End of training

README.md

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

all_results.json

@@ -1,21 +1,21 @@
 {
-    "accuracy": 0.9030662710187932,
+    "accuracy": 0.904055390702275,
     "epoch": 20.0,
-    "eval_accuracy": 0.8897243107769424,
-    "eval_f1": 0.8649122807017544,
-    "eval_loss": 0.2823803424835205,
-    "eval_precision": 0.8710116366366366,
-    "eval_recall": 0.8594744498999818,
-    "eval_runtime": 4.9357,
+    "eval_accuracy": 0.8822055137844611,
+    "eval_f1": 0.858259325044405,
+    "eval_loss": 0.3389217257499695,
+    "eval_precision": 0.8573798178418481,
+    "eval_recall": 0.8591562102200401,
+    "eval_runtime": 1.811,
     "eval_samples": 399,
-    "eval_samples_per_second": 80.84,
-    "eval_steps_per_second": 10.13,
-    "f1": 0.8843184599146265,
-    "precision": 0.8814837991934362,
-    "recall": 0.8873297438745447,
-    "train_loss": 0.23294218016452475,
-    "train_runtime": 1905.4788,
+    "eval_samples_per_second": 220.315,
+    "eval_steps_per_second": 27.608,
+    "f1": 0.8862491460015474,
+    "precision": 0.881173503483252,
+    "recall": 0.8919373664542485,
+    "train_loss": 0.22970127551282038,
+    "train_runtime": 621.3103,
     "train_samples": 3638,
-    "train_samples_per_second": 38.185,
-    "train_steps_per_second": 1.281
+    "train_samples_per_second": 117.107,
+    "train_steps_per_second": 3.927
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.8897243107769424,
-    "eval_f1": 0.8649122807017544,
-    "eval_loss": 0.2823803424835205,
-    "eval_precision": 0.8710116366366366,
-    "eval_recall": 0.8594744498999818,
-    "eval_runtime": 4.9357,
+    "eval_accuracy": 0.8822055137844611,
+    "eval_f1": 0.858259325044405,
+    "eval_loss": 0.3389217257499695,
+    "eval_precision": 0.8573798178418481,
+    "eval_recall": 0.8591562102200401,
+    "eval_runtime": 1.811,
     "eval_samples": 399,
-    "eval_samples_per_second": 80.84,
-    "eval_steps_per_second": 10.13
+    "eval_samples_per_second": 220.315,
+    "eval_steps_per_second": 27.608
 }

predict_results.json

@@ -1,6 +1,6 @@
 {
-    "accuracy": 0.9030662710187932,
-    "f1": 0.8843184599146265,
-    "precision": 0.8814837991934362,
-    "recall": 0.8873297438745447
+    "accuracy": 0.904055390702275,
+    "f1": 0.8862491460015474,
+    "precision": 0.881173503483252,
+    "recall": 0.8919373664542485
 }

predict_results.txt

@@ -13,14 +13,14 @@ index	prediction
 11	1
 12	1
 13	0
-14	0
+14	1
 15	1
 16	1
 17	1
 18	1
 19	1
 20	1
-21	0
+21	1
 22	1
 23	1
 24	1
@@ -30,13 +30,13 @@ index	prediction
 28	1
 29	1
 30	1
-31	0
+31	1
 32	1
 33	1
 34	1
 35	1
 36	1
-37	0
+37	1
 38	1
 39	1
 40	1
@@ -58,7 +58,7 @@ index	prediction
 56	1
 57	1
 58	1
-59	0
+59	1
 60	1
 61	1
 62	0
@@ -75,7 +75,7 @@ index	prediction
 73	1
 74	1
 75	1
-76	0
+76	1
 77	1
 78	0
 79	1
@@ -94,7 +94,7 @@ index	prediction
 92	1
 93	1
 94	1
-95	1
+95	0
 96	1
 97	1
 98	1
@@ -105,10 +105,10 @@ index	prediction
 103	1
 104	1
 105	0
-106	1
+106	0
 107	1
 108	1
-109	1
+109	0
 110	1
 111	1
 112	1
@@ -116,7 +116,7 @@ index	prediction
 114	1
 115	1
 116	1
-117	1
+117	0
 118	1
 119	1
 120	1
@@ -128,7 +128,7 @@ index	prediction
 126	1
 127	1
 128	1
-129	0
+129	1
 130	1
 131	1
 132	1
@@ -144,16 +144,16 @@ index	prediction
 142	1
 143	1
 144	1
-145	1
+145	0
 146	1
 147	1
 148	1
 149	1
 150	1
-151	1
+151	0
 152	1
 153	1
-154	0
+154	1
 155	1
 156	0
 157	1
@@ -162,7 +162,7 @@ index	prediction
 160	1
 161	0
 162	1
-163	1
