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": 0.8941641938674579,
+    "accuracy": 0.8961424332344213,
     "epoch": 20.0,
-    "eval_accuracy": 0.9022556390977443,
-    "eval_f1": 0.8829621606985718,
-    "eval_loss": 0.2765507400035858,
-    "eval_precision": 0.8802419354838709,
-    "eval_recall": 0.8858428805237315,
-    "eval_runtime": 5.0092,
+    "eval_accuracy": 0.8972431077694235,
+    "eval_f1": 0.8781334505389722,
+    "eval_loss": 0.2797949016094208,
+    "eval_precision": 0.872316715542522,
+    "eval_recall": 0.8847972358610656,
+    "eval_runtime": 1.8194,
     "eval_samples": 399,
-    "eval_samples_per_second": 79.653,
-    "eval_steps_per_second": 9.982,
-    "f1": 0.8742883638556821,
-    "precision": 0.8699655489205329,
-    "recall": 0.8790652032719297,
-    "train_loss": 0.23107414167435442,
-    "train_runtime": 1909.2167,
+    "eval_samples_per_second": 219.307,
+    "eval_steps_per_second": 27.482,
+    "f1": 0.8761730346384224,
+    "precision": 0.8730752146875925,
+    "recall": 0.8794911377392057,
+    "train_loss": 0.24180538537072355,
+    "train_runtime": 620.8783,
     "train_samples": 3638,
-    "train_samples_per_second": 38.11,
-    "train_steps_per_second": 1.278
+    "train_samples_per_second": 117.189,
+    "train_steps_per_second": 3.93
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.9022556390977443,
-    "eval_f1": 0.8829621606985718,
-    "eval_loss": 0.2765507400035858,
-    "eval_precision": 0.8802419354838709,
-    "eval_recall": 0.8858428805237315,
-    "eval_runtime": 5.0092,
+    "eval_accuracy": 0.8972431077694235,
+    "eval_f1": 0.8781334505389722,
+    "eval_loss": 0.2797949016094208,
+    "eval_precision": 0.872316715542522,
+    "eval_recall": 0.8847972358610656,
+    "eval_runtime": 1.8194,
     "eval_samples": 399,
-    "eval_samples_per_second": 79.653,
-    "eval_steps_per_second": 9.982
+    "eval_samples_per_second": 219.307,
+    "eval_steps_per_second": 27.482
 }

predict_results.json

@@ -1,6 +1,6 @@
 {
-    "accuracy": 0.8941641938674579,
-    "f1": 0.8742883638556821,
-    "precision": 0.8699655489205329,
-    "recall": 0.8790652032719297
+    "accuracy": 0.8961424332344213,
+    "f1": 0.8761730346384224,
+    "precision": 0.8730752146875925,
+    "recall": 0.8794911377392057
 }

predict_results.txt

@@ -1,6 +1,6 @@
 index	prediction
 0	1
-1	0
+1	1
 2	0
 3	1
 4	0
@@ -12,7 +12,7 @@ index	prediction
 10	1
 11	1
 12	1
-13	0
+13	1
 14	1
 15	0
 16	1
@@ -58,7 +58,7 @@ index	prediction
 56	1
 57	0
 58	1
-59	0
+59	1
 60	1
 61	1
 62	1
@@ -70,17 +70,17 @@ index	prediction
 68	1
 69	1
 70	1
-71	0
+71	1
 72	1
 73	1
 74	1
 75	1
 76	1
-77	1
+77	0
 78	1
 79	0
 80	1
-81	0
+81	1
 82	1
 83	1
 84	1
@@ -91,7 +91,7 @@ index	prediction
 89	1
 90	1
 91	1
-92	0
+92	1
 93	1
 94	1
 95	1
@@ -117,7 +117,7 @@ index	prediction
 115	1
 116	1
 117	1
-118	1
+118	0
 119	1
 120	1
 121	1
@@ -163,7 +163,7 @@ index	prediction
 161	1
 162	1
 163	1
-164	1
+164	0
 165	0
 166	1
 167	1
@@ -171,7 +171,7 @@ index	prediction
 169	1
 170	1
 171	1
-172	1
+172	0
 173	0
 174	1
 175	1
@@ -209,7 +209,7 @@ index	prediction
 207	1
 208	1
 209	1
-210	0
+210	1
 211	1
 212	1
 213	1
@@ -227,7 +227,7 @@ index	prediction
 225	1
 226	0
 227	0
-228	1
+228	0
 229	0
 230	0
 231	1
@@ -248,7 +248,7 @@ index	prediction
 246	1
 247	1
 248	1
-249	1
+249	0
 250	0
 251	1
 252	1
@@ -264,7 +264,7 @@ index	prediction
 262	1
 263	1
 264	1
-265	1
+265	0
 266	1
 267	1
 268	1
@@ -272,7 +272,7 @@ index	prediction
 270	1
 271	1
 272	1
-273	1
+273	0
 274	1
 275	1
 276	1
@@ -313,9 +313,9 @@ index	prediction
 311	0
 312	0
 313	0
-314	1
+314	0
 315	0
-316	1
+316	0
 317	0
 318	1
 319	0
@@ -343,9 +343,9 @@ index	prediction
 341	0
 342	0
 343	0
-344	1
+344	0
 345	0
-346	1
+346	0
 347	0
 348	0
 349	0
@@ -363,7 +363,7 @@ index	prediction
 361	0
 362	0
 363	0
-364	0
+364	1
 365	0
 366	1
 367	0
