End of training

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 14.0,
+    "eval_accuracy": 0.9872761545711592,
+    "eval_loss": 0.0543154738843441,
+    "eval_runtime": 66.8631,
+    "eval_samples_per_second": 31.737,
+    "eval_steps_per_second": 1.989,
+    "total_flos": 1.5805887922325717e+19,
+    "train_loss": 0.17515284487896893,
+    "train_runtime": 12010.6398,
+    "train_samples_per_second": 7.418,
+    "train_steps_per_second": 0.232
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 14.0,
+    "eval_accuracy": 0.9872761545711592,
+    "eval_loss": 0.0543154738843441,
+    "eval_runtime": 66.8631,
+    "eval_samples_per_second": 31.737,
+    "eval_steps_per_second": 1.989
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 14.0,
+    "total_flos": 1.5805887922325717e+19,
+    "train_loss": 0.17515284487896893,
+    "train_runtime": 12010.6398,
+    "train_samples_per_second": 7.418,
+    "train_steps_per_second": 0.232
+}

trainer_state.json

@@ -0,0 +1,3496 @@
+{
+  "best_metric": 0.9872761545711592,
+  "best_model_checkpoint": "convnextv2-large-1k-224-finetuned-BreastCancer-Classification-BreakHis-AH-60-20-20-Shuffled-3rd/checkpoint-2786",
+  "epoch": 14.0,
+  "eval_steps": 500,
+  "global_step": 2786,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.96810207336523e-08,
+      "loss": 0.6896,
+      "step": 5
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.993620414673046e-07,
+      "loss": 0.6883,
+      "step": 10
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.990430622009569e-07,
+      "loss": 0.6853,
+      "step": 15
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.987240829346092e-07,
+      "loss": 0.6958,
+      "step": 20
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.984051036682616e-07,
+      "loss": 0.6908,
+      "step": 25
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980861244019139e-07,
+      "loss": 0.6933,
+      "step": 30
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 6.977671451355662e-07,
+      "loss": 0.686,
+      "step": 35
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 7.974481658692184e-07,
+      "loss": 0.6897,
+      "step": 40
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.971291866028709e-07,
+      "loss": 0.6842,
+      "step": 45
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.968102073365232e-07,
+      "loss": 0.6886,
+      "step": 50
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.0964912280701754e-06,
+      "loss": 0.6794,
+      "step": 55
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.1961722488038277e-06,
+      "loss": 0.6829,
+      "step": 60
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.2958532695374803e-06,
+      "loss": 0.677,
+      "step": 65
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.3955342902711324e-06,
+      "loss": 0.6723,
+      "step": 70
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4952153110047847e-06,
+      "loss": 0.6743,
+      "step": 75
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5948963317384369e-06,
+      "loss": 0.6695,
+      "step": 80
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6945773524720894e-06,
+      "loss": 0.6711,
+      "step": 85
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.7942583732057418e-06,
+      "loss": 0.6671,
+      "step": 90
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.8939393939393941e-06,
+      "loss": 0.6599,
+      "step": 95
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9936204146730465e-06,
+      "loss": 0.6581,
+      "step": 100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.093301435406699e-06,
+      "loss": 0.6485,
+      "step": 105
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.1929824561403507e-06,
+      "loss": 0.6466,
+      "step": 110
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.292663476874003e-06,
+      "loss": 0.6416,
+      "step": 115
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.3923444976076554e-06,
+      "loss": 0.6462,
+      "step": 120
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.492025518341308e-06,
+      "loss": 0.6363,
+      "step": 125
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.5917065390749605e-06,
+      "loss": 0.6252,
+      "step": 130
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.691387559808613e-06,
+      "loss": 0.622,
+      "step": 135
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.7910685805422648e-06,
+      "loss": 0.6149,
+      "step": 140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.890749601275917e-06,
+      "loss": 0.6098,
+      "step": 145
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.9904306220095695e-06,
+      "loss": 0.5992,
+      "step": 150
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.090111642743222e-06,
+      "loss": 0.606,
+      "step": 155
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.1897926634768737e-06,
+      "loss": 0.5844,
+      "step": 160
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.2894736842105265e-06,
+      "loss": 0.5673,
+      "step": 165
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.389154704944179e-06,
+      "loss": 0.5577,
+      "step": 170
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.488835725677831e-06,
+      "loss": 0.5659,
+      "step": 175
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5885167464114835e-06,
+      "loss": 0.5354,
+      "step": 180
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.688197767145136e-06,
+      "loss": 0.5417,
+      "step": 185
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.7878787878787882e-06,
+      "loss": 0.5481,
+      "step": 190
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.88755980861244e-06,
+      "loss": 0.5284,
+      "step": 195
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9095193213949104,
+      "eval_loss": 0.5013382434844971,
+      "eval_runtime": 773.5063,
+      "eval_samples_per_second": 2.743,
+      "eval_steps_per_second": 0.172,
+      "step": 199
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.987240829346093e-06,
+      "loss": 0.5202,
+      "step": 200
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.086921850079746e-06,
+      "loss": 0.5187,
+      "step": 205
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.186602870813398e-06,
+      "loss": 0.5106,
+      "step": 210
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.2862838915470495e-06,
+      "loss": 0.4866,
+      "step": 215
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.3859649122807014e-06,
+      "loss": 0.4893,
+      "step": 220
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.485645933014354e-06,
+      "loss": 0.4467,
+      "step": 225
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.585326953748006e-06,
+      "loss": 0.4381,
+      "step": 230
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.685007974481659e-06,
+      "loss": 0.422,
+      "step": 235
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.784688995215311e-06,
+      "loss": 0.4099,
+      "step": 240
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.8843700159489636e-06,
+      "loss": 0.4288,
+      "step": 245
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.984051036682616e-06,
+      "loss": 0.3929,
+      "step": 250
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.083732057416268e-06,
+      "loss": 0.3995,
+      "step": 255
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.183413078149921e-06,
+      "loss": 0.3699,
+      "step": 260
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.283094098883573e-06,
+      "loss": 0.4035,
+      "step": 265
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.382775119617226e-06,
+      "loss": 0.3549,
+      "step": 270
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.482456140350877e-06,
+      "loss": 0.3314,
+      "step": 275
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5821371610845296e-06,
+      "loss": 0.352,
+      "step": 280
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.681818181818182e-06,
+      "loss": 0.3343,
+      "step": 285
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.781499202551834e-06,
+      "loss": 0.3346,
+      "step": 290
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.881180223285487e-06,
+      "loss": 0.3272,
+      "step": 295
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.980861244019139e-06,
+      "loss": 0.3045,
+      "step": 300
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 6.080542264752792e-06,
+      "loss": 0.334,
+      "step": 305
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 6.180223285486444e-06,
+      "loss": 0.2874,
+      "step": 310
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 6.2799043062200955e-06,
+      "loss": 0.3253,
+      "step": 315
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 6.3795853269537475e-06,
