End of training

README.md

@@ -3,9 +3,10 @@ library_name: transformers
 license: apache-2.0
 base_model: PlanTL-GOB-ES/bsc-bio-ehr-es
 tags:
+- token-classification
 - generated_from_trainer
 datasets:
-- drugtemist-fasttext-75-ner
+- Rodrigo1771/drugtemist-fasttext-75-ner
 metrics:
 - precision
 - recall
@@ -18,24 +19,24 @@ model-index:
       name: Token Classification
       type: token-classification
     dataset:
-      name: drugtemist-fasttext-75-ner
-      type: drugtemist-fasttext-75-ner
+      name: Rodrigo1771/drugtemist-fasttext-75-ner
+      type: Rodrigo1771/drugtemist-fasttext-75-ner
       config: DrugTEMIST NER
       split: validation
       args: DrugTEMIST NER
     metrics:
     - name: Precision
       type: precision
-      value: 0.9416590701914311
+      value: 0.9447963800904977
     - name: Recall
       type: recall
-      value: 0.9494485294117647
+      value: 0.9595588235294118
     - name: F1
       type: f1
-      value: 0.945537757437071
+      value: 0.9521203830369357
     - name: Accuracy
       type: accuracy
-      value: 0.9990525491919205
+      value: 0.9991418018042759
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -43,12 +44,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 # output
 
-This model is a fine-tuned version of [PlanTL-GOB-ES/bsc-bio-ehr-es](https://huggingface.co/PlanTL-GOB-ES/bsc-bio-ehr-es) on the drugtemist-fasttext-75-ner dataset.
+This model is a fine-tuned version of [PlanTL-GOB-ES/bsc-bio-ehr-es](https://huggingface.co/PlanTL-GOB-ES/bsc-bio-ehr-es) on the Rodrigo1771/drugtemist-fasttext-75-ner dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0052
-- Precision: 0.9417
-- Recall: 0.9494
-- F1: 0.9455
+- Loss: 0.0044
+- Precision: 0.9448
+- Recall: 0.9596
+- F1: 0.9521
 - Accuracy: 0.9991
 
 ## Model description

all_results.json

@@ -1,26 +1,26 @@
 {
-    "epoch": 9.99288256227758,
-    "eval_accuracy": 0.9758743323218038,
-    "eval_f1": 0.8053395240858967,
-    "eval_loss": 0.17113561928272247,
-    "eval_precision": 0.7991246256622898,
-    "eval_recall": 0.8116518483855872,
-    "eval_runtime": 14.1813,
+    "epoch": 10.0,
+    "eval_accuracy": 0.9991418018042759,
+    "eval_f1": 0.9521203830369357,
+    "eval_loss": 0.004440780263394117,
+    "eval_precision": 0.9447963800904977,
+    "eval_recall": 0.9595588235294118,
+    "eval_runtime": 14.1123,
     "eval_samples": 6810,
-    "eval_samples_per_second": 480.21,
-    "eval_steps_per_second": 60.079,
-    "predict_accuracy": 0.9755136059661639,
-    "predict_f1": 0.8018404526518684,
-    "predict_loss": 0.161112979054451,
-    "predict_precision": 0.7862455798073406,
-    "predict_recall": 0.818066480588683,
-    "predict_runtime": 29.1457,
-    "predict_samples_per_second": 501.411,
-    "predict_steps_per_second": 62.685,
-    "total_flos": 2.130317416831723e+16,
-    "train_loss": 0.017706579814779112,
-    "train_runtime": 1901.5665,
-    "train_samples": 44938,
-    "train_samples_per_second": 236.321,
-    "train_steps_per_second": 3.692
+    "eval_samples_per_second": 482.558,
+    "eval_steps_per_second": 60.373,
+    "predict_accuracy": 0.9987636326290384,
+    "predict_f1": 0.9231204070096098,
+    "predict_loss": 0.006668921560049057,
+    "predict_precision": 0.8967600219659527,
+    "predict_recall": 0.9510774606872452,
+    "predict_runtime": 28.5821,
+    "predict_samples_per_second": 511.3,
+    "predict_steps_per_second": 63.921,
+    "total_flos": 1.5694885147146138e+16,
+    "train_loss": 0.002533856062142209,
+    "train_runtime": 1422.7831,
+    "train_samples": 31229,
+    "train_samples_per_second": 219.492,
+    "train_steps_per_second": 3.43
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 9.99288256227758,
-    "eval_accuracy": 0.9758743323218038,
-    "eval_f1": 0.8053395240858967,
-    "eval_loss": 0.17113561928272247,
-    "eval_precision": 0.7991246256622898,
-    "eval_recall": 0.8116518483855872,
-    "eval_runtime": 14.1813,
+    "epoch": 10.0,
+    "eval_accuracy": 0.9991418018042759,
+    "eval_f1": 0.9521203830369357,
+    "eval_loss": 0.004440780263394117,
+    "eval_precision": 0.9447963800904977,
+    "eval_recall": 0.9595588235294118,
+    "eval_runtime": 14.1123,
     "eval_samples": 6810,
-    "eval_samples_per_second": 480.21,
-    "eval_steps_per_second": 60.079
+    "eval_samples_per_second": 482.558,
+    "eval_steps_per_second": 60.373
 }

predict_results.json

@@ -1,10 +1,10 @@
 {
-    "predict_accuracy": 0.9755136059661639,
-    "predict_f1": 0.8018404526518684,
-    "predict_loss": 0.161112979054451,
-    "predict_precision": 0.7862455798073406,
-    "predict_recall": 0.818066480588683,
-    "predict_runtime": 29.1457,
-    "predict_samples_per_second": 501.411,
-    "predict_steps_per_second": 62.685
+    "predict_accuracy": 0.9987636326290384,
+    "predict_f1": 0.9231204070096098,
+    "predict_loss": 0.006668921560049057,
+    "predict_precision": 0.8967600219659527,
+    "predict_recall": 0.9510774606872452,
+    "predict_runtime": 28.5821,
+    "predict_samples_per_second": 511.3,
+    "predict_steps_per_second": 63.921
 }

tb/events.out.tfevents.1725915390.0ada7e7d1d89.13010.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03d8cb7f565d31640d583d92b327a96be3b79169b4e94e4f8f1d5458aa42d308
+size 560

train.log

@@ -1463,3 +1463,51 @@ Training completed. Do not forget to share your model on huggingface.co/models =
 {'eval_loss': 0.005196314305067062, 'eval_precision': 0.9416590701914311, 'eval_recall': 0.9494485294117647, 'eval_f1': 0.945537757437071, 'eval_accuracy': 0.9990525491919205, 'eval_runtime': 14.6459, 'eval_samples_per_second': 464.976, 'eval_steps_per_second': 58.173, 'epoch': 10.0}
 {'train_runtime': 1422.7831, 'train_samples_per_second': 219.492, 'train_steps_per_second': 3.43, 'train_loss': 0.002533856062142209, 'epoch': 10.0}
 