@@ -400,7 +400,7 @@ index	prediction
 398	1
 399	0
 400	0
-401	0
+401	1
 402	1
 403	0
 404	0
@@ -419,8 +419,8 @@ index	prediction
 417	0
 418	0
 419	0
-420	1
-421	1
+420	0
+421	0
 422	0
 423	0
 424	0
@@ -486,7 +486,7 @@ index	prediction
 484	0
 485	0
 486	0
-487	0
+487	1
 488	0
 489	0
 490	0
@@ -517,7 +517,7 @@ index	prediction
 515	0
 516	0
 517	0
-518	0
+518	1
 519	0
 520	0
 521	0
@@ -537,11 +537,11 @@ index	prediction
 535	0
 536	0
 537	0
-538	1
+538	0
 539	0
 540	0
 541	0
-542	0
+542	1
 543	0
 544	0
 545	0
@@ -625,10 +625,10 @@ index	prediction
 623	0
 624	0
 625	0
-626	1
+626	0
 627	0
 628	0
-629	1
+629	0
 630	0
 631	0
 632	0
@@ -726,8 +726,8 @@ index	prediction
 724	0
 725	0
 726	0
-727	1
-728	1
+727	0
+728	0
 729	0
 730	0
 731	0
@@ -773,7 +773,7 @@ index	prediction
 771	0
 772	0
 773	0
-774	0
+774	1
 775	0
 776	0
 777	1
@@ -794,12 +794,12 @@ index	prediction
 792	0
 793	0
 794	0
-795	1
+795	0
 796	0
 797	0
 798	0
 799	0
-800	1
+800	0
 801	0
 802	0
 803	0
@@ -831,7 +831,7 @@ index	prediction
 829	0
 830	0
 831	0
-832	0
+832	1
 833	1
 834	0
 835	0
@@ -847,7 +847,7 @@ index	prediction
 845	0
 846	0
 847	0
-848	0
+848	1
 849	0
 850	0
 851	0
@@ -949,7 +949,7 @@ index	prediction
 947	0
 948	0
 949	0
-950	0
+950	1
 951	0
 952	0
 953	0
@@ -968,7 +968,7 @@ index	prediction
 966	0
 967	0
 968	0
-969	0
+969	1
 970	0
 971	0
 972	0

runs/Jun03_12-49-09_a358b85c7679/events.out.tfevents.1717419586.a358b85c7679.87787.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ecd50dd70d947628baa78aee72e620ed1aefb36a63b5c224b663cffb13395e7
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.23107414167435442,
-    "train_runtime": 1909.2167,
+    "train_loss": 0.24180538537072355,
+    "train_runtime": 620.8783,
     "train_samples": 3638,
-    "train_samples_per_second": 38.11,
-    "train_steps_per_second": 1.278
+    "train_samples_per_second": 117.189,
+    "train_steps_per_second": 3.93
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 4.234379768371582,
+      "grad_norm": 4.232791900634766,
       "learning_rate": 4.75e-05,
-      "loss": 0.5459,
+      "loss": 0.5568,
       "step": 122
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7393483709273183,
-      "eval_f1": 0.6458660476548099,
-      "eval_loss": 0.47629594802856445,
-      "eval_precision": 0.6804065499717673,
-      "eval_recall": 0.63557010365521,
-      "eval_runtime": 4.8279,
-      "eval_samples_per_second": 82.644,
-      "eval_steps_per_second": 10.356,
+      "eval_accuracy": 0.7243107769423559,
+      "eval_f1": 0.6144317942230656,
+      "eval_loss": 0.48217353224754333,
+      "eval_precision": 0.65566534914361,
+      "eval_recall": 0.6074286233860702,
+      "eval_runtime": 1.8124,
+      "eval_samples_per_second": 220.155,
+      "eval_steps_per_second": 27.588,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 4.299708843231201,
+      "grad_norm": 3.6621267795562744,
       "learning_rate": 4.5e-05,
-      "loss": 0.4528,
+      "loss": 0.4661,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7844611528822055,
-      "eval_f1": 0.7677594888732471,
-      "eval_loss": 0.43055105209350586,
-      "eval_precision": 0.7630357142857143,
-      "eval_recall": 0.8125113657028551,
-      "eval_runtime": 5.0594,
-      "eval_samples_per_second": 78.863,
-      "eval_steps_per_second": 9.883,
+      "eval_accuracy": 0.7543859649122807,
+      "eval_f1": 0.7304054054054054,
+      "eval_loss": 0.44529902935028076,
+      "eval_precision": 0.7240563585317666,
+      "eval_recall": 0.7612293144208038,
+      "eval_runtime": 1.7908,
+      "eval_samples_per_second": 222.81,
+      "eval_steps_per_second": 27.921,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 3.729501962661743,
+      "grad_norm": 3.223342180252075,
       "learning_rate": 4.25e-05,
-      "loss": 0.3653,
+      "loss": 0.3875,
