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": 0.9090009891196835,
     "epoch": 20.0,
-    "eval_accuracy": 0.9022556390977443,
-    "eval_f1": 0.8811928811928812,
-    "eval_loss": 0.29284632205963135,
-    "eval_precision": 0.8842105263157894,
-    "eval_recall": 0.878341516639389,
-    "eval_runtime": 5.4499,
+    "eval_accuracy": 0.8872180451127819,
+    "eval_f1": 0.8662440310793597,
+    "eval_loss": 0.3288058042526245,
+    "eval_precision": 0.8606158357771261,
+    "eval_recall": 0.872704128023277,
+    "eval_runtime": 2.0252,
     "eval_samples": 399,
-    "eval_samples_per_second": 73.212,
-    "eval_steps_per_second": 9.174,
-    "f1": 0.8907735522904062,
-    "precision": 0.8900210970464135,
-    "recall": 0.8915373175070833,
-    "train_loss": 0.2130153269064231,
-    "train_runtime": 2293.0977,
+    "eval_samples_per_second": 197.016,
+    "eval_steps_per_second": 24.689,
+    "f1": 0.8918102806521253,
+    "precision": 0.8875983617873084,
+    "recall": 0.8964202678916009,
+    "train_loss": 0.2203434850348801,
+    "train_runtime": 746.002,
     "train_samples": 3638,
-    "train_samples_per_second": 31.73,
-    "train_steps_per_second": 1.064
+    "train_samples_per_second": 97.533,
+    "train_steps_per_second": 3.271
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.9022556390977443,
-    "eval_f1": 0.8811928811928812,
-    "eval_loss": 0.29284632205963135,
-    "eval_precision": 0.8842105263157894,
-    "eval_recall": 0.878341516639389,
-    "eval_runtime": 5.4499,
+    "eval_accuracy": 0.8872180451127819,
+    "eval_f1": 0.8662440310793597,
+    "eval_loss": 0.3288058042526245,
+    "eval_precision": 0.8606158357771261,
+    "eval_recall": 0.872704128023277,
+    "eval_runtime": 2.0252,
     "eval_samples": 399,
-    "eval_samples_per_second": 73.212,
-    "eval_steps_per_second": 9.174
+    "eval_samples_per_second": 197.016,
+    "eval_steps_per_second": 24.689
 }

predict_results.json

@@ -1,6 +1,6 @@
 {
     "accuracy": 0.9090009891196835,
-    "f1": 0.8907735522904062,
-    "precision": 0.8900210970464135,
-    "recall": 0.8915373175070833
+    "f1": 0.8918102806521253,
+    "precision": 0.8875983617873084,
+    "recall": 0.8964202678916009
 }

predict_results.txt

@@ -21,7 +21,7 @@ index	prediction
 19	1
 20	1
 21	1
-22	0
+22	1
 23	1
 24	1
 25	0
@@ -40,12 +40,12 @@ index	prediction
 38	1
 39	1
 40	1
-41	0
+41	1
 42	1
 43	1
 44	1
 45	0
-46	0
+46	1
 47	1
 48	1
 49	1
@@ -67,7 +67,7 @@ index	prediction
 65	1
 66	1
 67	0
-68	1
+68	0
 69	1
 70	1
 71	1
@@ -75,7 +75,7 @@ index	prediction
 73	1
 74	1
 75	0
-76	0
+76	1
 77	1
 78	1
 79	1
@@ -104,9 +104,9 @@ index	prediction
 102	0
 103	1
 104	1
-105	0
+105	1
 106	1
-107	1
+107	0
 108	1
 109	1
 110	1
@@ -166,7 +166,7 @@ index	prediction
 164	1
 165	1
 166	1
-167	0
+167	1
 168	1
 169	1
 170	0
@@ -180,18 +180,18 @@ index	prediction
 178	1
 179	0
 180	1
-181	0
+181	1
 182	1
 183	1
 184	1
-185	1
+185	0
 186	1
 187	1
 188	1
 189	1
 190	1
 191	1
-192	0
+192	1
 193	1
 194	1
 195	1
@@ -228,7 +228,7 @@ index	prediction
 226	1
 227	1
 228	1
-229	0
+229	1
 230	1
 231	1
 232	1
@@ -236,7 +236,7 @@ index	prediction
 234	1
 235	0
 236	1
-237	0
+237	1
 238	1
 239	1
 240	1
@@ -247,21 +247,21 @@ index	prediction
 245	1
 246	1
 247	1
-248	0
+248	1
 249	1
 250	1
 251	1
 252	1
-253	0
+253	1
 254	1
 255	1
 256	1
-257	1
+257	0
 258	1
 259	1
 260	1
 261	1
-262	1
+262	0
 263	1
 264	1
 265	1
@@ -277,13 +277,13 @@ index	prediction
 275	0
 276	0
 277	1
-278	1
+278	0
 279	1
 280	1
 281	0
 282	1
 283	1
-284	1
+284	0
 285	1
 286	1
 287	1
@@ -322,13 +322,13 @@ index	prediction
 320	0
 321	0
 322	0
-323	1
+323	0
 324	1
 325	0
 326	0
 327	0
 328	0
-329	0
+329	1
 330	0
 331	0
 332	0
@@ -346,7 +346,7 @@ index	prediction
 344	0
 345	0
