updated model for stratified split

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "Models/bio-bert-base-spanish-wwm-uncased",
+  "_name_or_path": "./Models/bio-bert-base-spanish-wwm-uncased",
   "architectures": [
     "BertForSequenceClassification"
   ],
@@ -57,7 +57,7 @@
   "position_embedding_type": "absolute",
   "problem_type": "multi_label_classification",
   "torch_dtype": "float32",
-  "transformers_version": "4.19.2",
+  "transformers_version": "4.24.0",
   "type_vocab_size": 2,
   "use_cache": true,
   "vocab_size": 38468

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0103645d5238bcb778b35c2824a75ea14df52a5d1ca8f63d459afc3d5e85b14b
+oid sha256:ebb137e877bb4f592417f468e35466c05294f419aabe8caca2d8707adfedd6d4
 size 924893085

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:44f222b4ff2590dd569bf9514d0d4dd8041986c6f7f873e806c7331078d28a2d
+oid sha256:7539ebd87020112228b621477123c3a9606ce46f30ea21b470a6c797f221788f
 size 462459309

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c55efde8c43192f9ebfd10429b6dd25e1fb3b68a3f3f8425f98d0b4dddee3a67
-size 14567
+oid sha256:fb9a82346993ff4d4270db79f4c3f93be6a696b9a7ff41e62ddbad8e1099f69f
+size 14439

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:93585dd7fdae2ab5eced8160060697e84df90b6641ec48615080b3af11c23cb6
+oid sha256:a829f1b672b920ee77b0097d3441199318b1e0b1f4de9df50935e3325f55daa6
 size 623

special_tokens_map.json

@@ -1 +1,7 @@
-{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -1 +1,16 @@
-{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": false, "do_basic_tokenize": true, "never_split": null, "model_max_length": 512, "special_tokens_map_file": "Models/bio-bert-base-spanish-wwm-uncased/special_tokens_map.json", "name_or_path": "Models/bio-bert-base-spanish-wwm-uncased", "tokenizer_class": "BertTokenizer"}
+{
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "name_or_path": "Models/bio-bert-base-spanish-wwm-uncased",
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": "Models/bio-bert-base-spanish-wwm-uncased/special_tokens_map.json",
+  "strip_accents": false,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

trainer_state.json

@@ -1,312 +1,396 @@
 {
-  "best_metric": 0.5278386153994516,
-  "best_model_checkpoint": "./CARES/checkpoints/bio-ber/checkpoint-2679",
-  "epoch": 19.0,
-  "global_step": 2679,
-  "is_hyper_param_search": false,
+  "best_metric": 0.6252569868800233,
+  "best_model_checkpoint": "./CARES/checkpoints/bio-ber-stratified/run-1/checkpoint-3408",
+  "epoch": 24.170212765957448,
+  "global_step": 3408,
+  "is_hyper_param_search": true,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.0,
-      "eval_loss": 0.2276119738817215,
-      "eval_macro_f1": 0.09383679038919833,
-      "eval_macro_precision": 0.12681117070705897,
-      "eval_macro_recall": 0.07667770066244797,
-      "eval_micro_f1": 0.36897880539499034,
-      "eval_micro_precision": 0.7053406998158379,
-      "eval_micro_recall": 0.24983692106979777,
-      "eval_runtime": 2.8509,
-      "eval_samples_per_second": 338.842,
-      "eval_steps_per_second": 21.397,
-      "step": 141
+      "epoch": 1.01,
+      "eval_loss": 0.2541882395744324,
+      "eval_macro_f1": 0.026901004304160685,
+      "eval_macro_precision": 0.048828125,
+      "eval_macro_recall": 0.018564356435643563,
+      "eval_micro_f1": 0.17543859649122806,
+      "eval_micro_precision": 0.78125,
+      "eval_micro_recall": 0.09881422924901186,
+      "eval_runtime": 9.8993,
+      "eval_samples_per_second": 98.088,
+      "eval_steps_per_second": 3.132,
+      "step": 142
     },
     {
-      "epoch": 2.0,
