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": 0.9109792284866469,
+    "accuracy": 0.9060336300692384,
     "epoch": 20.0,
-    "eval_accuracy": 0.9022556390977443,
-    "eval_f1": 0.8817957385392532,
-    "eval_loss": 0.8104944229125977,
-    "eval_precision": 0.8827677592299257,
-    "eval_recall": 0.8808419712675032,
-    "eval_runtime": 4.7093,
+    "eval_accuracy": 0.8972431077694235,
+    "eval_f1": 0.8792560061999484,
+    "eval_loss": 0.8335620164871216,
+    "eval_precision": 0.8707622232472325,
+    "eval_recall": 0.889798145117294,
+    "eval_runtime": 1.6549,
     "eval_samples": 399,
-    "eval_samples_per_second": 84.725,
-    "eval_steps_per_second": 10.617,
-    "f1": 0.8920886346170267,
-    "precision": 0.8953297623033144,
-    "recall": 0.8890334817436486,
-    "train_loss": 0.05662053943168922,
-    "train_runtime": 2686.8503,
+    "eval_samples_per_second": 241.101,
+    "eval_steps_per_second": 30.213,
+    "f1": 0.8885945244345052,
+    "precision": 0.8834872799509323,
+    "recall": 0.8943164810753316,
+    "train_loss": 0.05526667458356404,
+    "train_runtime": 862.9394,
     "train_samples": 3638,
-    "train_samples_per_second": 27.08,
-    "train_steps_per_second": 0.908
+    "train_samples_per_second": 84.316,
+    "train_steps_per_second": 2.828
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.9022556390977443,
-    "eval_f1": 0.8817957385392532,
-    "eval_loss": 0.8104944229125977,
-    "eval_precision": 0.8827677592299257,
-    "eval_recall": 0.8808419712675032,
-    "eval_runtime": 4.7093,
+    "eval_accuracy": 0.8972431077694235,
+    "eval_f1": 0.8792560061999484,
+    "eval_loss": 0.8335620164871216,
+    "eval_precision": 0.8707622232472325,
+    "eval_recall": 0.889798145117294,
+    "eval_runtime": 1.6549,
     "eval_samples": 399,
-    "eval_samples_per_second": 84.725,
-    "eval_steps_per_second": 10.617
+    "eval_samples_per_second": 241.101,
+    "eval_steps_per_second": 30.213
 }

predict_results.json

@@ -1,6 +1,6 @@
 {
-    "accuracy": 0.9109792284866469,
-    "f1": 0.8920886346170267,
-    "precision": 0.8953297623033144,
-    "recall": 0.8890334817436486
+    "accuracy": 0.9060336300692384,
+    "f1": 0.8885945244345052,
+    "precision": 0.8834872799509323,
+    "recall": 0.8943164810753316
 }

predict_results.txt

@@ -1,14 +1,14 @@
 index	prediction
 0	1
-1	1
+1	0
 2	1
 3	1
-4	0
+4	1
 5	1
 6	1
 7	1
 8	0
-9	0
+9	1
 10	1
 11	1
 12	1
@@ -24,7 +24,7 @@ index	prediction
 22	1
 23	1
 24	0
-25	0
+25	1
 26	1
 27	1
 28	1
@@ -33,23 +33,23 @@ index	prediction
 31	1
 32	1
 33	1
-34	1
+34	0
 35	1
 36	1
 37	1
-38	1
-39	0
+38	0
+39	1
 40	1
 41	1
 42	1
 43	1
 44	1
-45	0
+45	1
 46	1
 47	1
 48	1
 49	0
-50	0
+50	1
 51	1
 52	0
 53	1
@@ -58,7 +58,7 @@ index	prediction
 56	1
 57	0
 58	0
-59	0
+59	1
 60	1
 61	1
 62	1
@@ -76,10 +76,10 @@ index	prediction
 74	1
 75	1
 76	1
-77	0
+77	1
 78	1
 79	1
-80	1
+80	0
 81	1
 82	1
 83	1
@@ -96,11 +96,11 @@ index	prediction
 94	1
 95	1
 96	1
-97	0
+97	1
 98	1
-99	1
+99	0
 100	1
-101	0
+101	1
 102	1
 103	1
 104	1
@@ -112,12 +112,12 @@ index	prediction
 110	1
 111	1
 112	1
-113	0
+113	1
 114	1
 115	1
 116	1
 117	1
-118	0
+118	1
 119	1
 120	1
 121	1
@@ -141,7 +141,7 @@ index	prediction
 139	1
 140	1
 141	1
-142	0
+142	1
 143	1
 144	1
 145	1
@@ -149,7 +149,7 @@ index	prediction
 147	1
 148	1
 149	1
-150	0
+150	1
 151	1
 152	1
 153	1
@@ -167,11 +167,11 @@ index	prediction
 165	0
 166	1
 167	1
-168	1
+168	0
 169	1
-170	1
+170	0
 171	1
-172	0
+172	1
 173	0
 174	1