+163	0
 164	1
 165	1
 166	0
@@ -186,10 +186,10 @@ index	prediction
 184	1
 185	1
 186	1
-187	1
+187	0
 188	1
 189	0
-190	1
+190	0
 191	1
 192	1
 193	0
@@ -217,7 +217,7 @@ index	prediction
 215	0
 216	1
 217	0
-218	0
+218	1
 219	1
 220	1
 221	1
@@ -232,7 +232,7 @@ index	prediction
 230	0
 231	1
 232	1
-233	0
+233	1
 234	1
 235	1
 236	1
@@ -272,7 +272,7 @@ index	prediction
 270	1
 271	0
 272	1
-273	0
+273	1
 274	1
 275	1
 276	1
@@ -286,9 +286,9 @@ index	prediction
 284	1
 285	1
 286	1
-287	0
+287	1
 288	1
-289	0
+289	1
 290	1
 291	1
 292	1
@@ -302,7 +302,7 @@ index	prediction
 300	0
 301	0
 302	0
-303	0
+303	1
 304	0
 305	0
 306	0
@@ -312,7 +312,7 @@ index	prediction
 310	0
 311	0
 312	0
-313	1
+313	0
 314	0
 315	0
 316	0
@@ -338,10 +338,10 @@ index	prediction
 336	0
 337	0
 338	0
-339	1
+339	0
 340	0
 341	0
-342	1
+342	0
 343	0
 344	0
 345	0
@@ -363,7 +363,7 @@ index	prediction
 361	0
 362	0
 363	0
-364	1
+364	0
 365	0
 366	0
 367	0
@@ -377,7 +377,7 @@ index	prediction
 375	0
 376	0
 377	0
-378	1
+378	0
 379	0
 380	0
 381	0
@@ -419,9 +419,9 @@ index	prediction
 417	0
 418	0
 419	0
-420	0
+420	1
 421	0
-422	0
+422	1
 423	0
 424	0
 425	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	1
-473	0
+473	1
 474	1
 475	0
 476	1
@@ -538,7 +538,7 @@ index	prediction
 536	0
 537	0
 538	0
-539	0
+539	1
 540	0
 541	0
 542	0
@@ -601,7 +601,7 @@ index	prediction
 599	0
 600	0
 601	0
-602	0
+602	1
 603	0
 604	0
 605	1
@@ -621,7 +621,7 @@ index	prediction
 619	0
 620	0
 621	0
-622	0
+622	1
 623	0
 624	0
 625	0
@@ -672,8 +672,8 @@ index	prediction
 670	0
 671	0
 672	0
-673	0
-674	0
+673	1
+674	1
 675	0
 676	0
 677	0
@@ -688,7 +688,7 @@ index	prediction
 686	0
 687	0
 688	0
-689	1
+689	0
 690	0
 691	0
 692	0
@@ -740,7 +740,7 @@ index	prediction
 738	0
 739	0
 740	0
-741	0
+741	1
 742	0
 743	0
 744	0
@@ -784,7 +784,7 @@ index	prediction
 782	0
 783	0
 784	0
-785	0
+785	1
 786	0
 787	0
 788	0
@@ -797,19 +797,19 @@ index	prediction
 795	1
 796	0
 797	0
-798	1
-799	1
+798	0
+799	0
 800	0
 801	0
-802	0
+802	1
 803	0
 804	0
 805	0
 806	0
 807	0
 808	0
-809	0
-810	1
+809	1
+810	0
 811	0
 812	0
 813	0
@@ -831,7 +831,7 @@ index	prediction
 829	0
 830	0
 831	1
-832	1
+832	0
 833	0
 834	0
 835	0
@@ -859,7 +859,7 @@ index	prediction
 857	1
 858	0
 859	0
-860	0
+860	1
 861	0
 862	0
 863	0
@@ -885,7 +885,7 @@ index	prediction
 883	0
 884	0
 885	0
-886	1
+886	0
 887	0
 888	0
 889	0
@@ -912,7 +912,7 @@ index	prediction
 910	0
 911	0
 912	1
-913	1
+913	0
 914	0
 915	0
 916	0
@@ -923,12 +923,12 @@ index	prediction
 921	1
 922	0
 923	0
-924	1
+924	0
 925	0
 926	1
 927	0
 928	0
-929	0
+929	1
 930	0
 931	0
 932	0
@@ -946,7 +946,7 @@ index	prediction
 944	0
 945	1
 946	0
-947	0
+947	1
 948	0
 949	0
 950	0
@@ -966,7 +966,7 @@ index	prediction
 964	0
 965	0
 966	0
-967	0
+967	1
 968	0
 969	0
 970	0

runs/Jun03_14-28-41_a358b85c7679/events.out.tfevents.1717425557.a358b85c7679.140888.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa2ba2793098dc75e6508b9c16edb8322d39b79e4617f6dc1fa57a06ee056028
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.23294218016452475,
-    "train_runtime": 1905.4788,
+    "train_loss": 0.22970127551282038,
+    "train_runtime": 621.3103,
     "train_samples": 3638,
-    "train_samples_per_second": 38.185,
-    "train_steps_per_second": 1.281
+    "train_samples_per_second": 117.107,
+    "train_steps_per_second": 3.927
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 4.96969747543335,