-      "eval_loss": 0.19956032931804657,
-      "eval_macro_f1": 0.13389070512548568,
-      "eval_macro_precision": 0.1949068871295339,
-      "eval_macro_recall": 0.11313811988335658,
-      "eval_micro_f1": 0.4958460865763008,
-      "eval_micro_precision": 0.7519893899204244,
-      "eval_micro_recall": 0.3698630136986301,
-      "eval_runtime": 2.8244,
-      "eval_samples_per_second": 342.022,
-      "eval_steps_per_second": 21.598,
-      "step": 282
+      "epoch": 2.01,
+      "eval_loss": 0.22154481709003448,
+      "eval_macro_f1": 0.10623663256202004,
+      "eval_macro_precision": 0.18484581306311026,
+      "eval_macro_recall": 0.09432864376066943,
+      "eval_micro_f1": 0.41890639481000924,
+      "eval_micro_precision": 0.70625,
+      "eval_micro_recall": 0.2977602108036891,
+      "eval_runtime": 9.9434,
+      "eval_samples_per_second": 97.653,
+      "eval_steps_per_second": 3.118,
+      "step": 284
     },
     {
-      "epoch": 3.0,
-      "eval_loss": 0.18670037388801575,
-      "eval_macro_f1": 0.22751166658204755,
-      "eval_macro_precision": 0.46178526182432433,
-      "eval_macro_recall": 0.18375879789138674,
-      "eval_micro_f1": 0.5680033071517155,
-      "eval_micro_precision": 0.7753950338600452,
-      "eval_micro_recall": 0.4481409001956947,
-      "eval_runtime": 2.8618,
-      "eval_samples_per_second": 337.545,
-      "eval_steps_per_second": 21.315,
-      "step": 423
+      "epoch": 3.02,
+      "eval_loss": 0.18873167037963867,
+      "eval_macro_f1": 0.20450440261358283,
+      "eval_macro_precision": 0.3100231872943203,
+      "eval_macro_recall": 0.1738961275122085,
+      "eval_micro_f1": 0.568944099378882,
+      "eval_micro_precision": 0.7658862876254181,
+      "eval_micro_recall": 0.4525691699604743,
+      "eval_runtime": 9.9071,
+      "eval_samples_per_second": 98.011,
+      "eval_steps_per_second": 3.129,
+      "step": 426
     },
     {
       "epoch": 3.55,
-      "learning_rate": 4.822695035460993e-05,
-      "loss": 0.2081,
+      "learning_rate": 4.1013980584550545e-05,
+      "loss": 0.2716,
       "step": 500
     },
     {
-      "epoch": 4.0,
-      "eval_loss": 0.18518751859664917,
-      "eval_macro_f1": 0.34478852490640566,
-      "eval_macro_precision": 0.5134144581169723,
-      "eval_macro_recall": 0.2886504191522892,
-      "eval_micro_f1": 0.6130150173276858,
-      "eval_micro_precision": 0.7481203007518797,
-      "eval_micro_recall": 0.5192433137638617,
-      "eval_runtime": 2.8317,
-      "eval_samples_per_second": 341.135,
-      "eval_steps_per_second": 21.542,
-      "step": 564
+      "epoch": 4.03,
+      "eval_loss": 0.17541566491127014,
+      "eval_macro_f1": 0.2919005682162893,
+      "eval_macro_precision": 0.3812561939105582,
+      "eval_macro_recall": 0.25249891158193927,
+      "eval_micro_f1": 0.6301369863013698,
+      "eval_micro_precision": 0.7762777242044359,
+      "eval_micro_recall": 0.5303030303030303,
+      "eval_runtime": 9.944,
+      "eval_samples_per_second": 97.647,
+      "eval_steps_per_second": 3.117,
+      "step": 568
     },
     {
-      "epoch": 5.0,
-      "eval_loss": 0.18300610780715942,
-      "eval_macro_f1": 0.4238271955718499,
-      "eval_macro_precision": 0.6346282008083806,
-      "eval_macro_recall": 0.36978265721033565,
-      "eval_micro_f1": 0.6232315711094564,
-      "eval_micro_precision": 0.7259323503902863,
-      "eval_micro_recall": 0.5459882583170255,
-      "eval_runtime": 2.8313,
-      "eval_samples_per_second": 341.185,
-      "eval_steps_per_second": 21.545,
-      "step": 705
+      "epoch": 5.04,
+      "eval_loss": 0.1692640334367752,
