Add AL_Test Model with adapted_tokenizer

README.md

@@ -10,22 +10,22 @@ metrics:
 - f1
 - accuracy
 model-index:
-- name: BiomedNLP_BiomedBERT_base
+- name: CeLLaTe-AL-Testbase_adapted_tok
   results: []
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# BiomedNLP_BiomedBERT_base
+# CeLLaTe-AL-Testbase_adapted_tok
 
 This model is a fine-tuned version of [microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext](https://huggingface.co/microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4573
-- Precision: 0.5991
-- Recall: 0.5668
-- F1: 0.5825
-- Accuracy: 0.9301
+- Loss: 0.4629
+- Precision: 0.6069
+- Recall: 0.5631
+- F1: 0.5842
+- Accuracy: 0.9307
 
 ## Model description
 

all_results.json

@@ -1,17 +1,17 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.9311370980866249,
-    "eval_f1": 0.5635515913237381,
-    "eval_loss": 0.32190775871276855,
-    "eval_precision": 0.585016835016835,
-    "eval_recall": 0.543605788032851,
-    "eval_runtime": 4.6723,
+    "eval_accuracy": 0.9306721223816056,
+    "eval_f1": 0.5841584158415841,
+    "eval_loss": 0.46293044090270996,
+    "eval_precision": 0.6069066862601029,
+    "eval_recall": 0.5630538513974097,
+    "eval_runtime": 4.4514,
     "eval_samples": 952,
-    "eval_samples_per_second": 203.754,
-    "eval_steps_per_second": 12.842,
+    "eval_samples_per_second": 213.865,
+    "eval_steps_per_second": 13.479,
     "total_flos": 1090015519875414.0,
-    "train_loss": 0.08582915891771731,
-    "train_runtime": 509.1667,
-    "train_samples_per_second": 72.157,
-    "train_steps_per_second": 4.517
+    "train_loss": 0.08596961084915244,
+    "train_runtime": 505.247,
+    "train_samples_per_second": 72.717,
+    "train_steps_per_second": 4.552
 }

eval_results.json

@@ -1,17 +1,17 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.9311370980866249,
-    "eval_f1": 0.5635515913237381,
-    "eval_loss": 0.32190775871276855,
-    "eval_precision": 0.585016835016835,
-    "eval_recall": 0.543605788032851,
-    "eval_runtime": 4.6723,
+    "eval_accuracy": 0.9306721223816056,
+    "eval_f1": 0.5841584158415841,
+    "eval_loss": 0.46293044090270996,
+    "eval_precision": 0.6069066862601029,
+    "eval_recall": 0.5630538513974097,
+    "eval_runtime": 4.4514,
     "eval_samples": 952,
-    "eval_samples_per_second": 203.754,
-    "eval_steps_per_second": 12.842,
+    "eval_samples_per_second": 213.865,
+    "eval_steps_per_second": 13.479,
     "total_flos": 1090015519875414.0,
-    "train_loss": 0.08582915891771731,
-    "train_runtime": 509.1667,
-    "train_samples_per_second": 72.157,
-    "train_steps_per_second": 4.517
+    "train_loss": 0.08596961084915244,
+    "train_runtime": 505.247,
+    "train_samples_per_second": 72.717,
+    "train_steps_per_second": 4.552
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
     "total_flos": 1090015519875414.0,
-    "train_loss": 0.08582915891771731,
-    "train_runtime": 509.1667,
-    "train_samples_per_second": 72.157,
-    "train_steps_per_second": 4.517
+    "train_loss": 0.08596961084915244,
+    "train_runtime": 505.247,
+    "train_samples_per_second": 72.717,
+    "train_steps_per_second": 4.552
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.5637284701114489,
-  "best_model_checkpoint": "/nfs/production/literature/amina-mardiyyah/new_data/OT-Entity-Extraction-Pipeline/model_outputs/hf/CeLLaTe-AL-Test_1.0/base/BiomedNLP_BiomedBERT_base/checkpoint-690",
+  "best_metric": 0.5841584158415841,
+  "best_model_checkpoint": "/nfs/production/literature/amina-mardiyyah/new_data/OT-Entity-Extraction-Pipeline/model_outputs/hf/CeLLaTe-AL-Test_1.0/base/BiomedNLP_BiomedBERT_base/checkpoint-2185",
   "epoch": 20.0,
   "eval_steps": 500,
   "global_step": 2300,
@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 3.141930341720581,