+***** train metrics *****
+  epoch                    =       10.0
+  total_flos               = 14617000GF
+  train_loss               =     0.0025
+  train_runtime            = 0:23:42.78
+  train_samples            =      31229
+  train_samples_per_second =    219.492
+  train_steps_per_second   =       3.43
+09/09/2024 20:56:16 - INFO - __main__ -   *** Evaluate ***
+[INFO|trainer.py:811] 2024-09-09 20:56:16,801 >> The following columns in the evaluation set don't have a corresponding argument in `RobertaForTokenClassification.forward` and have been ignored: ner_tags, tokens, id. If ner_tags, tokens, id are not expected by `RobertaForTokenClassification.forward`,  you can safely ignore this message.
+[INFO|trainer.py:3819] 2024-09-09 20:56:16,803 >> 
+***** Running Evaluation *****
+[INFO|trainer.py:3821] 2024-09-09 20:56:16,803 >>   Num examples = 6810
+[INFO|trainer.py:3824] 2024-09-09 20:56:16,803 >>   Batch size = 8
+
  0%|          | 0/852 [00:00<?, ?it/s]
  1%|          | 10/852 [00:00<00:08, 93.57it/s]
  2%|▏         | 20/852 [00:00<00:10, 80.55it/s]
  3%|▎         | 29/852 [00:00<00:10, 80.58it/s]
  4%|▍         | 38/852 [00:00<00:10, 81.11it/s]
  6%|▌         | 47/852 [00:00<00:09, 81.65it/s]
  7%|▋         | 56/852 [00:00<00:09, 83.14it/s]
  8%|▊         | 65/852 [00:00<00:10, 78.65it/s]
  9%|▊         | 73/852 [00:00<00:10, 77.04it/s]
 10%|▉         | 81/852 [00:01<00:09, 77.70it/s]
 11%|█         | 90/852 [00:01<00:09, 79.30it/s]
 12%|█▏        | 98/852 [00:01<00:09, 79.14it/s]
 13%|█▎        | 107/852 [00:01<00:09, 79.14it/s]
 14%|█▎        | 116/852 [00:01<00:09, 80.00it/s]
 15%|█▍        | 125/852 [00:01<00:08, 81.72it/s]
 16%|█▌        | 134/852 [00:01<00:09, 76.77it/s]
 17%|█▋        | 142/852 [00:01<00:09, 74.23it/s]
 18%|█▊        | 150/852 [00:01<00:09, 75.01it/s]
 19%|█▊        | 159/852 [00:02<00:08, 78.54it/s]
 20%|█▉        | 168/852 [00:02<00:08, 79.78it/s]
 21%|██        | 177/852 [00:02<00:08, 80.22it/s]
 22%|██▏       | 186/852 [00:02<00:08, 81.84it/s]
 23%|██▎       | 195/852 [00:02<00:08, 81.79it/s]
 24%|██▍       | 204/852 [00:02<00:07, 82.81it/s]
 25%|██▌       | 213/852 [00:02<00:07, 81.58it/s]
 26%|██▌       | 222/852 [00:02<00:07, 81.05it/s]
 27%|██▋       | 231/852 [00:02<00:07, 82.05it/s]
 28%|██▊       | 240/852 [00:02<00:07, 81.07it/s]
 29%|██▉       | 249/852 [00:03<00:07, 80.20it/s]
 30%|███       | 258/852 [00:03<00:07, 81.36it/s]
 31%|███▏      | 267/852 [00:03<00:07, 81.25it/s]
 32%|███▏      | 276/852 [00:03<00:07, 81.52it/s]
 33%|███▎      | 285/852 [00:03<00:06, 81.87it/s]
 35%|███▍      | 294/852 [00:03<00:06, 80.80it/s]
 36%|███▌      | 303/852 [00:03<00:07, 78.20it/s]
 37%|███▋      | 312/852 [00:03<00:06, 77.55it/s]
 38%|███▊      | 321/852 [00:04<00:06, 79.73it/s]
 39%|███▊      | 330/852 [00:04<00:06, 79.60it/s]
 40%|███▉      | 338/852 [00:04<00:06, 79.58it/s]
 41%|████      | 347/852 [00:04<00:06, 79.98it/s]
 42%|████▏     | 356/852 [00:04<00:06, 79.10it/s]
 43%|████▎     | 365/852 [00:04<00:06, 79.97it/s]
 44%|████▍     | 374/852 [00:04<00:05, 79.68it/s]
 45%|████▍     | 382/852 [00:04<00:05, 78.69it/s]
 46%|████▌     | 391/852 [00:04<00:05, 79.19it/s]
 47%|████▋     | 399/852 [00:04<00:05, 78.98it/s]
 48%|████▊     | 407/852 [00:05<00:05, 77.08it/s]
 49%|████▉     | 416/852 [00:05<00:05, 78.27it/s]
 50%|████▉     | 424/852 [00:05<00:05, 78.61it/s]
 51%|█████     | 432/852 [00:05<00:05, 78.46it/s]
 52%|█████▏    | 441/852 [00:05<00:05, 80.76it/s]
 53%|█████▎    | 450/852 [00:05<00:04, 81.55it/s]
 54%|█████▍    | 459/852 [00:05<00:04, 82.07it/s]
 55%|█████▍    | 468/852 [00:05<00:04, 79.16it/s]
 56%|█████▌    | 476/852 [00:05<00:04, 76.30it/s]
 57%|█████▋    | 484/852 [00:06<00:04, 76.49it/s]
 58%|█████▊    | 493/852 [00:06<00:04, 78.76it/s]
 59%|█████▉    | 502/852 [00:06<00:04, 80.51it/s]
 60%|█████▉    | 511/852 [00:06<00:04, 80.85it/s]
 61%|██████    | 520/852 [00:06<00:04, 82.11it/s]
 62%|██████▏   | 529/852 [00:06<00:04, 79.93it/s]
 63%|██████▎   | 538/852 [00:06<00:03, 81.28it/s]
 64%|██████▍   | 547/852 [00:06<00:03, 81.59it/s]
 65%|██████▌   | 556/852 [00:06<00:03, 78.91it/s]
 66%|██████▋   | 565/852 [00:07<00:03, 81.03it/s]
 67%|██████▋   | 574/852 [00:07<00:03, 81.36it/s]
 68%|██████▊   | 583/852 [00:07<00:03, 80.74it/s]
 69%|██████▉   | 592/852 [00:07<00:03, 80.30it/s]
 71%|███████   | 601/852 [00:07<00:03, 80.88it/s]
 72%|███████▏  | 610/852 [00:07<00:02, 81.38it/s]
 73%|███████▎  | 619/852 [00:07<00:02, 79.27it/s]
 74%|███████▎  | 627/852 [00:07<00:02, 79.27it/s]
 75%|███████▍  | 635/852 [00:07<00:02, 75.84it/s]