+      "eval_macro_f1": 0.4143233067428529,
+      "eval_macro_precision": 0.5073151086851961,
+      "eval_macro_recall": 0.3678263076835473,
+      "eval_micro_f1": 0.6415811478525276,
+      "eval_micro_precision": 0.7583108715184187,
+      "eval_micro_recall": 0.5559947299077734,
+      "eval_runtime": 9.9093,
+      "eval_samples_per_second": 97.989,
+      "eval_steps_per_second": 3.128,
+      "step": 710
     },
     {
-      "epoch": 6.0,
-      "eval_loss": 0.18070942163467407,
-      "eval_macro_f1": 0.427572174915214,
-      "eval_macro_precision": 0.5867883838688037,
-      "eval_macro_recall": 0.3661335629128818,
-      "eval_micro_f1": 0.6416332482683194,
-      "eval_micro_precision": 0.7272727272727273,
-      "eval_micro_recall": 0.5740378343118069,
-      "eval_runtime": 2.822,
-      "eval_samples_per_second": 342.307,
-      "eval_steps_per_second": 21.616,
-      "step": 846
+      "epoch": 6.04,
+      "eval_loss": 0.1651107370853424,
+      "eval_macro_f1": 0.4093940822021235,
+      "eval_macro_precision": 0.5308122229847152,
+      "eval_macro_recall": 0.358915994433547,
+      "eval_micro_f1": 0.6615737203972498,
+      "eval_micro_precision": 0.7872727272727272,
+      "eval_micro_recall": 0.5704874835309618,
+      "eval_runtime": 9.8986,
+      "eval_samples_per_second": 98.095,
+      "eval_steps_per_second": 3.132,
+      "step": 852
     },
     {
-      "epoch": 7.0,
-      "eval_loss": 0.18207068741321564,
-      "eval_macro_f1": 0.47285593173212287,
-      "eval_macro_precision": 0.5718512640389848,
-      "eval_macro_recall": 0.4270343269721778,
-      "eval_micro_f1": 0.6605374823196605,
-      "eval_micro_precision": 0.7212355212355213,
-      "eval_micro_recall": 0.6092628832354859,
-      "eval_runtime": 2.8204,
-      "eval_samples_per_second": 342.507,
-      "eval_steps_per_second": 21.628,
-      "step": 987
+      "epoch": 7.05,
+      "eval_loss": 0.1729801893234253,
+      "eval_macro_f1": 0.42556225648133283,
+      "eval_macro_precision": 0.58718674661248,
+      "eval_macro_recall": 0.3723531112715551,
+      "eval_micro_f1": 0.6549062844542448,
+      "eval_micro_precision": 0.7406483790523691,
+      "eval_micro_recall": 0.5869565217391305,
+      "eval_runtime": 9.912,
+      "eval_samples_per_second": 97.962,
+      "eval_steps_per_second": 3.128,
+      "step": 994
     },
     {
       "epoch": 7.09,
-      "learning_rate": 4.645390070921986e-05,
-      "loss": 0.115,
+      "learning_rate": 3.950611365129501e-05,
+      "loss": 0.1249,
       "step": 1000
     },
     {
-      "epoch": 8.0,
-      "eval_loss": 0.18614797294139862,
-      "eval_macro_f1": 0.48637001730423257,
-      "eval_macro_precision": 0.6057345339339811,
-      "eval_macro_recall": 0.4271528815196963,
-      "eval_micro_f1": 0.6618128867855844,
-      "eval_micro_precision": 0.7487644151565074,
-      "eval_micro_recall": 0.5929549902152642,
-      "eval_runtime": 2.8241,
-      "eval_samples_per_second": 342.057,
-      "eval_steps_per_second": 21.6,
-      "step": 1128
+      "epoch": 8.06,
+      "eval_loss": 0.17469000816345215,
+      "eval_macro_f1": 0.4273735214628892,
+      "eval_macro_precision": 0.6238596091113585,
+      "eval_macro_recall": 0.375472703111614,
+      "eval_micro_f1": 0.6517101875689592,
+      "eval_micro_precision": 0.7377185678601166,
+      "eval_micro_recall": 0.5836627140974967,
+      "eval_runtime": 9.8904,
+      "eval_samples_per_second": 98.176,
+      "eval_steps_per_second": 3.134,
+      "step": 1136
     },
     {
-      "epoch": 9.0,
-      "eval_loss": 0.18139992654323578,
-      "eval_macro_f1": 0.4968758987364566,