+      "loss": 0.3083,
+      "step": 320
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 6.4792663476874e-06,
+      "loss": 0.2564,
+      "step": 325
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 6.578947368421053e-06,
+      "loss": 0.3236,
+      "step": 330
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 6.678628389154705e-06,
+      "loss": 0.3053,
+      "step": 335
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 6.778309409888358e-06,
+      "loss": 0.3211,
+      "step": 340
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 6.87799043062201e-06,
+      "loss": 0.2978,
+      "step": 345
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 6.977671451355662e-06,
+      "loss": 0.283,
+      "step": 350
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.077352472089314e-06,
+      "loss": 0.3428,
+      "step": 355
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.177033492822967e-06,
+      "loss": 0.2284,
+      "step": 360
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.276714513556619e-06,
+      "loss": 0.2762,
+      "step": 365
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 7.376395534290272e-06,
+      "loss": 0.2744,
+      "step": 370
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 7.476076555023924e-06,
+      "loss": 0.2261,
+      "step": 375
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.2657,
+      "step": 380
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 7.675438596491228e-06,
+      "loss": 0.2806,
+      "step": 385
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 7.77511961722488e-06,
+      "loss": 0.2418,
+      "step": 390
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.874800637958534e-06,
+      "loss": 0.2084,
+      "step": 395
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9524033930254477,
+      "eval_loss": 0.2076038420200348,
+      "eval_runtime": 66.1775,
+      "eval_samples_per_second": 32.065,
+      "eval_steps_per_second": 2.01,
+      "step": 398
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.974481658692186e-06,
+      "loss": 0.1872,
+      "step": 400
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 8.074162679425838e-06,
+      "loss": 0.2704,
+      "step": 405
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 8.173843700159491e-06,
+      "loss": 0.2772,
+      "step": 410
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 8.273524720893143e-06,
+      "loss": 0.2041,
+      "step": 415
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 8.373205741626795e-06,
+      "loss": 0.1876,
+      "step": 420
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 8.472886762360447e-06,
+      "loss": 0.2421,
+      "step": 425
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 8.572567783094099e-06,
+      "loss": 0.1965,
+      "step": 430
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 8.672248803827751e-06,
+      "loss": 0.2317,
+      "step": 435
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 8.771929824561403e-06,
+      "loss": 0.237,
+      "step": 440
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 8.871610845295056e-06,
+      "loss": 0.1576,
+      "step": 445
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.971291866028708e-06,
+      "loss": 0.2396,
+      "step": 450
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 9.07097288676236e-06,
+      "loss": 0.1799,
+      "step": 455
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 9.170653907496012e-06,
+      "loss": 0.2594,
+      "step": 460
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 9.270334928229666e-06,
+      "loss": 0.1516,
+      "step": 465
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 9.370015948963318e-06,
+      "loss": 0.1638,
+      "step": 470
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 9.46969696969697e-06,
+      "loss": 0.211,
+      "step": 475
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.569377990430622e-06,
+      "loss": 0.2195,
+      "step": 480
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.669059011164275e-06,
+      "loss": 0.2009,
+      "step": 485
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.768740031897927e-06,
+      "loss": 0.2175,
+      "step": 490
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 9.868421052631579e-06,
+      "loss": 0.2132,
+      "step": 495
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 9.968102073365233e-06,
+      "loss": 0.172,
+      "step": 500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.0067783094098885e-05,
+      "loss": 0.2102,
+      "step": 505
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0167464114832537e-05,
+      "loss": 0.2038,
+      "step": 510
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.0267145135566188e-05,
+      "loss": 0.1683,
+      "step": 515
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.0366826156299842e-05,
+      "loss": 0.2149,
+      "step": 520
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.0466507177033494e-05,
+      "loss": 0.2511,
+      "step": 525
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.0566188197767146e-05,
+      "loss": 0.182,
+      "step": 530
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.0665869218500798e-05,
+      "loss": 0.1928,
+      "step": 535
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.0765550239234451e-05,
+      "loss": 0.2304,
+      "step": 540
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.0865231259968102e-05,
+      "loss": 0.2004,
+      "step": 545
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.0964912280701754e-05,
+      "loss": 0.1663,
+      "step": 550
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1064593301435407e-05,
+      "loss": 0.2223,
+      "step": 555
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.1164274322169059e-05,
+      "loss": 0.173,
+      "step": 560
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1263955342902711e-05,
+      "loss": 0.1882,
+      "step": 565
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1363636363636365e-05,
+      "loss": 0.2424,
+      "step": 570
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.1463317384370017e-05,
+      "loss": 0.1171,
+      "step": 575
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.1562998405103668e-05,
+      "loss": 0.1704,
+      "step": 580
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.166267942583732e-05,
+      "loss": 0.1806,
+      "step": 585
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.1762360446570974e-05,
+      "loss": 0.214,
+      "step": 590
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.1862041467304626e-05,
+      "loss": 0.1274,
+      "step": 595
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9566446748350612,
+      "eval_loss": 0.1459183692932129,
+      "eval_runtime": 66.1612,
+      "eval_samples_per_second": 32.073,
+      "eval_steps_per_second": 2.01,
+      "step": 597
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.1961722488038278e-05,
+      "loss": 0.2133,
+      "step": 600
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.206140350877193e-05,
+      "loss": 0.181,
+      "step": 605
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.2161084529505583e-05,
+      "loss": 0.1283,
+      "step": 610
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.2260765550239235e-05,
+      "loss": 0.13,
+      "step": 615
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.2360446570972887e-05,
+      "loss": 0.1608,
+      "step": 620
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.246012759170654e-05,
+      "loss": 0.2181,
+      "step": 625
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.2559808612440191e-05,
+      "loss": 0.232,
+      "step": 630
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.2659489633173843e-05,
+      "loss": 0.1208,
+      "step": 635
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.2759170653907495e-05,
+      "loss": 0.269,
+      "step": 640
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.2858851674641149e-05,
+      "loss": 0.1324,
+      "step": 645
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.29585326953748e-05,
+      "loss": 0.2421,
+      "step": 650
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.3058213716108452e-05,
+      "loss": 0.1482,
+      "step": 655
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 0.2451,
+      "step": 660
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.3257575757575758e-05,
+      "loss": 0.2553,
+      "step": 665
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.335725677830941e-05,
+      "loss": 0.1594,
+      "step": 670
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.3456937799043062e-05,
+      "loss": 0.066,
+      "step": 675
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.3556618819776715e-05,
+      "loss": 0.1135,
+      "step": 680