 75%|███████▌  | 643/852 [00:08<00:02, 75.01it/s]
 77%|███████▋  | 652/852 [00:08<00:02, 78.03it/s]
 78%|███████▊  | 661/852 [00:08<00:02, 79.05it/s]
 79%|███████▊  | 670/852 [00:08<00:02, 79.91it/s]
 80%|███████▉  | 679/852 [00:08<00:02, 80.51it/s]
 81%|████████  | 688/852 [00:08<00:02, 81.73it/s]
 82%|████████▏ | 697/852 [00:08<00:01, 80.82it/s]
 83%|████████▎ | 706/852 [00:08<00:01, 82.01it/s]
 84%|████████▍ | 715/852 [00:08<00:01, 81.67it/s]
 85%|████████▍ | 724/852 [00:09<00:01, 81.05it/s]
 86%|████████▌ | 733/852 [00:09<00:01, 81.95it/s]
 87%|████████▋ | 742/852 [00:09<00:01, 82.90it/s]
 88%|████████▊ | 751/852 [00:09<00:01, 82.87it/s]
 89%|████████▉ | 760/852 [00:09<00:01, 84.18it/s]
 90%|█████████ | 769/852 [00:09<00:01, 82.73it/s]
 91%|█████████▏| 778/852 [00:09<00:00, 82.09it/s]
 92%|█████████▏| 787/852 [00:09<00:00, 80.10it/s]
 93%|█████████▎| 796/852 [00:09<00:00, 79.31it/s]
 94%|█████████▍| 805/852 [00:10<00:00, 80.66it/s]
 96%|█████████▌| 814/852 [00:10<00:00, 79.72it/s]
 97%|█████████▋| 823/852 [00:10<00:00, 81.12it/s]
 98%|█████████▊| 832/852 [00:10<00:00, 81.37it/s]
 99%|█████████▊| 841/852 [00:10<00:00, 81.30it/s]
+***** eval metrics *****
+  epoch                   =       10.0
+  eval_accuracy           =     0.9991
+  eval_f1                 =     0.9521
+  eval_loss               =     0.0044
+  eval_precision          =     0.9448
+  eval_recall             =     0.9596
+  eval_runtime            = 0:00:14.11
+  eval_samples            =       6810
+  eval_samples_per_second =    482.558
+  eval_steps_per_second   =     60.373
+09/09/2024 20:56:30 - INFO - __main__ -   *** Predict ***
+[INFO|trainer.py:811] 2024-09-09 20:56:30,918 >> The following columns in the test set don't have a corresponding argument in `RobertaForTokenClassification.forward` and have been ignored: ner_tags, tokens, id. If ner_tags, tokens, id are not expected by `RobertaForTokenClassification.forward`,  you can safely ignore this message.
+[INFO|trainer.py:3819] 2024-09-09 20:56:30,920 >> 
+***** Running Prediction *****
+[INFO|trainer.py:3821] 2024-09-09 20:56:30,920 >>   Num examples = 14614
+[INFO|trainer.py:3824] 2024-09-09 20:56:30,920 >>   Batch size = 8
+
  0%|          | 0/1827 [00:00<?, ?it/s]
  0%|          | 9/1827 [00:00<00:22, 81.47it/s]
  1%|          | 18/1827 [00:00<00:24, 73.10it/s]
  1%|▏         | 27/1827 [00:00<00:23, 77.38it/s]
  2%|▏         | 35/1827 [00:00<00:22, 78.12it/s]
  2%|▏         | 44/1827 [00:00<00:22, 79.84it/s]
  3%|▎         | 53/1827 [00:00<00:22, 79.92it/s]
  3%|▎         | 62/1827 [00:00<00:22, 78.68it/s]
  4%|▍         | 71/1827 [00:00<00:22, 79.81it/s]
  4%|▍         | 80/1827 [00:01<00:21, 82.54it/s]
  5%|▍         | 89/1827 [00:01<00:21, 81.34it/s]
  5%|▌         | 98/1827 [00:01<00:21, 82.29it/s]
  6%|▌         | 107/1827 [00:01<00:20, 84.00it/s]
  6%|▋         | 116/1827 [00:01<00:20, 82.38it/s]
  7%|▋         | 125/1827 [00:01<00:20, 82.00it/s]
  7%|▋         | 134/1827 [00:01<00:20, 83.17it/s]
  8%|▊         | 143/1827 [00:01<00:20, 83.98it/s]
  8%|▊         | 152/1827 [00:01<00:20, 83.56it/s]
  9%|▉         | 161/1827 [00:01<00:20, 79.98it/s]
  9%|▉         | 170/1827 [00:02<00:20, 80.90it/s]
 10%|▉         | 179/1827 [00:02<00:20, 82.04it/s]
 10%|█         | 188/1827 [00:02<00:19, 82.53it/s]
 11%|█         | 197/1827 [00:02<00:19, 82.41it/s]
 11%|█▏        | 206/1827 [00:02<00:20, 81.02it/s]
 12%|█▏        | 215/1827 [00:02<00:20, 79.98it/s]
 12%|█▏        | 224/1827 [00:02<00:19, 81.33it/s]
 13%|█▎        | 233/1827 [00:02<00:19, 79.96it/s]
 13%|█▎        | 242/1827 [00:02<00:20, 79.19it/s]
 14%|█▎        | 250/1827 [00:03<00:19, 79.35it/s]
 14%|█▍        | 259/1827 [00:03<00:19, 79.72it/s]
 15%|█▍        | 268/1827 [00:03<00:19, 80.61it/s]
 15%|█▌        | 277/1827 [00:03<00:18, 82.50it/s]
 16%|█▌        | 286/1827 [00:03<00:18, 83.69it/s]
 16%|█▌        | 295/1827 [00:03<00:18, 83.07it/s]
 17%|█▋        | 304/1827 [00:03<00:18, 81.73it/s]
 17%|█▋        | 313/1827 [00:03<00:18, 82.33it/s]
 18%|█▊        | 322/1827 [00:03<00:18, 82.60it/s]
 18%|█▊        | 331/1827 [00:04<00:18, 82.65it/s]
 19%|█▊        | 340/1827 [00:04<00:17, 83.95it/s]
 19%|█▉        | 349/1827 [00:04<00:18, 79.59it/s]
 20%|█▉        | 358/1827 [00:04<00:18, 81.31it/s]
 20%|██        | 367/1827 [00:04<00:17, 82.07it/s]
 21%|██        | 376/1827 [00:04<00:17, 81.81it/s]
 21%|██        | 385/1827 [00:04<00:17, 81.80it/s]
 22%|██▏       | 394/1827 [00:04<00:17, 80.82it/s]
 22%|██▏       | 403/1827 [00:04<00:17, 82.60it/s]
 23%|██▎       | 412/1827 [00:05<00:17, 83.11it/s]
 23%|██▎       | 421/1827 [00:05<00:17, 80.56it/s]
