End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- id
 license: mit
 base_model: indolem/indobert-base-uncased
 tags:

all_results.json

@@ -2,20 +2,20 @@
     "accuracy": 0.9149357072205737,
     "epoch": 20.0,
     "eval_accuracy": 0.9022556390977443,
-    "eval_f1": 0.8793019197207679,
-    "eval_loss": 0.7049560546875,
-    "eval_precision": 0.8893184421534936,
-    "eval_recall": 0.8708401527550463,
-    "eval_runtime": 4.7519,
+    "eval_f1": 0.8799463033398397,
+    "eval_loss": 0.803022563457489,
+    "eval_precision": 0.8874803397294746,
+    "eval_recall": 0.8733406073831607,
+    "eval_runtime": 1.6501,
     "eval_samples": 399,
-    "eval_samples_per_second": 83.966,
-    "eval_steps_per_second": 10.522,
-    "f1": 0.8976980712934288,
-    "precision": 0.8976980712934288,
-    "recall": 0.8976980712934288,
-    "train_loss": 0.05485985055565834,
-    "train_runtime": 2706.9642,
+    "eval_samples_per_second": 241.808,
+    "eval_steps_per_second": 30.302,
+    "f1": 0.8960423740972787,
+    "precision": 0.9029638385430347,
+    "recall": 0.8898853506782006,
+    "train_loss": 0.05501617935226589,
+    "train_runtime": 867.5592,
     "train_samples": 3638,
-    "train_samples_per_second": 26.879,
-    "train_steps_per_second": 0.901
+    "train_samples_per_second": 83.867,
+    "train_steps_per_second": 2.812
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
     "eval_accuracy": 0.9022556390977443,
-    "eval_f1": 0.8793019197207679,
-    "eval_loss": 0.7049560546875,
-    "eval_precision": 0.8893184421534936,
-    "eval_recall": 0.8708401527550463,
-    "eval_runtime": 4.7519,
+    "eval_f1": 0.8799463033398397,
+    "eval_loss": 0.803022563457489,
+    "eval_precision": 0.8874803397294746,
+    "eval_recall": 0.8733406073831607,
+    "eval_runtime": 1.6501,
     "eval_samples": 399,
-    "eval_samples_per_second": 83.966,
-    "eval_steps_per_second": 10.522
+    "eval_samples_per_second": 241.808,
+    "eval_steps_per_second": 30.302
 }

predict_results.json

@@ -1,6 +1,6 @@
 {
     "accuracy": 0.9149357072205737,
-    "f1": 0.8976980712934288,
-    "precision": 0.8976980712934288,
-    "recall": 0.8976980712934288
+    "f1": 0.8960423740972787,
+    "precision": 0.9029638385430347,
+    "recall": 0.8898853506782006
 }

predict_results.txt

@@ -12,19 +12,19 @@ index	prediction
 10	1
 11	1
 12	1
-13	0
+13	1
 14	1
 15	1
-16	0
+16	1
 17	1
 18	0
 19	1
 20	1
-21	1
+21	0
 22	1
 23	1
 24	1
-25	1
+25	0
 26	1
 27	1
 28	1
@@ -60,8 +60,8 @@ index	prediction
 58	1
 59	1
 60	1
-61	1
-62	1
+61	0
+62	0
 63	1
 64	1
 65	1
@@ -90,7 +90,7 @@ index	prediction
 88	1
 89	1
 90	0
-91	1
+91	0
 92	1
 93	1
 94	1
@@ -105,7 +105,7 @@ index	prediction
 103	1
 104	1
 105	0
-106	1
+106	0
 107	1
 108	1
 109	0
@@ -127,9 +127,9 @@ index	prediction
 125	1
 126	1
 127	1
-128	1
-129	1
-130	1
+128	0
+129	0
+130	0
 131	1
 132	1
 133	1
@@ -138,8 +138,8 @@ index	prediction
 136	1
 137	1
 138	1
-139	0
-140	0
+139	1
+140	1
 141	1
 142	1
 143	1
@@ -167,7 +167,7 @@ index	prediction
 165	1
 166	1
 167	1
-168	1
+168	0
 169	1
 170	1
 171	1
@@ -186,13 +186,13 @@ index	prediction
 184	1
 185	1
 186	1
-187	0
+187	1
 188	1
 189	0
-190	1
+190	0
 191	1
 192	1
-193	1
+193	0
 194	1
 195	1
 196	1
@@ -217,7 +217,7 @@ index	prediction
 215	1
 216	1
 217	0
-218	1
+218	0
 219	1
 220	1
 221	1
@@ -247,7 +247,7 @@ index	prediction