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.3656299840510367e-05,
+      "loss": 0.2081,
+      "step": 685
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.375598086124402e-05,
+      "loss": 0.1919,
+      "step": 690
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.3855661881977671e-05,
+      "loss": 0.1001,
+      "step": 695
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.3955342902711325e-05,
+      "loss": 0.1908,
+      "step": 700
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1.4055023923444977e-05,
+      "loss": 0.1767,
+      "step": 705
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1.4154704944178629e-05,
+      "loss": 0.1658,
+      "step": 710
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.425438596491228e-05,
+      "loss": 0.1454,
+      "step": 715
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.4354066985645934e-05,
+      "loss": 0.1557,
+      "step": 720
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.4453748006379586e-05,
+      "loss": 0.1802,
+      "step": 725
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.4553429027113238e-05,
+      "loss": 0.1028,
+      "step": 730
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.4653110047846892e-05,
+      "loss": 0.1706,
+      "step": 735
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.4752791068580543e-05,
+      "loss": 0.1371,
+      "step": 740
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.4852472089314195e-05,
+      "loss": 0.1392,
+      "step": 745
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.4952153110047847e-05,
+      "loss": 0.1314,
+      "step": 750
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.5051834130781501e-05,
+      "loss": 0.2435,
+      "step": 755
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.1421,
+      "step": 760
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.5251196172248805e-05,
+      "loss": 0.1292,
+      "step": 765
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.5350877192982457e-05,
+      "loss": 0.1207,
+      "step": 770
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.545055821371611e-05,
+      "loss": 0.2106,
+      "step": 775
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.555023923444976e-05,
+      "loss": 0.1644,
+      "step": 780
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.5649920255183416e-05,
+      "loss": 0.1659,
+      "step": 785
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.5749601275917068e-05,
+      "loss": 0.096,
+      "step": 790
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.584928229665072e-05,
+      "loss": 0.1618,
+      "step": 795
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9382657869934025,
+      "eval_loss": 0.1533595770597458,
+      "eval_runtime": 66.239,
+      "eval_samples_per_second": 32.035,
+      "eval_steps_per_second": 2.008,
+      "step": 796
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.594896331738437e-05,
+      "loss": 0.1428,
+      "step": 800
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 1.6048644338118024e-05,
+      "loss": 0.1488,
+      "step": 805
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1.6148325358851675e-05,
+      "loss": 0.0972,
+      "step": 810
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.6248006379585327e-05,
+      "loss": 0.1187,
+      "step": 815
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.6347687400318983e-05,
+      "loss": 0.0871,
+      "step": 820
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.6447368421052635e-05,
+      "loss": 0.1333,
+      "step": 825
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.6547049441786287e-05,
+      "loss": 0.148,
+      "step": 830
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.664673046251994e-05,
+      "loss": 0.1552,
+      "step": 835
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.674641148325359e-05,
+      "loss": 0.1198,
+      "step": 840
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.684609250398724e-05,
+      "loss": 0.1764,
+      "step": 845
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.6945773524720894e-05,
+      "loss": 0.1048,
+      "step": 850
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.7045454545454546e-05,
+      "loss": 0.065,
+      "step": 855
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.7145135566188198e-05,
+      "loss": 0.1051,
+      "step": 860
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.724481658692185e-05,
+      "loss": 0.1513,
+      "step": 865
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 1.7344497607655502e-05,
+      "loss": 0.173,
+      "step": 870
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1.7444178628389154e-05,
+      "loss": 0.2399,
+      "step": 875
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1.7543859649122806e-05,
+      "loss": 0.1987,
+      "step": 880
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1.7643540669856458e-05,
+      "loss": 0.16,
+      "step": 885
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1.7743221690590113e-05,
+      "loss": 0.0938,
+      "step": 890
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.7842902711323765e-05,
+      "loss": 0.1978,
+      "step": 895
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1.7942583732057417e-05,
+      "loss": 0.1199,
+      "step": 900
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 1.804226475279107e-05,
+      "loss": 0.1267,
+      "step": 905
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.814194577352472e-05,
+      "loss": 0.0988,
+      "step": 910
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.8241626794258373e-05,
+      "loss": 0.2077,
+      "step": 915
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.8341307814992024e-05,
+      "loss": 0.0723,
+      "step": 920
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.844098883572568e-05,
+      "loss": 0.107,
+      "step": 925
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.8540669856459332e-05,
+      "loss": 0.0574,
+      "step": 930
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.8640350877192984e-05,
+      "loss": 0.1195,
+      "step": 935
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.8740031897926636e-05,
+      "loss": 0.0668,
+      "step": 940
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.8839712918660287e-05,
+      "loss": 0.1729,
+      "step": 945
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.893939393939394e-05,
+      "loss": 0.1988,
+      "step": 950
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.903907496012759e-05,
+      "loss": 0.1572,
+      "step": 955
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.9138755980861243e-05,
+      "loss": 0.1638,
+      "step": 960
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.92384370015949e-05,
+      "loss": 0.0853,
+      "step": 965
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.933811802232855e-05,
+      "loss": 0.1221,
+      "step": 970
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.9437799043062202e-05,
+      "loss": 0.2462,
+      "step": 975
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.9537480063795854e-05,
+      "loss": 0.3388,
+      "step": 980
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.9637161084529506e-05,
+      "loss": 0.1064,
+      "step": 985
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.9736842105263158e-05,
+      "loss": 0.0871,
+      "step": 990
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.983652312599681e-05,
+      "loss": 0.2118,
+      "step": 995
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9726672950047125,
+      "eval_loss": 0.08765223622322083,
+      "eval_runtime": 66.1349,
+      "eval_samples_per_second": 32.086,
+      "eval_steps_per_second": 2.011,
+      "step": 995
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1.9936204146730465e-05,
+      "loss": 0.221,
+      "step": 1000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.0035885167464117e-05,
+      "loss": 0.1202,
+      "step": 1005
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.013556618819777e-05,
+      "loss": 0.2251,
+      "step": 1010
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 2.023524720893142e-05,
+      "loss": 0.0988,
+      "step": 1015
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.0334928229665073e-05,
+      "loss": 0.1135,
+      "step": 1020
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 2.0434609250398725e-05,
+      "loss": 0.312,
+      "step": 1025
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 2.0534290271132377e-05,
+      "loss": 0.1456,
+      "step": 1030
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 2.0633971291866032e-05,
+      "loss": 0.0716,
+      "step": 1035
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 2.0733652312599684e-05,
+      "loss": 0.1508,
+      "step": 1040