 24%|██▎       | 430/1827 [00:05<00:17, 79.87it/s]
 24%|██▍       | 439/1827 [00:05<00:17, 79.77it/s]
 24%|██▍       | 447/1827 [00:05<00:17, 79.67it/s]
 25%|██▍       | 456/1827 [00:05<00:16, 80.72it/s]
 25%|██▌       | 465/1827 [00:05<00:16, 82.42it/s]
 26%|██▌       | 474/1827 [00:05<00:16, 80.58it/s]
 26%|██▋       | 483/1827 [00:05<00:16, 79.48it/s]
 27%|██▋       | 491/1827 [00:06<00:17, 78.12it/s]
 27%|██▋       | 500/1827 [00:06<00:16, 78.98it/s]
 28%|██▊       | 509/1827 [00:06<00:16, 79.79it/s]
 28%|██▊       | 518/1827 [00:06<00:16, 80.70it/s]
 29%|██▉       | 527/1827 [00:06<00:15, 81.95it/s]
 29%|██▉       | 536/1827 [00:06<00:15, 81.48it/s]
 30%|██▉       | 545/1827 [00:06<00:15, 81.56it/s]
 30%|███       | 554/1827 [00:06<00:15, 80.14it/s]
 31%|███       | 563/1827 [00:06<00:15, 80.95it/s]
 31%|███▏      | 572/1827 [00:07<00:15, 82.12it/s]
 32%|███▏      | 581/1827 [00:07<00:15, 81.74it/s]
 32%|███▏      | 590/1827 [00:07<00:15, 81.57it/s]
 33%|███▎      | 599/1827 [00:07<00:15, 81.15it/s]
 33%|███▎      | 608/1827 [00:07<00:14, 81.57it/s]
 34%|███▍      | 617/1827 [00:07<00:15, 80.16it/s]
 34%|███▍      | 626/1827 [00:07<00:14, 82.03it/s]
 35%|███▍      | 635/1827 [00:07<00:14, 82.57it/s]
 35%|███▌      | 644/1827 [00:07<00:14, 83.46it/s]
 36%|███▌      | 653/1827 [00:08<00:15, 77.96it/s]
 36%|███▌      | 662/1827 [00:08<00:14, 79.18it/s]
 37%|███▋      | 671/1827 [00:08<00:14, 79.75it/s]
 37%|███▋      | 680/1827 [00:08<00:14, 80.75it/s]
 38%|███▊      | 689/1827 [00:08<00:13, 82.05it/s]
 38%|███▊      | 698/1827 [00:08<00:13, 83.26it/s]
 39%|███▊      | 707/1827 [00:08<00:13, 82.35it/s]
 39%|███▉      | 716/1827 [00:08<00:13, 83.46it/s]
 40%|███▉      | 725/1827 [00:08<00:13, 84.47it/s]
 40%|████      | 734/1827 [00:09<00:12, 84.78it/s]
 41%|████      | 743/1827 [00:09<00:12, 84.36it/s]
 41%|████      | 752/1827 [00:09<00:12, 84.90it/s]
 42%|████▏     | 761/1827 [00:09<00:12, 84.81it/s]
 42%|████▏     | 770/1827 [00:09<00:12, 85.40it/s]
 43%|████▎     | 779/1827 [00:09<00:12, 82.72it/s]
 43%|████▎     | 788/1827 [00:09<00:12, 82.68it/s]
 44%|████▎     | 797/1827 [00:09<00:13, 78.45it/s]
 44%|████▍     | 806/1827 [00:09<00:12, 80.13it/s]
 45%|████▍     | 815/1827 [00:10<00:12, 81.93it/s]
 45%|████▌     | 824/1827 [00:10<00:12, 81.92it/s]
 46%|████▌     | 833/1827 [00:10<00:12, 80.09it/s]
 46%|████▌     | 842/1827 [00:10<00:12, 80.84it/s]
 47%|████▋     | 851/1827 [00:10<00:11, 82.26it/s]
 47%|████▋     | 860/1827 [00:10<00:11, 82.48it/s]
 48%|████▊     | 869/1827 [00:10<00:11, 80.82it/s]
 48%|████▊     | 878/1827 [00:10<00:11, 79.77it/s]
 48%|████▊     | 886/1827 [00:10<00:11, 79.80it/s]
 49%|████▉     | 895/1827 [00:11<00:11, 80.41it/s]
 49%|████▉     | 904/1827 [00:11<00:11, 82.22it/s]
 50%|████▉     | 913/1827 [00:11<00:10, 83.47it/s]
 50%|█████     | 922/1827 [00:11<00:10, 83.69it/s]
 51%|█████     | 931/1827 [00:11<00:11, 81.45it/s]
 51%|█████▏    | 940/1827 [00:11<00:10, 82.66it/s]
 52%|█████▏    | 949/1827 [00:11<00:10, 80.89it/s]
 52%|█████▏    | 958/1827 [00:11<00:10, 81.33it/s]
 53%|█████▎    | 967/1827 [00:11<00:10, 82.58it/s]
 53%|█████▎    | 976/1827 [00:11<00:10, 81.82it/s]
 54%|█████▍    | 985/1827 [00:12<00:10, 82.23it/s]
 54%|█████▍    | 994/1827 [00:12<00:10, 83.02it/s]
 55%|█████▍    | 1003/1827 [00:12<00:09, 83.11it/s]
 55%|█████▌    | 1012/1827 [00:12<00:09, 83.22it/s]
 56%|█████▌    | 1021/1827 [00:12<00:09, 84.03it/s]
 56%|█████▋    | 1030/1827 [00:12<00:09, 85.02it/s]
 57%|█████▋    | 1039/1827 [00:12<00:09, 82.97it/s]
 57%|█████▋    | 1048/1827 [00:12<00:09, 83.08it/s]
 58%|█████▊    | 1057/1827 [00:12<00:09, 83.42it/s]
 58%|█████▊    | 1066/1827 [00:13<00:09, 83.10it/s]
 59%|█████▉    | 1075/1827 [00:13<00:08, 84.02it/s]
 59%|█████▉    | 1084/1827 [00:13<00:08, 84.75it/s]
 60%|█████▉    | 1093/1827 [00:13<00:08, 85.02it/s]
 60%|██████    | 1102/1827 [00:13<00:08, 84.64it/s]
 61%|██████    | 1111/1827 [00:13<00:08, 84.52it/s]
 61%|██████▏   | 1120/1827 [00:13<00:08, 83.58it/s]
 62%|██████▏   | 1129/1827 [00:13<00:08, 83.47it/s]
 62%|██████▏   | 1138/1827 [00:13<00:08, 83.97it/s]
 63%|██████▎   | 1147/1827 [00:14<00:08, 84.60it/s]
 63%|██████▎   | 1156/1827 [00:14<00:07, 84.68it/s]
 64%|██████▍   | 1165/1827 [00:14<00:08, 81.69it/s]
 64%|██████▍   | 1174/1827 [00:14<00:07, 83.04it/s]
 65%|██████▍   | 1183/1827 [00:14<00:07, 81.23it/s]
 65%|██████▌   | 1192/1827 [00:14<00:07, 82.81it/s]
 66%|██████▌   | 1201/1827 [00:14<00:07, 83.12it/s]
 66%|██████▌   | 1210/1827 [00:14<00:07, 82.78it/s]
 67%|██████▋   | 1219/1827 [00:14<00:07, 81.18it/s]