       "step": 366
     },
     {
       "epoch": 3.0,
       "eval_accuracy": 0.8621553884711779,
-      "eval_f1": 0.8217317487266553,
-      "eval_loss": 0.3334697186946869,
-      "eval_precision": 0.853331681065005,
-      "eval_recall": 0.802464084378978,
-      "eval_runtime": 5.0455,
-      "eval_samples_per_second": 79.08,
-      "eval_steps_per_second": 9.91,
+      "eval_f1": 0.8238834717707957,
+      "eval_loss": 0.3446956276893616,
+      "eval_precision": 0.8487520627062706,
+      "eval_recall": 0.8074649936352064,
+      "eval_runtime": 1.7891,
+      "eval_samples_per_second": 223.02,
+      "eval_steps_per_second": 27.947,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 2.0170559883117676,
+      "grad_norm": 3.7967684268951416,
       "learning_rate": 4e-05,
-      "loss": 0.2987,
+      "loss": 0.318,
       "step": 488
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8546365914786967,
-      "eval_f1": 0.8246499363520641,
-      "eval_loss": 0.335675448179245,
-      "eval_precision": 0.8246499363520641,
-      "eval_recall": 0.8246499363520641,
-      "eval_runtime": 5.2896,
-      "eval_samples_per_second": 75.431,
-      "eval_steps_per_second": 9.452,
+      "eval_accuracy": 0.849624060150376,
+      "eval_f1": 0.8267427049559981,
+      "eval_loss": 0.34423500299453735,
+      "eval_precision": 0.8158466596088483,
+      "eval_recall": 0.8436079287143117,
+      "eval_runtime": 1.786,
+      "eval_samples_per_second": 223.404,
+      "eval_steps_per_second": 27.996,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.7316147089004517,
+      "grad_norm": 0.4756224453449249,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.2746,
+      "loss": 0.2855,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8546365914786967,
-      "eval_f1": 0.8338931955211025,
-      "eval_loss": 0.3400849401950836,
-      "eval_precision": 0.8216641926439057,
-      "eval_recall": 0.8546553918894344,
-      "eval_runtime": 5.081,
-      "eval_samples_per_second": 78.527,
-      "eval_steps_per_second": 9.841,
+      "eval_accuracy": 0.849624060150376,
+      "eval_f1": 0.826007326007326,
+      "eval_loss": 0.3348763883113861,
+      "eval_precision": 0.8157828282828283,
+      "eval_recall": 0.8411074740861975,
+      "eval_runtime": 1.7834,
+      "eval_samples_per_second": 223.733,
+      "eval_steps_per_second": 28.037,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 1.9782907962799072,
+      "grad_norm": 1.1933702230453491,
       "learning_rate": 3.5e-05,
-      "loss": 0.2477,
+      "loss": 0.2638,
       "step": 732
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.849624060150376,
-      "eval_f1": 0.8308270676691729,
-      "eval_loss": 0.33233991265296936,
-      "eval_precision": 0.8176375838926174,
-      "eval_recall": 0.858610656482997,
-      "eval_runtime": 5.0582,
-      "eval_samples_per_second": 78.882,
-      "eval_steps_per_second": 9.885,
+      "eval_accuracy": 0.8370927318295739,
+      "eval_f1": 0.8177454831659652,
+      "eval_loss": 0.35479673743247986,
+      "eval_precision": 0.805161943319838,
+      "eval_recall": 0.8472449536279323,
+      "eval_runtime": 1.7837,
+      "eval_samples_per_second": 223.688,
+      "eval_steps_per_second": 28.031,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 1.8530523777008057,
+      "grad_norm": 0.3131154775619507,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.24,
+      "loss": 0.2397,
       "step": 854
     },
     {
       "epoch": 7.0,
       "eval_accuracy": 0.8646616541353384,
-      "eval_f1": 0.8447157518450185,
-      "eval_loss": 0.3171246647834778,
-      "eval_precision": 0.8325401217487549,
-      "eval_recall": 0.864248045099109,
-      "eval_runtime": 5.0489,
-      "eval_samples_per_second": 79.028,
-      "eval_steps_per_second": 9.903,
+      "eval_f1": 0.8434065934065934,