 346	0
-347	0
+347	1
 348	0
 349	0
 350	0
@@ -382,7 +382,7 @@ index	prediction
 380	0
 381	0
 382	1
-383	0
+383	1
 384	0
 385	0
 386	0
@@ -470,7 +470,7 @@ index	prediction
 468	0
 469	0
 470	0
-471	0
+471	1
 472	0
 473	0
 474	0
@@ -490,19 +490,19 @@ index	prediction
 488	0
 489	0
 490	0
-491	1
+491	0
 492	0
 493	0
 494	0
 495	0
 496	0
 497	0
-498	1
+498	0
 499	0
 500	0
 501	0
 502	0
-503	1
+503	0
 504	0
 505	0
 506	0
@@ -587,11 +587,11 @@ index	prediction
 585	0
 586	0
 587	0
-588	0
+588	1
 589	0
 590	0
 591	1
-592	0
+592	1
 593	0
 594	0
 595	0
@@ -642,7 +642,7 @@ index	prediction
 640	0
 641	0
 642	0
-643	0
+643	1
 644	0
 645	0
 646	0
@@ -678,7 +678,7 @@ index	prediction
 676	0
 677	0
 678	0
-679	0
+679	1
 680	0
 681	0
 682	0
@@ -953,7 +953,7 @@ index	prediction
 951	0
 952	0
 953	0
-954	0
+954	1
 955	0
 956	0
 957	0

runs/Jun03_15-48-01_a358b85c7679/events.out.tfevents.1717430443.a358b85c7679.183185.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25b94d9bcf029836ebd4c3bcf6a7b0e59fc0deaf499f3fd73103deebfdc1809a
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.2130153269064231,
-    "train_runtime": 2293.0977,
+    "train_loss": 0.2203434850348801,
+    "train_runtime": 746.002,
     "train_samples": 3638,
-    "train_samples_per_second": 31.73,
-    "train_steps_per_second": 1.064
+    "train_samples_per_second": 97.533,
+    "train_steps_per_second": 3.271
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 4.303664207458496,
+      "grad_norm": 6.6789751052856445,
       "learning_rate": 4.75e-05,
-      "loss": 0.5535,
+      "loss": 0.5528,
       "step": 122
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7293233082706767,
-      "eval_f1": 0.6372727272727272,
-      "eval_loss": 0.49923330545425415,
-      "eval_precision": 0.6645702306079665,
-      "eval_recall": 0.6284779050736498,
-      "eval_runtime": 5.6404,
-      "eval_samples_per_second": 70.74,
-      "eval_steps_per_second": 8.865,
+      "eval_accuracy": 0.7167919799498746,
+      "eval_f1": 0.6444270944235455,
+      "eval_loss": 0.5050782561302185,
+      "eval_precision": 0.6520598138245197,
+      "eval_recall": 0.639616293871613,
+      "eval_runtime": 1.9903,
+      "eval_samples_per_second": 200.468,
+      "eval_steps_per_second": 25.121,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 4.080423831939697,
+      "grad_norm": 4.4306488037109375,
       "learning_rate": 4.5e-05,
-      "loss": 0.444,
+      "loss": 0.4431,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8170426065162907,
-      "eval_f1": 0.7960536910871955,
-      "eval_loss": 0.4052737355232239,
-      "eval_precision": 0.7846938775510204,
-      "eval_recall": 0.8255591925804692,
-      "eval_runtime": 5.5824,
-      "eval_samples_per_second": 71.475,
-      "eval_steps_per_second": 8.957,
+      "eval_accuracy": 0.7969924812030075,
+      "eval_f1": 0.7789473684210526,
+      "eval_loss": 0.4333195090293884,
+      "eval_precision": 0.7711038961038961,
+      "eval_recall": 0.8188761593016912,
+      "eval_runtime": 1.9992,
+      "eval_samples_per_second": 199.58,
+      "eval_steps_per_second": 25.01,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 4.110426425933838,
+      "grad_norm": 3.0677573680877686,
       "learning_rate": 4.25e-05,
-      "loss": 0.3464,
+      "loss": 0.3491,
       "step": 366
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8421052631578947,
-      "eval_f1": 0.7905197629940748,
-      "eval_loss": 0.3424628674983978,
-      "eval_precision": 0.8345238095238094,
-      "eval_recall": 0.7682760501909438,
-      "eval_runtime": 5.558,
-      "eval_samples_per_second": 71.789,
-      "eval_steps_per_second": 8.996,
+      "eval_accuracy": 0.8671679197994987,