 175	1
@@ -204,7 +204,7 @@ index	prediction
 202	1
 203	1
 204	1
-205	0
+205	1
 206	1
 207	0
 208	1
@@ -216,7 +216,7 @@ index	prediction
 214	0
 215	1
 216	0
-217	0
+217	1
 218	1
 219	1
 220	0
@@ -228,8 +228,8 @@ index	prediction
 226	0
 227	0
 228	1
-229	1
-230	1
+229	0
+230	0
 231	1
 232	1
 233	1
@@ -272,7 +272,7 @@ index	prediction
 270	1
 271	1
 272	1
-273	0
+273	1
 274	1
 275	1
 276	1
@@ -365,7 +365,7 @@ index	prediction
 363	0
 364	0
 365	0
-366	0
+366	1
 367	0
 368	0
 369	0
@@ -391,7 +391,7 @@ index	prediction
 389	0
 390	0
 391	0
-392	1
+392	0
 393	0
 394	0
 395	0
@@ -401,7 +401,7 @@ index	prediction
 399	0
 400	0
 401	0
-402	0
+402	1
 403	0
 404	0
 405	0
@@ -420,14 +420,14 @@ index	prediction
 418	0
 419	0
 420	1
-421	1
+421	0
 422	0
 423	0
 424	0
 425	0
 426	0
 427	0
-428	0
+428	1
 429	0
 430	0
 431	0
@@ -446,17 +446,17 @@ index	prediction
 444	0
 445	0
 446	0
-447	1
+447	0
 448	0
 449	0
 450	0
 451	0
-452	1
+452	0
 453	0
 454	0
 455	0
 456	0
-457	1
+457	0
 458	0
 459	0
 460	0
@@ -510,7 +510,7 @@ index	prediction
 508	0
 509	0
 510	0
-511	1
+511	0
 512	0
 513	0
 514	0
@@ -521,7 +521,7 @@ index	prediction
 519	0
 520	0
 521	0
-522	0
+522	1
 523	0
 524	0
 525	0
@@ -535,7 +535,7 @@ index	prediction
 533	0
 534	0
 535	0
-536	0
+536	1
 537	0
 538	1
 539	0
@@ -559,7 +559,7 @@ index	prediction
 557	0
 558	0
 559	0
-560	1
+560	0
 561	0
 562	0
 563	0
@@ -595,7 +595,7 @@ index	prediction
 593	0
 594	0
 595	0
-596	0
+596	1
 597	0
 598	0
 599	0
@@ -607,9 +607,9 @@ index	prediction
 605	0
 606	0
 607	0
-608	0
+608	1
 609	0
-610	1
+610	0
 611	0
 612	0
 613	0
@@ -625,20 +625,20 @@ index	prediction
 623	0
 624	0
 625	0
-626	0
+626	1
 627	0
-628	0
+628	1
 629	0
 630	0
 631	0
 632	0
-633	0
+633	1
 634	0
 635	0
 636	0
 637	0
 638	0
-639	0
+639	1
 640	0
 641	0
 642	0
@@ -649,7 +649,7 @@ index	prediction
 647	0
 648	0
 649	0
-650	0
+650	1
 651	0
 652	1
 653	0
@@ -668,11 +668,11 @@ index	prediction
 666	0
 667	0
 668	0
-669	0
+669	1
 670	0
 671	0
 672	0
-673	0
+673	1
 674	0
 675	0
 676	0
@@ -702,7 +702,7 @@ index	prediction
 700	0
 701	0
 702	0
-703	0
+703	1
 704	0
 705	0
 706	0
@@ -726,7 +726,7 @@ index	prediction
 724	0
 725	0
 726	0
-727	0
+727	1
 728	1
 729	0
 730	0
@@ -738,7 +738,7 @@ index	prediction
 736	0
 737	0
 738	0
-739	0
+739	1
 740	0
 741	0
 742	0
@@ -769,7 +769,7 @@ index	prediction
 767	0
 768	0
 769	0
-770	0
+770	1
 771	0
 772	0
 773	0
@@ -837,7 +837,7 @@ index	prediction
 835	0
 836	0
 837	0
-838	1
+838	0
 839	0
 840	0
 841	1
@@ -861,7 +861,7 @@ index	prediction
 859	0
 860	0
 861	0
-862	0
+862	1
 863	0
 864	0
 865	0
@@ -885,7 +885,7 @@ index	prediction
 883	0
 884	0
 885	0
-886	0
+886	1
 887	0
 888	0
 889	0
@@ -933,13 +933,13 @@ index	prediction
 931	0
 932	0
 933	0
-934	1
+934	0
 935	0
 936	0
 937	0
 938	0
 939	0
-940	0
+940	1
 941	0
 942	0
 943	1
@@ -949,11 +949,11 @@ index	prediction
 947	0
 948	0
 949	0
-950	0
+950	1
 951	0
 952	0
 953	0
-954	0
+954	1
 955	1
 956	0
 957	0
@@ -968,7 +968,7 @@ index	prediction
 966	0
 967	0
 968	0
-969	0
+969	1
 970	0
 971	0
 972	0
@@ -984,13 +984,13 @@ index	prediction
 982	0
 983	0
 984	0
-985	0
+985	1
 986	1
 987	0
 988	0
 989	0
 990	0
-991	1
+991	0
 992	0
 993	0
 994	0

runs/Jun03_09-42-28_a358b85c7679/events.out.tfevents.1717408644.a358b85c7679.12601.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc2d609c87bd801d879bee42d0a8edaf96cff90b8084f8ad411431b44b674449
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.05662053943168922,
-    "train_runtime": 2686.8503,
+    "train_loss": 0.05526667458356404,