+      "grad_norm": 3.2613918781280518,
       "learning_rate": 9.652173913043478e-06,
-      "loss": 1.0965,
+      "loss": 1.0711,
       "step": 115
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8700857880175761,
-      "eval_f1": 0.2011861313868613,
-      "eval_loss": 0.4151777923107147,
-      "eval_precision": 0.2413793103448276,
-      "eval_recall": 0.1724677356276887,
-      "eval_runtime": 4.6552,
-      "eval_samples_per_second": 204.501,
-      "eval_steps_per_second": 12.889,
+      "eval_accuracy": 0.8598795712924,
+      "eval_f1": 0.19120043454644214,
+      "eval_loss": 0.43796488642692566,
+      "eval_precision": 0.20393974507531865,
+      "eval_recall": 0.17995910020449898,
+      "eval_runtime": 4.4304,
+      "eval_samples_per_second": 214.881,
+      "eval_steps_per_second": 13.543,
       "step": 115
     },
     {
       "epoch": 2.0,
-      "grad_norm": 5.598133563995361,
+      "grad_norm": 7.965872764587402,
       "learning_rate": 1.965217391304348e-05,
-      "loss": 0.2292,
+      "loss": 0.2579,
       "step": 230
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9119103526840723,
-      "eval_f1": 0.46394984326018807,
-      "eval_loss": 0.2859836220741272,
-      "eval_precision": 0.43893928820655964,
-      "eval_recall": 0.4919827923347673,
-      "eval_runtime": 4.7635,
-      "eval_samples_per_second": 199.851,
-      "eval_steps_per_second": 12.596,
+      "eval_accuracy": 0.9111663915560412,
+      "eval_f1": 0.4053305642090689,
+      "eval_loss": 0.28297552466392517,
+      "eval_precision": 0.41174402250351616,
+      "eval_recall": 0.3991138377641445,
+      "eval_runtime": 4.5407,
+      "eval_samples_per_second": 209.658,
+      "eval_steps_per_second": 13.214,
       "step": 230
     },
     {
       "epoch": 3.0,
-      "grad_norm": 1.903611183166504,
+      "grad_norm": 1.755993366241455,
       "learning_rate": 1.892753623188406e-05,
-      "loss": 0.1037,
+      "loss": 0.1083,
       "step": 345
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9199544323809081,
-      "eval_f1": 0.5066666666666667,
-      "eval_loss": 0.2779094874858856,
-      "eval_precision": 0.49387300408466395,
-      "eval_recall": 0.5201407899882675,
-      "eval_runtime": 4.7455,
-      "eval_samples_per_second": 200.612,
-      "eval_steps_per_second": 12.644,
+      "eval_accuracy": 0.9221630669797503,
+      "eval_f1": 0.5042134831460674,
+      "eval_loss": 0.2840645909309387,
+      "eval_precision": 0.5199131064446053,
+      "eval_recall": 0.4894342194955692,
+      "eval_runtime": 4.4831,
+      "eval_samples_per_second": 212.354,
+      "eval_steps_per_second": 13.384,
       "step": 345
     },
     {
       "epoch": 4.0,
-      "grad_norm": 1.9366143941879272,
+      "grad_norm": 2.016618013381958,
       "learning_rate": 1.781642512077295e-05,
-      "loss": 0.0684,
+      "loss": 0.0707,
       "step": 460
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9274405412317206,
-      "eval_f1": 0.5361736334405145,
-      "eval_loss": 0.29712674021720886,
-      "eval_precision": 0.5514675485737909,
-      "eval_recall": 0.5217051231912397,
-      "eval_runtime": 4.733,
-      "eval_samples_per_second": 201.143,
-      "eval_steps_per_second": 12.677,
+      "eval_accuracy": 0.923953223444075,
+      "eval_f1": 0.538422903063787,
+      "eval_loss": 0.289853036403656,
+      "eval_precision": 0.5291214215202369,
+      "eval_recall": 0.5480572597137015,
+      "eval_runtime": 4.5051,
+      "eval_samples_per_second": 211.318,
+      "eval_steps_per_second": 13.318,
       "step": 460
     },
     {
       "epoch": 5.0,
-      "grad_norm": 1.2101763486862183,
+      "grad_norm": 1.379032015800476,
       "learning_rate": 1.6705314009661836e-05,
-      "loss": 0.0463,