 245	1
 246	1
 247	1
-248	0
+248	1
 249	1
 250	1
 251	1
@@ -270,27 +270,27 @@ index	prediction
 268	1
 269	1
 270	1
-271	0
+271	1
 272	1
 273	1
 274	1
 275	1
-276	1
+276	0
 277	1
 278	1
 279	1
-280	1
+280	0
 281	1
 282	0
 283	1
-284	1
+284	0
 285	1
 286	1
 287	1
 288	1
 289	0
 290	1
-291	0
+291	1
 292	1
 293	1
 294	1
@@ -349,14 +349,14 @@ index	prediction
 347	0
 348	0
 349	0
-350	0
+350	1
 351	0
 352	0
 353	0
 354	0
 355	0
 356	0
-357	0
+357	1
 358	0
 359	0
 360	0
@@ -365,7 +365,7 @@ index	prediction
 363	0
 364	0
 365	0
-366	1
+366	0
 367	0
 368	0
 369	0
@@ -377,7 +377,7 @@ index	prediction
 375	0
 376	0
 377	0
-378	1
+378	0
 379	0
 380	0
 381	0
@@ -407,7 +407,7 @@ index	prediction
 405	0
 406	0
 407	0
-408	1
+408	0
 409	0
 410	0
 411	0
@@ -419,7 +419,7 @@ index	prediction
 417	0
 418	0
 419	0
-420	0
+420	1
 421	0
 422	0
 423	0
@@ -441,7 +441,7 @@ index	prediction
 439	0
 440	0
 441	0
-442	1
+442	0
 443	0
 444	0
 445	0
@@ -463,7 +463,7 @@ index	prediction
 461	0
 462	0
 463	0
-464	1
+464	0
 465	0
 466	0
 467	0
@@ -472,7 +472,7 @@ index	prediction
 470	0
 471	0
 472	0
-473	1
+473	0
 474	1
 475	0
 476	1
@@ -590,13 +590,13 @@ index	prediction
 588	0
 589	0
 590	0
-591	0
+591	1
 592	0
 593	0
 594	0
 595	0
 596	0
-597	0
+597	1
 598	0
 599	0
 600	0
@@ -653,7 +653,7 @@ index	prediction
 651	0
 652	0
 653	0
-654	0
+654	1
 655	0
 656	0
 657	0
@@ -665,7 +665,7 @@ index	prediction
 663	0
 664	0
 665	0
-666	1
+666	0
 667	0
 668	0
 669	0
@@ -700,7 +700,7 @@ index	prediction
 698	0
 699	0
 700	0
-701	1
+701	0
 702	0
 703	0
 704	0
@@ -776,7 +776,7 @@ index	prediction
 774	0
 775	0
 776	0
-777	1
+777	0
 778	0
 779	0
 780	0
@@ -787,7 +787,7 @@ index	prediction
 785	0
 786	0
 787	0
-788	1
+788	0
 789	0
 790	0
 791	0
@@ -797,11 +797,11 @@ index	prediction
 795	1
 796	0
 797	0
-798	1
+798	0
 799	0
 800	0
 801	0
-802	1
+802	0
 803	0
 804	0
 805	0
@@ -827,7 +827,7 @@ index	prediction
 825	0
 826	0
 827	0
-828	1
+828	0
 829	0
 830	0
 831	1
@@ -874,7 +874,7 @@ index	prediction
 872	0
 873	0
 874	0
-875	0
+875	1
 876	0
 877	0
 878	0
@@ -885,7 +885,7 @@ index	prediction
 883	0
 884	0
 885	0
-886	0
+886	1
 887	0
 888	0
 889	0
@@ -911,7 +911,7 @@ index	prediction
 909	0
 910	0
 911	0
-912	1
+912	0
 913	1
 914	0
 915	0
@@ -940,7 +940,7 @@ index	prediction
 938	0
 939	0
 940	0
-941	1
+941	0
 942	0
 943	0
 944	0
@@ -1003,7 +1003,7 @@ index	prediction
 1001	0
 1002	0
 1003	0
-1004	1
+1004	0
 1005	0
 1006	0
 1007	0

runs/Jun03_10-28-14_a358b85c7679/events.out.tfevents.1717411391.a358b85c7679.30286.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7c81ad822ad450bc8ed62964362077ff8695d1b3a2d53cddc0f6d031b25630c
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.05485985055565834,
-    "train_runtime": 2706.9642,
+    "train_loss": 0.05501617935226589,
+    "train_runtime": 867.5592,
     "train_samples": 3638,
-    "train_samples_per_second": 26.879,
-    "train_steps_per_second": 0.901
+    "train_samples_per_second": 83.867,
+    "train_steps_per_second": 2.812
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 10.745368957519531,
+      "grad_norm": 8.300418853759766,
       "learning_rate": 4.75e-05,
-      "loss": 0.3971,
+      "loss": 0.4092,
       "step": 122
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.851257564821498,
-      "eval_loss": 0.27043774724006653,
-      "eval_precision": 0.8714882943143812,