+    "train_runtime": 862.9394,
     "train_samples": 3638,
-    "train_samples_per_second": 27.08,
-    "train_steps_per_second": 0.908
+    "train_samples_per_second": 84.316,
+    "train_steps_per_second": 2.828
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 21.900392532348633,
+      "grad_norm": 78.255126953125,
       "learning_rate": 4.75e-05,
-      "loss": 0.4267,
+      "loss": 0.3942,
       "step": 122
     },
     {
       "epoch": 1.0,
       "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8409196624360422,
-      "eval_loss": 0.35864609479904175,
-      "eval_precision": 0.8892469089546646,
-      "eval_recall": 0.8148754318967084,
-      "eval_runtime": 4.8207,
-      "eval_samples_per_second": 82.769,
-      "eval_steps_per_second": 10.372,
+      "eval_f1": 0.8419489007724301,
+      "eval_loss": 0.3128369450569153,
+      "eval_precision": 0.8857758620689655,
+      "eval_recall": 0.8173758865248226,
+      "eval_runtime": 1.6299,
+      "eval_samples_per_second": 244.801,
+      "eval_steps_per_second": 30.677,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 23.599449157714844,
+      "grad_norm": 42.82415771484375,
       "learning_rate": 4.5e-05,
-      "loss": 0.2234,
+      "loss": 0.2168,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8696741854636592,
-      "eval_f1": 0.8539284708532808,
-      "eval_loss": 0.36683306097984314,
-      "eval_precision": 0.8394707327494125,
-      "eval_recall": 0.8852973267866885,
-      "eval_runtime": 4.9471,
-      "eval_samples_per_second": 80.653,
-      "eval_steps_per_second": 10.107,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8676337535436396,
+      "eval_loss": 0.3043781518936157,
+      "eval_precision": 0.8658613445378152,
+      "eval_recall": 0.8694762684124386,
+      "eval_runtime": 1.6375,
+      "eval_samples_per_second": 243.661,
+      "eval_steps_per_second": 30.534,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 8.989884376525879,
+      "grad_norm": 0.2970781624317169,
       "learning_rate": 4.25e-05,
-      "loss": 0.126,
+      "loss": 0.1372,
       "step": 366
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8755702215614461,
-      "eval_loss": 0.4554330110549927,
-      "eval_precision": 0.8632157235517781,
-      "eval_recall": 0.8937534097108566,
-      "eval_runtime": 4.9562,
-      "eval_samples_per_second": 80.505,
-      "eval_steps_per_second": 10.088,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8595250288055307,
+      "eval_loss": 0.5317866802215576,
+      "eval_precision": 0.885164197446576,
+      "eval_recall": 0.8419712675031824,
+      "eval_runtime": 1.6412,
+      "eval_samples_per_second": 243.114,
+      "eval_steps_per_second": 30.465,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 15.863175392150879,
+      "grad_norm": 0.16418644785881042,
       "learning_rate": 4e-05,
-      "loss": 0.0886,
+      "loss": 0.0957,
       "step": 488
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.9072681704260651,
-      "eval_f1": 0.8854915648632926,
-      "eval_loss": 0.44412538409233093,
-      "eval_precision": 0.8956662848415425,
-      "eval_recall": 0.8768867066739408,
-      "eval_runtime": 4.9565,
-      "eval_samples_per_second": 80.5,
-      "eval_steps_per_second": 10.088,
+      "eval_accuracy": 0.8947368421052632,
+      "eval_f1": 0.8765906680805938,
+      "eval_loss": 0.47654101252555847,
+      "eval_precision": 0.8675710594315245,
