForTokenClassification.forward`,  you can safely ignore this message.
+[INFO|trainer.py:3819] 2024-09-09 16:18:12,182 >> 
+***** Running Evaluation *****
+[INFO|trainer.py:3821] 2024-09-09 16:18:12,182 >>   Num examples = 6798
+[INFO|trainer.py:3824] 2024-09-09 16:18:12,182 >>   Batch size = 8
+
  0%|          | 0/850 [00:00<?, ?it/s]
  1%|          | 9/850 [00:00<00:09, 84.85it/s]
  2%|▏         | 18/850 [00:00<00:12, 69.16it/s]
  3%|▎         | 26/850 [00:00<00:12, 67.03it/s]
  4%|▍         | 34/850 [00:00<00:11, 69.34it/s]
  5%|▍         | 42/850 [00:00<00:12, 66.38it/s]
  6%|▌         | 49/850 [00:00<00:14, 55.56it/s]
  7%|▋         | 57/850 [00:00<00:13, 60.68it/s]
  8%|▊         | 64/850 [00:01<00:13, 59.50it/s]
  8%|▊         | 71/850 [00:01<00:14, 53.38it/s]
  9%|▉         | 77/850 [00:01<00:14, 54.52it/s]
 10%|▉         | 83/850 [00:01<00:14, 54.60it/s]
 11%|█         | 91/850 [00:01<00:12, 59.33it/s]
 12%|█▏        | 98/850 [00:01<00:12, 60.45it/s]
 12%|█▏        | 106/850 [00:01<00:11, 64.35it/s]
 13%|█▎        | 113/850 [00:01<00:11, 63.96it/s]
 14%|█▍        | 120/850 [00:01<00:11, 64.78it/s]
 15%|█▍        | 127/850 [00:02<00:11, 65.12it/s]
 16%|█▌        | 134/850 [00:02<00:11, 61.87it/s]
 17%|█▋        | 142/850 [00:02<00:11, 64.23it/s]
 18%|█▊        | 149/850 [00:02<00:11, 60.11it/s]
 18%|█▊        | 156/850 [00:02<00:11, 60.66it/s]
 19%|█▉        | 163/850 [00:02<00:10, 62.78it/s]
 20%|██        | 170/850 [00:02<00:10, 63.49it/s]
 21%|██        | 177/850 [00:02<00:10, 65.22it/s]
 22%|██▏       | 184/850 [00:02<00:10, 62.35it/s]
 22%|██▏       | 191/850 [00:03<00:10, 60.53it/s]
 23%|██▎       | 198/850 [00:03<00:10, 62.55it/s]
 24%|██▍       | 205/850 [00:03<00:10, 61.27it/s]
 25%|██▍       | 212/850 [00:03<00:10, 61.68it/s]
 26%|██▌       | 219/850 [00:03<00:10, 60.02it/s]
 27%|██▋       | 227/850 [00:03<00:09, 63.15it/s]
 28%|██▊       | 234/850 [00:03<00:09, 64.86it/s]
 28%|██▊       | 241/850 [00:03<00:09, 63.81it/s]
 29%|██▉       | 248/850 [00:04<00:09, 61.38it/s]
 30%|███       | 256/850 [00:04<00:09, 65.97it/s]
 31%|███       | 263/850 [00:04<00:09, 65.18it/s]
 32%|███▏      | 271/850 [00:04<00:08, 66.94it/s]
 33%|███▎      | 279/850 [00:04<00:08, 68.88it/s]
 34%|███▍      | 287/850 [00:04<00:08, 69.84it/s]
 35%|███▍      | 295/850 [00:04<00:08, 69.34it/s]
 36%|███▌      | 302/850 [00:04<00:08, 66.76it/s]
 36%|███▋      | 309/850 [00:04<00:08, 67.41it/s]
 37%|███▋      | 316/850 [00:04<00:07, 67.96it/s]
 38%|███▊      | 323/850 [00:05<00:07, 68.48it/s]
 39%|███▉      | 330/850 [00:05<00:07, 67.10it/s]
 40%|███▉      | 337/850 [00:05<00:07, 66.09it/s]
 40%|████      | 344/850 [00:05<00:07, 66.02it/s]
 41%|████▏     | 351/850 [00:05<00:07, 66.82it/s]
 42%|████▏     | 358/850 [00:05<00:07, 64.35it/s]
 43%|████▎     | 365/850 [00:05<00:07, 63.66it/s]
 44%|████▍     | 372/850 [00:05<00:07, 65.24it/s]
 45%|████▍     | 379/850 [00:05<00:07, 59.31it/s]
 45%|████▌     | 386/850 [00:06<00:07, 59.44it/s]
 46%|████▌     | 393/850 [00:06<00:07, 58.18it/s]
 47%|████▋     | 399/850 [00:06<00:07, 56.50it/s]
 48%|████▊     | 407/850 [00:06<00:07, 62.14it/s]
 49%|████▉     | 415/850 [00:06<00:06, 66.41it/s]
 50%|████▉     | 423/850 [00:06<00:06, 68.67it/s]
 51%|█████     | 431/850 [00:06<00:05, 70.17it/s]
 52%|█████▏    | 439/850 [00:06<00:06, 66.16it/s]
 52%|█████▏    | 446/850 [00:07<00:06, 63.58it/s]
 53%|█████▎    | 453/850 [00:07<00:06, 61.74it/s]
 54%|█████▍    | 461/850 [00:07<00:05, 64.99it/s]
 55%|█████▌    | 469/850 [00:07<00:05, 67.95it/s]
