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": 0.9010880316518298,
+    "accuracy": 0.9090009891196835,
     "epoch": 20.0,
-    "eval_accuracy": 0.8822055137844611,
-    "eval_f1": 0.8589543987905864,
-    "eval_loss": 0.32526540756225586,
-    "eval_precision": 0.8564068100358423,
-    "eval_recall": 0.8616566648481543,
-    "eval_runtime": 5.0035,
+    "eval_accuracy": 0.8847117794486216,
+    "eval_f1": 0.8609292598654301,
+    "eval_loss": 0.29736196994781494,
+    "eval_precision": 0.8609292598654301,
+    "eval_recall": 0.8609292598654301,
+    "eval_runtime": 1.8087,
     "eval_samples": 399,
-    "eval_samples_per_second": 79.744,
-    "eval_steps_per_second": 9.993,
-    "f1": 0.8826220220127247,
-    "precision": 0.8778956201166501,
-    "recall": 0.8878803994841722,
-    "train_loss": 0.21670899273919278,
-    "train_runtime": 1925.8785,
+    "eval_samples_per_second": 220.599,
+    "eval_steps_per_second": 27.644,
+    "f1": 0.8922124878320122,
+    "precision": 0.8867530544919537,
+    "recall": 0.8983734480454079,
+    "train_loss": 0.23053875125822473,
+    "train_runtime": 628.042,
     "train_samples": 3638,
-    "train_samples_per_second": 37.78,
-    "train_steps_per_second": 1.267
+    "train_samples_per_second": 115.852,
+    "train_steps_per_second": 3.885
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.8822055137844611,
-    "eval_f1": 0.8589543987905864,
-    "eval_loss": 0.32526540756225586,
-    "eval_precision": 0.8564068100358423,
-    "eval_recall": 0.8616566648481543,
-    "eval_runtime": 5.0035,
+    "eval_accuracy": 0.8847117794486216,
+    "eval_f1": 0.8609292598654301,
+    "eval_loss": 0.29736196994781494,
+    "eval_precision": 0.8609292598654301,
+    "eval_recall": 0.8609292598654301,
+    "eval_runtime": 1.8087,
     "eval_samples": 399,
-    "eval_samples_per_second": 79.744,
-    "eval_steps_per_second": 9.993
+    "eval_samples_per_second": 220.599,
+    "eval_steps_per_second": 27.644
 }

predict_results.json

@@ -1,6 +1,6 @@
 {
-    "accuracy": 0.9010880316518298,
-    "f1": 0.8826220220127247,
-    "precision": 0.8778956201166501,
-    "recall": 0.8878803994841722
+    "accuracy": 0.9090009891196835,
+    "f1": 0.8922124878320122,
+    "precision": 0.8867530544919537,
+    "recall": 0.8983734480454079
 }

predict_results.txt

@@ -5,7 +5,7 @@ index	prediction
 3	1
 4	1
 5	1
-6	0
+6	1
 7	1
 8	0
 9	1
@@ -36,7 +36,7 @@ index	prediction
 34	1
 35	1
 36	1
-37	1
+37	0
 38	1
 39	1
 40	1
@@ -75,21 +75,21 @@ index	prediction
 73	1
 74	1
 75	1
-76	1
+76	0
 77	1
 78	0
 79	1
 80	1
 81	1
 82	1
-83	0
+83	1
 84	1
 85	1
 86	1
 87	1
 88	1
 89	1
-90	1
+90	0
 91	1
 92	1
 93	1
@@ -128,7 +128,7 @@ index	prediction
 126	1
 127	1
 128	1
-129	0
+129	1
 130	1
 131	1
 132	1
@@ -144,13 +144,13 @@ index	prediction
 142	1
 143	1
 144	1
-145	0
+145	1
 146	1
 147	1
-148	0
+148	1
 149	1
 150	1
-151	0
+151	1
 152	1
 153	1
 154	1
@@ -165,10 +165,10 @@ index	prediction
 163	1
 164	1
 165	1
-166	1
+166	0
 167	0
 168	1
-169	0
+169	1
 170	1
 171	1
 172	1
@@ -186,7 +186,7 @@ index	prediction
 184	1
 185	1
 186	1
-187	1
+187	0
 188	1
 189	0
 190	1
@@ -211,7 +211,7 @@ index	prediction
 209	1
 210	1
 211	1
-212	0
+212	1
 213	1
 214	1
 215	0
@@ -224,7 +224,7 @@ index	prediction
 222	1
 223	1
 224	1
-225	0
+225	1
 226	1
 227	0
 228	1
@@ -247,7 +247,7 @@ index	prediction
 245	1
 246	1
 247	1
-248	0
+248	1
 249	1
 250	1
 251	1
@@ -267,7 +267,7 @@ index	prediction
 265	1
 266	1
 267	1
-268	1
+268	0
 269	1
 270	1
 271	1
@@ -281,12 +281,12 @@ index	prediction
 279	1
 280	1
 281	1
-282	0
+282	1
 283	1
-284	0
+284	1
 285	1
 286	1
-287	1
+287	0
 288	1
 289	0
 290	1
@@ -365,7 +365,7 @@ index	prediction
 363	0
 364	0
 365	0
-366	1
+366	0
 367	0
 368	0
 369	0
@@ -385,7 +385,7 @@ index	prediction
 383	0
 384	1
 385	0
-386	0
+386	1
 387	0
 388	0
 389	0
@@ -397,7 +397,7 @@ index	prediction
 395	0
 396	0
 397	0