-      "eval_recall": 0.8366521185670122,
-      "eval_runtime": 4.9491,
-      "eval_samples_per_second": 80.621,
-      "eval_steps_per_second": 10.103,
+      "eval_accuracy": 0.8521303258145363,
+      "eval_f1": 0.7892045810686176,
+      "eval_loss": 0.3456897437572479,
+      "eval_precision": 0.8929804104477612,
+      "eval_recall": 0.7553646117475905,
+      "eval_runtime": 1.6637,
+      "eval_samples_per_second": 239.829,
+      "eval_steps_per_second": 30.054,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 7.182548999786377,
+      "grad_norm": 9.286004066467285,
       "learning_rate": 4.5e-05,
-      "loss": 0.214,
+      "loss": 0.2282,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9197994987468672,
-      "eval_f1": 0.9032551372976905,
-      "eval_loss": 0.23814059793949127,
-      "eval_precision": 0.9032551372976905,
-      "eval_recall": 0.9032551372976905,
-      "eval_runtime": 4.9784,
-      "eval_samples_per_second": 80.146,
-      "eval_steps_per_second": 10.043,
+      "eval_accuracy": 0.8922305764411027,
+      "eval_f1": 0.8676331036823873,
+      "eval_loss": 0.258427232503891,
+      "eval_precision": 0.8749292230261088,
+      "eval_recall": 0.8612474995453718,
+      "eval_runtime": 1.6592,
+      "eval_samples_per_second": 240.478,
+      "eval_steps_per_second": 30.135,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 93.81620025634766,
+      "grad_norm": 13.956122398376465,
       "learning_rate": 4.25e-05,
-      "loss": 0.1301,
+      "loss": 0.138,
       "step": 366
     },
     {
       "epoch": 3.0,
       "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8439374185136896,
-      "eval_loss": 0.38320210576057434,
-      "eval_precision": 0.8794955044955045,
-      "eval_recall": 0.822376795781051,
-      "eval_runtime": 5.0811,
-      "eval_samples_per_second": 78.526,
-      "eval_steps_per_second": 9.84,
+      "eval_f1": 0.8429546050905274,
+      "eval_loss": 0.44172462821006775,
+      "eval_precision": 0.882529902138456,
+      "eval_recall": 0.8198763411529368,
+      "eval_runtime": 1.6567,
+      "eval_samples_per_second": 240.84,
+      "eval_steps_per_second": 30.18,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 0.22968685626983643,
+      "grad_norm": 0.09012622386217117,
       "learning_rate": 4e-05,
-      "loss": 0.0904,
+      "loss": 0.0837,
       "step": 488
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8782051282051282,
-      "eval_loss": 0.3873072564601898,
-      "eval_precision": 0.8661616161616161,
-      "eval_recall": 0.8955264593562466,
-      "eval_runtime": 4.9496,
-      "eval_samples_per_second": 80.613,
-      "eval_steps_per_second": 10.102,
+      "eval_accuracy": 0.9022556390977443,
+      "eval_f1": 0.8793019197207679,
+      "eval_loss": 0.4037090241909027,
+      "eval_precision": 0.8893184421534936,
+      "eval_recall": 0.8708401527550463,
+      "eval_runtime": 1.6737,
+      "eval_samples_per_second": 238.388,
+      "eval_steps_per_second": 29.873,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.057639382779598236,
+      "grad_norm": 5.347772121429443,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.0587,
+      "loss": 0.0426,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.9147869674185464,
-      "eval_f1": 0.8956140350877193,
-      "eval_loss": 0.4033690392971039,