 67%|██████▋   | 1228/1827 [00:15<00:07, 83.02it/s]
 68%|██████▊   | 1237/1827 [00:15<00:07, 83.76it/s]
 68%|██████▊   | 1246/1827 [00:15<00:06, 84.24it/s]
 69%|██████▊   | 1255/1827 [00:15<00:07, 81.04it/s]
 69%|██████▉   | 1264/1827 [00:15<00:06, 81.41it/s]
 70%|██████▉   | 1273/1827 [00:15<00:06, 81.98it/s]
 70%|███████   | 1282/1827 [00:15<00:06, 83.21it/s]
 71%|███████   | 1291/1827 [00:15<00:06, 83.31it/s]
 71%|███████   | 1300/1827 [00:15<00:06, 83.92it/s]
 72%|███████▏  | 1309/1827 [00:15<00:06, 84.33it/s]
 72%|███████▏  | 1318/1827 [00:16<00:05, 84.99it/s]
 73%|███████▎  | 1327/1827 [00:16<00:05, 84.67it/s]
 73%|███████▎  | 1336/1827 [00:16<00:05, 85.18it/s]
 74%|███████▎  | 1345/1827 [00:16<00:05, 84.30it/s]
 74%|███████▍  | 1354/1827 [00:16<00:05, 84.28it/s]
 75%|███████▍  | 1363/1827 [00:16<00:05, 84.42it/s]
 75%|███████▌  | 1372/1827 [00:16<00:05, 83.96it/s]
 76%|███████▌  | 1381/1827 [00:16<00:05, 84.35it/s]
 76%|███████▌  | 1390/1827 [00:16<00:05, 83.32it/s]
 77%|███████▋  | 1399/1827 [00:17<00:05, 81.72it/s]
 77%|███████▋  | 1408/1827 [00:17<00:05, 82.44it/s]
 78%|███████▊  | 1417/1827 [00:17<00:04, 83.25it/s]
 78%|███████▊  | 1426/1827 [00:17<00:04, 81.63it/s]
 79%|███████▊  | 1435/1827 [00:17<00:04, 81.84it/s]
 79%|███████▉  | 1444/1827 [00:17<00:04, 79.43it/s]
 79%|███████▉  | 1452/1827 [00:17<00:04, 78.88it/s]
 80%|███████▉  | 1460/1827 [00:17<00:04, 79.00it/s]
 80%|████████  | 1469/1827 [00:17<00:04, 77.76it/s]
 81%|████████  | 1478/1827 [00:18<00:04, 78.91it/s]
 81%|████████▏ | 1487/1827 [00:18<00:04, 79.81it/s]
 82%|████████▏ | 1495/1827 [00:18<00:04, 77.98it/s]
 82%|████████▏ | 1503/1827 [00:18<00:04, 78.22it/s]
 83%|████████▎ | 1512/1827 [00:18<00:03, 79.84it/s]
 83%|████████▎ | 1520/1827 [00:18<00:03, 79.72it/s]
 84%|████████▎ | 1528/1827 [00:18<00:03, 78.62it/s]
 84%|████████▍ | 1536/1827 [00:18<00:03, 77.91it/s]
 85%|████████▍ | 1545/1827 [00:18<00:03, 78.92it/s]
 85%|████████▌ | 1554/1827 [00:19<00:03, 79.49it/s]
 85%|████████▌ | 1562/1827 [00:19<00:03, 79.02it/s]
 86%|████████▌ | 1570/1827 [00:19<00:03, 78.77it/s]
 86%|████████▋ | 1579/1827 [00:19<00:03, 80.51it/s]
 87%|████████▋ | 1588/1827 [00:19<00:03, 79.62it/s]
 87%|████████▋ | 1597/1827 [00:19<00:02, 79.98it/s]
 88%|████████▊ | 1606/1827 [00:19<00:02, 81.13it/s]
 88%|████████▊ | 1615/1827 [00:19<00:02, 80.28it/s]
 89%|████████▉ | 1624/1827 [00:19<00:02, 74.42it/s]
 89%|████████▉ | 1632/1827 [00:20<00:02, 74.97it/s]
 90%|████████▉ | 1641/1827 [00:20<00:02, 76.63it/s]
 90%|█████████ | 1650/1827 [00:20<00:02, 78.43it/s]
 91%|█████████ | 1659/1827 [00:20<00:02, 79.16it/s]
 91%|█████████ | 1667/1827 [00:20<00:02, 78.81it/s]
 92%|█████████▏| 1676/1827 [00:20<00:01, 80.04it/s]
 92%|█████████▏| 1685/1827 [00:20<00:01, 81.58it/s]
 93%|█████████▎| 1694/1827 [00:20<00:01, 79.80it/s]
 93%|█████████▎| 1703/1827 [00:20<00:01, 80.70it/s]
 94%|█████████▎| 1712/1827 [00:20<00:01, 82.14it/s]
 94%|█████████▍| 1721/1827 [00:21<00:01, 79.30it/s]
 95%|█████████▍| 1730/1827 [00:21<00:01, 81.07it/s]
 95%|█████████▌| 1739/1827 [00:21<00:01, 82.15it/s]
 96%|█████████▌| 1748/1827 [00:21<00:00, 83.61it/s]
 96%|█████████▌| 1757/1827 [00:21<00:00, 83.43it/s]
 97%|█████████▋| 1766/1827 [00:21<00:00, 83.99it/s]
 97%|█████████▋| 1775/1827 [00:21<00:00, 84.37it/s]
 98%|█████████▊| 1784/1827 [00:21<00:00, 85.02it/s]
 98%|█████████▊| 1793/1827 [00:21<00:00, 82.53it/s]
 99%|█████████▊| 1802/1827 [00:22<00:00, 82.32it/s]
 99%|█████████▉| 1811/1827 [00:22<00:00, 81.76it/s]
+[INFO|trainer.py:3503] 2024-09-09 20:57:00,376 >> Saving model checkpoint to /content/dissertation/scripts/ner/output
+[INFO|configuration_utils.py:472] 2024-09-09 20:57:00,377 >> Configuration saved in /content/dissertation/scripts/ner/output/config.json
+[INFO|modeling_utils.py:2799] 2024-09-09 20:57:01,738 >> Model weights saved in /content/dissertation/scripts/ner/output/model.safetensors
+[INFO|tokenization_utils_base.py:2684] 2024-09-09 20:57:01,739 >> tokenizer config file saved in /content/dissertation/scripts/ner/output/tokenizer_config.json
+[INFO|tokenization_utils_base.py:2693] 2024-09-09 20:57:01,739 >> Special tokens file saved in /content/dissertation/scripts/ner/output/special_tokens_map.json
+***** predict metrics *****
+  predict_accuracy           =     0.9988
+  predict_f1                 =     0.9231
+  predict_loss               =     0.0067
+  predict_precision          =     0.8968
+  predict_recall             =     0.9511
+  predict_runtime            = 0:00:28.58
+  predict_samples_per_second =      511.3
+  predict_steps_per_second   =     63.921
+