 56%|█████▌    | 477/850 [00:07<00:05, 69.70it/s]
 57%|█████▋    | 485/850 [00:07<00:05, 66.98it/s]
 58%|█████▊    | 492/850 [00:07<00:05, 65.57it/s]
 59%|█████▉    | 500/850 [00:07<00:05, 67.64it/s]
 60%|█████▉    | 508/850 [00:07<00:04, 69.00it/s]
 61%|██████    | 515/850 [00:08<00:04, 67.70it/s]
 61%|██████▏   | 522/850 [00:08<00:04, 68.14it/s]
 62%|██████▏   | 529/850 [00:08<00:05, 63.43it/s]
 63%|██████▎   | 536/850 [00:08<00:05, 62.14it/s]
 64%|██████▍   | 543/850 [00:08<00:05, 60.12it/s]
 65%|██████▍   | 550/850 [00:08<00:05, 58.62it/s]
 66%|██████▌   | 557/850 [00:08<00:04, 59.23it/s]
 66%|██████▌   | 563/850 [00:08<00:05, 55.59it/s]
 67%|██████▋   | 570/850 [00:08<00:04, 57.72it/s]
 68%|██████▊   | 577/850 [00:09<00:04, 57.68it/s]
 69%|██████▊   | 584/850 [00:09<00:04, 59.76it/s]
 70%|██████▉   | 591/850 [00:09<00:04, 60.25it/s]
 70%|███████   | 598/850 [00:09<00:04, 60.81it/s]
 71%|███████   | 605/850 [00:09<00:04, 60.53it/s]
 72%|███████▏  | 612/850 [00:09<00:03, 62.53it/s]
 73%|███████▎  | 619/850 [00:09<00:03, 64.26it/s]
 74%|███████▎  | 626/850 [00:09<00:03, 65.14it/s]
 74%|███████▍  | 633/850 [00:09<00:03, 63.37it/s]
 75%|███████▌  | 640/850 [00:10<00:03, 64.52it/s]
 76%|███████▌  | 648/850 [00:10<00:02, 67.99it/s]
 77%|███████▋  | 655/850 [00:10<00:02, 68.40it/s]
 78%|███████▊  | 662/850 [00:10<00:02, 68.11it/s]
 79%|███████▊  | 669/850 [00:10<00:02, 68.30it/s]
 80%|███████▉  | 676/850 [00:10<00:02, 64.43it/s]
 80%|████████  | 683/850 [00:10<00:02, 64.38it/s]
 81%|████████  | 690/850 [00:10<00:02, 63.19it/s]
 82%|████████▏ | 697/850 [00:10<00:02, 63.47it/s]
 83%|████████▎ | 704/850 [00:11<00:02, 62.55it/s]
 84%|████████▎ | 711/850 [00:11<00:02, 59.59it/s]
 84%|████████▍ | 717/850 [00:11<00:02, 58.91it/s]
 85%|████████▌ | 724/850 [00:11<00:02, 59.52it/s]
 86%|████████▌ | 730/850 [00:11<00:02, 57.96it/s]
 87%|████████▋ | 737/850 [00:11<00:01, 60.64it/s]
 88%|████████▊ | 744/850 [00:11<00:01, 57.91it/s]
 88%|████████▊ | 750/850 [00:11<00:01, 55.74it/s]
 89%|████████▉ | 757/850 [00:12<00:01, 59.42it/s]
 90%|████████▉ | 764/850 [00:12<00:01, 61.63it/s]
 91%|█████████ | 771/850 [00:12<00:01, 59.98it/s]
 92%|█████████▏| 778/850 [00:12<00:01, 59.34it/s]
 92%|█████████▏| 784/850 [00:12<00:01, 56.57it/s]
 93%|█████████▎| 790/850 [00:12<00:01, 54.23it/s]
 94%|█████████▍| 797/850 [00:12<00:00, 58.26it/s]
 94%|█████████▍| 803/850 [00:12<00:00, 58.70it/s]
 95%|█████████▌| 809/850 [00:12<00:00, 58.22it/s]
 96%|█████████▌| 815/850 [00:13<00:00, 57.34it/s]
 97%|█████████▋| 823/850 [00:13<00:00, 62.38it/s]
 98%|█████████▊| 830/850 [00:13<00:00, 60.31it/s]
 98%|█████████▊| 837/850 [00:13<00:00, 61.46it/s]
 99%|█████████▉| 844/850 [00:13<00:00, 62.52it/s]
+***** eval metrics *****
+  epoch                   =     9.9897
+  eval_accuracy           =     0.9986
+  eval_f1                 =     0.9231
+  eval_loss               =     0.0094
+  eval_precision          =     0.9169
+  eval_recall             =     0.9293
+  eval_runtime            = 0:00:17.51
+  eval_samples            =       6798
+  eval_samples_per_second =    388.236
+  eval_steps_per_second   =     48.544
+09/09/2024 16:18:29 - INFO - __main__ -   *** Predict ***
+[INFO|trainer.py:811] 2024-09-09 16:18:29,694 >> The following columns in the test set don't have a corresponding argument in `BertForTokenClassification.forward` and have been ignored: id, tokens, ner_tags. If id, tokens, ner_tags are not expected by `BertForTokenClassification.forward`,  you can safely ignore this message.