-      "eval_precision": 0.9022147147147147,
-      "eval_recall": 0.8897072194944535,
-      "eval_runtime": 4.9468,
-      "eval_samples_per_second": 80.659,
-      "eval_steps_per_second": 10.108,
+      "eval_accuracy": 0.9047619047619048,
+      "eval_f1": 0.8873149414352814,
+      "eval_loss": 0.5462044477462769,
+      "eval_precision": 0.8806277372262774,
+      "eval_recall": 0.8951172940534643,
+      "eval_runtime": 1.6751,
+      "eval_samples_per_second": 238.188,
+      "eval_steps_per_second": 29.848,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 0.12450123578310013,
+      "grad_norm": 0.011684279888868332,
       "learning_rate": 3.5e-05,
-      "loss": 0.0496,
+      "loss": 0.0502,
       "step": 732
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8755702215614461,
-      "eval_loss": 0.5248555541038513,
-      "eval_precision": 0.8632157235517781,
-      "eval_recall": 0.8937534097108566,
-      "eval_runtime": 4.9627,
-      "eval_samples_per_second": 80.399,
-      "eval_steps_per_second": 10.075,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8707140332272888,
+      "eval_loss": 0.5626452565193176,
+      "eval_precision": 0.8618432385874246,
+      "eval_recall": 0.8819785415530097,
+      "eval_runtime": 1.6815,
+      "eval_samples_per_second": 237.294,
+      "eval_steps_per_second": 29.736,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.029339170083403587,
+      "grad_norm": 0.015737071633338928,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.0362,
+      "loss": 0.0242,
       "step": 854
     },
     {
       "epoch": 7.0,
       "eval_accuracy": 0.9072681704260651,
-      "eval_f1": 0.8842272010790718,
-      "eval_loss": 0.533008873462677,
-      "eval_precision": 0.8999348109517601,
-      "eval_recall": 0.8718857974177123,
-      "eval_runtime": 4.956,
-      "eval_samples_per_second": 80.508,
-      "eval_steps_per_second": 10.089,
+      "eval_f1": 0.8848664457009163,
+      "eval_loss": 0.6241247653961182,
+      "eval_precision": 0.8977236138837015,
+      "eval_recall": 0.8743862520458265,
+      "eval_runtime": 1.6765,
+      "eval_samples_per_second": 238.002,
+      "eval_steps_per_second": 29.825,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 0.012251710519194603,
+      "grad_norm": 0.004997015465050936,
       "learning_rate": 3e-05,
-      "loss": 0.0223,
+      "loss": 0.0217,
       "step": 976
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8769602202215754,
-      "eval_loss": 0.6168325543403625,
-      "eval_precision": 0.8742831541218639,
-      "eval_recall": 0.8797963266048372,
-      "eval_runtime": 4.9435,
-      "eval_samples_per_second": 80.712,
-      "eval_steps_per_second": 10.114,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8692251105268142,
+      "eval_loss": 0.7096332907676697,
+      "eval_precision": 0.8579132638693325,
+      "eval_recall": 0.885206401163848,
+      "eval_runtime": 1.6742,
+      "eval_samples_per_second": 238.324,
+      "eval_steps_per_second": 29.865,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 0.027291299775242805,
+      "grad_norm": 0.004219838418066502,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.0191,
+      "loss": 0.0229,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8674628282189181,
-      "eval_loss": 0.7103565335273743,