-      "eval_macro_precision": 0.6188063911104451,
-      "eval_macro_recall": 0.4425857583556203,
-      "eval_micro_f1": 0.6716259298618491,
-      "eval_micro_precision": 0.7348837209302326,
-      "eval_micro_recall": 0.6183953033268101,
-      "eval_runtime": 2.8195,
-      "eval_samples_per_second": 342.612,
-      "eval_steps_per_second": 21.635,
-      "step": 1269
+      "epoch": 9.06,
+      "eval_loss": 0.17877863347530365,
+      "eval_macro_f1": 0.475201225203798,
+      "eval_macro_precision": 0.5739651831374122,
+      "eval_macro_recall": 0.4350214360169413,
+      "eval_micro_f1": 0.6738227146814404,
+      "eval_micro_precision": 0.7102189781021898,
+      "eval_micro_recall": 0.6409749670619236,
+      "eval_runtime": 9.9151,
+      "eval_samples_per_second": 97.932,
+      "eval_steps_per_second": 3.127,
+      "step": 1278
     },
     {
-      "epoch": 10.0,
-      "eval_loss": 0.19238035380840302,
-      "eval_macro_f1": 0.4762674411195111,
-      "eval_macro_precision": 0.5687316855202853,
-      "eval_macro_recall": 0.42920385227971614,
-      "eval_micro_f1": 0.6690066690066689,
-      "eval_micro_precision": 0.7241641337386018,
-      "eval_micro_recall": 0.6216568819308546,
-      "eval_runtime": 2.8152,
-      "eval_samples_per_second": 343.137,
-      "eval_steps_per_second": 21.668,
-      "step": 1410
+      "epoch": 10.07,
+      "eval_loss": 0.17759235203266144,
+      "eval_macro_f1": 0.47628022362120415,
+      "eval_macro_precision": 0.6058635009097308,
+      "eval_macro_recall": 0.42583522254336503,
+      "eval_micro_f1": 0.6828591256072173,
+      "eval_micro_precision": 0.7214076246334311,
+      "eval_micro_recall": 0.6482213438735178,
+      "eval_runtime": 9.9061,
+      "eval_samples_per_second": 98.021,
+      "eval_steps_per_second": 3.129,
+      "step": 1420
     },
     {
       "epoch": 10.64,
-      "learning_rate": 4.468085106382979e-05,
-      "loss": 0.0668,
+      "learning_rate": 3.7998246718039476e-05,
+      "loss": 0.0762,
       "step": 1500
     },
     {
-      "epoch": 11.0,
-      "eval_loss": 0.20003741979599,
-      "eval_macro_f1": 0.5041732295700452,
-      "eval_macro_precision": 0.5819236883447072,
-      "eval_macro_recall": 0.458955350441925,
-      "eval_micro_f1": 0.6811645036829183,
-      "eval_micro_precision": 0.7367223065250379,
-      "eval_micro_recall": 0.6333985649054142,
-      "eval_runtime": 2.8186,
-      "eval_samples_per_second": 342.72,
-      "eval_steps_per_second": 21.642,
-      "step": 1551
+      "epoch": 11.08,
+      "eval_loss": 0.1776473969221115,
+      "eval_macro_f1": 0.5089905336706273,
+      "eval_macro_precision": 0.591647351441912,
+      "eval_macro_recall": 0.47047423582847886,
+      "eval_micro_f1": 0.688728024819028,
+      "eval_micro_precision": 0.7223427331887202,
+      "eval_micro_recall": 0.658102766798419,
+      "eval_runtime": 9.9428,
+      "eval_samples_per_second": 97.658,
+      "eval_steps_per_second": 3.118,
+      "step": 1562
     },
     {
-      "epoch": 12.0,
-      "eval_loss": 0.21071653068065643,
-      "eval_macro_f1": 0.5060970351192071,
-      "eval_macro_precision": 0.6216233664621571,
-      "eval_macro_recall": 0.46859394229849854,
-      "eval_micro_f1": 0.6627946703108984,
-      "eval_micro_precision": 0.6958393113342898,
-      "eval_micro_recall": 0.6327462491846053,
-      "eval_runtime": 2.817,
-      "eval_samples_per_second": 342.914,
-      "eval_steps_per_second": 21.654,
-      "step": 1692
+      "epoch": 12.09,
+      "eval_loss": 0.18189238011837006,
+      "eval_macro_f1": 0.48004944819202805,