+      "eval_recall": 0.888025095471904,
+      "eval_runtime": 1.6551,
+      "eval_samples_per_second": 241.073,
+      "eval_steps_per_second": 30.21,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.03718271106481552,
+      "grad_norm": 0.4955180287361145,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.0611,
+      "loss": 0.0674,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.8839406001224739,
-      "eval_loss": 0.4922772943973541,
-      "eval_precision": 0.8880654743486602,
-      "eval_recall": 0.880114566284779,
-      "eval_runtime": 4.9305,
-      "eval_samples_per_second": 80.925,
-      "eval_steps_per_second": 10.141,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8728804559453431,
+      "eval_loss": 0.552257239818573,
+      "eval_precision": 0.8576773985140519,
+      "eval_recall": 0.9027095835606473,
+      "eval_runtime": 1.6807,
+      "eval_samples_per_second": 237.402,
+      "eval_steps_per_second": 29.75,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 0.006236851681023836,
+      "grad_norm": 0.03946012258529663,
       "learning_rate": 3.5e-05,
-      "loss": 0.0366,
+      "loss": 0.0535,
       "step": 732
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.8813841488792438,
-      "eval_loss": 0.6796092391014099,
-      "eval_precision": 0.8748029197080291,
-      "eval_recall": 0.8890707401345699,
-      "eval_runtime": 4.9528,
-      "eval_samples_per_second": 80.56,
-      "eval_steps_per_second": 10.095,
+      "eval_accuracy": 0.9072681704260651,
+      "eval_f1": 0.8878574955372402,
+      "eval_loss": 0.5158531069755554,
+      "eval_precision": 0.8888448885098087,
+      "eval_recall": 0.8868885251863976,
+      "eval_runtime": 1.6465,
+      "eval_samples_per_second": 242.338,
+      "eval_steps_per_second": 30.368,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.006574722938239574,
+      "grad_norm": 0.008608223870396614,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.0358,
+      "loss": 0.027,
       "step": 854
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.8820775261324042,
-      "eval_loss": 0.5746134519577026,
-      "eval_precision": 0.8934835488413775,
-      "eval_recall": 0.8726132024004365,
-      "eval_runtime": 4.962,
-      "eval_samples_per_second": 80.411,
-      "eval_steps_per_second": 10.077,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8642908431276217,
+      "eval_loss": 0.5940884351730347,
+      "eval_precision": 0.8633964654080464,
+      "eval_recall": 0.8652027641389344,
+      "eval_runtime": 1.6485,
+      "eval_samples_per_second": 242.043,
+      "eval_steps_per_second": 30.331,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 0.0049128723330795765,
+      "grad_norm": 0.010127891786396503,
       "learning_rate": 3e-05,
-      "loss": 0.0272,
+      "loss": 0.0223,
       "step": 976
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8736504011098378,
-      "eval_loss": 0.5952700972557068,
-      "eval_precision": 0.8718487394957983,
-      "eval_recall": 0.8755228223313329,
-      "eval_runtime": 4.9497,
-      "eval_samples_per_second": 80.612,
-      "eval_steps_per_second": 10.102,
+      "eval_accuracy": 0.8796992481203008,
+      "eval_f1": 0.8548827059465357,
+      "eval_loss": 0.7166243195533752,
+      "eval_precision": 0.8548827059465357,
+      "eval_recall": 0.8548827059465357,
+      "eval_runtime": 1.6562,
+      "eval_samples_per_second": 240.913,