+      "grad_norm": 5.112319469451904,
       "learning_rate": 4.75e-05,
-      "loss": 0.5526,
+      "loss": 0.5509,
       "step": 122
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7142857142857143,
-      "eval_f1": 0.6246286393345217,
-      "eval_loss": 0.5089075565338135,
-      "eval_precision": 0.6432360742705571,
-      "eval_recall": 0.6178396072013094,
-      "eval_runtime": 5.1308,
-      "eval_samples_per_second": 77.766,
-      "eval_steps_per_second": 9.745,
+      "eval_accuracy": 0.7393483709273183,
+      "eval_f1": 0.6507070707070707,
+      "eval_loss": 0.4983255863189697,
+      "eval_precision": 0.6800605637083625,
+      "eval_recall": 0.6405710129114385,
+      "eval_runtime": 1.7657,
+      "eval_samples_per_second": 225.971,
+      "eval_steps_per_second": 28.317,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 3.059352159500122,
+      "grad_norm": 3.6866044998168945,
       "learning_rate": 4.5e-05,
-      "loss": 0.4578,
+      "loss": 0.4511,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7994987468671679,
-      "eval_f1": 0.771804403774664,
-      "eval_loss": 0.4193674325942993,
-      "eval_precision": 0.7619509396853147,
-      "eval_recall": 0.7906437534097108,
-      "eval_runtime": 5.0426,
-      "eval_samples_per_second": 79.125,
-      "eval_steps_per_second": 9.915,
+      "eval_accuracy": 0.7769423558897243,
+      "eval_f1": 0.7593078346448687,
+      "eval_loss": 0.4377373456954956,
+      "eval_precision": 0.7546743295019157,
+      "eval_recall": 0.8021913075104565,
+      "eval_runtime": 1.769,
+      "eval_samples_per_second": 225.555,
+      "eval_steps_per_second": 28.265,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 3.294614553451538,
+      "grad_norm": 3.584764242172241,
       "learning_rate": 4.25e-05,
-      "loss": 0.3632,
+      "loss": 0.368,
       "step": 366
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8370927318295739,
-      "eval_f1": 0.7866573438078395,
-      "eval_loss": 0.3468063175678253,
-      "eval_precision": 0.8221645717626425,
-      "eval_recall": 0.7672304055282779,
-      "eval_runtime": 5.0451,
-      "eval_samples_per_second": 79.086,
-      "eval_steps_per_second": 9.911,
+      "eval_accuracy": 0.8571428571428571,
+      "eval_f1": 0.8196102381877741,
+      "eval_loss": 0.32603567838668823,
+      "eval_precision": 0.8381270903010034,
+      "eval_recall": 0.8064193489725404,
+      "eval_runtime": 1.7715,
+      "eval_samples_per_second": 225.23,
+      "eval_steps_per_second": 28.224,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 3.894381046295166,
+      "grad_norm": 2.8483095169067383,
       "learning_rate": 4e-05,
-      "loss": 0.3063,
+      "loss": 0.3019,
       "step": 488
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.8589543987905864,
-      "eval_loss": 0.2975314259529114,
-      "eval_precision": 0.8564068100358423,
-      "eval_recall": 0.8616566648481543,
-      "eval_runtime": 5.045,
-      "eval_samples_per_second": 79.087,
-      "eval_steps_per_second": 9.911,
+      "eval_accuracy": 0.8646616541353384,
+      "eval_f1": 0.8333281762485303,
+      "eval_loss": 0.30364951491355896,
+      "eval_precision": 0.8410471369819678,
+      "eval_recall": 0.8267412256773959,
+      "eval_runtime": 1.7702,
+      "eval_samples_per_second": 225.393,
+      "eval_steps_per_second": 28.245,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 3.5420114994049072,