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 9.99288256227758,
-    "total_flos": 2.130317416831723e+16,
-    "train_loss": 0.017706579814779112,
-    "train_runtime": 1901.5665,
-    "train_samples": 44938,
-    "train_samples_per_second": 236.321,
-    "train_steps_per_second": 3.692
+    "epoch": 10.0,
+    "total_flos": 1.5694885147146138e+16,
+    "train_loss": 0.002533856062142209,
+    "train_runtime": 1422.7831,
+    "train_samples": 31229,
+    "train_samples_per_second": 219.492,
+    "train_steps_per_second": 3.43
 }

trainer_state.json

@@ -1,243 +1,208 @@
 {
-  "best_metric": 0.8053395240858967,
-  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-7020",
-  "epoch": 9.99288256227758,
+  "best_metric": 0.9521203830369357,
+  "best_model_checkpoint": "/content/dissertation/scripts/ner/output/checkpoint-2440",
+  "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 7020,
+  "global_step": 4880,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.7117437722419929,
-      "grad_norm": 0.739061713218689,
-      "learning_rate": 4.643874643874644e-05,
-      "loss": 0.1165,
+      "epoch": 1.0,
+      "eval_accuracy": 0.9987641945981572,
+      "eval_f1": 0.9355123674911661,
+      "eval_loss": 0.003916793502867222,
+      "eval_precision": 0.9005102040816326,
+      "eval_recall": 0.9733455882352942,
+      "eval_runtime": 13.936,
+      "eval_samples_per_second": 488.661,
+      "eval_steps_per_second": 61.136,
+      "step": 488
+    },
+    {
+      "epoch": 1.0245901639344261,
+      "grad_norm": 0.24995087087154388,
+      "learning_rate": 4.487704918032787e-05,
+      "loss": 0.0189,
       "step": 500
     },
     {
-      "epoch": 0.999288256227758,
-      "eval_accuracy": 0.9735468988149999,
-      "eval_f1": 0.7736125182933693,
-      "eval_loss": 0.08085697889328003,
-      "eval_precision": 0.7454979388153613,
-      "eval_recall": 0.8039307440336921,
-      "eval_runtime": 14.5471,
-      "eval_samples_per_second": 468.136,
-      "eval_steps_per_second": 58.569,
-      "step": 702
-    },
-    {
-      "epoch": 1.4234875444839858,
-      "grad_norm": 0.5600239038467407,
-      "learning_rate": 4.287749287749288e-05,
-      "loss": 0.0461,
+      "epoch": 2.0,
+      "eval_accuracy": 0.9989495654084337,
+      "eval_f1": 0.9413886384129847,
+      "eval_loss": 0.0032420416828244925,
+      "eval_precision": 0.9238938053097345,
+      "eval_recall": 0.9595588235294118,
+      "eval_runtime": 14.0058,
+      "eval_samples_per_second": 486.228,
+      "eval_steps_per_second": 60.832,
+      "step": 976
+    },
+    {
+      "epoch": 2.0491803278688523,
+      "grad_norm": 0.04940846189856529,
+      "learning_rate": 3.975409836065574e-05,
+      "loss": 0.0027,
       "step": 1000
     },
     {
-      "epoch": 2.0,
-      "eval_accuracy": 0.9746797204333557,
-      "eval_f1": 0.7832803271240344,
-      "eval_loss": 0.09561743587255478,
-      "eval_precision": 0.7611479028697572,
-      "eval_recall": 0.8067384183434722,
-      "eval_runtime": 14.2524,
-      "eval_samples_per_second": 477.815,
-      "eval_steps_per_second": 59.779,
-      "step": 1405
+      "epoch": 3.0,
+      "eval_accuracy": 0.9989358342373021,
+      "eval_f1": 0.9402784014369105,
+      "eval_loss": 0.0043699671514332294,
+      "eval_precision": 0.9192273924495171,
+      "eval_recall": 0.9623161764705882,
+      "eval_runtime": 14.0607,
+      "eval_samples_per_second": 484.329,
+      "eval_steps_per_second": 60.594,
+      "step": 1464
     },
     {
-      "epoch": 2.135231316725979,
-      "grad_norm": 0.4771190881729126,
-      "learning_rate": 3.931623931623932e-05,
-      "loss": 0.0273,
+      "epoch": 3.0737704918032787,
+      "grad_norm": 0.01636342518031597,
+      "learning_rate": 3.463114754098361e-05,
+      "loss": 0.0015,
       "step": 1500
     },
     {
-      "epoch": 2.8469750889679717,
-      "grad_norm": 0.29212868213653564,
-      "learning_rate": 3.575498575498576e-05,
-      "loss": 0.0165,
+      "epoch": 4.0,
+      "eval_accuracy": 0.9990594147774864,
+      "eval_f1": 0.9445208619899129,
+      "eval_loss": 0.0036185304634273052,
+      "eval_precision": 0.9423604757548033,
+      "eval_recall": 0.9466911764705882,
+      "eval_runtime": 14.26,
+      "eval_samples_per_second": 477.56,
+      "eval_steps_per_second": 59.748,
+      "step": 1952
+    },
+    {
+      "epoch": 4.098360655737705,