+      "loss": 0.0472,
       "step": 575
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.9276265315137284,
-      "eval_f1": 0.5296671490593343,
-      "eval_loss": 0.3128739893436432,
-      "eval_precision": 0.5618421052631579,
-      "eval_recall": 0.5009777082518576,
-      "eval_runtime": 4.7226,
-      "eval_samples_per_second": 201.582,
-      "eval_steps_per_second": 12.705,
+      "eval_accuracy": 0.9297421709715668,
+      "eval_f1": 0.5575865128660159,
+      "eval_loss": 0.29540660977363586,
+      "eval_precision": 0.5816364309514994,
+      "eval_recall": 0.5354464894342195,
+      "eval_runtime": 4.4726,
+      "eval_samples_per_second": 212.852,
+      "eval_steps_per_second": 13.415,
       "step": 575
     },
     {
       "epoch": 6.0,
-      "grad_norm": 2.3417909145355225,
+      "grad_norm": 2.1465651988983154,
       "learning_rate": 1.5594202898550727e-05,
-      "loss": 0.0319,
+      "loss": 0.0336,
       "step": 690
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.931113849301374,
-      "eval_f1": 0.5637284701114489,
-      "eval_loss": 0.3199135661125183,
-      "eval_precision": 0.5849453322119428,
-      "eval_recall": 0.5439968713335941,
-      "eval_runtime": 5.02,
-      "eval_samples_per_second": 189.643,
-      "eval_steps_per_second": 11.952,
+      "eval_accuracy": 0.9296026782600609,
+      "eval_f1": 0.5614724803431023,
+      "eval_loss": 0.31345757842063904,
+      "eval_precision": 0.5901577761081893,
+      "eval_recall": 0.5354464894342195,
+      "eval_runtime": 4.7406,
+      "eval_samples_per_second": 200.819,
+      "eval_steps_per_second": 12.657,
       "step": 690
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.4602757394313812,
+      "grad_norm": 0.2758278548717499,
       "learning_rate": 1.4483091787439615e-05,
-      "loss": 0.0257,
+      "loss": 0.0232,
       "step": 805
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.9257201311231488,
-      "eval_f1": 0.525974025974026,
-      "eval_loss": 0.37688568234443665,
-      "eval_precision": 0.5466048080978491,
-      "eval_recall": 0.5068439577630035,
-      "eval_runtime": 4.7781,
-      "eval_samples_per_second": 199.241,
-      "eval_steps_per_second": 12.557,
+      "eval_accuracy": 0.9299049124683235,
+      "eval_f1": 0.5750452079566004,
+      "eval_loss": 0.3596416115760803,
+      "eval_precision": 0.6124807395993837,
+      "eval_recall": 0.5419222903885481,
+      "eval_runtime": 4.5648,
+      "eval_samples_per_second": 208.554,
+      "eval_steps_per_second": 13.144,
       "step": 805
     },
     {
       "epoch": 8.0,
-      "grad_norm": 1.866155743598938,
+      "grad_norm": 0.687205970287323,
       "learning_rate": 1.3371980676328504e-05,
-      "loss": 0.0223,
+      "loss": 0.0178,
       "step": 920
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.9242089600818357,
-      "eval_f1": 0.5371711780404118,
-      "eval_loss": 0.356906533241272,
-      "eval_precision": 0.523986612123466,
-      "eval_recall": 0.5510363707469691,
-      "eval_runtime": 4.6714,
-      "eval_samples_per_second": 203.793,
-      "eval_steps_per_second": 12.844,
+      "eval_accuracy": 0.9284634877827633,
+      "eval_f1": 0.5786835950551287,
+      "eval_loss": 0.3411566913127899,
+      "eval_precision": 0.5674967234600262,
+      "eval_recall": 0.5903203817314246,
+      "eval_runtime": 4.4171,
+      "eval_samples_per_second": 215.528,
+      "eval_steps_per_second": 13.584,
       "step": 920
     },
     {
       "epoch": 9.0,
-      "grad_norm": 1.9551316499710083,
+      "grad_norm": 0.688730776309967,
       "learning_rate": 1.2260869565217393e-05,
-      "loss": 0.0176,
+      "loss": 0.0181,
       "step": 1035
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.9249529212098668,