+      "eval_loss": 0.3253527581691742,
+      "eval_precision": 0.8325757575757575,
+      "eval_recall": 0.8592471358428806,
+      "eval_runtime": 1.7938,
+      "eval_samples_per_second": 222.438,
+      "eval_steps_per_second": 27.874,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 3.7897450923919678,
+      "grad_norm": 5.750446319580078,
       "learning_rate": 3e-05,
-      "loss": 0.2069,
+      "loss": 0.2428,
       "step": 976
     },
     {
       "epoch": 8.0,
       "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8683279483657071,
-      "eval_loss": 0.27698931097984314,
-      "eval_precision": 0.873366724738676,
-      "eval_recall": 0.863747954173486,
-      "eval_runtime": 5.0659,
-      "eval_samples_per_second": 78.762,
-      "eval_steps_per_second": 9.87,
+      "eval_f1": 0.8654532336864889,
+      "eval_loss": 0.2798740565776825,
+      "eval_precision": 0.8804269882659713,
+      "eval_recall": 0.8537461356610292,
+      "eval_runtime": 1.7849,
+      "eval_samples_per_second": 223.546,
+      "eval_steps_per_second": 28.013,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 5.382436275482178,
+      "grad_norm": 5.672217845916748,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.2197,
+      "loss": 0.2229,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.8671679197994987,
-      "eval_f1": 0.8491192020377148,
-      "eval_loss": 0.3091033399105072,
-      "eval_precision": 0.8356209150326797,
-      "eval_recall": 0.8735224586288416,
-      "eval_runtime": 5.2322,
-      "eval_samples_per_second": 76.258,
-      "eval_steps_per_second": 9.556,
+      "eval_accuracy": 0.8721804511278195,
+      "eval_f1": 0.8484099018899409,
+      "eval_loss": 0.29030779004096985,
+      "eval_precision": 0.8430645161290322,
+      "eval_recall": 0.8545644662665939,
+      "eval_runtime": 1.7837,
+      "eval_samples_per_second": 223.689,
+      "eval_steps_per_second": 28.031,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 3.9839751720428467,
+      "grad_norm": 2.1360011100769043,
       "learning_rate": 2.5e-05,
-      "loss": 0.2005,
+      "loss": 0.2144,
       "step": 1220
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8811928811928812,
-      "eval_loss": 0.2552439570426941,
-      "eval_precision": 0.8842105263157894,
-      "eval_recall": 0.878341516639389,
-      "eval_runtime": 5.0419,
-      "eval_samples_per_second": 79.136,
-      "eval_steps_per_second": 9.917,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8769602202215754,
+      "eval_loss": 0.2583388686180115,
+      "eval_precision": 0.8742831541218639,
+      "eval_recall": 0.8797963266048372,
+      "eval_runtime": 1.7944,
+      "eval_samples_per_second": 222.362,
+      "eval_steps_per_second": 27.865,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 0.8527041673660278,
+      "grad_norm": 4.677872657775879,
       "learning_rate": 2.25e-05,
-      "loss": 0.1867,
+      "loss": 0.1967,
       "step": 1342
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.8867831541218638,
-      "eval_loss": 0.2726534903049469,
-      "eval_precision": 0.88158359844468,
-      "eval_recall": 0.8926168394253501,
-      "eval_runtime": 5.0607,
-      "eval_samples_per_second": 78.843,
-      "eval_steps_per_second": 9.88,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.8622085718274466,
+      "eval_loss": 0.27431806921958923,
+      "eval_precision": 0.8530168716042322,
+      "eval_recall": 0.8741589379887251,
+      "eval_runtime": 1.7823,
+      "eval_samples_per_second": 223.874,
+      "eval_steps_per_second": 28.054,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 2.9348244667053223,
+      "grad_norm": 2.564518451690674,
       "learning_rate": 2e-05,
-      "loss": 0.1722,
+      "loss": 0.1855,
       "step": 1464
