{
  "best_metric": 0.6788575409265064,
  "best_model_checkpoint": "finetuned-FER2013/checkpoint-4040",
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 4040,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "learning_rate": 1.2376237623762377e-07,
      "loss": 2.073,
      "step": 10
    },
    {
      "epoch": 0.1,
      "learning_rate": 2.4752475247524754e-07,
      "loss": 2.1018,
      "step": 20
    },
    {
      "epoch": 0.15,
      "learning_rate": 3.712871287128713e-07,
      "loss": 2.0777,
      "step": 30
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.950495049504951e-07,
      "loss": 2.0483,
      "step": 40
    },
    {
      "epoch": 0.25,
      "learning_rate": 6.188118811881188e-07,
      "loss": 2.0325,
      "step": 50
    },
    {
      "epoch": 0.3,
      "learning_rate": 7.425742574257426e-07,
      "loss": 1.9965,
      "step": 60
    },
    {
      "epoch": 0.35,
      "learning_rate": 8.663366336633663e-07,
      "loss": 1.9529,
      "step": 70
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.900990099009902e-07,
      "loss": 1.9238,
      "step": 80
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.113861386138614e-06,
      "loss": 1.9034,
      "step": 90
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.2376237623762377e-06,
      "loss": 1.8819,
      "step": 100
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.3613861386138616e-06,
      "loss": 1.8242,
      "step": 110
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.4851485148514852e-06,
      "loss": 1.8252,
      "step": 120
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.6089108910891091e-06,
      "loss": 1.7892,
      "step": 130
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.7326732673267326e-06,
      "loss": 1.7766,
      "step": 140
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.8564356435643566e-06,
      "loss": 1.7322,
      "step": 150
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.9801980198019803e-06,
      "loss": 1.6977,
      "step": 160
    },
    {
      "epoch": 0.84,
      "learning_rate": 2.103960396039604e-06,
      "loss": 1.6751,
      "step": 170
    },
    {
      "epoch": 0.89,
      "learning_rate": 2.227722772277228e-06,
      "loss": 1.6337,
      "step": 180
    },
    {
      "epoch": 0.94,
      "learning_rate": 2.3514851485148514e-06,
      "loss": 1.5947,
      "step": 190
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 1.5466,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4500174155346569,
      "eval_loss": 1.502170443534851,
      "eval_runtime": 38.4158,
      "eval_samples_per_second": 74.735,
      "eval_steps_per_second": 2.343,
      "step": 202
    },
    {
      "epoch": 1.04,
      "learning_rate": 2.5990099009900993e-06,
      "loss": 1.5667,
      "step": 210
    },
    {
      "epoch": 1.09,
      "learning_rate": 2.7227722772277232e-06,
      "loss": 1.551,
      "step": 220
    },
    {
      "epoch": 1.14,
      "learning_rate": 2.8465346534653464e-06,
      "loss": 1.5325,
      "step": 230
    },
    {
      "epoch": 1.19,
      "learning_rate": 2.9702970297029703e-06,
      "loss": 1.5274,
      "step": 240
    },
    {
      "epoch": 1.24,
      "learning_rate": 3.0940594059405943e-06,
      "loss": 1.4687,
      "step": 250
    },
    {
      "epoch": 1.29,
      "learning_rate": 3.2178217821782182e-06,
      "loss": 1.4933,
      "step": 260
    },
    {
      "epoch": 1.34,
      "learning_rate": 3.341584158415842e-06,
      "loss": 1.4407,
      "step": 270
    },
    {
      "epoch": 1.39,
      "learning_rate": 3.4653465346534653e-06,
      "loss": 1.4817,
      "step": 280
    },
    {
      "epoch": 1.44,
      "learning_rate": 3.5891089108910892e-06,
      "loss": 1.4594,
      "step": 290
    },
    {
      "epoch": 1.49,
      "learning_rate": 3.712871287128713e-06,
      "loss": 1.4494,
      "step": 300
    },
    {
      "epoch": 1.53,
      "learning_rate": 3.836633663366337e-06,