-      "eval_f1": 0.549402390438247,
-      "eval_loss": 0.3771890699863434,
-      "eval_precision": 0.559886317498985,
-      "eval_recall": 0.5393038717246773,
-      "eval_runtime": 4.7822,
-      "eval_samples_per_second": 199.072,
-      "eval_steps_per_second": 12.547,
+      "eval_accuracy": 0.92957942947481,
+      "eval_f1": 0.5768194070080863,
+      "eval_loss": 0.36207419633865356,
+      "eval_precision": 0.6100342075256556,
+      "eval_recall": 0.5470347648261759,
+      "eval_runtime": 4.5105,
+      "eval_samples_per_second": 211.064,
+      "eval_steps_per_second": 13.302,
       "step": 1035
     },
     {
       "epoch": 10.0,
-      "grad_norm": 0.21579106152057648,
+      "grad_norm": 2.2689576148986816,
       "learning_rate": 1.114975845410628e-05,
-      "loss": 0.0129,
+      "loss": 0.0132,
       "step": 1150
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.9298119173273196,
-      "eval_f1": 0.5549613784907903,
-      "eval_loss": 0.3583073616027832,
-      "eval_precision": 0.562199036918138,
-      "eval_recall": 0.5479077043410246,
-      "eval_runtime": 4.7887,
-      "eval_samples_per_second": 198.801,
-      "eval_steps_per_second": 12.529,
+      "eval_accuracy": 0.9301838978913352,
+      "eval_f1": 0.5817049974061905,
+      "eval_loss": 0.36382588744163513,
+      "eval_precision": 0.5903825903825903,
+      "eval_recall": 0.5732788002726653,
+      "eval_runtime": 4.5676,
+      "eval_samples_per_second": 208.426,
+      "eval_steps_per_second": 13.136,
       "step": 1150
     },
     {
       "epoch": 11.0,
-      "grad_norm": 0.4364924728870392,
+      "grad_norm": 0.7564047574996948,
       "learning_rate": 1.003864734299517e-05,
-      "loss": 0.0112,
+      "loss": 0.0106,
       "step": 1265
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.9262316043986701,
-      "eval_f1": 0.5348368908578334,
-      "eval_loss": 0.39430442452430725,
-      "eval_precision": 0.5512660855126609,
-      "eval_recall": 0.5193586233867814,
-      "eval_runtime": 4.6852,
-      "eval_samples_per_second": 203.191,
-      "eval_steps_per_second": 12.806,
+      "eval_accuracy": 0.9303466393880919,
+      "eval_f1": 0.5804843304843305,
+      "eval_loss": 0.4133897125720978,
+      "eval_precision": 0.6077554064131245,
+      "eval_recall": 0.5555555555555556,
+      "eval_runtime": 4.4683,
+      "eval_samples_per_second": 213.056,
+      "eval_steps_per_second": 13.428,
       "step": 1265
     },
     {
       "epoch": 12.0,
-      "grad_norm": 1.8533910512924194,
+      "grad_norm": 2.226407051086426,
       "learning_rate": 8.927536231884059e-06,
-      "loss": 0.0096,
+      "loss": 0.0088,
       "step": 1380
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.9285099853532653,
-      "eval_f1": 0.5507480792559645,
-      "eval_loss": 0.4048748314380646,
-      "eval_precision": 0.5701130179991628,
-      "eval_recall": 0.5326554556120454,
-      "eval_runtime": 4.9847,
-      "eval_samples_per_second": 190.986,
-      "eval_steps_per_second": 12.037,
+      "eval_accuracy": 0.9297189221863158,
+      "eval_f1": 0.5792639773531494,
+      "eval_loss": 0.41973167657852173,
+      "eval_precision": 0.6022810890360559,
+      "eval_recall": 0.5579413769597819,
+      "eval_runtime": 4.4945,
+      "eval_samples_per_second": 211.813,
+      "eval_steps_per_second": 13.35,
       "step": 1380
     },
     {
       "epoch": 13.0,
-      "grad_norm": 0.022831812500953674,
+      "grad_norm": 0.0287578534334898,
       "learning_rate": 7.816425120772948e-06,
-      "loss": 0.0078,
+      "loss": 0.0071,
       "step": 1495
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.927138307023458,
-      "eval_f1": 0.5415120140652471,
-      "eval_loss": 0.42278555035591125,
-      "eval_precision": 0.5409836065573771,