-      "eval_precision": 0.8869858462356303,
-      "eval_recall": 0.8530187306783051,
-      "eval_runtime": 4.9624,
-      "eval_samples_per_second": 80.405,
-      "eval_steps_per_second": 10.076,
+      "eval_accuracy": 0.9122807017543859,
+      "eval_f1": 0.8954723392788977,
+      "eval_loss": 0.611499547958374,
+      "eval_precision": 0.8909569746108776,
+      "eval_recall": 0.9004364429896345,
+      "eval_runtime": 1.66,
+      "eval_samples_per_second": 240.365,
+      "eval_steps_per_second": 30.121,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 0.26321402192115784,
+      "grad_norm": 0.006534805987030268,
       "learning_rate": 2.5e-05,
-      "loss": 0.0147,
+      "loss": 0.0109,
       "step": 1220
     },
     {
       "epoch": 10.0,
       "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8775533117267087,
-      "eval_loss": 0.6125138998031616,
-      "eval_precision": 0.873246730188791,
-      "eval_recall": 0.8822967812329514,
-      "eval_runtime": 4.95,
-      "eval_samples_per_second": 80.606,
-      "eval_steps_per_second": 10.101,
+      "eval_f1": 0.8744522298370696,
+      "eval_loss": 0.7575166821479797,
+      "eval_precision": 0.8795731707317074,
+      "eval_recall": 0.8697945080923805,
+      "eval_runtime": 1.6781,
+      "eval_samples_per_second": 237.769,
+      "eval_steps_per_second": 29.796,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 0.00528654595836997,
+      "grad_norm": 0.00184684619307518,
       "learning_rate": 2.25e-05,
-      "loss": 0.0121,
+      "loss": 0.0068,
       "step": 1342
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.888343937787204,
-      "eval_loss": 0.6882844567298889,
-      "eval_precision": 0.8790267011197244,
-      "eval_recall": 0.9001182033096926,
-      "eval_runtime": 4.9822,
-      "eval_samples_per_second": 80.085,
-      "eval_steps_per_second": 10.036,
+      "eval_accuracy": 0.9072681704260651,
+      "eval_f1": 0.8861029031685659,
+      "eval_loss": 0.7536790370941162,
+      "eval_precision": 0.8937558980811576,
+      "eval_recall": 0.879387161302055,
+      "eval_runtime": 1.6764,
+      "eval_samples_per_second": 238.006,
+      "eval_steps_per_second": 29.825,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 0.007998102344572544,
+      "grad_norm": 0.01189060416072607,
       "learning_rate": 2e-05,
-      "loss": 0.0118,
+      "loss": 0.0131,
       "step": 1464
     },
     {
       "epoch": 12.0,
       "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8781334505389722,
-      "eval_loss": 0.6760030388832092,
-      "eval_precision": 0.872316715542522,
-      "eval_recall": 0.8847972358610656,
-      "eval_runtime": 4.9668,
-      "eval_samples_per_second": 80.334,
-      "eval_steps_per_second": 10.067,
+      "eval_f1": 0.8775533117267087,
+      "eval_loss": 0.7247006297111511,
+      "eval_precision": 0.873246730188791,
+      "eval_recall": 0.8822967812329514,
+      "eval_runtime": 1.6752,
+      "eval_samples_per_second": 238.178,
+      "eval_steps_per_second": 29.847,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 0.002160990610718727,
+      "grad_norm": 0.0015333497431129217,
       "learning_rate": 1.75e-05,
-      "loss": 0.0034,
+      "loss": 0.0101,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.8851154755410074,