+[INFO|trainer.py:3819] 2024-09-09 16:18:29,696 >> 
+***** Running Prediction *****
+[INFO|trainer.py:3821] 2024-09-09 16:18:29,696 >>   Num examples = 14605
+[INFO|trainer.py:3824] 2024-09-09 16:18:29,696 >>   Batch size = 8
+
  0%|          | 0/1826 [00:00<?, ?it/s]
  1%|          | 10/1826 [00:00<00:19, 94.11it/s]
  1%|          | 20/1826 [00:00<00:26, 68.33it/s]
  2%|▏         | 28/1826 [00:00<00:26, 67.41it/s]
  2%|▏         | 35/1826 [00:00<00:26, 67.50it/s]
  2%|▏         | 43/1826 [00:00<00:25, 70.74it/s]
  3%|▎         | 51/1826 [00:00<00:24, 72.57it/s]
  3%|▎         | 59/1826 [00:00<00:26, 67.19it/s]
  4%|▎         | 66/1826 [00:00<00:28, 61.91it/s]
  4%|▍         | 74/1826 [00:01<00:27, 64.76it/s]
  4%|▍         | 82/1826 [00:01<00:25, 67.85it/s]
  5%|▍         | 89/1826 [00:01<00:25, 66.96it/s]
  5%|▌         | 98/1826 [00:01<00:23, 72.93it/s]
  6%|▌         | 107/1826 [00:01<00:22, 76.04it/s]
  6%|▋         | 115/1826 [00:01<00:22, 74.77it/s]
  7%|▋         | 123/1826 [00:01<00:25, 67.64it/s]
  7%|▋         | 130/1826 [00:01<00:25, 65.55it/s]
  8%|▊         | 138/1826 [00:02<00:25, 66.69it/s]
  8%|▊         | 146/1826 [00:02<00:24, 69.01it/s]
  8%|▊         | 153/1826 [00:02<00:24, 69.21it/s]
  9%|▉         | 160/1826 [00:02<00:27, 60.78it/s]
  9%|▉         | 167/1826 [00:02<00:26, 61.52it/s]
 10%|▉         | 176/1826 [00:02<00:24, 68.18it/s]
 10%|█         | 184/1826 [00:02<00:23, 69.66it/s]
 11%|█         | 192/1826 [00:02<00:23, 69.46it/s]
 11%|█         | 200/1826 [00:02<00:24, 67.41it/s]
 11%|█▏        | 207/1826 [00:03<00:24, 66.24it/s]
 12%|█▏        | 214/1826 [00:03<00:24, 66.73it/s]
 12%|█▏        | 221/1826 [00:03<00:23, 67.00it/s]
 12%|█▏        | 228/1826 [00:03<00:25, 62.86it/s]
 13%|█▎        | 235/1826 [00:03<00:26, 59.84it/s]
 13%|█▎        | 242/1826 [00:03<00:25, 61.91it/s]
 14%|█▎        | 249/1826 [00:03<00:25, 61.34it/s]
 14%|█▍        | 256/1826 [00:03<00:25, 62.00it/s]
 14%|█▍        | 263/1826 [00:03<00:24, 62.93it/s]
 15%|█▍        | 271/1826 [00:04<00:23, 67.26it/s]
 15%|█▌        | 280/1826 [00:04<00:21, 72.72it/s]
 16%|█▌        | 288/1826 [00:04<00:21, 70.72it/s]
 16%|█▌        | 296/1826 [00:04<00:21, 69.79it/s]
 17%|█▋        | 304/1826 [00:04<00:22, 67.51it/s]
 17%|█▋        | 312/1826 [00:04<00:21, 69.04it/s]
 17%|█▋        | 319/1826 [00:04<00:21, 68.83it/s]
 18%|█▊        | 327/1826 [00:04<00:21, 70.40it/s]
 18%|█▊        | 335/1826 [00:04<00:21, 69.17it/s]
 19%|█▉        | 343/1826 [00:05<00:21, 70.48it/s]
 19%|█▉        | 351/1826 [00:05<00:22, 65.72it/s]
 20%|█▉        | 358/1826 [00:05<00:22, 64.30it/s]
 20%|██        | 366/1826 [00:05<00:21, 68.34it/s]
 20%|██        | 374/1826 [00:05<00:20, 69.37it/s]
 21%|██        | 382/1826 [00:05<00:20, 71.56it/s]
 21%|██▏       | 390/1826 [00:05<00:22, 64.90it/s]
 22%|██▏       | 398/1826 [00:05<00:21, 66.61it/s]
 22%|██▏       | 405/1826 [00:06<00:21, 66.68it/s]
 23%|██▎       | 412/1826 [00:06<00:22, 63.63it/s]
 23%|██▎       | 420/1826 [00:06<00:21, 65.85it/s]
 23%|██▎       | 427/1826 [00:06<00:22, 62.95it/s]
 24%|██▍       | 434/1826 [00:06<00:21, 63.77it/s]
 24%|██▍       | 441/1826 [00:06<00:22, 61.74it/s]
 25%|██▍       | 450/1826 [00:06<00:20, 68.00it/s]
 25%|██▌       | 457/1826 [00:06<00:20, 67.45it/s]
 26%|██▌       | 466/1826 [00:06<00:19, 69.66it/s]
 26%|██▌       | 473/1826 [00:07<00:21, 64.33it/s]
 26%|██▋       | 480/1826 [00:07<00:21, 63.73it/s]
 27%|██▋       | 487/1826 [00:07<00:21, 62.35it/s]
 27%|██▋       | 494/1826 [00:07<00:21, 62.04it/s]
 27%|██▋       | 502/1826 [00:07<00:20, 64.86it/s]
 28%|██▊       | 510/1826 [00:07<00:19, 67.59it/s]
 28%|██▊       | 518/1826 [00:07<00:18, 70.73it/s]
 29%|██▉       | 527/1826 [00:07<00:17, 73.95it/s]
 29%|██▉       | 535/1826 [00:07<00:17, 71.81it/s]
 30%|██▉       | 543/1826 [00:08<00:17, 71.49it/s]
 30%|███       | 551/1826 [00:08<00:19, 65.90it/s]
 31%|███       | 558/1826 [00:08<00:19, 64.66it/s]
 31%|███       | 567/1826 [00:08<00:17, 70.27it/s]
 32%|███▏      | 576/1826 [00:08<00:17, 73.36it/s]
 32%|███▏      | 584/1826 [00:08<00:18, 68.09it/s]
 32%|███▏      | 592/1826 [00:08<00:17, 70.54it/s]
 33%|███▎      | 600/1826 [00:08<00:17, 69.88it/s]
 33%|███▎      | 608/1826 [00:09<00:18, 65.58it/s]