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.2162,
+      "step": 1045
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.0933014354066988e-05,
+      "loss": 0.2184,
+      "step": 1050
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.103269537480064e-05,
+      "loss": 0.1873,
+      "step": 1055
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.1132376395534292e-05,
+      "loss": 0.1089,
+      "step": 1060
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.1232057416267944e-05,
+      "loss": 0.1074,
+      "step": 1065
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 2.1331738437001596e-05,
+      "loss": 0.1286,
+      "step": 1070
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.143141945773525e-05,
+      "loss": 0.1686,
+      "step": 1075
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 2.1531100478468903e-05,
+      "loss": 0.1312,
+      "step": 1080
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 2.1630781499202555e-05,
+      "loss": 0.1211,
+      "step": 1085
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.1730462519936203e-05,
+      "loss": 0.0526,
+      "step": 1090
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 2.1830143540669855e-05,
+      "loss": 0.0628,
+      "step": 1095
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.1929824561403507e-05,
+      "loss": 0.0398,
+      "step": 1100
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 2.2029505582137162e-05,
+      "loss": 0.2401,
+      "step": 1105
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.2129186602870814e-05,
+      "loss": 0.1126,
+      "step": 1110
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.2228867623604466e-05,
+      "loss": 0.1391,
+      "step": 1115
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.2328548644338118e-05,
+      "loss": 0.1835,
+      "step": 1120
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.242822966507177e-05,
+      "loss": 0.2572,
+      "step": 1125
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 2.2527910685805422e-05,
+      "loss": 0.1638,
+      "step": 1130
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.2627591706539074e-05,
+      "loss": 0.1415,
+      "step": 1135
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.1298,
+      "step": 1140
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.282695374800638e-05,
+      "loss": 0.102,
+      "step": 1145
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 2.2926634768740033e-05,
+      "loss": 0.1355,
+      "step": 1150
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.3026315789473685e-05,
+      "loss": 0.2175,
+      "step": 1155
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 2.3125996810207337e-05,
+      "loss": 0.1042,
+      "step": 1160
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 2.322567783094099e-05,
+      "loss": 0.1658,
+      "step": 1165
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.332535885167464e-05,
+      "loss": 0.1317,
+      "step": 1170
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 2.3425039872408293e-05,
+      "loss": 0.1368,
+      "step": 1175
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 2.3524720893141948e-05,
+      "loss": 0.0984,
+      "step": 1180
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 2.36244019138756e-05,
+      "loss": 0.1546,
+      "step": 1185
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 2.3724082934609252e-05,
+      "loss": 0.0306,
+      "step": 1190
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.96559849198869,
+      "eval_loss": 0.10479557514190674,
+      "eval_runtime": 66.2517,
+      "eval_samples_per_second": 32.029,
+      "eval_steps_per_second": 2.007,
+      "step": 1194
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.3823763955342904e-05,
+      "loss": 0.2805,
+      "step": 1195
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.3923444976076556e-05,
+      "loss": 0.1948,
+      "step": 1200
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 2.4023125996810208e-05,
+      "loss": 0.1294,
+      "step": 1205
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 2.412280701754386e-05,
+      "loss": 0.131,
+      "step": 1210
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 2.4222488038277515e-05,
+      "loss": 0.0771,
+      "step": 1215
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 2.4322169059011167e-05,
+      "loss": 0.056,
+      "step": 1220
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 2.442185007974482e-05,
+      "loss": 0.0424,
+      "step": 1225
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 2.452153110047847e-05,
+      "loss": 0.1259,
+      "step": 1230
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 2.4621212121212123e-05,
+      "loss": 0.1575,
+      "step": 1235
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 2.4720893141945774e-05,
+      "loss": 0.1067,
+      "step": 1240
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 2.4820574162679426e-05,
+      "loss": 0.0743,
+      "step": 1245
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 2.492025518341308e-05,
+      "loss": 0.1293,
+      "step": 1250
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 2.5019936204146734e-05,
+      "loss": 0.0878,
+      "step": 1255
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 2.5119617224880382e-05,
+      "loss": 0.0561,
+      "step": 1260
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 2.5219298245614037e-05,
+      "loss": 0.1468,
+      "step": 1265
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 2.5318979266347686e-05,
+      "loss": 0.1016,
+      "step": 1270
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 2.541866028708134e-05,
+      "loss": 0.1267,
+      "step": 1275
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 2.551834130781499e-05,
+      "loss": 0.0815,
+      "step": 1280
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 2.5618022328548645e-05,
+      "loss": 0.1103,
+      "step": 1285
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 2.5717703349282297e-05,
+      "loss": 0.0895,
+      "step": 1290
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 2.5817384370015952e-05,
+      "loss": 0.2504,
+      "step": 1295
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 2.59170653907496e-05,
+      "loss": 0.078,
+      "step": 1300
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 2.6016746411483256e-05,
+      "loss": 0.0883,
+      "step": 1305
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 2.6116427432216905e-05,
+      "loss": 0.1765,
+      "step": 1310
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 2.621610845295056e-05,
+      "loss": 0.0785,
+      "step": 1315
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.0979,
+      "step": 1320
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 2.6415470494417864e-05,
+      "loss": 0.0959,
+      "step": 1325
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 2.6515151515151516e-05,
+      "loss": 0.0935,
+      "step": 1330
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 2.661483253588517e-05,
+      "loss": 0.0987,
+      "step": 1335
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 2.671451355661882e-05,
+      "loss": 0.1285,
+      "step": 1340
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 2.6814194577352475e-05,
+      "loss": 0.0481,
+      "step": 1345
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 2.6913875598086123e-05,
+      "loss": 0.0909,
+      "step": 1350
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 2.701355661881978e-05,
+      "loss": 0.0919,
+      "step": 1355
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 2.711323763955343e-05,
+      "loss": 0.1335,
+      "step": 1360
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 2.7212918660287086e-05,
+      "loss": 0.1727,
+      "step": 1365
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 2.7312599681020735e-05,
+      "loss": 0.1956,
+      "step": 1370
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 2.741228070175439e-05,
+      "loss": 0.0489,
+      "step": 1375
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 2.751196172248804e-05,
+      "loss": 0.0656,
+      "step": 1380
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 2.7611642743221694e-05,
+      "loss": 0.0851,
+      "step": 1385
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 2.7711323763955342e-05,
+      "loss": 0.1012,
+      "step": 1390
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9754948162111216,
+      "eval_loss": 0.06740697473287582,
+      "eval_runtime": 66.4497,
+      "eval_samples_per_second": 31.934,
+      "eval_steps_per_second": 2.002,
+      "step": 1393
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 2.7811004784688998e-05,
+      "loss": 0.1534,
+      "step": 1395