+      "grad_norm": 2.774143934249878,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.2771,
+      "loss": 0.2668,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8721804511278195,
-      "eval_f1": 0.8524204953403198,
-      "eval_loss": 0.3194977939128876,
-      "eval_precision": 0.8409052784611943,
-      "eval_recall": 0.8695671940352792,
-      "eval_runtime": 5.0554,
-      "eval_samples_per_second": 78.926,
-      "eval_steps_per_second": 9.89,
+      "eval_accuracy": 0.8671679197994987,
+      "eval_f1": 0.8424651921601347,
+      "eval_loss": 0.31921207904815674,
+      "eval_precision": 0.8372140762463343,
+      "eval_recall": 0.8485179123476996,
+      "eval_runtime": 1.7714,
+      "eval_samples_per_second": 225.248,
+      "eval_steps_per_second": 28.227,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 2.310750961303711,
+      "grad_norm": 4.2327117919921875,
       "learning_rate": 3.5e-05,
-      "loss": 0.2463,
+      "loss": 0.2471,
       "step": 732
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.8771929824561403,
-      "eval_f1": 0.8536612749904566,
-      "eval_loss": 0.28968048095703125,
-      "eval_precision": 0.8496330709593418,
-      "eval_recall": 0.8581105655573741,
-      "eval_runtime": 5.0495,
-      "eval_samples_per_second": 79.017,
-      "eval_steps_per_second": 9.902,
+      "eval_accuracy": 0.8621553884711779,
+      "eval_f1": 0.8380263497804185,
+      "eval_loss": 0.30589351058006287,
+      "eval_precision": 0.830503344095941,
+      "eval_recall": 0.8474722676850337,
+      "eval_runtime": 1.7732,
+      "eval_samples_per_second": 225.015,
+      "eval_steps_per_second": 28.197,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 2.563045024871826,
+      "grad_norm": 0.5115749835968018,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.2317,
+      "loss": 0.2422,
       "step": 854
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8609292598654301,
-      "eval_loss": 0.27178412675857544,
-      "eval_precision": 0.8609292598654301,
-      "eval_recall": 0.8609292598654301,
-      "eval_runtime": 5.0509,
-      "eval_samples_per_second": 78.995,
-      "eval_steps_per_second": 9.899,
+      "eval_accuracy": 0.87468671679198,
+      "eval_f1": 0.8524146298159436,
+      "eval_loss": 0.2949831783771515,
+      "eval_precision": 0.8451250578971746,
+      "eval_recall": 0.8613384251682124,
+      "eval_runtime": 1.7731,
+      "eval_samples_per_second": 225.024,
+      "eval_steps_per_second": 28.198,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 1.676988959312439,
+      "grad_norm": 1.2918312549591064,
       "learning_rate": 3e-05,
-      "loss": 0.2182,
+      "loss": 0.2258,
       "step": 976
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8654398826979472,
-      "eval_loss": 0.26825374364852905,
-      "eval_precision": 0.8556454369374502,
-      "eval_recall": 0.8784324422622295,
-      "eval_runtime": 5.0526,
-      "eval_samples_per_second": 78.969,
-      "eval_steps_per_second": 9.896,
+      "eval_accuracy": 0.8721804511278195,
+      "eval_f1": 0.8454251965513313,
+      "eval_loss": 0.29280924797058105,
+      "eval_precision": 0.8463049835506276,
+      "eval_recall": 0.8445626477541371,
+      "eval_runtime": 1.7799,
+      "eval_samples_per_second": 224.171,
+      "eval_steps_per_second": 28.092,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 5.442895412445068,