 34%|███▎      | 615/1826 [00:09<00:18, 65.45it/s]
 34%|███▍      | 622/1826 [00:09<00:18, 66.26it/s]
 35%|███▍      | 632/1826 [00:09<00:16, 73.73it/s]
 35%|███▌      | 641/1826 [00:09<00:15, 76.81it/s]
 36%|███▌      | 649/1826 [00:09<00:16, 71.30it/s]
 36%|███▌      | 657/1826 [00:09<00:16, 69.22it/s]
 37%|███▋      | 667/1826 [00:09<00:15, 75.05it/s]
 37%|███▋      | 675/1826 [00:09<00:16, 71.22it/s]
 37%|███▋      | 683/1826 [00:10<00:16, 69.84it/s]
 38%|███▊      | 692/1826 [00:10<00:15, 73.46it/s]
 38%|███▊      | 700/1826 [00:10<00:15, 73.83it/s]
 39%|███▉      | 708/1826 [00:10<00:14, 75.07it/s]
 39%|███▉      | 718/1826 [00:10<00:13, 79.86it/s]
 40%|███▉      | 727/1826 [00:10<00:13, 78.55it/s]
 40%|████      | 735/1826 [00:10<00:14, 76.87it/s]
 41%|████      | 743/1826 [00:10<00:14, 75.47it/s]
 41%|████      | 752/1826 [00:10<00:13, 79.11it/s]
 42%|████▏     | 760/1826 [00:11<00:13, 77.59it/s]
 42%|████▏     | 769/1826 [00:11<00:13, 78.55it/s]
 43%|████▎     | 777/1826 [00:11<00:13, 77.65it/s]
 43%|████▎     | 785/1826 [00:11<00:14, 71.54it/s]
 43%|████▎     | 793/1826 [00:11<00:14, 73.07it/s]
 44%|████▍     | 801/1826 [00:11<00:15, 68.26it/s]
 44%|████▍     | 809/1826 [00:11<00:14, 69.66it/s]
 45%|████▍     | 817/1826 [00:11<00:14, 71.79it/s]
 45%|████▌     | 826/1826 [00:11<00:13, 76.14it/s]
 46%|████▌     | 834/1826 [00:12<00:14, 68.30it/s]
 46%|████▌     | 842/1826 [00:12<00:14, 69.69it/s]
 47%|████▋     | 851/1826 [00:12<00:13, 72.87it/s]
 47%|████▋     | 860/1826 [00:12<00:12, 77.24it/s]
 48%|████▊     | 868/1826 [00:12<00:13, 72.75it/s]
 48%|████▊     | 876/1826 [00:12<00:13, 72.25it/s]
 48%|████▊     | 884/1826 [00:12<00:13, 71.43it/s]
 49%|████▉     | 892/1826 [00:12<00:13, 68.24it/s]
 49%|████▉     | 902/1826 [00:13<00:12, 74.85it/s]
 50%|████▉     | 911/1826 [00:13<00:11, 77.59it/s]
 50%|█████     | 920/1826 [00:13<00:11, 80.11it/s]
 51%|█████     | 929/1826 [00:13<00:11, 79.42it/s]
 51%|█████▏    | 937/1826 [00:13<00:12, 73.54it/s]
 52%|█████▏    | 945/1826 [00:13<00:12, 68.71it/s]
 52%|█████▏    | 952/1826 [00:13<00:12, 68.31it/s]
 53%|█████▎    | 960/1826 [00:13<00:12, 69.39it/s]
 53%|█████▎    | 968/1826 [00:13<00:11, 71.85it/s]
 53%|█████▎    | 976/1826 [00:14<00:12, 67.88it/s]
 54%|█████▍    | 985/1826 [00:14<00:11, 71.98it/s]
 54%|█████▍    | 993/1826 [00:14<00:11, 69.95it/s]
 55%|█████▍    | 1001/1826 [00:14<00:12, 67.60it/s]
 55%|█████▌    | 1008/1826 [00:14<00:12, 68.06it/s]
 56%|█████▌    | 1016/1826 [00:14<00:11, 70.58it/s]
 56%|█████▌    | 1025/1826 [00:14<00:10, 74.99it/s]
 57%|█████▋    | 1033/1826 [00:14<00:10, 73.75it/s]
 57%|█████▋    | 1041/1826 [00:14<00:10, 71.80it/s]
 57%|█████▋    | 1049/1826 [00:15<00:11, 69.86it/s]
 58%|█████▊    | 1057/1826 [00:15<00:11, 68.46it/s]
 58%|█████▊    | 1066/1826 [00:15<00:10, 73.61it/s]
 59%|█████▉    | 1074/1826 [00:15<00:10, 74.53it/s]
 59%|█████▉    | 1083/1826 [00:15<00:09, 78.56it/s]
 60%|█████▉    | 1092/1826 [00:15<00:09, 79.51it/s]
 60%|██████    | 1101/1826 [00:15<00:09, 79.20it/s]
 61%|██████    | 1109/1826 [00:15<00:09, 74.39it/s]
 61%|██████    | 1117/1826 [00:15<00:09, 71.61it/s]
 62%|██████▏   | 1126/1826 [00:16<00:09, 73.71it/s]
 62%|██████▏   | 1134/1826 [00:16<00:09, 75.14it/s]
 63%|██████▎   | 1142/1826 [00:16<00:09, 72.85it/s]
 63%|██████▎   | 1150/1826 [00:16<00:09, 71.67it/s]
 63%|██████▎   | 1158/1826 [00:16<00:09, 71.29it/s]
 64%|██████▍   | 1166/1826 [00:16<00:09, 67.84it/s]
 64%|██████▍   | 1174/1826 [00:16<00:09, 69.65it/s]
 65%|██████▍   | 1182/1826 [00:16<00:09, 67.30it/s]
 65%|██████▌   | 1190/1826 [00:17<00:09, 68.54it/s]
 66%|██████▌   | 1198/1826 [00:17<00:08, 70.13it/s]
 66%|██████▌   | 1206/1826 [00:17<00:08, 70.01it/s]
 66%|██████▋   | 1214/1826 [00:17<00:09, 64.04it/s]
 67%|██████▋   | 1221/1826 [00:17<00:09, 62.76it/s]
 67%|██████▋   | 1230/1826 [00:17<00:08, 68.49it/s]
 68%|██████▊   | 1238/1826 [00:17<00:08, 69.88it/s]
 68%|██████▊   | 1246/1826 [00:17<00:08, 70.54it/s]
 69%|██████▊   | 1254/1826 [00:17<00:08, 64.45it/s]
 69%|██████▉   | 1262/1826 [00:18<00:08, 66.57it/s]
 70%|██████▉   | 1270/1826 [00:18<00:08, 69.08it/s]
 70%|██████▉   | 1278/1826 [00:18<00:07, 71.87it/s]