+      "eval_macro_precision": 0.6458220963071541,
+      "eval_macro_recall": 0.41943684599646036,
+      "eval_micro_f1": 0.6883162725026473,
+      "eval_micro_precision": 0.7414448669201521,
+      "eval_micro_recall": 0.642292490118577,
+      "eval_runtime": 9.9058,
+      "eval_samples_per_second": 98.024,
+      "eval_steps_per_second": 3.129,
+      "step": 1704
     },
     {
-      "epoch": 13.0,
-      "eval_loss": 0.20960789918899536,
-      "eval_macro_f1": 0.5055202034151333,
-      "eval_macro_precision": 0.6090230781362929,
-      "eval_macro_recall": 0.450320346412875,
-      "eval_micro_f1": 0.6799041424169804,
-      "eval_micro_precision": 0.7154178674351584,
-      "eval_micro_recall": 0.6477495107632094,
-      "eval_runtime": 4.7452,
-      "eval_samples_per_second": 203.576,
-      "eval_steps_per_second": 12.855,
-      "step": 1833
+      "epoch": 13.09,
+      "eval_loss": 0.19002576172351837,
+      "eval_macro_f1": 0.5199688644954614,
+      "eval_macro_precision": 0.5810572518525183,
+      "eval_macro_recall": 0.49026284720966035,
+      "eval_micro_f1": 0.6896090878717006,
+      "eval_micro_precision": 0.6996610169491525,
+      "eval_micro_recall": 0.6798418972332015,
+      "eval_runtime": 9.9018,
+      "eval_samples_per_second": 98.063,
+      "eval_steps_per_second": 3.131,
+      "step": 1846
     },
     {
-      "epoch": 14.0,
-      "eval_loss": 0.21599768102169037,
-      "eval_macro_f1": 0.5190622646217878,
-      "eval_macro_precision": 0.6551140465404754,
-      "eval_macro_recall": 0.47196058443148803,
-      "eval_micro_f1": 0.6803764377831997,
-      "eval_micro_precision": 0.7305389221556886,
-      "eval_micro_recall": 0.6366601435094585,
-      "eval_runtime": 2.819,
-      "eval_samples_per_second": 342.677,
-      "eval_steps_per_second": 21.639,
-      "step": 1974
+      "epoch": 14.1,
+      "eval_loss": 0.19194385409355164,
+      "eval_macro_f1": 0.520691157744967,
+      "eval_macro_precision": 0.5966466597800141,
+      "eval_macro_recall": 0.47615158217927345,
+      "eval_micro_f1": 0.6891228070175438,
+      "eval_micro_precision": 0.7372372372372372,
+      "eval_micro_recall": 0.6469038208168643,
+      "eval_runtime": 9.9153,
+      "eval_samples_per_second": 97.93,
+      "eval_steps_per_second": 3.126,
+      "step": 1988
     },
     {
       "epoch": 14.18,
-      "learning_rate": 4.2907801418439716e-05,
-      "loss": 0.0365,
+      "learning_rate": 3.649037978478394e-05,
+      "loss": 0.0449,
       "step": 2000
     },
     {
-      "epoch": 15.0,
-      "eval_loss": 0.21732226014137268,
-      "eval_macro_f1": 0.4982722788861393,
-      "eval_macro_precision": 0.5812880908852627,
-      "eval_macro_recall": 0.46102462821831114,
-      "eval_micro_f1": 0.6856754921928038,
-      "eval_micro_precision": 0.7147912243453645,
-      "eval_micro_recall": 0.6588388780169602,
-      "eval_runtime": 2.837,
-      "eval_samples_per_second": 340.498,
-      "eval_steps_per_second": 21.501,
-      "step": 2115
+      "epoch": 15.11,
+      "eval_loss": 0.192080557346344,
+      "eval_macro_f1": 0.5605505080291033,
+      "eval_macro_precision": 0.6623700312301602,
+      "eval_macro_recall": 0.5087278128204601,
+      "eval_micro_f1": 0.6957736639888229,
+      "eval_micro_precision": 0.7405204460966542,
+      "eval_micro_recall": 0.6561264822134387,
+      "eval_runtime": 9.9139,
+      "eval_samples_per_second": 97.943,
+      "eval_steps_per_second": 3.127,
+      "step": 2130
     },
     {
-      "epoch": 16.0,
-      "eval_loss": 0.23044872283935547,
-      "eval_macro_f1": 0.5067733640623313,