+      "grad_norm": 0.006754329428076744,
+      "learning_rate": 2.9508196721311478e-05,
+      "loss": 0.0007,
       "step": 2000
     },
     {
-      "epoch": 2.999288256227758,
-      "eval_accuracy": 0.974439424938553,
-      "eval_f1": 0.7853282741175118,
-      "eval_loss": 0.10571284592151642,
-      "eval_precision": 0.7721003843545106,
-      "eval_recall": 0.7990173139915769,
-      "eval_runtime": 14.2529,
-      "eval_samples_per_second": 477.797,
-      "eval_steps_per_second": 59.777,
-      "step": 2107
+      "epoch": 5.0,
+      "eval_accuracy": 0.9991418018042759,
+      "eval_f1": 0.9521203830369357,
+      "eval_loss": 0.004440780263394117,
+      "eval_precision": 0.9447963800904977,
+      "eval_recall": 0.9595588235294118,
+      "eval_runtime": 14.0071,
+      "eval_samples_per_second": 486.182,
+      "eval_steps_per_second": 60.826,
+      "step": 2440
     },
     {
-      "epoch": 3.5587188612099645,
-      "grad_norm": 0.08157803863286972,
-      "learning_rate": 3.2193732193732194e-05,
-      "loss": 0.011,
+      "epoch": 5.122950819672131,
+      "grad_norm": 0.0005009469459764659,
+      "learning_rate": 2.4385245901639343e-05,
+      "loss": 0.0004,
       "step": 2500
     },
     {
-      "epoch": 4.0,
-      "eval_accuracy": 0.9750504620539086,
-      "eval_f1": 0.7971327796108773,
-      "eval_loss": 0.12744221091270447,
-      "eval_precision": 0.775858250276855,
-      "eval_recall": 0.8196069255966308,
-      "eval_runtime": 14.6113,
-      "eval_samples_per_second": 466.079,
-      "eval_steps_per_second": 58.311,
-      "step": 2810
-    },
-    {
-      "epoch": 4.270462633451958,
-      "grad_norm": 0.45329615473747253,
-      "learning_rate": 2.863247863247863e-05,
-      "loss": 0.0079,
+      "epoch": 6.0,
+      "eval_accuracy": 0.9990456836063548,
+      "eval_f1": 0.9464368886818817,
+      "eval_loss": 0.005531433038413525,
+      "eval_precision": 0.959395656279509,
+      "eval_recall": 0.9338235294117647,
+      "eval_runtime": 14.0724,
+      "eval_samples_per_second": 483.926,
+      "eval_steps_per_second": 60.544,
+      "step": 2928
+    },
+    {
+      "epoch": 6.147540983606557,
+      "grad_norm": 0.7484572529792786,
+      "learning_rate": 1.9262295081967212e-05,
+      "loss": 0.0002,
       "step": 3000
     },
     {
-      "epoch": 4.98220640569395,
-      "grad_norm": 0.4696062207221985,
-      "learning_rate": 2.5071225071225073e-05,
-      "loss": 0.006,
-      "step": 3500
+      "epoch": 7.0,
+      "eval_accuracy": 0.9989907589218284,
+      "eval_f1": 0.9427393495190106,
+      "eval_loss": 0.004938796162605286,
+      "eval_precision": 0.9397260273972603,
+      "eval_recall": 0.9457720588235294,
+      "eval_runtime": 13.9575,
+      "eval_samples_per_second": 487.909,
+      "eval_steps_per_second": 61.042,
+      "step": 3416
     },
     {
-      "epoch": 4.999288256227758,
-      "eval_accuracy": 0.9744943496230793,
-      "eval_f1": 0.7975886853698122,
-      "eval_loss": 0.13583102822303772,
-      "eval_precision": 0.7904411764705882,
-      "eval_recall": 0.8048666354702855,
-      "eval_runtime": 14.3515,
-      "eval_samples_per_second": 474.516,
-      "eval_steps_per_second": 59.367,
-      "step": 3512
+      "epoch": 7.172131147540983,
+      "grad_norm": 0.00041869020787999034,
+      "learning_rate": 1.4139344262295081e-05,
+      "loss": 0.0002,
+      "step": 3500
     },
     {
-      "epoch": 5.693950177935943,
-      "grad_norm": 0.6243860721588135,
-      "learning_rate": 2.150997150997151e-05,
-      "loss": 0.0045,
+      "epoch": 8.0,
+      "eval_accuracy": 0.999073145948618,
+      "eval_f1": 0.9468864468864469,
+      "eval_loss": 0.005282656755298376,
+      "eval_precision": 0.9434306569343066,
+      "eval_recall": 0.9503676470588235,
+      "eval_runtime": 14.1087,
+      "eval_samples_per_second": 482.68,
+      "eval_steps_per_second": 60.388,
+      "step": 3904
+    },
+    {
+      "epoch": 8.19672131147541,
+      "grad_norm": 0.004756764974445105,
+      "learning_rate": 9.016393442622952e-06,
+      "loss": 0.0001,
       "step": 4000
     },
     {
-      "epoch": 6.0,
-      "eval_accuracy": 0.9746179301632636,
-      "eval_f1": 0.7948299953423382,
-      "eval_loss": 0.14197705686092377,
-      "eval_precision": 0.7911451089476125,
-      "eval_recall": 0.7985493682732803,
-      "eval_runtime": 14.3392,
-      "eval_samples_per_second": 474.921,