 70%|███████   | 1286/1826 [00:18<00:07, 72.39it/s]
 71%|███████   | 1295/1826 [00:18<00:06, 76.32it/s]
 71%|███████▏  | 1304/1826 [00:18<00:06, 79.67it/s]
 72%|███████▏  | 1313/1826 [00:18<00:06, 79.61it/s]
 72%|███████▏  | 1322/1826 [00:18<00:06, 78.03it/s]
 73%|███████▎  | 1330/1826 [00:18<00:06, 75.63it/s]
 73%|███████▎  | 1338/1826 [00:19<00:06, 73.90it/s]
 74%|███████▎  | 1346/1826 [00:19<00:06, 69.31it/s]
 74%|███████▍  | 1354/1826 [00:19<00:06, 70.19it/s]
 75%|███████▍  | 1362/1826 [00:19<00:06, 71.84it/s]
 75%|███████▌  | 1370/1826 [00:19<00:06, 67.49it/s]
 76%|███████▌  | 1379/1826 [00:19<00:06, 72.42it/s]
 76%|███████▌  | 1387/1826 [00:19<00:05, 74.47it/s]
 76%|███████▋  | 1395/1826 [00:19<00:05, 75.32it/s]
 77%|███████▋  | 1404/1826 [00:19<00:05, 79.20it/s]
 77%|███████▋  | 1413/1826 [00:20<00:05, 80.03it/s]
 78%|███████▊  | 1422/1826 [00:20<00:05, 76.15it/s]
 78%|███████▊  | 1430/1826 [00:20<00:05, 72.26it/s]
 79%|███████▉  | 1438/1826 [00:20<00:05, 68.01it/s]
 79%|███████▉  | 1446/1826 [00:20<00:05, 69.61it/s]
 80%|███████▉  | 1454/1826 [00:20<00:05, 71.49it/s]
 80%|████████  | 1462/1826 [00:20<00:05, 72.56it/s]
 81%|████████  | 1470/1826 [00:20<00:05, 66.00it/s]
 81%|████████  | 1477/1826 [00:21<00:05, 65.05it/s]
 81%|████████▏ | 1484/1826 [00:21<00:05, 65.40it/s]
 82%|████████▏ | 1491/1826 [00:21<00:05, 60.28it/s]
 82%|████████▏ | 1498/1826 [00:21<00:05, 61.97it/s]
 82%|████████▏ | 1505/1826 [00:21<00:05, 60.61it/s]
 83%|████████▎ | 1513/1826 [00:21<00:04, 64.51it/s]
 83%|████████▎ | 1520/1826 [00:21<00:04, 65.43it/s]
 84%|████████▎ | 1527/1826 [00:21<00:04, 62.45it/s]
 84%|████████▍ | 1534/1826 [00:21<00:04, 64.07it/s]
 84%|████████▍ | 1541/1826 [00:22<00:04, 63.72it/s]
 85%|████████▍ | 1548/1826 [00:22<00:04, 64.27it/s]
 85%|████████▌ | 1555/1826 [00:22<00:04, 65.72it/s]
 86%|████████▌ | 1562/1826 [00:22<00:03, 66.77it/s]
 86%|████████▌ | 1569/1826 [00:22<00:04, 63.11it/s]
 86%|████████▋ | 1576/1826 [00:22<00:03, 62.95it/s]
 87%|████████▋ | 1584/1826 [00:22<00:03, 65.44it/s]
 87%|████████▋ | 1591/1826 [00:22<00:03, 63.71it/s]
 88%|████████▊ | 1598/1826 [00:22<00:03, 63.28it/s]
 88%|████████▊ | 1607/1826 [00:23<00:03, 69.37it/s]
 88%|████████▊ | 1614/1826 [00:23<00:03, 68.35it/s]
 89%|████████▉ | 1621/1826 [00:23<00:03, 61.61it/s]
 89%|████████▉ | 1628/1826 [00:23<00:03, 56.69it/s]
 90%|████████▉ | 1636/1826 [00:23<00:03, 60.23it/s]
 90%|████████▉ | 1643/1826 [00:23<00:03, 60.95it/s]
 90%|█████████ | 1651/1826 [00:23<00:02, 65.46it/s]
 91%|█████████ | 1658/1826 [00:23<00:02, 64.85it/s]
 91%|█████████ | 1665/1826 [00:24<00:02, 65.01it/s]
 92%|█████████▏| 1673/1826 [00:24<00:02, 68.49it/s]
 92%|█████████▏| 1681/1826 [00:24<00:02, 70.31it/s]
 93%|█████████▎| 1690/1826 [00:24<00:01, 73.05it/s]
 93%|█████████▎| 1698/1826 [00:24<00:02, 63.59it/s]
 93%|█████████▎| 1706/1826 [00:24<00:01, 65.52it/s]
 94%|█████████▍| 1714/1826 [00:24<00:01, 68.18it/s]
 94%|█████████▍| 1721/1826 [00:24<00:01, 61.69it/s]
 95%|█████████▍| 1729/1826 [00:24<00:01, 65.81it/s]
 95%|█████████▌| 1737/1826 [00:25<00:01, 67.43it/s]
 96%|█████████▌| 1746/1826 [00:25<00:01, 71.34it/s]
 96%|█████████▌| 1754/1826 [00:25<00:01, 69.71it/s]
 96%|█████████▋| 1762/1826 [00:25<00:00, 68.18it/s]
 97%|█████████▋| 1769/1826 [00:25<00:00, 66.79it/s]
 97%|█████████▋| 1776/1826 [00:25<00:00, 65.37it/s]
 98%|█████████▊| 1784/1826 [00:25<00:00, 69.28it/s]
 98%|█████████▊| 1791/1826 [00:25<00:00, 64.98it/s]
 98%|█████████▊| 1798/1826 [00:26<00:00, 60.14it/s]
 99%|█████████▉| 1805/1826 [00:26<00:00, 62.25it/s]
 99%|█████████▉| 1813/1826 [00:26<00:00, 66.07it/s]
+[INFO|trainer.py:3503] 2024-09-09 16:19:03,621 >> Saving model checkpoint to /content/dissertation/scripts/ner/output
+[INFO|configuration_utils.py:472] 2024-09-09 16:19:03,622 >> Configuration saved in /content/dissertation/scripts/ner/output/config.json
+[INFO|modeling_utils.py:2799] 2024-09-09 16:19:04,830 >> Model weights saved in /content/dissertation/scripts/ner/output/model.safetensors
+[INFO|tokenization_utils_base.py:2684] 2024-09-09 16:19:04,831 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2693] 2024-09-09 16:19:04,832 >> Special tokens file saved in /content/dissertation/scripts/ner/output/special_tokens_map.json