+      "eval_steps_per_second": 30.19,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 0.006197785492986441,
+      "grad_norm": 0.005933025386184454,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.0231,
+      "loss": 0.0145,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.8751876876876876,
-      "eval_loss": 0.6506014466285706,
-      "eval_precision": 0.8891156462585035,
-      "eval_recall": 0.864066193853428,
-      "eval_runtime": 4.9626,
-      "eval_samples_per_second": 80.401,
-      "eval_steps_per_second": 10.075,
+      "eval_accuracy": 0.9022556390977443,
+      "eval_f1": 0.8829621606985718,
+      "eval_loss": 0.7022837996482849,
+      "eval_precision": 0.8802419354838709,
+      "eval_recall": 0.8858428805237315,
+      "eval_runtime": 1.6595,
+      "eval_samples_per_second": 240.429,
+      "eval_steps_per_second": 30.129,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 0.009196682833135128,
+      "grad_norm": 0.02505210041999817,
       "learning_rate": 2.5e-05,
-      "loss": 0.0141,
+      "loss": 0.0106,
       "step": 1220
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8823853973772722,
-      "eval_loss": 0.6854431629180908,
-      "eval_precision": 0.8814464081066409,
-      "eval_recall": 0.8833424258956174,
-      "eval_runtime": 4.945,
-      "eval_samples_per_second": 80.688,
-      "eval_steps_per_second": 10.111,
+      "eval_accuracy": 0.9047619047619048,
+      "eval_f1": 0.8839406001224739,
+      "eval_loss": 0.699307918548584,
+      "eval_precision": 0.8880654743486602,
+      "eval_recall": 0.880114566284779,
+      "eval_runtime": 1.6551,
+      "eval_samples_per_second": 241.07,
+      "eval_steps_per_second": 30.209,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 0.01001653354614973,
+      "grad_norm": 0.002501419745385647,
       "learning_rate": 2.25e-05,
-      "loss": 0.023,
+      "loss": 0.0093,
       "step": 1342
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8823853973772722,
-      "eval_loss": 0.7218220829963684,
-      "eval_precision": 0.8814464081066409,
-      "eval_recall": 0.8833424258956174,
-      "eval_runtime": 4.9612,
-      "eval_samples_per_second": 80.424,
-      "eval_steps_per_second": 10.078,
+      "eval_accuracy": 0.8947368421052632,
+      "eval_f1": 0.8703663593044124,
+      "eval_loss": 0.8273664712905884,
+      "eval_precision": 0.8789149003479912,
+      "eval_recall": 0.8630205491907619,
+      "eval_runtime": 1.6583,
+      "eval_samples_per_second": 240.615,
+      "eval_steps_per_second": 30.152,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 0.0031623237300664186,
+      "grad_norm": 0.012166227214038372,
       "learning_rate": 2e-05,
-      "loss": 0.0067,
+      "loss": 0.0086,
       "step": 1464
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8823853973772722,
-      "eval_loss": 0.7694610357284546,
-      "eval_precision": 0.8814464081066409,
-      "eval_recall": 0.8833424258956174,
-      "eval_runtime": 4.9888,
-      "eval_samples_per_second": 79.98,
-      "eval_steps_per_second": 10.023,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8744522298370696,
+      "eval_loss": 0.7971612215042114,
+      "eval_precision": 0.8795731707317074,
+      "eval_recall": 0.8697945080923805,
+      "eval_runtime": 1.6712,
+      "eval_samples_per_second": 238.744,
+      "eval_steps_per_second": 29.918,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 0.002341507002711296,
+      "grad_norm": 0.00197013420984149,
       "learning_rate": 1.75e-05,