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 2.791068580542265e-05,
+      "loss": 0.0807,
+      "step": 1400
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 2.8010366826156305e-05,
+      "loss": 0.0988,
+      "step": 1405
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 2.8110047846889953e-05,
+      "loss": 0.1152,
+      "step": 1410
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 2.820972886762361e-05,
+      "loss": 0.2121,
+      "step": 1415
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 2.8309409888357257e-05,
+      "loss": 0.0698,
+      "step": 1420
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 2.8409090909090912e-05,
+      "loss": 0.146,
+      "step": 1425
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 2.850877192982456e-05,
+      "loss": 0.0627,
+      "step": 1430
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 2.8608452950558216e-05,
+      "loss": 0.0741,
+      "step": 1435
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 2.8708133971291868e-05,
+      "loss": 0.1172,
+      "step": 1440
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 2.8807814992025524e-05,
+      "loss": 0.0642,
+      "step": 1445
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 2.8907496012759172e-05,
+      "loss": 0.0695,
+      "step": 1450
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 2.900717703349282e-05,
+      "loss": 0.0517,
+      "step": 1455
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 2.9106858054226476e-05,
+      "loss": 0.1438,
+      "step": 1460
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 2.9206539074960128e-05,
+      "loss": 0.1856,
+      "step": 1465
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 2.9306220095693783e-05,
+      "loss": 0.1129,
+      "step": 1470
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 2.940590111642743e-05,
+      "loss": 0.1274,
+      "step": 1475
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 2.9505582137161087e-05,
+      "loss": 0.066,
+      "step": 1480
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 2.9605263157894735e-05,
+      "loss": 0.0464,
+      "step": 1485
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 2.970494417862839e-05,
+      "loss": 0.0444,
+      "step": 1490
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 2.980462519936204e-05,
+      "loss": 0.1212,
+      "step": 1495
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 2.9904306220095695e-05,
+      "loss": 0.0917,
+      "step": 1500
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 3.0003987240829347e-05,
+      "loss": 0.0964,
+      "step": 1505
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 3.0103668261563002e-05,
+      "loss": 0.1538,
+      "step": 1510
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 3.020334928229665e-05,
+      "loss": 0.2752,
+      "step": 1515
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.1428,
+      "step": 1520
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 3.0402711323763954e-05,
+      "loss": 0.0793,
+      "step": 1525
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 3.050239234449761e-05,
+      "loss": 0.14,
+      "step": 1530
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 3.060207336523126e-05,
+      "loss": 0.0503,
+      "step": 1535
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 3.0701754385964913e-05,
+      "loss": 0.0705,
+      "step": 1540
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 3.080143540669856e-05,
+      "loss": 0.0764,
+      "step": 1545
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 3.090111642743222e-05,
+      "loss": 0.0533,
+      "step": 1550
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 3.100079744816587e-05,
+      "loss": 0.234,
+      "step": 1555
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 3.110047846889952e-05,
+      "loss": 0.0946,
+      "step": 1560
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 3.1200159489633176e-05,
+      "loss": 0.1723,
+      "step": 1565
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 3.129984051036683e-05,
+      "loss": 0.0569,
+      "step": 1570
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 3.139952153110048e-05,
+      "loss": 0.1702,
+      "step": 1575
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 3.1499202551834136e-05,
+      "loss": 0.2412,
+      "step": 1580
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 3.1598883572567784e-05,
+      "loss": 0.1177,
+      "step": 1585
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 3.169856459330144e-05,
+      "loss": 0.2079,
+      "step": 1590
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9731385485391141,
+      "eval_loss": 0.0661500096321106,
+      "eval_runtime": 66.1809,
+      "eval_samples_per_second": 32.064,
+      "eval_steps_per_second": 2.01,
+      "step": 1592
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 3.179824561403509e-05,
+      "loss": 0.0555,
+      "step": 1595
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 3.189792663476874e-05,
+      "loss": 0.0337,
+      "step": 1600
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 3.199760765550239e-05,
+      "loss": 0.1281,
+      "step": 1605
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 3.209728867623605e-05,
+      "loss": 0.1517,
+      "step": 1610
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 3.2196969696969696e-05,
+      "loss": 0.0803,
+      "step": 1615
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 3.229665071770335e-05,
+      "loss": 0.1442,
+      "step": 1620
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 3.2396331738437e-05,
+      "loss": 0.2514,
+      "step": 1625
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 3.2496012759170655e-05,
+      "loss": 0.0401,
+      "step": 1630
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 3.259569377990431e-05,
+      "loss": 0.0642,
+      "step": 1635
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 3.2695374800637965e-05,
+      "loss": 0.18,
+      "step": 1640
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 3.2795055821371614e-05,
+      "loss": 0.0759,
+      "step": 1645
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 3.289473684210527e-05,
+      "loss": 0.3765,
+      "step": 1650
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 3.299441786283892e-05,
+      "loss": 0.259,
+      "step": 1655
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 3.309409888357257e-05,
+      "loss": 0.0904,
+      "step": 1660
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 3.319377990430622e-05,
+      "loss": 0.1401,
+      "step": 1665
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 3.329346092503988e-05,
+      "loss": 0.0365,
+      "step": 1670
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 3.3393141945773525e-05,
+      "loss": 0.0753,
+      "step": 1675
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 3.349282296650718e-05,
+      "loss": 0.1211,
+      "step": 1680
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 3.359250398724083e-05,
+      "loss": 0.1027,
+      "step": 1685
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 3.369218500797448e-05,
+      "loss": 0.1879,
+      "step": 1690
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 3.379186602870813e-05,
+      "loss": 0.1507,
+      "step": 1695
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 3.389154704944179e-05,
+      "loss": 0.0764,
+      "step": 1700
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 3.3991228070175444e-05,
+      "loss": 0.0566,
+      "step": 1705
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 3.409090909090909e-05,
+      "loss": 0.1513,
+      "step": 1710
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 3.419059011164275e-05,
+      "loss": 0.0502,
+      "step": 1715
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 3.4290271132376396e-05,
+      "loss": 0.0252,
+      "step": 1720
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 3.438995215311005e-05,
+      "loss": 0.168,
+      "step": 1725
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 3.44896331738437e-05,
+      "loss": 0.0836,
+      "step": 1730
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 3.4589314194577355e-05,
+      "loss": 0.0736,
+      "step": 1735
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 3.4688995215311004e-05,
+      "loss": 0.0712,
+      "step": 1740
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 3.478867623604466e-05,
+      "loss": 0.103,
+      "step": 1745
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 3.488835725677831e-05,
+      "loss": 0.1001,
+      "step": 1750
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 3.498803827751196e-05,
+      "loss": 0.1186,
+      "step": 1755
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 3.508771929824561e-05,
+      "loss": 0.0598,