+***** predict metrics *****
+  predict_accuracy           =     0.9982
+  predict_f1                 =     0.8965
+  predict_loss               =     0.0127
+  predict_precision          =     0.8773
+  predict_recall             =     0.9166
+  predict_runtime            = 0:00:33.03
+  predict_samples_per_second =    442.053
+  predict_steps_per_second   =     55.268
+

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 9.988518943742825,
-    "total_flos": 1.1084127968547612e+16,
-    "train_loss": 0.0030981004957495064,
-    "train_runtime": 1801.8371,
-    "train_samples": 27841,
-    "train_samples_per_second": 154.515,
-    "train_steps_per_second": 2.414
+    "epoch": 9.989701338825952,
+    "total_flos": 1.7736053837017554e+16,
+    "train_loss": 0.002068036902580679,
+    "train_runtime": 2701.3395,
+    "train_samples": 31053,
+    "train_samples_per_second": 114.954,
+    "train_steps_per_second": 1.795
 }

trainer_state.json

@@ -1,201 +1,208 @@
 {
-  "best_metric": 0.9284386617100372,
-  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-4350",
-  "epoch": 9.988518943742825,
+  "best_metric": 0.9230769230769231,
+  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-4850",
+  "epoch": 9.989701338825952,
   "eval_steps": 500,
-  "global_step": 4350,
+  "global_step": 4850,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.9988518943742825,
-      "eval_accuracy": 0.9981371919738788,
-      "eval_f1": 0.895427795382526,
-      "eval_loss": 0.006041484884917736,
-      "eval_precision": 0.8705985915492958,
-      "eval_recall": 0.9217148182665424,
-      "eval_runtime": 15.0476,
-      "eval_samples_per_second": 461.603,
-      "eval_steps_per_second": 57.75,
-      "step": 435
-    },
-    {
-      "epoch": 1.148105625717566,
-      "grad_norm": 0.025844832882285118,
-      "learning_rate": 4.4252873563218394e-05,
-      "loss": 0.0156,
+      "epoch": 0.9989701338825953,
+      "eval_accuracy": 0.9984030122046221,
+      "eval_f1": 0.9132331555986428,
+      "eval_loss": 0.004271956626325846,
+      "eval_precision": 0.9145631067961165,
+      "eval_recall": 0.9119070667957405,
+      "eval_runtime": 17.5218,
+      "eval_samples_per_second": 387.973,
+      "eval_steps_per_second": 48.511,
+      "step": 485
+    },
+    {
+      "epoch": 1.0298661174047374,
+      "grad_norm": 0.35554030537605286,
+      "learning_rate": 4.484536082474227e-05,
+      "loss": 0.0127,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9984232411464292,
-      "eval_f1": 0.9032855159648312,
-      "eval_loss": 0.004733391106128693,
-      "eval_precision": 0.8970588235294118,
-      "eval_recall": 0.9095992544268406,
-      "eval_runtime": 15.0374,
-      "eval_samples_per_second": 461.915,
-      "eval_steps_per_second": 57.789,
-      "step": 871
-    },
-    {
-      "epoch": 2.296211251435132,
-      "grad_norm": 0.013152423314750195,
-      "learning_rate": 3.850574712643678e-05,
-      "loss": 0.0038,
+      "eval_accuracy": 0.9984679304076863,
+      "eval_f1": 0.9130645944633317,
+      "eval_loss": 0.005422212183475494,
+      "eval_precision": 0.9161793372319688,
+      "eval_recall": 0.9099709583736689,
+      "eval_runtime": 17.524,
+      "eval_samples_per_second": 387.925,
+      "eval_steps_per_second": 48.505,
+      "step": 971
+    },
+    {
+      "epoch": 2.059732234809475,
+      "grad_norm": 0.2354772835969925,
+      "learning_rate": 3.9690721649484535e-05,
+      "loss": 0.0033,
       "step": 1000
     },
     {
-      "epoch": 2.9988518943742823,
-      "eval_accuracy": 0.9985348700918149,
-      "eval_f1": 0.8998178506375226,
-      "eval_loss": 0.004836090374737978,
-      "eval_precision": 0.8797862867319679,
-      "eval_recall": 0.9207828518173345,
-      "eval_runtime": 15.0312,
-      "eval_samples_per_second": 462.105,
-      "eval_steps_per_second": 57.813,
-      "step": 1306
+      "epoch": 2.998970133882595,
+      "eval_accuracy": 0.9985912749935082,
+      "eval_f1": 0.9173955296404277,
+      "eval_loss": 0.0059292372316122055,
+      "eval_precision": 0.9209756097560976,
+      "eval_recall": 0.9138431752178122,
+      "eval_runtime": 17.4807,
+      "eval_samples_per_second": 388.886,