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8727838950061173,
-      "eval_loss": 0.27391675114631653,
-      "eval_precision": 0.8656898656898657,
-      "eval_recall": 0.8812511365702855,
-      "eval_runtime": 5.0491,
-      "eval_samples_per_second": 79.024,
-      "eval_steps_per_second": 9.903,
+      "eval_accuracy": 0.8771929824561403,
+      "eval_f1": 0.8563451067988272,
+      "eval_loss": 0.29132312536239624,
+      "eval_precision": 0.8473119816985988,
+      "eval_recall": 0.8681123840698308,
+      "eval_runtime": 1.7831,
+      "eval_samples_per_second": 223.773,
+      "eval_steps_per_second": 28.042,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 1.8993942737579346,
+      "grad_norm": 2.314499855041504,
       "learning_rate": 1.75e-05,
-      "loss": 0.161,
+      "loss": 0.1761,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.8765393898137261,
-      "eval_loss": 0.2714451551437378,
-      "eval_precision": 0.8852261942423283,
-      "eval_recall": 0.8690671031096563,
-      "eval_runtime": 5.0752,
-      "eval_samples_per_second": 78.618,
-      "eval_steps_per_second": 9.852,
+      "eval_accuracy": 0.9022556390977443,
+      "eval_f1": 0.8786430103333984,
+      "eval_loss": 0.26596176624298096,
+      "eval_precision": 0.8913001481099878,
+      "eval_recall": 0.8683396981269322,
+      "eval_runtime": 1.7906,
+      "eval_samples_per_second": 222.829,
+      "eval_steps_per_second": 27.923,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 5.6604695320129395,
+      "grad_norm": 7.584296226501465,
       "learning_rate": 1.5e-05,
-      "loss": 0.1684,
+      "loss": 0.1733,
       "step": 1708
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8781334505389722,
-      "eval_loss": 0.27736207842826843,
-      "eval_precision": 0.872316715542522,
-      "eval_recall": 0.8847972358610656,
-      "eval_runtime": 5.281,
-      "eval_samples_per_second": 75.554,
-      "eval_steps_per_second": 9.468,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.8622085718274466,
+      "eval_loss": 0.28683483600616455,
+      "eval_precision": 0.8530168716042322,
+      "eval_recall": 0.8741589379887251,
+      "eval_runtime": 1.7857,
+      "eval_samples_per_second": 223.438,
+      "eval_steps_per_second": 28.0,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 0.4771318733692169,
+      "grad_norm": 1.435178518295288,
       "learning_rate": 1.25e-05,
-      "loss": 0.1548,
+      "loss": 0.1582,
       "step": 1830
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.8802521008403361,
-      "eval_loss": 0.27421677112579346,
-      "eval_precision": 0.8767168083714847,
-      "eval_recall": 0.8840698308783415,
-      "eval_runtime": 5.0853,
-      "eval_samples_per_second": 78.462,
-      "eval_steps_per_second": 9.832,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8648373983739837,
+      "eval_loss": 0.28010857105255127,
+      "eval_precision": 0.8561154177433248,
+      "eval_recall": 0.8759319876341153,
+      "eval_runtime": 1.788,
+      "eval_samples_per_second": 223.159,
+      "eval_steps_per_second": 27.965,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 0.8061890006065369,
+      "grad_norm": 1.7755215167999268,
       "learning_rate": 1e-05,
-      "loss": 0.1526,
+      "loss": 0.1537,
       "step": 1952
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.8872180451127819,
-      "eval_f1": 0.8703282417939551,
-      "eval_loss": 0.29698580503463745,
-      "eval_precision": 0.8574462768615692,
-      "eval_recall": 0.8902073104200764,
-      "eval_runtime": 5.0518,
-      "eval_samples_per_second": 78.982,
-      "eval_steps_per_second": 9.898,
+      "eval_accuracy": 0.87468671679198,
+      "eval_f1": 0.8550061050061051,
+      "eval_loss": 0.30731528997421265,
+      "eval_precision": 0.8437710437710437,
+      "eval_recall": 0.8713402436806692,