+      "step": 1760
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 3.518740031897927e-05,
+      "loss": 0.0643,
+      "step": 1765
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 3.5287081339712915e-05,
+      "loss": 0.1206,
+      "step": 1770
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 3.538676236044657e-05,
+      "loss": 0.0932,
+      "step": 1775
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 3.5486443381180226e-05,
+      "loss": 0.1756,
+      "step": 1780
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 3.558612440191388e-05,
+      "loss": 0.0591,
+      "step": 1785
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 3.568580542264753e-05,
+      "loss": 0.087,
+      "step": 1790
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.9458058435438266,
+      "eval_loss": 0.11827990412712097,
+      "eval_runtime": 66.1948,
+      "eval_samples_per_second": 32.057,
+      "eval_steps_per_second": 2.009,
+      "step": 1791
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 3.5785486443381185e-05,
+      "loss": 0.187,
+      "step": 1795
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 3.5885167464114834e-05,
+      "loss": 0.107,
+      "step": 1800
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 3.598484848484849e-05,
+      "loss": 0.0999,
+      "step": 1805
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 3.608452950558214e-05,
+      "loss": 0.0598,
+      "step": 1810
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 3.618421052631579e-05,
+      "loss": 0.1376,
+      "step": 1815
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 3.628389154704944e-05,
+      "loss": 0.0898,
+      "step": 1820
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 3.6383572567783097e-05,
+      "loss": 0.0416,
+      "step": 1825
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 3.6483253588516745e-05,
+      "loss": 0.0715,
+      "step": 1830
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 3.65829346092504e-05,
+      "loss": 0.0467,
+      "step": 1835
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 3.668261562998405e-05,
+      "loss": 0.0446,
+      "step": 1840
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 3.6782296650717704e-05,
+      "loss": 0.1276,
+      "step": 1845
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 3.688197767145136e-05,
+      "loss": 0.1263,
+      "step": 1850
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 3.6981658692185015e-05,
+      "loss": 0.2921,
+      "step": 1855
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 3.7081339712918663e-05,
+      "loss": 0.2148,
+      "step": 1860
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 3.718102073365232e-05,
+      "loss": 0.343,
+      "step": 1865
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 3.728070175438597e-05,
+      "loss": 0.1294,
+      "step": 1870
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 3.738038277511962e-05,
+      "loss": 0.1261,
+      "step": 1875
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 3.748006379585327e-05,
+      "loss": 0.0339,
+      "step": 1880
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 3.7579744816586926e-05,
+      "loss": 0.0713,
+      "step": 1885
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 3.7679425837320575e-05,
+      "loss": 0.1052,
+      "step": 1890
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 3.777910685805423e-05,
+      "loss": 0.0637,
+      "step": 1895
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 3.787878787878788e-05,
+      "loss": 0.0649,
+      "step": 1900
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 3.7978468899521534e-05,
+      "loss": 0.1127,
+      "step": 1905
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 3.807814992025518e-05,
+      "loss": 0.0524,
+      "step": 1910
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 3.817783094098884e-05,
+      "loss": 0.1178,
+      "step": 1915
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 3.8277511961722486e-05,
+      "loss": 0.0226,
+      "step": 1920
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 3.837719298245614e-05,
+      "loss": 0.1353,
+      "step": 1925
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 3.84768740031898e-05,
+      "loss": 0.1414,
+      "step": 1930
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 3.8576555023923446e-05,
+      "loss": 0.2608,
+      "step": 1935
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 3.86762360446571e-05,
+      "loss": 0.1566,
+      "step": 1940
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 3.877591706539075e-05,
+      "loss": 0.0824,
+      "step": 1945
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 3.8875598086124405e-05,
+      "loss": 0.091,
+      "step": 1950
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 3.897527910685805e-05,
+      "loss": 0.1284,
+      "step": 1955
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 3.907496012759171e-05,
+      "loss": 0.1011,
+      "step": 1960
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 3.917464114832536e-05,
+      "loss": 0.0626,
+      "step": 1965
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 3.927432216905901e-05,
+      "loss": 0.2002,
+      "step": 1970
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 3.937400318979266e-05,
+      "loss": 0.1474,
+      "step": 1975
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 0.0989,
+      "step": 1980
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 3.9573365231259965e-05,
+      "loss": 0.1807,
+      "step": 1985
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 3.967304625199362e-05,
+      "loss": 0.1543,
+      "step": 1990
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9839773798303487,
+      "eval_loss": 0.06045444682240486,
+      "eval_runtime": 66.0471,
+      "eval_samples_per_second": 32.129,
+      "eval_steps_per_second": 2.014,
+      "step": 1990
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 3.9772727272727275e-05,
+      "loss": 0.1179,
+      "step": 1995
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 3.987240829346093e-05,
+      "loss": 0.0731,
+      "step": 2000
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 3.997208931419458e-05,
+      "loss": 0.0752,
+      "step": 2005
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 4.0071770334928235e-05,
+      "loss": 0.1055,
+      "step": 2010
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 4.017145135566188e-05,
+      "loss": 0.0622,
+      "step": 2015
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 4.027113237639554e-05,
+      "loss": 0.0227,
+      "step": 2020
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 4.037081339712919e-05,
+      "loss": 0.105,
+      "step": 2025
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 4.047049441786284e-05,
+      "loss": 0.107,
+      "step": 2030
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 4.057017543859649e-05,
+      "loss": 0.1032,
+      "step": 2035
+    },
+    {
+      "epoch": 10.25,
+      "learning_rate": 4.0669856459330146e-05,
+      "loss": 0.1356,
+      "step": 2040
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 4.0769537480063795e-05,
+      "loss": 0.1786,
+      "step": 2045
+    },
+    {
+      "epoch": 10.3,
+      "learning_rate": 4.086921850079745e-05,
+      "loss": 0.1046,
+      "step": 2050
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 4.09688995215311e-05,
+      "loss": 0.0877,
+      "step": 2055
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 4.1068580542264754e-05,
+      "loss": 0.1164,
+      "step": 2060
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 4.116826156299841e-05,
+      "loss": 0.072,
+      "step": 2065
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 4.1267942583732064e-05,
+      "loss": 0.041,
+      "step": 2070
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 4.136762360446571e-05,
+      "loss": 0.0274,
+      "step": 2075
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 4.146730462519937e-05,
+      "loss": 0.1184,
+      "step": 2080
+    },
+    {
+      "epoch": 10.48,
+      "learning_rate": 4.156698564593302e-05,
+      "loss": 0.2738,
+      "step": 2085
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.0581,
+      "step": 2090
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 4.176634768740032e-05,
+      "loss": 0.0789,
+      "step": 2095
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 4.1866028708133976e-05,
+      "loss": 0.1123,
+      "step": 2100
+    },
+    {
+      "epoch": 10.58,
+      "learning_rate": 4.1965709728867624e-05,
+      "loss": 0.0608,
+      "step": 2105
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 4.206539074960128e-05,
+      "loss": 0.138,
+      "step": 2110
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 4.216507177033493e-05,