+      "grad_norm": 5.160737037658691,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.2065,
+      "loss": 0.2054,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.8696741854636592,
-      "eval_f1": 0.844327731092437,
-      "eval_loss": 0.27725929021835327,
-      "eval_precision": 0.8411320530352577,
-      "eval_recall": 0.8477905073649754,
-      "eval_runtime": 5.0455,
-      "eval_samples_per_second": 79.08,
-      "eval_steps_per_second": 9.91,
+      "eval_accuracy": 0.8796992481203008,
+      "eval_f1": 0.8533986527862829,
+      "eval_loss": 0.30492648482322693,
+      "eval_precision": 0.8572003218020917,
+      "eval_recall": 0.8498817966903074,
+      "eval_runtime": 1.779,
+      "eval_samples_per_second": 224.288,
+      "eval_steps_per_second": 28.106,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 4.095264434814453,
+      "grad_norm": 3.917464017868042,
       "learning_rate": 2.5e-05,
-      "loss": 0.2012,
+      "loss": 0.2009,
       "step": 1220
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.852937255424767,
-      "eval_loss": 0.28407934308052063,
-      "eval_precision": 0.8674217731421121,
-      "eval_recall": 0.8416530278232406,
-      "eval_runtime": 5.0513,
-      "eval_samples_per_second": 78.99,
-      "eval_steps_per_second": 9.898,
+      "eval_accuracy": 0.87468671679198,
+      "eval_f1": 0.8488361520276414,
+      "eval_loss": 0.30127042531967163,
+      "eval_precision": 0.8488361520276414,
+      "eval_recall": 0.8488361520276414,
+      "eval_runtime": 1.7757,
+      "eval_samples_per_second": 224.7,
+      "eval_steps_per_second": 28.158,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 7.4546098709106445,
+      "grad_norm": 6.667805194854736,
       "learning_rate": 2.25e-05,
-      "loss": 0.1944,
+      "loss": 0.1755,
       "step": 1342
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8616171059774413,
-      "eval_loss": 0.27333080768585205,
-      "eval_precision": 0.859873949579832,
-      "eval_recall": 0.8634297144935443,
-      "eval_runtime": 5.0526,
-      "eval_samples_per_second": 78.968,
-      "eval_steps_per_second": 9.896,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.858259325044405,
+      "eval_loss": 0.30701279640197754,
+      "eval_precision": 0.8573798178418481,
+      "eval_recall": 0.8591562102200401,
+      "eval_runtime": 1.7942,
+      "eval_samples_per_second": 222.38,
+      "eval_steps_per_second": 27.867,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 8.82780647277832,
+      "grad_norm": 8.611730575561523,
       "learning_rate": 2e-05,
-      "loss": 0.176,
+      "loss": 0.1821,
       "step": 1464
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8724195749658803,
-      "eval_loss": 0.27092599868774414,
-      "eval_precision": 0.8848766823362741,
-      "eval_recall": 0.8622931442080378,
-      "eval_runtime": 5.0447,
-      "eval_samples_per_second": 79.094,
-      "eval_steps_per_second": 9.911,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.8568221901555235,
+      "eval_loss": 0.2995355427265167,
+      "eval_precision": 0.8596491228070176,
+      "eval_recall": 0.8541553009638116,
+      "eval_runtime": 1.7796,
+      "eval_samples_per_second": 224.202,
+      "eval_steps_per_second": 28.095,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 3.990729331970215,
+      "grad_norm": 2.71295428276062,
       "learning_rate": 1.75e-05,
-      "loss": 0.168,