+      "eval_f1": 0.8292627788498026,
+      "eval_loss": 0.32689258456230164,
+      "eval_precision": 0.8579176839902337,
+      "eval_recall": 0.8110110929259866,
+      "eval_runtime": 1.9988,
+      "eval_samples_per_second": 199.624,
+      "eval_steps_per_second": 25.015,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 1.6437464952468872,
+      "grad_norm": 2.5607471466064453,
       "learning_rate": 4e-05,
-      "loss": 0.2852,
+      "loss": 0.2886,
       "step": 488
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.8721804511278195,
-      "eval_f1": 0.8469505178365937,
-      "eval_loss": 0.3135569393634796,
-      "eval_precision": 0.844489247311828,
-      "eval_recall": 0.8495635570103655,
-      "eval_runtime": 5.5547,
-      "eval_samples_per_second": 71.83,
-      "eval_steps_per_second": 9.001,
+      "eval_f1": 0.8430067043674827,
+      "eval_loss": 0.2994827926158905,
+      "eval_precision": 0.8498269896193772,
+      "eval_recall": 0.8370612838697945,
+      "eval_runtime": 2.0374,
+      "eval_samples_per_second": 195.834,
+      "eval_steps_per_second": 24.541,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 3.6121034622192383,
+      "grad_norm": 2.971632957458496,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.2608,
+      "loss": 0.2671,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8721804511278195,
-      "eval_f1": 0.8469505178365937,
-      "eval_loss": 0.3060314953327179,
-      "eval_precision": 0.844489247311828,
-      "eval_recall": 0.8495635570103655,
-      "eval_runtime": 5.5535,
-      "eval_samples_per_second": 71.846,
-      "eval_steps_per_second": 9.003,
+      "eval_accuracy": 0.8696741854636592,
+      "eval_f1": 0.842789598108747,
+      "eval_loss": 0.29669511318206787,
+      "eval_precision": 0.842789598108747,
+      "eval_recall": 0.842789598108747,
+      "eval_runtime": 2.0319,
+      "eval_samples_per_second": 196.364,
+      "eval_steps_per_second": 24.607,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 2.0154571533203125,
+      "grad_norm": 4.395784854888916,
       "learning_rate": 3.5e-05,
-      "loss": 0.2415,
+      "loss": 0.2481,
       "step": 732
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.8646616541353384,
-      "eval_f1": 0.8447157518450185,
-      "eval_loss": 0.3100413978099823,
-      "eval_precision": 0.8325401217487549,
-      "eval_recall": 0.864248045099109,
-      "eval_runtime": 5.6089,
-      "eval_samples_per_second": 71.137,
-      "eval_steps_per_second": 8.914,
+      "eval_accuracy": 0.8546365914786967,
+      "eval_f1": 0.8364661654135338,
+      "eval_loss": 0.33456894755363464,
+      "eval_precision": 0.8229932885906039,
+      "eval_recall": 0.8646572104018913,
+      "eval_runtime": 2.0674,
+      "eval_samples_per_second": 192.992,
+      "eval_steps_per_second": 24.184,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.5025161504745483,
+      "grad_norm": 0.19698984920978546,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.2329,
+      "loss": 0.2339,
       "step": 854
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8642214594306682,
-      "eval_loss": 0.28597915172576904,
-      "eval_precision": 0.8566755442334414,
-      "eval_recall": 0.8734315330060011,
-      "eval_runtime": 5.5584,
-      "eval_samples_per_second": 71.783,
-      "eval_steps_per_second": 8.995,
+      "eval_accuracy": 0.8671679197994987,
+      "eval_f1": 0.8424651921601347,
+      "eval_loss": 0.2965049147605896,
+      "eval_precision": 0.8372140762463343,
+      "eval_recall": 0.8485179123476996,
+      "eval_runtime": 2.0387,
+      "eval_samples_per_second": 195.713,
+      "eval_steps_per_second": 24.525,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 2.4278862476348877,