      "loss": 1.4094,
      "step": 310
    },
    {
      "epoch": 1.58,
      "learning_rate": 3.960396039603961e-06,
      "loss": 1.4515,
      "step": 320
    },
    {
      "epoch": 1.63,
      "learning_rate": 4.084158415841584e-06,
      "loss": 1.3378,
      "step": 330
    },
    {
      "epoch": 1.68,
      "learning_rate": 4.207920792079208e-06,
      "loss": 1.3925,
      "step": 340
    },
    {
      "epoch": 1.73,
      "learning_rate": 4.331683168316832e-06,
      "loss": 1.3663,
      "step": 350
    },
    {
      "epoch": 1.78,
      "learning_rate": 4.455445544554456e-06,
      "loss": 1.3544,
      "step": 360
    },
    {
      "epoch": 1.83,
      "learning_rate": 4.57920792079208e-06,
      "loss": 1.3758,
      "step": 370
    },
    {
      "epoch": 1.88,
      "learning_rate": 4.702970297029703e-06,
      "loss": 1.3395,
      "step": 380
    },
    {
      "epoch": 1.93,
      "learning_rate": 4.826732673267327e-06,
      "loss": 1.328,
      "step": 390
    },
    {
      "epoch": 1.98,
      "learning_rate": 4.950495049504951e-06,
      "loss": 1.3372,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5632183908045977,
      "eval_loss": 1.1727259159088135,
      "eval_runtime": 38.6928,
      "eval_samples_per_second": 74.2,
      "eval_steps_per_second": 2.326,
      "step": 404
    },
    {
      "epoch": 2.03,
      "learning_rate": 4.991749174917492e-06,
      "loss": 1.313,
      "step": 410
    },
    {
      "epoch": 2.08,
      "learning_rate": 4.977997799779978e-06,
      "loss": 1.3004,
      "step": 420
    },
    {
      "epoch": 2.13,
      "learning_rate": 4.964246424642465e-06,
      "loss": 1.2099,
      "step": 430
    },
    {
      "epoch": 2.18,
      "learning_rate": 4.950495049504951e-06,
      "loss": 1.311,
      "step": 440
    },
    {
      "epoch": 2.23,
      "learning_rate": 4.936743674367437e-06,
      "loss": 1.2735,
      "step": 450
    },
    {
      "epoch": 2.28,
      "learning_rate": 4.922992299229923e-06,
      "loss": 1.2598,
      "step": 460
    },
    {
      "epoch": 2.33,
      "learning_rate": 4.90924092409241e-06,
      "loss": 1.2941,
      "step": 470
    },
    {
      "epoch": 2.38,
      "learning_rate": 4.8954895489548956e-06,
      "loss": 1.2459,
      "step": 480
    },
    {
      "epoch": 2.43,
      "learning_rate": 4.881738173817382e-06,
      "loss": 1.2597,
      "step": 490
    },
    {
      "epoch": 2.48,
      "learning_rate": 4.867986798679869e-06,
      "loss": 1.2726,
      "step": 500
    },
    {
      "epoch": 2.52,
      "learning_rate": 4.854235423542355e-06,
      "loss": 1.3236,
      "step": 510
    },
    {
      "epoch": 2.57,
      "learning_rate": 4.840484048404841e-06,
      "loss": 1.2716,
      "step": 520
    },
    {
      "epoch": 2.62,
      "learning_rate": 4.826732673267327e-06,
      "loss": 1.2441,
      "step": 530
    },
    {
      "epoch": 2.67,
      "learning_rate": 4.812981298129814e-06,
      "loss": 1.2843,
      "step": 540
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.7992299229923e-06,
      "loss": 1.1991,
      "step": 550
    },
    {
      "epoch": 2.77,
      "learning_rate": 4.785478547854786e-06,
      "loss": 1.2845,
      "step": 560
    },
    {
      "epoch": 2.82,
      "learning_rate": 4.771727172717272e-06,
      "loss": 1.2553,
      "step": 570
    },
    {
      "epoch": 2.87,
      "learning_rate": 4.757975797579759e-06,
      "loss": 1.2146,
      "step": 580
    },
    {
      "epoch": 2.92,
      "learning_rate": 4.7442244224422445e-06,
      "loss": 1.2662,
      "step": 590
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.730473047304731e-06,
      "loss": 1.2372,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6074538488331592,
      "eval_loss": 1.0635906457901,
      "eval_runtime": 38.5764,
      "eval_samples_per_second": 74.424,
      "eval_steps_per_second": 2.333,
      "step": 606