-      "loss": 0.0064,
+      "loss": 0.0106,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8602043795620438,
-      "eval_loss": 0.9004446268081665,
-      "eval_precision": 0.849624060150376,
-      "eval_recall": 0.8748863429714493,
-      "eval_runtime": 4.9345,
-      "eval_samples_per_second": 80.86,
-      "eval_steps_per_second": 10.133,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8787009231453675,
+      "eval_loss": 0.7591652870178223,
+      "eval_precision": 0.8714896214896215,
+      "eval_recall": 0.8872976904891798,
+      "eval_runtime": 1.6672,
+      "eval_samples_per_second": 239.329,
+      "eval_steps_per_second": 29.991,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 0.0026841196231544018,
+      "grad_norm": 0.0050615849904716015,
       "learning_rate": 1.5e-05,
-      "loss": 0.0103,
+      "loss": 0.0072,
       "step": 1708
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8835263209107715,
-      "eval_loss": 0.7978267073631287,
-      "eval_precision": 0.8791501449961532,
-      "eval_recall": 0.8883433351518457,
-      "eval_runtime": 4.9323,
-      "eval_samples_per_second": 80.895,
-      "eval_steps_per_second": 10.137,
+      "eval_accuracy": 0.899749373433584,
+      "eval_f1": 0.8813841488792438,
+      "eval_loss": 0.7834069728851318,
+      "eval_precision": 0.8748029197080291,
+      "eval_recall": 0.8890707401345699,
+      "eval_runtime": 1.6555,
+      "eval_samples_per_second": 241.019,
+      "eval_steps_per_second": 30.203,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 0.0031805976759642363,
+      "grad_norm": 0.002086851978674531,
       "learning_rate": 1.25e-05,
-      "loss": 0.0072,
+      "loss": 0.0098,
       "step": 1830
     },
     {
       "epoch": 15.0,
       "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.8790689216221131,
-      "eval_loss": 0.8251467943191528,
-      "eval_precision": 0.8790689216221131,
-      "eval_recall": 0.8790689216221131,
-      "eval_runtime": 4.9785,
-      "eval_samples_per_second": 80.144,
-      "eval_steps_per_second": 10.043,
+      "eval_f1": 0.8802521008403361,
+      "eval_loss": 0.8048883676528931,
+      "eval_precision": 0.8767168083714847,
+      "eval_recall": 0.8840698308783415,
+      "eval_runtime": 1.6591,
+      "eval_samples_per_second": 240.488,
+      "eval_steps_per_second": 30.136,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 0.0014117677928879857,
+      "grad_norm": 0.0012473827227950096,
       "learning_rate": 1e-05,
-      "loss": 0.0054,
+      "loss": 0.0058,
       "step": 1952
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8823853973772722,
-      "eval_loss": 0.7715300917625427,
-      "eval_precision": 0.8814464081066409,
-      "eval_recall": 0.8833424258956174,
-      "eval_runtime": 4.9483,
-      "eval_samples_per_second": 80.634,
-      "eval_steps_per_second": 10.104,
+      "eval_accuracy": 0.899749373433584,
+      "eval_f1": 0.8802521008403361,
+      "eval_loss": 0.7670984268188477,
+      "eval_precision": 0.8767168083714847,
+      "eval_recall": 0.8840698308783415,
+      "eval_runtime": 1.659,
+      "eval_samples_per_second": 240.503,
+      "eval_steps_per_second": 30.138,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 0.0014807094121351838,
+      "grad_norm": 0.00188881263602525,
       "learning_rate": 7.5e-06,
-      "loss": 0.0038,
+      "loss": 0.0035,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.9072681704260651,
-      "eval_f1": 0.8867007927797945,