+      "grad_norm": 1.805320143699646,
       "learning_rate": 3e-05,
-      "loss": 0.199,
+      "loss": 0.217,
       "step": 976
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8872180451127819,
-      "eval_f1": 0.8622036668943447,
-      "eval_loss": 0.2878971993923187,
-      "eval_precision": 0.8671602787456446,
-      "eval_recall": 0.8577014002545917,
-      "eval_runtime": 5.5998,
-      "eval_samples_per_second": 71.253,
-      "eval_steps_per_second": 8.929,
+      "eval_accuracy": 0.8771929824561403,
+      "eval_f1": 0.8483536940081443,
+      "eval_loss": 0.29290771484375,
+      "eval_precision": 0.8575792287132493,
+      "eval_recall": 0.8406073831605747,
+      "eval_runtime": 2.0519,
+      "eval_samples_per_second": 194.45,
+      "eval_steps_per_second": 24.367,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 8.504670143127441,
+      "grad_norm": 8.52855396270752,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.1939,
+      "loss": 0.1984,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8676337535436396,
-      "eval_loss": 0.28258949518203735,
-      "eval_precision": 0.8658613445378152,
-      "eval_recall": 0.8694762684124386,
-      "eval_runtime": 5.5681,
-      "eval_samples_per_second": 71.658,
-      "eval_steps_per_second": 8.98,
+      "eval_accuracy": 0.8796992481203008,
+      "eval_f1": 0.8533986527862829,
+      "eval_loss": 0.2778124511241913,
+      "eval_precision": 0.8572003218020917,
+      "eval_recall": 0.8498817966903074,
+      "eval_runtime": 2.0595,
+      "eval_samples_per_second": 193.734,
+      "eval_steps_per_second": 24.277,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 2.462061882019043,
+      "grad_norm": 4.634001731872559,
       "learning_rate": 2.5e-05,
-      "loss": 0.1806,
+      "loss": 0.1937,
       "step": 1220
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8439374185136896,
-      "eval_loss": 0.2981988787651062,
-      "eval_precision": 0.8794955044955045,
-      "eval_recall": 0.822376795781051,
-      "eval_runtime": 5.6592,
-      "eval_samples_per_second": 70.505,
-      "eval_steps_per_second": 8.835,
+      "eval_accuracy": 0.8771929824561403,
+      "eval_f1": 0.8458135188208294,
+      "eval_loss": 0.29049694538116455,
+      "eval_precision": 0.8627946127946129,
+      "eval_recall": 0.8331060192762321,
+      "eval_runtime": 2.0594,
+      "eval_samples_per_second": 193.746,
+      "eval_steps_per_second": 24.279,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 1.0077548027038574,
+      "grad_norm": 4.496700763702393,
       "learning_rate": 2.25e-05,
-      "loss": 0.1674,
+      "loss": 0.1734,
       "step": 1342
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8730223677032187,
-      "eval_loss": 0.2734816372394562,
-      "eval_precision": 0.8730223677032187,
-      "eval_recall": 0.8730223677032187,
-      "eval_runtime": 5.6293,
-      "eval_samples_per_second": 70.879,
-      "eval_steps_per_second": 8.882,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8649563392675828,
+      "eval_loss": 0.2918776869773865,
+      "eval_precision": 0.8623655913978494,
+      "eval_recall": 0.8677032187670486,
+      "eval_runtime": 2.0871,
+      "eval_samples_per_second": 191.171,
+      "eval_steps_per_second": 23.956,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 3.9673709869384766,
+      "grad_norm": 5.332368850708008,
       "learning_rate": 2e-05,
-      "loss": 0.1553,
+      "loss": 0.1561,
       "step": 1464
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8717238211879976,
-      "eval_loss": 0.2753015458583832,
-      "eval_precision": 0.8757194133300328,
-      "eval_recall": 0.8680214584469903,
-      "eval_runtime": 5.5504,