-398	1
+398	0
 399	0
 400	0
 401	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
@@ -472,7 +472,7 @@ index	prediction
 470	0
 471	0
 472	0
-473	0
+473	1
 474	1
 475	0
 476	1
@@ -566,7 +566,7 @@ index	prediction
 564	0
 565	1
 566	0
-567	0
+567	1
 568	0
 569	0
 570	0
@@ -590,12 +590,12 @@ index	prediction
 588	0
 589	0
 590	0
-591	0
+591	1
 592	0
 593	0
 594	0
 595	0
-596	1
+596	0
 597	1
 598	0
 599	0
@@ -617,7 +617,7 @@ index	prediction
 615	0
 616	0
 617	0
-618	1
+618	0
 619	0
 620	0
 621	0
@@ -648,7 +648,7 @@ index	prediction
 646	0
 647	0
 648	0
-649	1
+649	0
 650	0
 651	0
 652	0
@@ -672,8 +672,8 @@ index	prediction
 670	0
 671	0
 672	0
-673	1
-674	0
+673	0
+674	1
 675	0
 676	0
 677	0
@@ -696,11 +696,11 @@ index	prediction
 694	0
 695	0
 696	0
-697	1
+697	0
 698	0
 699	0
 700	0
-701	1
+701	0
 702	0
 703	0
 704	0
@@ -725,9 +725,9 @@ index	prediction
 723	0
 724	0
 725	0
-726	1
+726	0
 727	0
-728	1
+728	0
 729	0
 730	0
 731	0
@@ -801,7 +801,7 @@ index	prediction
 799	0
 800	0
 801	0
-802	0
+802	1
 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
@@ -853,13 +853,13 @@ index	prediction
 851	0
 852	0
 853	0
-854	0
+854	1
 855	0
 856	0
 857	0
 858	0
 859	0
-860	0
+860	1
 861	0
 862	0
 863	0
@@ -912,7 +912,7 @@ index	prediction
 910	0
 911	0
 912	0
-913	0
+913	1
 914	0
 915	0
 916	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	1
+967	0
 968	0
 969	0
 970	0
@@ -1001,7 +1001,7 @@ index	prediction
 999	1
 1000	0
 1001	0
-1002	1
+1002	0
 1003	0
 1004	1
 1005	0

runs/Jun03_14-39-35_a358b85c7679/events.out.tfevents.1717426219.a358b85c7679.146701.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f777f3ea51e241fe7c17b7ea22ee042d0ba1bc84101b1e2eee6519c6d20376ed
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.21670899273919278,
-    "train_runtime": 1925.8785,
+    "train_loss": 0.23053875125822473,
+    "train_runtime": 628.042,
     "train_samples": 3638,
-    "train_samples_per_second": 37.78,
-    "train_steps_per_second": 1.267
+    "train_samples_per_second": 115.852,
+    "train_steps_per_second": 3.885
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 6.052820205688477,
+      "grad_norm": 3.532687187194824,
       "learning_rate": 4.75e-05,
-      "loss": 0.5445,
+      "loss": 0.551,
       "step": 122
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7518796992481203,
-      "eval_f1": 0.6708167704192605,
-      "eval_loss": 0.48936620354652405,
-      "eval_precision": 0.6988095238095238,
-      "eval_recall": 0.6594380796508457,
-      "eval_runtime": 5.1391,
-      "eval_samples_per_second": 77.64,
-      "eval_steps_per_second": 9.729,
+      "eval_accuracy": 0.7243107769423559,
+      "eval_f1": 0.6144317942230656,
+      "eval_loss": 0.5008909106254578,
+      "eval_precision": 0.65566534914361,
+      "eval_recall": 0.6074286233860702,
+      "eval_runtime": 1.7719,
+      "eval_samples_per_second": 225.176,
+      "eval_steps_per_second": 28.218,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 3.3890905380249023,
+      "grad_norm": 3.3572421073913574,
       "learning_rate": 4.5e-05,
-      "loss": 0.432,
+      "loss": 0.4528,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8195488721804511,
-      "eval_f1": 0.7954312407428505,
-      "eval_loss": 0.36909204721450806,
-      "eval_precision": 0.7842791202823785,
-      "eval_recall": 0.8173304237134025,
-      "eval_runtime": 5.0614,
-      "eval_samples_per_second": 78.833,
-      "eval_steps_per_second": 9.879,
+      "eval_accuracy": 0.7819548872180451,
+      "eval_f1": 0.7578021978021978,
+      "eval_loss": 0.4117695391178131,
+      "eval_precision": 0.7486942070275404,
+      "eval_recall": 0.7857337697763229,
+      "eval_runtime": 1.781,
+      "eval_samples_per_second": 224.038,
+      "eval_steps_per_second": 28.075,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 2.7538790702819824,