+      "loss": 0.1384,
+      "step": 2115
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 4.2264752791068584e-05,
+      "loss": 0.1021,
+      "step": 2120
+    },
+    {
+      "epoch": 10.68,
+      "learning_rate": 4.236443381180223e-05,
+      "loss": 0.1238,
+      "step": 2125
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 4.246411483253589e-05,
+      "loss": 0.0837,
+      "step": 2130
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 4.2563795853269536e-05,
+      "loss": 0.0611,
+      "step": 2135
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 4.266347687400319e-05,
+      "loss": 0.0489,
+      "step": 2140
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 4.2763157894736847e-05,
+      "loss": 0.1189,
+      "step": 2145
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 4.28628389154705e-05,
+      "loss": 0.1002,
+      "step": 2150
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 4.296251993620415e-05,
+      "loss": 0.0508,
+      "step": 2155
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 4.3062200956937806e-05,
+      "loss": 0.1133,
+      "step": 2160
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 4.3161881977671454e-05,
+      "loss": 0.125,
+      "step": 2165
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 4.326156299840511e-05,
+      "loss": 0.0678,
+      "step": 2170
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 4.336124401913876e-05,
+      "loss": 0.0494,
+      "step": 2175
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 4.346092503987241e-05,
+      "loss": 0.018,
+      "step": 2180
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 4.356060606060606e-05,
+      "loss": 0.0788,
+      "step": 2185
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.9868049010367578,
+      "eval_loss": 0.05569874495267868,
+      "eval_runtime": 66.0864,
+      "eval_samples_per_second": 32.109,
+      "eval_steps_per_second": 2.013,
+      "step": 2189
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 4.366028708133971e-05,
+      "loss": 0.0769,
+      "step": 2190
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 4.3759968102073366e-05,
+      "loss": 0.0467,
+      "step": 2195
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 4.3859649122807014e-05,
+      "loss": 0.0651,
+      "step": 2200
+    },
+    {
+      "epoch": 11.08,
+      "learning_rate": 4.395933014354067e-05,
+      "loss": 0.0585,
+      "step": 2205
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 4.4059011164274325e-05,
+      "loss": 0.0348,
+      "step": 2210
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 4.415869218500798e-05,
+      "loss": 0.0733,
+      "step": 2215
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 4.425837320574163e-05,
+      "loss": 0.0373,
+      "step": 2220
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 4.4358054226475284e-05,
+      "loss": 0.1054,
+      "step": 2225
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 4.445773524720893e-05,
+      "loss": 0.0776,
+      "step": 2230
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 4.455741626794259e-05,
+      "loss": 0.1979,
+      "step": 2235
+    },
+    {
+      "epoch": 11.26,
+      "learning_rate": 4.4657097288676236e-05,
+      "loss": 0.0852,
+      "step": 2240
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 4.475677830940989e-05,
+      "loss": 0.1613,
+      "step": 2245
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 4.485645933014354e-05,
+      "loss": 0.0458,
+      "step": 2250
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 4.4956140350877196e-05,
+      "loss": 0.1051,
+      "step": 2255
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 4.5055821371610844e-05,
+      "loss": 0.0443,
+      "step": 2260
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 4.51555023923445e-05,
+      "loss": 0.0704,
+      "step": 2265
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 4.525518341307815e-05,
+      "loss": 0.0637,
+      "step": 2270
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 4.53548644338118e-05,
+      "loss": 0.1003,
+      "step": 2275
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.0275,
+      "step": 2280
+    },
+    {
+      "epoch": 11.48,
+      "learning_rate": 4.555422647527911e-05,
+      "loss": 0.0417,
+      "step": 2285
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 4.565390749601276e-05,
+      "loss": 0.0953,
+      "step": 2290
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 4.575358851674642e-05,
+      "loss": 0.2359,
+      "step": 2295
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 4.5853269537480066e-05,
+      "loss": 0.0955,
+      "step": 2300
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 4.595295055821372e-05,
+      "loss": 0.0316,
+      "step": 2305
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 4.605263157894737e-05,
+      "loss": 0.1141,
+      "step": 2310
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 4.6152312599681025e-05,
+      "loss": 0.0744,
+      "step": 2315
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 4.6251993620414674e-05,
+      "loss": 0.0027,
+      "step": 2320
+    },
+    {
+      "epoch": 11.68,
+      "learning_rate": 4.635167464114833e-05,
+      "loss": 0.1356,
+      "step": 2325
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 4.645135566188198e-05,
+      "loss": 0.1447,
+      "step": 2330
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 4.655103668261563e-05,
+      "loss": 0.1148,
+      "step": 2335
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 4.665071770334928e-05,
+      "loss": 0.0077,
+      "step": 2340
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 4.675039872408294e-05,
+      "loss": 0.0557,
+      "step": 2345
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 4.6850079744816585e-05,
+      "loss": 0.1106,
+      "step": 2350
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 4.694976076555024e-05,
+      "loss": 0.1177,
+      "step": 2355
+    },
+    {
+      "epoch": 11.86,
+      "learning_rate": 4.7049441786283896e-05,
+      "loss": 0.0267,
+      "step": 2360
+    },
+    {
+      "epoch": 11.88,
+      "learning_rate": 4.714912280701755e-05,
+      "loss": 0.1233,
+      "step": 2365
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 4.72488038277512e-05,
+      "loss": 0.0681,
+      "step": 2370
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 4.7348484848484855e-05,
+      "loss": 0.0687,
+      "step": 2375
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 4.7448165869218504e-05,
+      "loss": 0.0903,
+      "step": 2380
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 4.754784688995216e-05,
+      "loss": 0.0604,
+      "step": 2385
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.9868049010367578,
+      "eval_loss": 0.04614832252264023,
+      "eval_runtime": 66.1088,
+      "eval_samples_per_second": 32.099,
+      "eval_steps_per_second": 2.012,
+      "step": 2388
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 4.764752791068581e-05,
+      "loss": 0.0806,
+      "step": 2390
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 4.774720893141946e-05,
+      "loss": 0.1088,
+      "step": 2395
+    },
+    {
+      "epoch": 12.06,
+      "learning_rate": 4.784688995215311e-05,
+      "loss": 0.0385,
+      "step": 2400
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 4.794657097288677e-05,
+      "loss": 0.0434,
+      "step": 2405
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 4.8046251993620415e-05,
+      "loss": 0.004,
+      "step": 2410
+    },
+    {
+      "epoch": 12.14,
+      "learning_rate": 4.8145933014354064e-05,
+      "loss": 0.0189,
+      "step": 2415
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 4.824561403508772e-05,
+      "loss": 0.0846,
+      "step": 2420
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 4.8345295055821374e-05,
+      "loss": 0.3582,
+      "step": 2425
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 4.844497607655503e-05,
+      "loss": 0.2713,
+      "step": 2430
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 4.854465709728868e-05,
+      "loss": 0.2434,
+      "step": 2435
+    },
+    {
+      "epoch": 12.26,
+      "learning_rate": 4.8644338118022334e-05,
+      "loss": 0.0674,
+      "step": 2440
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 4.874401913875598e-05,
+      "loss": 0.1085,
+      "step": 2445
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 4.884370015948964e-05,
+      "loss": 0.0215,
+      "step": 2450
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 4.8943381180223286e-05,
+      "loss": 0.0041,
+      "step": 2455
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 4.904306220095694e-05,
+      "loss": 0.1379,
+      "step": 2460
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 4.914274322169059e-05,