-      "eval_samples_per_second": 71.887,
-      "eval_steps_per_second": 9.008,
+      "eval_accuracy": 0.8796992481203008,
+      "eval_f1": 0.8556004584112431,
+      "eval_loss": 0.31261545419692993,
+      "eval_precision": 0.8538865546218487,
+      "eval_recall": 0.85738316057465,
+      "eval_runtime": 2.0745,
+      "eval_samples_per_second": 192.331,
+      "eval_steps_per_second": 24.102,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 3.968949794769287,
+      "grad_norm": 0.6808285713195801,
       "learning_rate": 1.75e-05,
-      "loss": 0.1431,
+      "loss": 0.1579,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8661961395983623,
-      "eval_loss": 0.2937251627445221,
-      "eval_precision": 0.8784532165625604,
-      "eval_recall": 0.8562465902891435,
-      "eval_runtime": 5.5697,
-      "eval_samples_per_second": 71.638,
-      "eval_steps_per_second": 8.977,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.8560793854229822,
+      "eval_loss": 0.3164937198162079,
+      "eval_precision": 0.8609538327526132,
+      "eval_recall": 0.8516548463356974,
+      "eval_runtime": 2.07,
+      "eval_samples_per_second": 192.753,
+      "eval_steps_per_second": 24.155,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 8.566404342651367,
+      "grad_norm": 11.75323486328125,
       "learning_rate": 1.5e-05,
-      "loss": 0.1417,
+      "loss": 0.1459,
       "step": 1708
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.9072681704260651,
-      "eval_f1": 0.8910359080340997,
-      "eval_loss": 0.29110613465309143,
-      "eval_precision": 0.8822647601476015,
-      "eval_recall": 0.9018912529550827,
-      "eval_runtime": 5.5933,
-      "eval_samples_per_second": 71.335,
-      "eval_steps_per_second": 8.939,
+      "eval_accuracy": 0.8972431077694235,
+      "eval_f1": 0.8787009231453675,
+      "eval_loss": 0.3108453154563904,
+      "eval_precision": 0.8714896214896215,
+      "eval_recall": 0.8872976904891798,
+      "eval_runtime": 2.0849,
+      "eval_samples_per_second": 191.373,
+      "eval_steps_per_second": 23.982,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 0.1758796125650406,
+      "grad_norm": 0.21581852436065674,
       "learning_rate": 1.25e-05,
-      "loss": 0.1236,
+      "loss": 0.1399,
       "step": 1830
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8817957385392532,
-      "eval_loss": 0.2955999970436096,
-      "eval_precision": 0.8827677592299257,
-      "eval_recall": 0.8808419712675032,
-      "eval_runtime": 5.5574,
-      "eval_samples_per_second": 71.797,
-      "eval_steps_per_second": 8.997,
+      "eval_accuracy": 0.8922305764411027,
+      "eval_f1": 0.8727838950061173,
+      "eval_loss": 0.3192116618156433,
+      "eval_precision": 0.8656898656898657,
+      "eval_recall": 0.8812511365702855,
+      "eval_runtime": 2.0655,
+      "eval_samples_per_second": 193.177,
+      "eval_steps_per_second": 24.208,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 0.7015694975852966,
+      "grad_norm": 1.523335337638855,
       "learning_rate": 1e-05,
-      "loss": 0.1304,
+      "loss": 0.1377,
       "step": 1952
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.884617951284618,
-      "eval_loss": 0.3010990023612976,
-      "eval_precision": 0.8772893772893773,
-      "eval_recall": 0.8933442444080741,
-      "eval_runtime": 5.5644,
-      "eval_samples_per_second": 71.705,
-      "eval_steps_per_second": 8.986,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8642214594306682,
+      "eval_loss": 0.32000479102134705,
+      "eval_precision": 0.8566755442334414,
+      "eval_recall": 0.8734315330060011,
+      "eval_runtime": 2.0706,
+      "eval_samples_per_second": 192.702,