+      "grad_norm": 2.5434460639953613,
       "learning_rate": 4.25e-05,
-      "loss": 0.3342,
+      "loss": 0.3588,
       "step": 366
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.849624060150376,
-      "eval_f1": 0.7956919505700826,
-      "eval_loss": 0.33010444045066833,
-      "eval_precision": 0.85742518351214,
-      "eval_recall": 0.7685942898708855,
-      "eval_runtime": 5.0669,
-      "eval_samples_per_second": 78.747,
-      "eval_steps_per_second": 9.868,
+      "eval_accuracy": 0.8521303258145363,
+      "eval_f1": 0.8063505120717311,
+      "eval_loss": 0.3427870571613312,
+      "eval_precision": 0.844033177433499,
+      "eval_recall": 0.7853700672849608,
+      "eval_runtime": 1.7815,
+      "eval_samples_per_second": 223.975,
+      "eval_steps_per_second": 28.067,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 3.051776170730591,
+      "grad_norm": 4.183424949645996,
       "learning_rate": 4e-05,
-      "loss": 0.2846,
+      "loss": 0.3192,
       "step": 488
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8502252252252251,
-      "eval_loss": 0.2885676324367523,
-      "eval_precision": 0.863265306122449,
-      "eval_recall": 0.8398799781778505,
-      "eval_runtime": 5.0493,
-      "eval_samples_per_second": 79.021,
-      "eval_steps_per_second": 9.902,
+      "eval_accuracy": 0.8546365914786967,
+      "eval_f1": 0.8246499363520641,
+      "eval_loss": 0.31169387698173523,
+      "eval_precision": 0.8246499363520641,
+      "eval_recall": 0.8246499363520641,
+      "eval_runtime": 1.8089,
+      "eval_samples_per_second": 220.578,
+      "eval_steps_per_second": 27.641,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 3.2031478881835938,
+      "grad_norm": 1.7957929372787476,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.2621,
+      "loss": 0.2714,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.87468671679198,
-      "eval_f1": 0.8488361520276414,
-      "eval_loss": 0.27282822132110596,
-      "eval_precision": 0.8488361520276414,
-      "eval_recall": 0.8488361520276414,
-      "eval_runtime": 5.0626,
-      "eval_samples_per_second": 78.814,
-      "eval_steps_per_second": 9.876,
+      "eval_accuracy": 0.8671679197994987,
+      "eval_f1": 0.8446181767415888,
+      "eval_loss": 0.3037036657333374,
+      "eval_precision": 0.835902201887332,
+      "eval_recall": 0.8560192762320422,
+      "eval_runtime": 1.8052,
+      "eval_samples_per_second": 221.025,
+      "eval_steps_per_second": 27.697,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 1.391174554824829,
+      "grad_norm": 3.427933692932129,
       "learning_rate": 3.5e-05,
-      "loss": 0.2419,
+      "loss": 0.257,
       "step": 732
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.87468671679198,
-      "eval_f1": 0.8503151260504201,
-      "eval_loss": 0.27526727318763733,
-      "eval_precision": 0.8470628455912955,
-      "eval_recall": 0.8538370612838698,
-      "eval_runtime": 5.0607,
-      "eval_samples_per_second": 78.843,
-      "eval_steps_per_second": 9.88,
+      "eval_accuracy": 0.8771929824561403,
+      "eval_f1": 0.8475258334958082,
+      "eval_loss": 0.2832619249820709,
+      "eval_precision": 0.8591828192414193,
+      "eval_recall": 0.8381069285324605,
+      "eval_runtime": 1.8047,
+      "eval_samples_per_second": 221.086,
+      "eval_steps_per_second": 27.705,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.6138697862625122,
+      "grad_norm": 1.5598583221435547,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.2132,
+      "loss": 0.2405,
       "step": 854
     },
     {
       "epoch": 7.0,
       "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8622899159663866,
-      "eval_loss": 0.27531638741493225,
-      "eval_precision": 0.8589244307033712,
-      "eval_recall": 0.8659301691216585,
-      "eval_runtime": 5.0703,
-      "eval_samples_per_second": 78.693,