-      "eval_recall": 0.5420414548298788,
-      "eval_runtime": 4.7419,
-      "eval_samples_per_second": 200.765,
-      "eval_steps_per_second": 12.653,
+      "eval_accuracy": 0.9301141515355823,
+      "eval_f1": 0.5827538247566064,
+      "eval_loss": 0.41622841358184814,
+      "eval_precision": 0.5947480482611781,
+      "eval_recall": 0.5712338104976141,
+      "eval_runtime": 4.5215,
+      "eval_samples_per_second": 210.547,
+      "eval_steps_per_second": 13.27,
       "step": 1495
     },
     {
       "epoch": 14.0,
-      "grad_norm": 2.755345582962036,
+      "grad_norm": 0.4656817615032196,
       "learning_rate": 6.705314009661836e-06,
-      "loss": 0.0068,
+      "loss": 0.0067,
       "step": 1610
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.9264408434659289,
-      "eval_f1": 0.541335453100159,
-      "eval_loss": 0.41575905680656433,
-      "eval_precision": 0.5503030303030303,
-      "eval_recall": 0.5326554556120454,
-      "eval_runtime": 4.73,
-      "eval_samples_per_second": 201.268,
-      "eval_steps_per_second": 12.685,
+      "eval_accuracy": 0.9305093808848488,
+      "eval_f1": 0.578420864594498,
+      "eval_loss": 0.41959309577941895,
+      "eval_precision": 0.6077327327327328,
+      "eval_recall": 0.5518064076346285,
+      "eval_runtime": 4.5082,
+      "eval_samples_per_second": 211.171,
+      "eval_steps_per_second": 13.309,
       "step": 1610
     },
     {
       "epoch": 15.0,
-      "grad_norm": 0.009739220142364502,
+      "grad_norm": 0.01097877323627472,
       "learning_rate": 5.594202898550725e-06,
-      "loss": 0.0057,
+      "loss": 0.005,
       "step": 1725
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.9306256248111037,
-      "eval_f1": 0.5625126646403241,
-      "eval_loss": 0.41495904326438904,
-      "eval_precision": 0.5836837678721615,
-      "eval_recall": 0.5428236214313649,
-      "eval_runtime": 4.7302,
-      "eval_samples_per_second": 201.262,
-      "eval_steps_per_second": 12.685,
+      "eval_accuracy": 0.929509683119057,
+      "eval_f1": 0.5774378585086042,
+      "eval_loss": 0.43912333250045776,
+      "eval_precision": 0.5892160340546293,
+      "eval_recall": 0.5661213360599864,
+      "eval_runtime": 4.4886,
+      "eval_samples_per_second": 212.091,
+      "eval_steps_per_second": 13.367,
       "step": 1725
     },
     {
       "epoch": 16.0,
-      "grad_norm": 0.20581257343292236,
+      "grad_norm": 0.04738329350948334,
       "learning_rate": 4.483091787439614e-06,
-      "loss": 0.0051,
+      "loss": 0.0047,
       "step": 1840
     },
     {
       "epoch": 16.0,
       "eval_accuracy": 0.929509683119057,
-      "eval_f1": 0.5591397849462366,
-      "eval_loss": 0.427743524312973,
-      "eval_precision": 0.5695740365111562,
-      "eval_recall": 0.5490809542432538,
-      "eval_runtime": 4.7523,
-      "eval_samples_per_second": 200.322,
-      "eval_steps_per_second": 12.625,
+      "eval_f1": 0.5775160228650615,
+      "eval_loss": 0.4424164593219757,
+      "eval_precision": 0.5871785840084537,
+      "eval_recall": 0.5681663258350375,
+      "eval_runtime": 4.5013,
+      "eval_samples_per_second": 211.495,
+      "eval_steps_per_second": 13.329,
       "step": 1840
     },
     {
       "epoch": 17.0,
-      "grad_norm": 1.4635635614395142,
+      "grad_norm": 1.582375168800354,
       "learning_rate": 3.371980676328503e-06,
-      "loss": 0.0048,
+      "loss": 0.0042,
       "step": 1955
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.9296026782600609,
-      "eval_f1": 0.5549841772151899,
-      "eval_loss": 0.4313228130340576,
-      "eval_precision": 0.5614245698279312,
-      "eval_recall": 0.5486898709425108,
-      "eval_runtime": 4.9134,
-      "eval_samples_per_second": 193.757,
-      "eval_steps_per_second": 12.212,