+      "eval_runtime": 1.7883,
+      "eval_samples_per_second": 223.118,
+      "eval_steps_per_second": 27.96,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 2.3830015659332275,
+      "grad_norm": 3.1951849460601807,
       "learning_rate": 7.5e-06,
-      "loss": 0.1467,
+      "loss": 0.1537,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8707140332272888,
-      "eval_loss": 0.2729242146015167,
-      "eval_precision": 0.8618432385874246,
-      "eval_recall": 0.8819785415530097,
-      "eval_runtime": 5.0506,
-      "eval_samples_per_second": 79.001,
-      "eval_steps_per_second": 9.9,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8781334505389722,
+      "eval_loss": 0.2702249586582184,
+      "eval_precision": 0.872316715542522,
+      "eval_recall": 0.8847972358610656,
+      "eval_runtime": 1.7845,
+      "eval_samples_per_second": 223.597,
+      "eval_steps_per_second": 28.02,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 4.170547962188721,
+      "grad_norm": 3.639573097229004,
       "learning_rate": 5e-06,
-      "loss": 0.1484,
+      "loss": 0.1461,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8781334505389722,
-      "eval_loss": 0.27389636635780334,
-      "eval_precision": 0.872316715542522,
-      "eval_recall": 0.8847972358610656,
-      "eval_runtime": 5.0897,
-      "eval_samples_per_second": 78.393,
-      "eval_steps_per_second": 9.824,
+      "eval_accuracy": 0.8947368421052632,
+      "eval_f1": 0.8760282890453928,
+      "eval_loss": 0.29228049516677856,
+      "eval_precision": 0.8682260305697083,
+      "eval_recall": 0.8855246408437898,
+      "eval_runtime": 1.8077,
+      "eval_samples_per_second": 220.726,
+      "eval_steps_per_second": 27.66,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 3.1922006607055664,
+      "grad_norm": 3.177137613296509,
       "learning_rate": 2.5e-06,
-      "loss": 0.1434,
+      "loss": 0.1449,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.879667048676036,
-      "eval_loss": 0.27286645770072937,
-      "eval_precision": 0.8778361344537815,
-      "eval_recall": 0.8815693762502272,
-      "eval_runtime": 5.3802,
-      "eval_samples_per_second": 74.161,
-      "eval_steps_per_second": 9.293,
+      "eval_accuracy": 0.8947368421052632,
+      "eval_f1": 0.8754533563232059,
+      "eval_loss": 0.27906104922294617,
+      "eval_precision": 0.8689781021897811,
+      "eval_recall": 0.8830241862156756,
+      "eval_runtime": 1.7931,
+      "eval_samples_per_second": 222.52,
+      "eval_steps_per_second": 27.885,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 1.1650974750518799,
+      "grad_norm": 0.6333515048027039,
       "learning_rate": 0.0,
-      "loss": 0.1354,
+      "loss": 0.1502,
       "step": 2440
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8829621606985718,
-      "eval_loss": 0.2765507400035858,
-      "eval_precision": 0.8802419354838709,
-      "eval_recall": 0.8858428805237315,
-      "eval_runtime": 5.0601,
-      "eval_samples_per_second": 78.853,
-      "eval_steps_per_second": 9.881,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8781334505389722,
+      "eval_loss": 0.2797949016094208,
+      "eval_precision": 0.872316715542522,
+      "eval_recall": 0.8847972358610656,
+      "eval_runtime": 1.7912,
+      "eval_samples_per_second": 222.75,
+      "eval_steps_per_second": 27.914,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
       "total_flos": 8444128359504000.0,
-      "train_loss": 0.23107414167435442,
-      "train_runtime": 1909.2167,
-      "train_samples_per_second": 38.11,
-      "train_steps_per_second": 1.278
+      "train_loss": 0.24180538537072355,
+      "train_runtime": 620.8783,
+      "train_samples_per_second": 117.189,
+      "train_steps_per_second": 3.93
     }
   ],
   "logging_steps": 500,