-      "eval_macro_precision": 0.6404358198207177,
-      "eval_macro_recall": 0.46429191921900426,
-      "eval_micro_f1": 0.6748382703438883,
-      "eval_micro_precision": 0.7058404558404558,
-      "eval_micro_recall": 0.6464448793215917,
-      "eval_runtime": 2.8419,
-      "eval_samples_per_second": 339.909,
-      "eval_steps_per_second": 21.464,
-      "step": 2256
+      "epoch": 16.11,
+      "eval_loss": 0.1960616409778595,
+      "eval_macro_f1": 0.53306733167955,
+      "eval_macro_precision": 0.5965001445588101,
+      "eval_macro_recall": 0.49021955336676715,
+      "eval_micro_f1": 0.698961937716263,
+      "eval_micro_precision": 0.7361516034985423,
+      "eval_micro_recall": 0.6653491436100132,
+      "eval_runtime": 9.942,
+      "eval_samples_per_second": 97.667,
+      "eval_steps_per_second": 3.118,
+      "step": 2272
     },
     {
-      "epoch": 17.0,
-      "eval_loss": 0.23040802776813507,
-      "eval_macro_f1": 0.5182045437347171,
-      "eval_macro_precision": 0.6102649650627312,
-      "eval_macro_recall": 0.474946805831835,
-      "eval_micro_f1": 0.6864349011588274,
-      "eval_micro_precision": 0.7187723054960742,
-      "eval_micro_recall": 0.6568819308545336,
-      "eval_runtime": 2.8516,
-      "eval_samples_per_second": 338.756,
-      "eval_steps_per_second": 21.391,
-      "step": 2397
+      "epoch": 17.12,
+      "eval_loss": 0.20587308704853058,
+      "eval_macro_f1": 0.5199165971603507,
+      "eval_macro_precision": 0.5554999099415838,
+      "eval_macro_recall": 0.501101816922742,
+      "eval_micro_f1": 0.6840148698884758,
+      "eval_micro_precision": 0.7022900763358778,
+      "eval_micro_recall": 0.6666666666666666,
+      "eval_runtime": 9.9026,
+      "eval_samples_per_second": 98.055,
+      "eval_steps_per_second": 3.131,
+      "step": 2414
     },
     {
       "epoch": 17.73,
-      "learning_rate": 4.1134751773049644e-05,
-      "loss": 0.021,
+      "learning_rate": 3.498251285152841e-05,
+      "loss": 0.0256,
       "step": 2500
     },
     {
-      "epoch": 18.0,
-      "eval_loss": 0.23820702731609344,
-      "eval_macro_f1": 0.5109008179961498,
-      "eval_macro_precision": 0.6241805280991644,
-      "eval_macro_recall": 0.4711108204223084,
-      "eval_micro_f1": 0.6758481692979509,
-      "eval_micro_precision": 0.6966759002770083,
-      "eval_micro_recall": 0.6562296151337247,
-      "eval_runtime": 2.8209,
-      "eval_samples_per_second": 342.448,
-      "eval_steps_per_second": 21.625,
-      "step": 2538
+      "epoch": 18.13,
+      "eval_loss": 0.20113714039325714,
+      "eval_macro_f1": 0.529618760993627,
+      "eval_macro_precision": 0.6032995030406657,
+      "eval_macro_recall": 0.4871012589073676,
+      "eval_micro_f1": 0.6982167352537724,
+      "eval_micro_precision": 0.7281831187410587,
+      "eval_micro_recall": 0.6706192358366272,
+      "eval_runtime": 9.899,
+      "eval_samples_per_second": 98.091,
+      "eval_steps_per_second": 3.132,
+      "step": 2556
     },
     {
-      "epoch": 19.0,
-      "eval_loss": 0.23519271612167358,
-      "eval_macro_f1": 0.5278386153994516,
-      "eval_macro_precision": 0.6066088954753265,
-      "eval_macro_recall": 0.48873540160049156,
-      "eval_micro_f1": 0.6875645883568722,
-      "eval_micro_precision": 0.7284671532846715,
-      "eval_micro_recall": 0.6510110893672537,
-      "eval_runtime": 2.8218,
-      "eval_samples_per_second": 342.33,
-      "eval_steps_per_second": 21.617,
-      "step": 2679
+      "epoch": 19.13,
+      "eval_loss": 0.20477713644504547,
+      "eval_macro_f1": 0.5176514067210329,
+      "eval_macro_precision": 0.5857117797598407,