+      "eval_steps_per_second": 48.625,
+      "step": 1456
     },
     {
-      "epoch": 3.444316877152698,
-      "grad_norm": 0.13433514535427094,
-      "learning_rate": 3.275862068965517e-05,
-      "loss": 0.0026,
+      "epoch": 3.089598352214212,
+      "grad_norm": 0.026833873242139816,
+      "learning_rate": 3.4536082474226805e-05,
+      "loss": 0.0016,
       "step": 1500
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9986464990372004,
-      "eval_f1": 0.9222988505747127,
-      "eval_loss": 0.004586596507579088,
-      "eval_precision": 0.9101633393829401,
-      "eval_recall": 0.934762348555452,
-      "eval_runtime": 15.3559,
-      "eval_samples_per_second": 452.335,
-      "eval_steps_per_second": 56.591,
-      "step": 1742
-    },
-    {
-      "epoch": 4.592422502870264,
-      "grad_norm": 0.04434504359960556,
-      "learning_rate": 2.7011494252873566e-05,
-      "loss": 0.0016,
+      "eval_accuracy": 0.9985523240716697,
+      "eval_f1": 0.9135212613473483,
+      "eval_loss": 0.006239830516278744,
+      "eval_precision": 0.9018867924528302,
+      "eval_recall": 0.925459825750242,
+      "eval_runtime": 17.8043,
+      "eval_samples_per_second": 381.818,
+      "eval_steps_per_second": 47.741,
+      "step": 1942
+    },
+    {
+      "epoch": 4.11946446961895,
+      "grad_norm": 0.1513855904340744,
+      "learning_rate": 2.9381443298969075e-05,
+      "loss": 0.001,
       "step": 2000
     },
     {
-      "epoch": 4.998851894374282,
-      "eval_accuracy": 0.9986395222281137,
-      "eval_f1": 0.9255269320843091,
-      "eval_loss": 0.005914350505918264,
-      "eval_precision": 0.9303201506591338,
-      "eval_recall": 0.9207828518173345,
-      "eval_runtime": 15.1878,
-      "eval_samples_per_second": 457.342,
-      "eval_steps_per_second": 57.217,
-      "step": 2177
+      "epoch": 4.998970133882596,
+      "eval_accuracy": 0.9986042586341211,
+      "eval_f1": 0.920328343795268,
+      "eval_loss": 0.006359434220939875,
+      "eval_precision": 0.918111753371869,
+      "eval_recall": 0.9225556631171346,
+      "eval_runtime": 17.4628,
+      "eval_samples_per_second": 389.284,
+      "eval_steps_per_second": 48.675,
+      "step": 2427
     },
     {
-      "epoch": 5.7405281285878305,
-      "grad_norm": 0.004046172834932804,
-      "learning_rate": 2.1264367816091954e-05,
-      "loss": 0.0012,
+      "epoch": 5.1493305870236865,
+      "grad_norm": 0.007741741370409727,
+      "learning_rate": 2.422680412371134e-05,
+      "loss": 0.0006,
       "step": 2500
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.9986883598917199,
-      "eval_f1": 0.9282385834109972,
-      "eval_loss": 0.005369984544813633,
-      "eval_precision": 0.9282385834109972,
-      "eval_recall": 0.9282385834109972,
-      "eval_runtime": 15.0665,
-      "eval_samples_per_second": 461.023,
-      "eval_steps_per_second": 57.678,
-      "step": 2613
-    },
-    {
-      "epoch": 6.888633754305396,
-      "grad_norm": 0.20072227716445923,
-      "learning_rate": 1.5517241379310346e-05,
-      "loss": 0.001,
+      "eval_accuracy": 0.9985523240716697,
+      "eval_f1": 0.9180009704027171,
+      "eval_loss": 0.00812861043959856,
+      "eval_precision": 0.9202334630350194,
+      "eval_recall": 0.9157792836398838,
+      "eval_runtime": 17.5273,
+      "eval_samples_per_second": 387.852,
+      "eval_steps_per_second": 48.496,
+      "step": 2913
+    },
+    {
+      "epoch": 6.179196704428424,
+      "grad_norm": 0.06708419322967529,
+      "learning_rate": 1.9072164948453608e-05,
+      "loss": 0.0004,
       "step": 3000
     },
     {
-      "epoch": 6.998851894374282,
-      "eval_accuracy": 0.9983325426283035,
-      "eval_f1": 0.9087640449438201,
-      "eval_loss": 0.006764214485883713,
-      "eval_precision": 0.8776041666666666,
-      "eval_recall": 0.9422180801491147,
-      "eval_runtime": 15.0682,
-      "eval_samples_per_second": 460.97,
-      "eval_steps_per_second": 57.671,
-      "step": 3048
+      "epoch": 6.998970133882596,
+      "eval_accuracy": 0.9985653077122826,
+      "eval_f1": 0.920328343795268,
+      "eval_loss": 0.008457792922854424,
+      "eval_precision": 0.918111753371869,