-      "eval_loss": 0.7163070440292358,
-      "eval_precision": 0.8851154755410074,
-      "eval_recall": 0.8851154755410074,
-      "eval_runtime": 4.9658,
-      "eval_samples_per_second": 80.349,
-      "eval_steps_per_second": 10.069,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8763538792940554,
+      "eval_loss": 0.7927835583686829,
+      "eval_precision": 0.8754297605404427,
+      "eval_recall": 0.877295871976723,
+      "eval_runtime": 1.6839,
+      "eval_samples_per_second": 236.944,
+      "eval_steps_per_second": 29.692,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 0.09897086024284363,
+      "grad_norm": 0.0037907068617641926,
       "learning_rate": 1.5e-05,
-      "loss": 0.0064,
+      "loss": 0.0061,
       "step": 1708
     },
     {
       "epoch": 14.0,
       "eval_accuracy": 0.9072681704260651,
       "eval_f1": 0.8884169154604891,
-      "eval_loss": 0.7180221080780029,
+      "eval_loss": 0.784883975982666,
       "eval_precision": 0.8874630556728391,
       "eval_recall": 0.8893889798145117,
-      "eval_runtime": 4.9417,
-      "eval_samples_per_second": 80.741,
-      "eval_steps_per_second": 10.118,
+      "eval_runtime": 1.6745,
+      "eval_samples_per_second": 238.284,
+      "eval_steps_per_second": 29.86,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 0.004853605292737484,
+      "grad_norm": 0.0031544596422463655,
       "learning_rate": 1.25e-05,
-      "loss": 0.008,
+      "loss": 0.0135,
       "step": 1830
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.8807643913180245,
-      "eval_loss": 0.6914934515953064,
-      "eval_precision": 0.8978286538966151,
-      "eval_recall": 0.8676122931442081,
-      "eval_runtime": 4.9615,
-      "eval_samples_per_second": 80.42,
-      "eval_steps_per_second": 10.078,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8731122745782431,
+      "eval_loss": 0.781574010848999,
+      "eval_precision": 0.8829705994654449,
+      "eval_recall": 0.864793598836152,
+      "eval_runtime": 1.6721,
+      "eval_samples_per_second": 238.624,
+      "eval_steps_per_second": 29.903,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 0.0016086915275081992,
+      "grad_norm": 0.0015803646529093385,
       "learning_rate": 1e-05,
-      "loss": 0.0052,
+      "loss": 0.0081,
       "step": 1952
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.9072681704260651,
-      "eval_f1": 0.8884169154604891,
-      "eval_loss": 0.6778000593185425,
-      "eval_precision": 0.8874630556728391,
-      "eval_recall": 0.8893889798145117,
-      "eval_runtime": 4.9627,
-      "eval_samples_per_second": 80.4,
-      "eval_steps_per_second": 10.075,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8757339815412664,
+      "eval_loss": 0.7727145552635193,
+      "eval_precision": 0.8766906299500427,
+      "eval_recall": 0.8747954173486088,
+      "eval_runtime": 1.6771,
+      "eval_samples_per_second": 237.909,
+      "eval_steps_per_second": 29.813,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 0.0021244632080197334,
+      "grad_norm": 0.00223415601067245,
       "learning_rate": 7.5e-06,
-      "loss": 0.0066,
+      "loss": 0.0027,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.9072681704260651,
-      "eval_f1": 0.8884169154604891,
-      "eval_loss": 0.6992811560630798,
-      "eval_precision": 0.8874630556728391,