+      "eval_steps_per_second": 24.148,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 0.19915825128555298,
+      "grad_norm": 9.250269889831543,
       "learning_rate": 7.5e-06,
-      "loss": 0.1164,
+      "loss": 0.1259,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.879667048676036,
-      "eval_loss": 0.29428762197494507,
-      "eval_precision": 0.8778361344537815,
-      "eval_recall": 0.8815693762502272,
-      "eval_runtime": 5.5462,
-      "eval_samples_per_second": 71.942,
-      "eval_steps_per_second": 9.015,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8622899159663866,
+      "eval_loss": 0.3303412199020386,
+      "eval_precision": 0.8589244307033712,
+      "eval_recall": 0.8659301691216585,
+      "eval_runtime": 2.084,
+      "eval_samples_per_second": 191.461,
+      "eval_steps_per_second": 23.993,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 9.03445816040039,
+      "grad_norm": 7.081236362457275,
       "learning_rate": 5e-06,
-      "loss": 0.1144,
+      "loss": 0.1293,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8775533117267087,
-      "eval_loss": 0.2937219738960266,
-      "eval_precision": 0.873246730188791,
-      "eval_recall": 0.8822967812329514,
-      "eval_runtime": 5.5516,
-      "eval_samples_per_second": 71.872,
-      "eval_steps_per_second": 9.006,
+      "eval_accuracy": 0.8922305764411027,
+      "eval_f1": 0.8727838950061173,
+      "eval_loss": 0.3265065848827362,
+      "eval_precision": 0.8656898656898657,
+      "eval_recall": 0.8812511365702855,
+      "eval_runtime": 2.0665,
+      "eval_samples_per_second": 193.079,
+      "eval_steps_per_second": 24.195,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 5.45957612991333,
+      "grad_norm": 1.5093868970870972,
       "learning_rate": 2.5e-06,
-      "loss": 0.1198,
+      "loss": 0.1187,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8737897035111135,
-      "eval_loss": 0.29848915338516235,
-      "eval_precision": 0.8812047813777917,
-      "eval_recall": 0.8672940534642661,
-      "eval_runtime": 5.5952,
-      "eval_samples_per_second": 71.31,
-      "eval_steps_per_second": 8.936,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8622899159663866,
+      "eval_loss": 0.33049851655960083,
+      "eval_precision": 0.8589244307033712,
+      "eval_recall": 0.8659301691216585,
+      "eval_runtime": 2.0389,
+      "eval_samples_per_second": 195.69,
+      "eval_steps_per_second": 24.523,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 3.5000033378601074,
+      "grad_norm": 0.4667631983757019,
       "learning_rate": 0.0,
-      "loss": 0.1104,
+      "loss": 0.1302,
       "step": 2440
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8811928811928812,
-      "eval_loss": 0.29284632205963135,
-      "eval_precision": 0.8842105263157894,
-      "eval_recall": 0.878341516639389,
-      "eval_runtime": 5.5669,
-      "eval_samples_per_second": 71.674,
-      "eval_steps_per_second": 8.982,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8662440310793597,
+      "eval_loss": 0.3288058042526245,
+      "eval_precision": 0.8606158357771261,
+      "eval_recall": 0.872704128023277,
+      "eval_runtime": 2.0647,
+      "eval_samples_per_second": 193.253,
+      "eval_steps_per_second": 24.217,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
       "total_flos": 8551203605328000.0,
-      "train_loss": 0.2130153269064231,
-      "train_runtime": 2293.0977,
-      "train_samples_per_second": 31.73,
-      "train_steps_per_second": 1.064
+      "train_loss": 0.2203434850348801,
+      "train_runtime": 746.002,
+      "train_samples_per_second": 97.533,
+      "train_steps_per_second": 3.271
     }
   ],
   "logging_steps": 500,