+      "eval_recall": 0.9225556631171346,
+      "eval_runtime": 17.4726,
+      "eval_samples_per_second": 389.067,
+      "eval_steps_per_second": 48.648,
+      "step": 3398
     },
     {
-      "epoch": 8.0,
-      "eval_accuracy": 0.9986744062735468,
-      "eval_f1": 0.9283720930232559,
-      "eval_loss": 0.0063699171878397465,
-      "eval_precision": 0.9266480965645311,
-      "eval_recall": 0.9301025163094129,
-      "eval_runtime": 15.0386,
-      "eval_samples_per_second": 461.877,
-      "eval_steps_per_second": 57.785,
-      "step": 3484
-    },
-    {
-      "epoch": 8.036739380022961,
-      "grad_norm": 0.0006512438994832337,
-      "learning_rate": 9.770114942528738e-06,
-      "loss": 0.0006,
+      "epoch": 7.209062821833162,
+      "grad_norm": 0.0032746351789683104,
+      "learning_rate": 1.3917525773195878e-05,
+      "loss": 0.0003,
       "step": 3500
     },
     {
-      "epoch": 8.998851894374283,
-      "eval_accuracy": 0.9987441743644127,
-      "eval_f1": 0.928273947246645,
-      "eval_loss": 0.006663947831839323,
-      "eval_precision": 0.921875,
-      "eval_recall": 0.934762348555452,
-      "eval_runtime": 15.3138,
-      "eval_samples_per_second": 453.576,
-      "eval_steps_per_second": 56.746,
-      "step": 3919
-    },
-    {
-      "epoch": 9.184845005740527,
-      "grad_norm": 0.007138502784073353,
-      "learning_rate": 4.022988505747127e-06,
-      "loss": 0.0004,
+      "epoch": 8.0,
+      "eval_accuracy": 0.9985847831732018,
+      "eval_f1": 0.9210653753026635,
+      "eval_loss": 0.00870645884424448,
+      "eval_precision": 0.9215116279069767,
+      "eval_recall": 0.920619554695063,
+      "eval_runtime": 17.5192,
+      "eval_samples_per_second": 388.031,
+      "eval_steps_per_second": 48.518,
+      "step": 3884
+    },
+    {
+      "epoch": 8.2389289392379,
+      "grad_norm": 0.0013693332439288497,
+      "learning_rate": 8.762886597938144e-06,
+      "loss": 0.0002,
       "step": 4000
     },
     {
-      "epoch": 9.988518943742825,
-      "eval_accuracy": 0.9986674294644602,
-      "eval_f1": 0.9284386617100372,
-      "eval_loss": 0.0068497806787490845,
-      "eval_precision": 0.9258572752548656,
-      "eval_recall": 0.9310344827586207,
-      "eval_runtime": 15.1753,
-      "eval_samples_per_second": 457.717,
-      "eval_steps_per_second": 57.264,
-      "step": 4350
-    },
-    {
-      "epoch": 9.988518943742825,
-      "step": 4350,
-      "total_flos": 1.1084127968547612e+16,
-      "train_loss": 0.0030981004957495064,
-      "train_runtime": 1801.8371,
-      "train_samples_per_second": 154.515,
-      "train_steps_per_second": 2.414
+      "epoch": 8.998970133882596,
+      "eval_accuracy": 0.9985198649701377,
+      "eval_f1": 0.9206500956022945,
+      "eval_loss": 0.008967469446361065,
+      "eval_precision": 0.9093484419263456,
+      "eval_recall": 0.9322362052274927,
+      "eval_runtime": 17.7573,
+      "eval_samples_per_second": 382.828,
+      "eval_steps_per_second": 47.868,
+      "step": 4369
+    },
+    {
+      "epoch": 9.268795056642636,
+      "grad_norm": 0.0005869244341738522,
+      "learning_rate": 3.608247422680412e-06,
+      "loss": 0.0001,
+      "step": 4500
+    },
+    {
+      "epoch": 9.989701338825952,
+      "eval_accuracy": 0.9986302259153467,
+      "eval_f1": 0.9230769230769231,
+      "eval_loss": 0.009366312995553017,
+      "eval_precision": 0.9169054441260746,
+      "eval_recall": 0.9293320425943853,
+      "eval_runtime": 17.4762,
+      "eval_samples_per_second": 388.987,
+      "eval_steps_per_second": 48.638,
+      "step": 4850
+    },
+    {
+      "epoch": 9.989701338825952,
+      "step": 4850,
+      "total_flos": 1.7736053837017554e+16,
+      "train_loss": 0.002068036902580679,
+      "train_runtime": 2701.3395,
+      "train_samples_per_second": 114.954,
+      "train_steps_per_second": 1.795
     }
   ],
   "logging_steps": 500,
-  "max_steps": 4350,
+  "max_steps": 4850,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
@@ -211,7 +218,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.1084127968547612e+16,
+  "total_flos": 1.7736053837017554e+16,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null