+      "loss": 0.1652,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8736504011098378,
-      "eval_loss": 0.26514801383018494,
-      "eval_precision": 0.8718487394957983,
-      "eval_recall": 0.8755228223313329,
-      "eval_runtime": 5.0943,
-      "eval_samples_per_second": 78.323,
-      "eval_steps_per_second": 9.815,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.866029197080292,
+      "eval_loss": 0.3272043764591217,
+      "eval_precision": 0.8552631578947368,
+      "eval_recall": 0.8809328968903437,
+      "eval_runtime": 1.7775,
+      "eval_samples_per_second": 224.467,
+      "eval_steps_per_second": 28.129,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 2.973822593688965,
+      "grad_norm": 5.373868942260742,
       "learning_rate": 1.5e-05,
-      "loss": 0.1644,
+      "loss": 0.1566,
       "step": 1708
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8727838950061173,
-      "eval_loss": 0.2710956037044525,
-      "eval_precision": 0.8656898656898657,
-      "eval_recall": 0.8812511365702855,
-      "eval_runtime": 5.0742,
-      "eval_samples_per_second": 78.633,
-      "eval_steps_per_second": 9.854,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8718540145985401,
+      "eval_loss": 0.33357149362564087,
+      "eval_precision": 0.8609022556390977,
+      "eval_recall": 0.886979450809238,
+      "eval_runtime": 1.7836,
+      "eval_samples_per_second": 223.703,
+      "eval_steps_per_second": 28.033,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 2.690046548843384,
+      "grad_norm": 5.369639873504639,
       "learning_rate": 1.25e-05,
-      "loss": 0.1541,
+      "loss": 0.1634,
       "step": 1830
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8683279483657071,
-      "eval_loss": 0.2790246605873108,
-      "eval_precision": 0.873366724738676,
-      "eval_recall": 0.863747954173486,
-      "eval_runtime": 5.055,
-      "eval_samples_per_second": 78.932,
-      "eval_steps_per_second": 9.891,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8622899159663866,
+      "eval_loss": 0.314995676279068,
+      "eval_precision": 0.8589244307033712,
+      "eval_recall": 0.8659301691216585,
+      "eval_runtime": 1.777,
+      "eval_samples_per_second": 224.539,
+      "eval_steps_per_second": 28.138,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 1.9607453346252441,
+      "grad_norm": 6.779192924499512,
       "learning_rate": 1e-05,
-      "loss": 0.1574,
+      "loss": 0.1496,
       "step": 1952
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8656154317207594,
-      "eval_loss": 0.27665647864341736,
-      "eval_precision": 0.869546382820719,
-      "eval_recall": 0.861974904528096,
-      "eval_runtime": 5.06,
-      "eval_samples_per_second": 78.853,
-      "eval_steps_per_second": 9.881,
+      "eval_accuracy": 0.8922305764411027,
+      "eval_f1": 0.8696722245432793,
+      "eval_loss": 0.3320792317390442,
+      "eval_precision": 0.8706135006701596,
+      "eval_recall": 0.8687488634297145,
+      "eval_runtime": 1.7833,
+      "eval_samples_per_second": 223.741,
+      "eval_steps_per_second": 28.038,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 0.14830851554870605,
+      "grad_norm": 1.1515932083129883,
       "learning_rate": 7.5e-06,
-      "loss": 0.1508,
+      "loss": 0.1355,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8676337535436396,
-      "eval_loss": 0.2825946509838104,