+      "loss": 0.111,
+      "step": 2465
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 4.9242424242424245e-05,
+      "loss": 0.0911,
+      "step": 2470
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 4.9342105263157894e-05,
+      "loss": 0.0739,
+      "step": 2475
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 4.944178628389155e-05,
+      "loss": 0.094,
+      "step": 2480
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 4.95414673046252e-05,
+      "loss": 0.1502,
+      "step": 2485
+    },
+    {
+      "epoch": 12.51,
+      "learning_rate": 4.964114832535885e-05,
+      "loss": 0.1727,
+      "step": 2490
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 4.97408293460925e-05,
+      "loss": 0.1593,
+      "step": 2495
+    },
+    {
+      "epoch": 12.56,
+      "learning_rate": 4.984051036682616e-05,
+      "loss": 0.1629,
+      "step": 2500
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 4.994019138755981e-05,
+      "loss": 0.1052,
+      "step": 2505
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 4.964028776978418e-05,
+      "loss": 0.0405,
+      "step": 2510
+    },
+    {
+      "epoch": 12.64,
+      "learning_rate": 4.874100719424461e-05,
+      "loss": 0.1163,
+      "step": 2515
+    },
+    {
+      "epoch": 12.66,
+      "learning_rate": 4.784172661870504e-05,
+      "loss": 0.1759,
+      "step": 2520
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 4.6942446043165475e-05,
+      "loss": 0.0619,
+      "step": 2525
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 4.60431654676259e-05,
+      "loss": 0.2245,
+      "step": 2530
+    },
+    {
+      "epoch": 12.74,
+      "learning_rate": 4.5143884892086334e-05,
+      "loss": 0.4108,
+      "step": 2535
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 4.424460431654677e-05,
+      "loss": 0.1267,
+      "step": 2540
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 4.33453237410072e-05,
+      "loss": 0.0401,
+      "step": 2545
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 4.244604316546763e-05,
+      "loss": 0.0667,
+      "step": 2550
+    },
+    {
+      "epoch": 12.84,
+      "learning_rate": 4.154676258992806e-05,
+      "loss": 0.2116,
+      "step": 2555
+    },
+    {
+      "epoch": 12.86,
+      "learning_rate": 4.064748201438849e-05,
+      "loss": 0.2168,
+      "step": 2560
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 3.9748201438848925e-05,
+      "loss": 0.1982,
+      "step": 2565
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 3.884892086330936e-05,
+      "loss": 0.1961,
+      "step": 2570
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 3.794964028776979e-05,
+      "loss": 0.1808,
+      "step": 2575
+    },
+    {
+      "epoch": 12.96,
+      "learning_rate": 3.7050359712230217e-05,
+      "loss": 0.1237,
+      "step": 2580
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 3.615107913669065e-05,
+      "loss": 0.0306,
+      "step": 2585
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.9853911404335532,
+      "eval_loss": 0.04759230092167854,
+      "eval_runtime": 66.3415,
+      "eval_samples_per_second": 31.986,
+      "eval_steps_per_second": 2.005,
+      "step": 2587
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 3.5251798561151075e-05,
+      "loss": 0.0663,
+      "step": 2590
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 3.435251798561151e-05,
+      "loss": 0.069,
+      "step": 2595
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 3.345323741007194e-05,
+      "loss": 0.0514,
+      "step": 2600
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 3.2553956834532374e-05,
+      "loss": 0.037,
+      "step": 2605
+    },
+    {
+      "epoch": 13.12,
+      "learning_rate": 3.165467625899281e-05,
+      "loss": 0.0133,
+      "step": 2610
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 3.075539568345324e-05,
+      "loss": 0.0275,
+      "step": 2615
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 2.985611510791367e-05,
+      "loss": 0.0972,
+      "step": 2620
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 2.89568345323741e-05,
+      "loss": 0.0956,
+      "step": 2625
+    },
+    {
+      "epoch": 13.22,
+      "learning_rate": 2.805755395683453e-05,
+      "loss": 0.0767,
+      "step": 2630
+    },
+    {
+      "epoch": 13.24,
+      "learning_rate": 2.7158273381294964e-05,
+      "loss": 0.15,
+      "step": 2635
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 2.6258992805755394e-05,
+      "loss": 0.0419,
+      "step": 2640
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 2.5359712230215827e-05,
+      "loss": 0.0507,
+      "step": 2645
+    },
+    {
+      "epoch": 13.32,
+      "learning_rate": 2.446043165467626e-05,
+      "loss": 0.0616,
+      "step": 2650
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 2.3561151079136692e-05,
+      "loss": 0.044,
+      "step": 2655
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 2.2661870503597125e-05,
+      "loss": 0.0768,
+      "step": 2660
+    },
+    {
+      "epoch": 13.39,
+      "learning_rate": 2.1762589928057555e-05,
+      "loss": 0.0692,
+      "step": 2665
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 2.0863309352517988e-05,
+      "loss": 0.0125,
+      "step": 2670
+    },
+    {
+      "epoch": 13.44,
+      "learning_rate": 1.996402877697842e-05,
+      "loss": 0.0041,
+      "step": 2675
+    },
+    {
+      "epoch": 13.47,
+      "learning_rate": 1.906474820143885e-05,
+      "loss": 0.0146,
+      "step": 2680
+    },
+    {
+      "epoch": 13.49,
+      "learning_rate": 1.8165467625899283e-05,
+      "loss": 0.0644,
+      "step": 2685
+    },
+    {
+      "epoch": 13.52,
+      "learning_rate": 1.7266187050359716e-05,
+      "loss": 0.1609,
+      "step": 2690
+    },
+    {
+      "epoch": 13.54,
+      "learning_rate": 1.6366906474820145e-05,
+      "loss": 0.0555,
+      "step": 2695
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 1.5467625899280578e-05,
+      "loss": 0.0434,
+      "step": 2700
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 1.4568345323741009e-05,
+      "loss": 0.0426,
+      "step": 2705
+    },
+    {
+      "epoch": 13.62,
+      "learning_rate": 1.366906474820144e-05,
+      "loss": 0.0058,
+      "step": 2710
+    },
+    {
+      "epoch": 13.64,
+      "learning_rate": 1.2769784172661871e-05,
+      "loss": 0.0529,
+      "step": 2715
+    },
+    {
+      "epoch": 13.67,
+      "learning_rate": 1.1870503597122303e-05,
+      "loss": 0.0317,
+      "step": 2720
+    },
+    {
+      "epoch": 13.69,
+      "learning_rate": 1.0971223021582734e-05,
+      "loss": 0.0088,
+      "step": 2725
+    },
+    {
+      "epoch": 13.72,
+      "learning_rate": 1.0071942446043167e-05,
+      "loss": 0.0026,
+      "step": 2730
+    },
+    {
+      "epoch": 13.74,
+      "learning_rate": 9.172661870503598e-06,
+      "loss": 0.0342,
+      "step": 2735
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 8.273381294964029e-06,
+      "loss": 0.0583,
+      "step": 2740
+    },
+    {
+      "epoch": 13.79,
+      "learning_rate": 7.374100719424461e-06,
+      "loss": 0.0081,
+      "step": 2745
+    },
+    {
+      "epoch": 13.82,
+      "learning_rate": 6.474820143884892e-06,
+      "loss": 0.0015,
+      "step": 2750
+    },
+    {
+      "epoch": 13.84,
+      "learning_rate": 5.575539568345324e-06,
+      "loss": 0.1185,
+      "step": 2755
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 4.676258992805756e-06,
+      "loss": 0.0596,
+      "step": 2760
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 3.7769784172661873e-06,
+      "loss": 0.0218,
+      "step": 2765
+    },
+    {
+      "epoch": 13.92,
+      "learning_rate": 2.877697841726619e-06,
+      "loss": 0.1141,
+      "step": 2770
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 1.9784172661870504e-06,
+      "loss": 0.0319,
+      "step": 2775
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 1.0791366906474822e-06,
+      "loss": 0.0341,
+      "step": 2780
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 1.7985611510791368e-07,
+      "loss": 0.0365,
+      "step": 2785
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.9872761545711592,
+      "eval_loss": 0.0543154738843441,
+      "eval_runtime": 65.952,
+      "eval_samples_per_second": 32.175,
+      "eval_steps_per_second": 2.017,
+      "step": 2786
+    },
+    {
+      "epoch": 14.0,
+      "step": 2786,
+      "total_flos": 1.5805887922325717e+19,
+      "train_loss": 0.17515284487896893,
+      "train_runtime": 12010.6398,
+      "train_samples_per_second": 7.418,
+      "train_steps_per_second": 0.232
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 2786,
+  "num_train_epochs": 14,
+  "save_steps": 500,
+  "total_flos": 1.5805887922325717e+19,
+  "trial_name": null,
+  "trial_params": null
+}