-      "eval_steps_per_second": 9.861,
+      "eval_f1": 0.8616171059774413,
+      "eval_loss": 0.28606978058815,
+      "eval_precision": 0.859873949579832,
+      "eval_recall": 0.8634297144935443,
+      "eval_runtime": 1.8066,
+      "eval_samples_per_second": 220.859,
+      "eval_steps_per_second": 27.677,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 1.8498824834823608,
+      "grad_norm": 1.1964571475982666,
       "learning_rate": 3e-05,
-      "loss": 0.2055,
+      "loss": 0.2163,
       "step": 976
     },
     {
       "epoch": 8.0,
       "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8541488972828073,
-      "eval_loss": 0.27910253405570984,
-      "eval_precision": 0.8559859154929578,
-      "eval_recall": 0.8523822513184216,
-      "eval_runtime": 5.0606,
-      "eval_samples_per_second": 78.844,
-      "eval_steps_per_second": 9.88,
+      "eval_f1": 0.8556004584112431,
+      "eval_loss": 0.295376181602478,
+      "eval_precision": 0.8538865546218487,
+      "eval_recall": 0.85738316057465,
+      "eval_runtime": 1.8106,
+      "eval_samples_per_second": 220.364,
+      "eval_steps_per_second": 27.614,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 6.801692485809326,
+      "grad_norm": 4.184772491455078,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.1903,
+      "loss": 0.2135,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.8621553884711779,
-      "eval_f1": 0.8306935047100303,
-      "eval_loss": 0.3009396195411682,
-      "eval_precision": 0.8372758729160114,
-      "eval_recall": 0.8249681760320058,
-      "eval_runtime": 5.0892,
-      "eval_samples_per_second": 78.402,
-      "eval_steps_per_second": 9.825,
+      "eval_accuracy": 0.87468671679198,
+      "eval_f1": 0.8472902633190447,
+      "eval_loss": 0.29421547055244446,
+      "eval_precision": 0.8510272912927781,
+      "eval_recall": 0.8438352427714131,
+      "eval_runtime": 1.8081,
+      "eval_samples_per_second": 220.677,
+      "eval_steps_per_second": 27.654,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 5.017825126647949,
+      "grad_norm": 4.817326545715332,
       "learning_rate": 2.5e-05,
-      "loss": 0.1852,
+      "loss": 0.2001,
       "step": 1220
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.8671679197994987,
-      "eval_f1": 0.8341632880321839,
-      "eval_loss": 0.3085351884365082,
-      "eval_precision": 0.8479139504563233,
-      "eval_recall": 0.8235133660665576,
-      "eval_runtime": 5.0614,
-      "eval_samples_per_second": 78.832,
-      "eval_steps_per_second": 9.879,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.8537492688633261,
+      "eval_loss": 0.3002479076385498,
+      "eval_precision": 0.8656062850151329,
+      "eval_recall": 0.8441534824513548,
+      "eval_runtime": 1.8147,
+      "eval_samples_per_second": 219.877,
+      "eval_steps_per_second": 27.553,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 3.8085110187530518,
+      "grad_norm": 2.364151954650879,
       "learning_rate": 2.25e-05,
-      "loss": 0.1758,
+      "loss": 0.1825,
       "step": 1342
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8556004584112431,
-      "eval_loss": 0.28515923023223877,
-      "eval_precision": 0.8538865546218487,
-      "eval_recall": 0.85738316057465,
-      "eval_runtime": 5.053,
-      "eval_samples_per_second": 78.964,
-      "eval_steps_per_second": 9.895,
+      "eval_accuracy": 0.8922305764411027,
+      "eval_f1": 0.8676331036823873,
+      "eval_loss": 0.30105578899383545,
+      "eval_precision": 0.8749292230261088,
+      "eval_recall": 0.8612474995453718,
+      "eval_runtime": 1.7869,
+      "eval_samples_per_second": 223.295,
+      "eval_steps_per_second": 27.982,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 4.22409725189209,
+      "grad_norm": 4.355996608734131,
       "learning_rate": 2e-05,