-      "eval_precision": 0.8658613445378152,
-      "eval_recall": 0.8694762684124386,
-      "eval_runtime": 5.0609,
-      "eval_samples_per_second": 78.84,
-      "eval_steps_per_second": 9.88,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8616171059774413,
+      "eval_loss": 0.32759982347488403,
+      "eval_precision": 0.859873949579832,
+      "eval_recall": 0.8634297144935443,
+      "eval_runtime": 1.7782,
+      "eval_samples_per_second": 224.387,
+      "eval_steps_per_second": 28.119,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 1.0990123748779297,
+      "grad_norm": 1.6571087837219238,
       "learning_rate": 5e-06,
-      "loss": 0.1463,
+      "loss": 0.1477,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.8872180451127819,
-      "eval_f1": 0.8614765038536611,
-      "eval_loss": 0.2823769748210907,
-      "eval_precision": 0.8686536646744258,
-      "eval_recall": 0.8552009456264775,
-      "eval_runtime": 5.0926,
-      "eval_samples_per_second": 78.349,
-      "eval_steps_per_second": 9.818,
+      "eval_accuracy": 0.8796992481203008,
+      "eval_f1": 0.8563025210084034,
+      "eval_loss": 0.33653610944747925,
+      "eval_precision": 0.8529936381473334,
+      "eval_recall": 0.8598836152027641,
+      "eval_runtime": 1.7851,
+      "eval_samples_per_second": 223.518,
+      "eval_steps_per_second": 28.01,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 0.10962895303964615,
+      "grad_norm": 2.6701011657714844,
       "learning_rate": 2.5e-06,
-      "loss": 0.1467,
+      "loss": 0.1317,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8572517421602788,
-      "eval_loss": 0.2876257598400116,
-      "eval_precision": 0.8679426449878376,
-      "eval_recall": 0.8484269867248591,
-      "eval_runtime": 5.0519,
-      "eval_samples_per_second": 78.981,
-      "eval_steps_per_second": 9.897,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.858259325044405,
+      "eval_loss": 0.3385031819343567,
+      "eval_precision": 0.8573798178418481,
+      "eval_recall": 0.8591562102200401,
+      "eval_runtime": 1.7765,
+      "eval_samples_per_second": 224.597,
+      "eval_steps_per_second": 28.145,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 4.516838550567627,
+      "grad_norm": 3.197312593460083,
       "learning_rate": 0.0,
-      "loss": 0.1399,
+      "loss": 0.1267,
       "step": 2440
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8649122807017544,
-      "eval_loss": 0.2823803424835205,
-      "eval_precision": 0.8710116366366366,
-      "eval_recall": 0.8594744498999818,
-      "eval_runtime": 5.0556,
-      "eval_samples_per_second": 78.922,
-      "eval_steps_per_second": 9.89,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.858259325044405,
+      "eval_loss": 0.3389217257499695,
+      "eval_precision": 0.8573798178418481,
+      "eval_recall": 0.8591562102200401,
+      "eval_runtime": 1.7779,
+      "eval_samples_per_second": 224.423,
+      "eval_steps_per_second": 28.123,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
       "total_flos": 8444128359504000.0,
-      "train_loss": 0.23294218016452475,
-      "train_runtime": 1905.4788,
-      "train_samples_per_second": 38.185,
-      "train_steps_per_second": 1.281
+      "train_loss": 0.22970127551282038,
+      "train_runtime": 621.3103,
+      "train_samples_per_second": 117.107,
+      "train_steps_per_second": 3.927
     }
   ],
   "logging_steps": 500,