-      "eval_recall": 0.8893889798145117,
-      "eval_runtime": 4.9649,
-      "eval_samples_per_second": 80.364,
-      "eval_steps_per_second": 10.071,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8763538792940554,
+      "eval_loss": 0.81281578540802,
+      "eval_precision": 0.8754297605404427,
+      "eval_recall": 0.877295871976723,
+      "eval_runtime": 1.6614,
+      "eval_samples_per_second": 240.156,
+      "eval_steps_per_second": 30.095,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 0.0017945035360753536,
+      "grad_norm": 0.0023393542505800724,
       "learning_rate": 5e-06,
-      "loss": 0.0053,
+      "loss": 0.0041,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.8814283033033032,
-      "eval_loss": 0.6966382265090942,
-      "eval_precision": 0.895578231292517,
-      "eval_recall": 0.8701127477723223,
-      "eval_runtime": 4.9631,
-      "eval_samples_per_second": 80.393,
-      "eval_steps_per_second": 10.074,
+      "eval_accuracy": 0.9022556390977443,
+      "eval_f1": 0.8817957385392532,
+      "eval_loss": 0.8081415891647339,
+      "eval_precision": 0.8827677592299257,
+      "eval_recall": 0.8808419712675032,
+      "eval_runtime": 1.6587,
+      "eval_samples_per_second": 240.553,
+      "eval_steps_per_second": 30.145,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 0.0012101498432457447,
+      "grad_norm": 0.0010473760776221752,
       "learning_rate": 2.5e-06,
-      "loss": 0.0022,
+      "loss": 0.0018,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.8765393898137261,
-      "eval_loss": 0.7112175822257996,
-      "eval_precision": 0.8852261942423283,
-      "eval_recall": 0.8690671031096563,
-      "eval_runtime": 4.9613,
-      "eval_samples_per_second": 80.422,
-      "eval_steps_per_second": 10.078,
+      "eval_accuracy": 0.9022556390977443,
+      "eval_f1": 0.8793019197207679,
+      "eval_loss": 0.8038576245307922,
+      "eval_precision": 0.8893184421534936,
+      "eval_recall": 0.8708401527550463,
+      "eval_runtime": 1.6566,
+      "eval_samples_per_second": 240.849,
+      "eval_steps_per_second": 30.182,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 0.009784560650587082,
+      "grad_norm": 0.0015798051608726382,
       "learning_rate": 0.0,
-      "loss": 0.0042,
+      "loss": 0.0025,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8793019197207679,
-      "eval_loss": 0.7049560546875,
-      "eval_precision": 0.8893184421534936,
-      "eval_recall": 0.8708401527550463,
-      "eval_runtime": 4.9243,
-      "eval_samples_per_second": 81.027,
-      "eval_steps_per_second": 10.154,
+      "eval_f1": 0.8799463033398397,
+      "eval_loss": 0.803022563457489,
+      "eval_precision": 0.8874803397294746,
+      "eval_recall": 0.8733406073831607,
+      "eval_runtime": 1.6611,
+      "eval_samples_per_second": 240.198,
+      "eval_steps_per_second": 30.1,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
       "total_flos": 7584162436176000.0,
-      "train_loss": 0.05485985055565834,
-      "train_runtime": 2706.9642,
-      "train_samples_per_second": 26.879,
-      "train_steps_per_second": 0.901
+      "train_loss": 0.05501617935226589,
+      "train_runtime": 867.5592,
+      "train_samples_per_second": 83.867,
+      "train_steps_per_second": 2.812
     }
   ],
   "logging_steps": 500,