-      "loss": 0.1617,
+      "loss": 0.1765,
       "step": 1464
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.8872180451127819,
-      "eval_f1": 0.8642908431276217,
-      "eval_loss": 0.30107688903808594,
-      "eval_precision": 0.8633964654080464,
-      "eval_recall": 0.8652027641389344,
-      "eval_runtime": 5.0864,
-      "eval_samples_per_second": 78.444,
-      "eval_steps_per_second": 9.83,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8656154317207594,
+      "eval_loss": 0.28576698899269104,
+      "eval_precision": 0.869546382820719,
+      "eval_recall": 0.861974904528096,
+      "eval_runtime": 1.7863,
+      "eval_samples_per_second": 223.361,
+      "eval_steps_per_second": 27.99,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 4.442287921905518,
+      "grad_norm": 1.647818922996521,
       "learning_rate": 1.75e-05,
-      "loss": 0.1581,
+      "loss": 0.1674,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8703223612108386,
-      "eval_loss": 0.3050229847431183,
-      "eval_precision": 0.8694131129742446,
-      "eval_recall": 0.8712493180578287,
-      "eval_runtime": 5.0713,
-      "eval_samples_per_second": 78.678,
-      "eval_steps_per_second": 9.859,
+      "eval_accuracy": 0.8947368421052632,
+      "eval_f1": 0.8748655913978494,
+      "eval_loss": 0.29319024085998535,
+      "eval_precision": 0.86983032873807,
+      "eval_recall": 0.8805237315875614,
+      "eval_runtime": 1.7923,
+      "eval_samples_per_second": 222.613,
+      "eval_steps_per_second": 27.896,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 3.8908846378326416,
+      "grad_norm": 2.9686295986175537,
       "learning_rate": 1.5e-05,
-      "loss": 0.149,
+      "loss": 0.1597,
       "step": 1708
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8744925055412909,
-      "eval_loss": 0.3143095374107361,
-      "eval_precision": 0.8639342866870956,
-      "eval_recall": 0.8887525004546282,
-      "eval_runtime": 5.0958,
-      "eval_samples_per_second": 78.299,
-      "eval_steps_per_second": 9.812,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8668668668668669,
+      "eval_loss": 0.2937151789665222,
+      "eval_precision": 0.8598901098901099,
+      "eval_recall": 0.8752045826513912,
+      "eval_runtime": 1.8072,
+      "eval_samples_per_second": 220.782,
+      "eval_steps_per_second": 27.667,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 3.6980299949645996,
+      "grad_norm": 2.176563262939453,
       "learning_rate": 1.25e-05,
-      "loss": 0.1386,
+      "loss": 0.1564,
       "step": 1830
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.8790689216221131,
-      "eval_loss": 0.302778035402298,
-      "eval_precision": 0.8790689216221131,
-      "eval_recall": 0.8790689216221131,
-      "eval_runtime": 5.0849,
-      "eval_samples_per_second": 78.468,
-      "eval_steps_per_second": 9.833,
+      "eval_accuracy": 0.8947368421052632,
+      "eval_f1": 0.8717238211879976,
+      "eval_loss": 0.2963174879550934,
+      "eval_precision": 0.8757194133300328,
+      "eval_recall": 0.8680214584469903,
+      "eval_runtime": 1.8185,
+      "eval_samples_per_second": 219.414,
+      "eval_steps_per_second": 27.495,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 2.6725683212280273,
+      "grad_norm": 0.7651334404945374,
       "learning_rate": 1e-05,
-      "loss": 0.1465,
+      "loss": 0.142,
       "step": 1952
     },
     {
       "epoch": 16.0,
       "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8696722245432793,
-      "eval_loss": 0.3111669719219208,
-      "eval_precision": 0.8706135006701596,
-      "eval_recall": 0.8687488634297145,
-      "eval_runtime": 5.0592,
-      "eval_samples_per_second": 78.866,
-      "eval_steps_per_second": 9.883,
+      "eval_f1": 0.8683279483657071,
+      "eval_loss": 0.302460253238678,
+      "eval_precision": 0.873366724738676,