+      "eval_accuracy": 0.9292771952665473,
+      "eval_f1": 0.5731922398589064,
+      "eval_loss": 0.46323156356811523,
+      "eval_precision": 0.5939327485380117,
+      "eval_recall": 0.5538513974096796,
+      "eval_runtime": 4.4828,
+      "eval_samples_per_second": 212.369,
+      "eval_steps_per_second": 13.385,
       "step": 1955
     },
     {
       "epoch": 18.0,
-      "grad_norm": 0.020520634949207306,
+      "grad_norm": 0.008421148173511028,
       "learning_rate": 2.2608695652173913e-06,
-      "loss": 0.004,
+      "loss": 0.0037,
       "step": 2070
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.9297421709715668,
-      "eval_f1": 0.5621019108280255,
-      "eval_loss": 0.4391424357891083,
-      "eval_precision": 0.5723550871503851,
-      "eval_recall": 0.5522096206491983,
-      "eval_runtime": 4.7507,
-      "eval_samples_per_second": 200.39,
-      "eval_steps_per_second": 12.63,
+      "eval_accuracy": 0.9299746588240765,
+      "eval_f1": 0.5827953236782412,
+      "eval_loss": 0.45411384105682373,
+      "eval_precision": 0.5970682874508402,
+      "eval_recall": 0.569188820722563,
+      "eval_runtime": 4.5178,
+      "eval_samples_per_second": 210.72,
+      "eval_steps_per_second": 13.281,
       "step": 2070
     },
     {
       "epoch": 19.0,
-      "grad_norm": 1.6688083410263062,
+      "grad_norm": 0.15860314667224884,
       "learning_rate": 1.1497584541062802e-06,
-      "loss": 0.0036,
+      "loss": 0.0041,
       "step": 2185
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.9293236928370493,
-      "eval_f1": 0.5589764094362255,
-      "eval_loss": 0.44707345962524414,
-      "eval_precision": 0.5717791411042945,
-      "eval_recall": 0.5467344544387954,
-      "eval_runtime": 4.7037,
-      "eval_samples_per_second": 202.393,
-      "eval_steps_per_second": 12.756,
+      "eval_accuracy": 0.9306721223816056,
+      "eval_f1": 0.5841584158415841,
+      "eval_loss": 0.4611005485057831,
+      "eval_precision": 0.6069066862601029,
+      "eval_recall": 0.5630538513974097,
+      "eval_runtime": 4.4985,
+      "eval_samples_per_second": 211.625,
+      "eval_steps_per_second": 13.338,
       "step": 2185
     },
     {
       "epoch": 20.0,
-      "grad_norm": 0.006341221742331982,
+      "grad_norm": 0.006469877902418375,
       "learning_rate": 3.864734299516908e-08,
-      "loss": 0.0035,
+      "loss": 0.0034,
       "step": 2300
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.9296956734010648,
-      "eval_f1": 0.5620888977476579,
-      "eval_loss": 0.4408387243747711,
-      "eval_precision": 0.573170731707317,
-      "eval_recall": 0.5514274540477122,
-      "eval_runtime": 4.7859,
-      "eval_samples_per_second": 198.919,
-      "eval_steps_per_second": 12.537,
+      "eval_accuracy": 0.9300909027503313,
+      "eval_f1": 0.5824868651488616,
+      "eval_loss": 0.4573433995246887,
+      "eval_precision": 0.5990634005763689,
+      "eval_recall": 0.5668029993183368,
+      "eval_runtime": 4.5637,
+      "eval_samples_per_second": 208.603,
+      "eval_steps_per_second": 13.147,
       "step": 2300
     },
     {
       "epoch": 20.0,
       "step": 2300,
       "total_flos": 1090015519875414.0,
-      "train_loss": 0.08582915891771731,
-      "train_runtime": 509.1667,
-      "train_samples_per_second": 72.157,
-      "train_steps_per_second": 4.517
+      "train_loss": 0.08596961084915244,
+      "train_runtime": 505.247,
+      "train_samples_per_second": 72.717,
+      "train_steps_per_second": 4.552
     }
   ],
   "logging_steps": 500,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:58de44b767f87a9d735be0bdb274e63ffb309a02b75583fe7b93e9368959757a
-size 5560
+oid sha256:93af11601083697640d357770973619e84f8d43004b863c6086988760681029f
+size 5688