-      "eval_loss": 0.7821339964866638,
-      "eval_precision": 0.89198606271777,
-      "eval_recall": 0.8818876159301692,
-      "eval_runtime": 4.951,
-      "eval_samples_per_second": 80.589,
-      "eval_steps_per_second": 10.099,
+      "eval_accuracy": 0.9022556390977443,
+      "eval_f1": 0.8856624319419237,
+      "eval_loss": 0.8084732294082642,
+      "eval_precision": 0.8758364312267658,
+      "eval_recall": 0.8983451536643026,
+      "eval_runtime": 1.6569,
+      "eval_samples_per_second": 240.816,
+      "eval_steps_per_second": 30.177,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 0.0014356797328218818,
+      "grad_norm": 0.0014366944087669253,
       "learning_rate": 5e-06,
-      "loss": 0.0021,
+      "loss": 0.0052,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8763538792940554,
-      "eval_loss": 0.8211472630500793,
-      "eval_precision": 0.8754297605404427,
-      "eval_recall": 0.877295871976723,
-      "eval_runtime": 4.9494,
-      "eval_samples_per_second": 80.616,
-      "eval_steps_per_second": 10.102,
+      "eval_accuracy": 0.899749373433584,
+      "eval_f1": 0.8808243727598566,
+      "eval_loss": 0.7721081972122192,
+      "eval_precision": 0.875706963591375,
+      "eval_recall": 0.8865702855064557,
+      "eval_runtime": 1.6546,
+      "eval_samples_per_second": 241.143,
+      "eval_steps_per_second": 30.218,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 0.0011136590037494898,
+      "grad_norm": 0.0011094665387645364,
       "learning_rate": 2.5e-06,
-      "loss": 0.0022,
+      "loss": 0.0028,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.8790689216221131,
-      "eval_loss": 0.8161725401878357,
-      "eval_precision": 0.8790689216221131,
-      "eval_recall": 0.8790689216221131,
-      "eval_runtime": 4.9488,
-      "eval_samples_per_second": 80.626,
-      "eval_steps_per_second": 10.104,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8792560061999484,
+      "eval_loss": 0.8358559608459473,
+      "eval_precision": 0.8707622232472325,
+      "eval_recall": 0.889798145117294,
+      "eval_runtime": 1.6584,
+      "eval_samples_per_second": 240.592,
+      "eval_steps_per_second": 30.149,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 0.0013613783521577716,
+      "grad_norm": 0.0015741140814498067,
       "learning_rate": 0.0,
-      "loss": 0.0027,
+      "loss": 0.0033,
       "step": 2440
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8817957385392532,
-      "eval_loss": 0.8104944229125977,
-      "eval_precision": 0.8827677592299257,
-      "eval_recall": 0.8808419712675032,
-      "eval_runtime": 4.9834,
-      "eval_samples_per_second": 80.065,
-      "eval_steps_per_second": 10.033,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8792560061999484,
+      "eval_loss": 0.8335620164871216,
+      "eval_precision": 0.8707622232472325,
+      "eval_recall": 0.889798145117294,
+      "eval_runtime": 1.6776,
+      "eval_samples_per_second": 237.834,
+      "eval_steps_per_second": 29.804,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
       "total_flos": 7584162436176000.0,
-      "train_loss": 0.05662053943168922,
-      "train_runtime": 2686.8503,
-      "train_samples_per_second": 27.08,
-      "train_steps_per_second": 0.908
+      "train_loss": 0.05526667458356404,
+      "train_runtime": 862.9394,
+      "train_samples_per_second": 84.316,
+      "train_steps_per_second": 2.828
     }
   ],
   "logging_steps": 500,