+      "eval_macro_recall": 0.47641342794341435,
+      "eval_micro_f1": 0.6941015089163237,
+      "eval_micro_precision": 0.7238912732474965,
+      "eval_micro_recall": 0.6666666666666666,
+      "eval_runtime": 9.9038,
+      "eval_samples_per_second": 98.043,
+      "eval_steps_per_second": 3.13,
+      "step": 2698
+    },
+    {
+      "epoch": 20.14,
+      "eval_loss": 0.2074529379606247,
+      "eval_macro_f1": 0.5704867306049771,
+      "eval_macro_precision": 0.6916588415347732,
+      "eval_macro_recall": 0.5189916858370269,
+      "eval_micro_f1": 0.7134187457855697,
+      "eval_micro_precision": 0.7306629834254144,
+      "eval_micro_recall": 0.696969696969697,
+      "eval_runtime": 9.8989,
+      "eval_samples_per_second": 98.092,
+      "eval_steps_per_second": 3.132,
+      "step": 2840
+    },
+    {
+      "epoch": 21.15,
+      "eval_loss": 0.21805770695209503,
+      "eval_macro_f1": 0.5306833888134577,
+      "eval_macro_precision": 0.6107555764603518,
+      "eval_macro_recall": 0.4823641731674541,
+      "eval_micro_f1": 0.709366391184573,
+      "eval_micro_precision": 0.7431457431457431,
+      "eval_micro_recall": 0.6785243741765481,
+      "eval_runtime": 9.95,
+      "eval_samples_per_second": 97.588,
+      "eval_steps_per_second": 3.116,
+      "step": 2982
+    },
+    {
+      "epoch": 21.28,
+      "learning_rate": 3.347464591827287e-05,
+      "loss": 0.015,
+      "step": 3000
+    },
+    {
+      "epoch": 22.16,
+      "eval_loss": 0.21425750851631165,
+      "eval_macro_f1": 0.5900896049733363,
+      "eval_macro_precision": 0.7164448328913706,
+      "eval_macro_recall": 0.5339604602685475,
+      "eval_micro_f1": 0.7065292096219932,
+      "eval_micro_precision": 0.7385057471264368,
+      "eval_micro_recall": 0.6772068511198946,
+      "eval_runtime": 9.9181,
+      "eval_samples_per_second": 97.901,
+      "eval_steps_per_second": 3.126,
+      "step": 3124
+    },
+    {
+      "epoch": 23.16,
+      "eval_loss": 0.21841417253017426,
+      "eval_macro_f1": 0.5731327712532719,
+      "eval_macro_precision": 0.6766904212874656,
+      "eval_macro_recall": 0.5216860842370524,
+      "eval_micro_f1": 0.7131730443616662,
+      "eval_micro_precision": 0.7337979094076655,
+      "eval_micro_recall": 0.6936758893280632,
+      "eval_runtime": 9.8943,
+      "eval_samples_per_second": 98.137,
+      "eval_steps_per_second": 3.133,
+      "step": 3266
+    },
+    {
+      "epoch": 24.17,
+      "eval_loss": 0.22420497238636017,
+      "eval_macro_f1": 0.6252569868800233,
+      "eval_macro_precision": 0.7327575817423139,
+      "eval_macro_recall": 0.5742398722065618,
+      "eval_micro_f1": 0.7081471295978001,
+      "eval_micro_precision": 0.7404744787922358,
+      "eval_micro_recall": 0.6785243741765481,
+      "eval_runtime": 9.961,
+      "eval_samples_per_second": 97.48,
+      "eval_steps_per_second": 3.112,
+      "step": 3408
     }
   ],
   "max_steps": 14100,
   "num_train_epochs": 100,
-  "total_flos": 1.0688023305003456e+16,
+  "total_flos": 1.1986190383959552e+16,
   "trial_name": null,
-  "trial_params": null
+  "trial_params": {
+    "adam_epsilon": 1.241521755885265e-07,
+    "learning_rate": 4.181013432530947e-05,
+    "per_device_eval_batch_size": 32,
+    "per_device_train_batch_size": 16,
+    "seed": 321,
+    "warmup_steps": 236,
+    "weight_decay": 3.190500833235664e-11
+  }
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d2a1147ed9cfb8d56c0709711a5e7f2cc6001c6110bc1058d46761f68db1f8cd
-size 3183
+oid sha256:67f59b494d3ca6dddabd8ddbd8a1784f33689fc7327a434b7a6651a9dfcc91ab
+size 3439