+      "eval_recall": 0.863747954173486,
+      "eval_runtime": 1.8132,
+      "eval_samples_per_second": 220.059,
+      "eval_steps_per_second": 27.576,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 0.7393283247947693,
+      "grad_norm": 0.19514738023281097,
       "learning_rate": 7.5e-06,
-      "loss": 0.1307,
+      "loss": 0.143,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8629480286738351,
-      "eval_loss": 0.31975144147872925,
-      "eval_precision": 0.8580770590314599,
-      "eval_recall": 0.8684306237497728,
-      "eval_runtime": 5.0777,
-      "eval_samples_per_second": 78.579,
-      "eval_steps_per_second": 9.847,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8682773109243698,
+      "eval_loss": 0.2950632870197296,
+      "eval_precision": 0.864855223259409,
+      "eval_recall": 0.8719767230405528,
+      "eval_runtime": 1.8087,
+      "eval_samples_per_second": 220.597,
+      "eval_steps_per_second": 27.644,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 0.382880836725235,
+      "grad_norm": 0.15066123008728027,
       "learning_rate": 5e-06,
-      "loss": 0.1231,
+      "loss": 0.1315,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8609292598654301,
-      "eval_loss": 0.32525473833084106,
-      "eval_precision": 0.8609292598654301,
-      "eval_recall": 0.8609292598654301,
-      "eval_runtime": 5.096,
-      "eval_samples_per_second": 78.296,
-      "eval_steps_per_second": 9.812,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.858259325044405,
+      "eval_loss": 0.3012860417366028,
+      "eval_precision": 0.8573798178418481,
+      "eval_recall": 0.8591562102200401,
+      "eval_runtime": 1.8058,
+      "eval_samples_per_second": 220.949,
+      "eval_steps_per_second": 27.688,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 0.1525256484746933,
+      "grad_norm": 0.04642534255981445,
       "learning_rate": 2.5e-06,
-      "loss": 0.1344,
+      "loss": 0.1378,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8541488972828073,
-      "eval_loss": 0.3289809823036194,
-      "eval_precision": 0.8559859154929578,
-      "eval_recall": 0.8523822513184216,
-      "eval_runtime": 5.0693,
-      "eval_samples_per_second": 78.709,
-      "eval_steps_per_second": 9.863,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8629148629148629,
+      "eval_loss": 0.30377450585365295,
+      "eval_precision": 0.8657894736842104,
+      "eval_recall": 0.860201854882706,
+      "eval_runtime": 1.8121,
+      "eval_samples_per_second": 220.187,
+      "eval_steps_per_second": 27.592,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 4.0254364013671875,
+      "grad_norm": 3.5789785385131836,
       "learning_rate": 0.0,
-      "loss": 0.1229,
+      "loss": 0.1333,
       "step": 2440
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.8589543987905864,
-      "eval_loss": 0.32526540756225586,
-      "eval_precision": 0.8564068100358423,
-      "eval_recall": 0.8616566648481543,
-      "eval_runtime": 5.0827,
-      "eval_samples_per_second": 78.502,
-      "eval_steps_per_second": 9.837,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8609292598654301,
+      "eval_loss": 0.29736196994781494,
+      "eval_precision": 0.8609292598654301,
+      "eval_recall": 0.8609292598654301,
+      "eval_runtime": 1.814,
+      "eval_samples_per_second": 219.959,
+      "eval_steps_per_second": 27.564,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
       "total_flos": 8444128359504000.0,
-      "train_loss": 0.21670899273919278,
-      "train_runtime": 1925.8785,
-      "train_samples_per_second": 37.78,
-      "train_steps_per_second": 1.267
+      "train_loss": 0.23053875125822473,
+      "train_runtime": 628.042,
+      "train_samples_per_second": 115.852,
+      "train_steps_per_second": 3.885
     }
   ],
   "logging_steps": 500,