-      "eval_steps_per_second": 59.417,
-      "step": 4215
+      "epoch": 9.0,
+      "eval_accuracy": 0.9990525491919205,
+      "eval_f1": 0.946404031149794,
+      "eval_loss": 0.0050343358889222145,
+      "eval_precision": 0.94337899543379,
+      "eval_recall": 0.9494485294117647,
+      "eval_runtime": 14.373,
+      "eval_samples_per_second": 473.804,
+      "eval_steps_per_second": 59.278,
+      "step": 4392
     },
     {
-      "epoch": 6.405693950177936,
-      "grad_norm": 0.2963380813598633,
-      "learning_rate": 1.794871794871795e-05,
-      "loss": 0.0037,
+      "epoch": 9.221311475409836,
+      "grad_norm": 0.00033881032140925527,
+      "learning_rate": 3.89344262295082e-06,
+      "loss": 0.0001,
       "step": 4500
     },
     {
-      "epoch": 6.999288256227758,
-      "eval_accuracy": 0.9748925535858953,
-      "eval_f1": 0.796227293898463,
-      "eval_loss": 0.160146102309227,
-      "eval_precision": 0.7925359295317571,
-      "eval_recall": 0.7999532054281703,
-      "eval_runtime": 14.175,
-      "eval_samples_per_second": 480.423,
-      "eval_steps_per_second": 60.106,
-      "step": 4917
-    },
-    {
-      "epoch": 7.117437722419929,
-      "grad_norm": 0.4029097557067871,
-      "learning_rate": 1.4387464387464389e-05,
-      "loss": 0.0028,
-      "step": 5000
-    },
-    {
-      "epoch": 7.829181494661921,
-      "grad_norm": 0.04551521688699722,
-      "learning_rate": 1.0826210826210826e-05,
-      "loss": 0.0022,
-      "step": 5500
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.9758331388084089,
-      "eval_f1": 0.8050679995350458,
-      "eval_loss": 0.1620824635028839,
-      "eval_precision": 0.7999537999538,
-      "eval_recall": 0.8102480112306972,
-      "eval_runtime": 14.4773,
-      "eval_samples_per_second": 470.391,
-      "eval_steps_per_second": 58.851,
-      "step": 5620
-    },
-    {
-      "epoch": 8.540925266903916,
-      "grad_norm": 0.3445014953613281,
-      "learning_rate": 7.264957264957266e-06,
-      "loss": 0.0016,
-      "step": 6000
-    },
-    {
-      "epoch": 8.999288256227757,
-      "eval_accuracy": 0.9757713485383168,
-      "eval_f1": 0.8028807062376583,
-      "eval_loss": 0.16807720065116882,
-      "eval_precision": 0.7972318339100346,
-      "eval_recall": 0.8086102012166588,
-      "eval_runtime": 14.1973,
-      "eval_samples_per_second": 479.67,
-      "eval_steps_per_second": 60.012,
-      "step": 6322
-    },
-    {
-      "epoch": 9.252669039145907,
-      "grad_norm": 0.12732785940170288,
-      "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.0013,
-      "step": 6500
-    },
-    {
-      "epoch": 9.9644128113879,
-      "grad_norm": 0.05971187725663185,
-      "learning_rate": 1.4245014245014247e-07,
-      "loss": 0.0013,
-      "step": 7000
-    },
-    {
-      "epoch": 9.99288256227758,
-      "eval_accuracy": 0.9758743323218038,
-      "eval_f1": 0.8053395240858967,
-      "eval_loss": 0.17113561928272247,
-      "eval_precision": 0.7991246256622898,
-      "eval_recall": 0.8116518483855872,
-      "eval_runtime": 14.5918,
-      "eval_samples_per_second": 466.7,
-      "eval_steps_per_second": 58.389,
-      "step": 7020
-    },
-    {
-      "epoch": 9.99288256227758,
-      "step": 7020,
-      "total_flos": 2.130317416831723e+16,
-      "train_loss": 0.017706579814779112,
-      "train_runtime": 1901.5665,
-      "train_samples_per_second": 236.321,
-      "train_steps_per_second": 3.692
+      "epoch": 10.0,
+      "eval_accuracy": 0.9990525491919205,
+      "eval_f1": 0.945537757437071,
+      "eval_loss": 0.005196314305067062,
+      "eval_precision": 0.9416590701914311,
+      "eval_recall": 0.9494485294117647,
+      "eval_runtime": 14.6459,
+      "eval_samples_per_second": 464.976,
+      "eval_steps_per_second": 58.173,
+      "step": 4880
+    },
+    {
+      "epoch": 10.0,
+      "step": 4880,
+      "total_flos": 1.5694885147146138e+16,
+      "train_loss": 0.002533856062142209,
+      "train_runtime": 1422.7831,
+      "train_samples_per_second": 219.492,
+      "train_steps_per_second": 3.43
     }
   ],
   "logging_steps": 500,
-  "max_steps": 7020,
+  "max_steps": 4880,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
@@ -253,7 +218,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.130317416831723e+16,
+  "total_flos": 1.5694885147146138e+16,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null