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.716721672167217e-06,
      "loss": 1.2289,
      "step": 610
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.702970297029703e-06,
      "loss": 1.2239,
      "step": 620
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.689218921892189e-06,
      "loss": 1.2212,
      "step": 630
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.675467546754676e-06,
      "loss": 1.2033,
      "step": 640
    },
    {
      "epoch": 3.22,
      "learning_rate": 4.661716171617162e-06,
      "loss": 1.2295,
      "step": 650
    },
    {
      "epoch": 3.27,
      "learning_rate": 4.6479647964796485e-06,
      "loss": 1.2271,
      "step": 660
    },
    {
      "epoch": 3.32,
      "learning_rate": 4.634213421342134e-06,
      "loss": 1.2193,
      "step": 670
    },
    {
      "epoch": 3.37,
      "learning_rate": 4.620462046204621e-06,
      "loss": 1.2324,
      "step": 680
    },
    {
      "epoch": 3.42,
      "learning_rate": 4.606710671067108e-06,
      "loss": 1.1368,
      "step": 690
    },
    {
      "epoch": 3.47,
      "learning_rate": 4.592959295929593e-06,
      "loss": 1.1451,
      "step": 700
    },
    {
      "epoch": 3.51,
      "learning_rate": 4.57920792079208e-06,
      "loss": 1.1696,
      "step": 710
    },
    {
      "epoch": 3.56,
      "learning_rate": 4.565456545654566e-06,
      "loss": 1.2064,
      "step": 720
    },
    {
      "epoch": 3.61,
      "learning_rate": 4.5517051705170525e-06,
      "loss": 1.2161,
      "step": 730
    },
    {
      "epoch": 3.66,
      "learning_rate": 4.537953795379538e-06,
      "loss": 1.1665,
      "step": 740
    },
    {
      "epoch": 3.71,
      "learning_rate": 4.524202420242025e-06,
      "loss": 1.165,
      "step": 750
    },
    {
      "epoch": 3.76,
      "learning_rate": 4.510451045104511e-06,
      "loss": 1.232,
      "step": 760
    },
    {
      "epoch": 3.81,
      "learning_rate": 4.4966996699669966e-06,
      "loss": 1.2113,
      "step": 770
    },
    {
      "epoch": 3.86,
      "learning_rate": 4.482948294829483e-06,
      "loss": 1.1504,
      "step": 780
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.469196919691969e-06,
      "loss": 1.1573,
      "step": 790
    },
    {
      "epoch": 3.96,
      "learning_rate": 4.455445544554456e-06,
      "loss": 1.2096,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6116335771508186,
      "eval_loss": 1.019976019859314,
      "eval_runtime": 38.619,
      "eval_samples_per_second": 74.342,
      "eval_steps_per_second": 2.33,
      "step": 808
    },
    {
      "epoch": 4.01,
      "learning_rate": 4.4416941694169415e-06,
      "loss": 1.1387,
      "step": 810
    },
    {
      "epoch": 4.06,
      "learning_rate": 4.427942794279428e-06,
      "loss": 1.1161,
      "step": 820
    },
    {
      "epoch": 4.11,
      "learning_rate": 4.414191419141915e-06,
      "loss": 1.1883,
      "step": 830
    },
    {
      "epoch": 4.16,
      "learning_rate": 4.400440044004401e-06,
      "loss": 1.1215,
      "step": 840
    },
    {
      "epoch": 4.21,
      "learning_rate": 4.386688668866887e-06,
      "loss": 1.1195,
      "step": 850
    },
    {
      "epoch": 4.26,
      "learning_rate": 4.372937293729374e-06,
      "loss": 1.1991,
      "step": 860
    },
    {
      "epoch": 4.31,
      "learning_rate": 4.35918591859186e-06,
      "loss": 1.1645,
      "step": 870
    },
    {
      "epoch": 4.36,
      "learning_rate": 4.345434543454346e-06,
      "loss": 1.1446,
      "step": 880
    },
    {
      "epoch": 4.41,
      "learning_rate": 4.331683168316832e-06,
      "loss": 1.1308,
      "step": 890
    },
    {
      "epoch": 4.46,
      "learning_rate": 4.317931793179319e-06,
      "loss": 1.1758,
      "step": 900
    },
    {
      "epoch": 4.5,
      "learning_rate": 4.304180418041805e-06,
      "loss": 1.1254,
      "step": 910
    },
    {
      "epoch": 4.55,
      "learning_rate": 4.29042904290429e-06,
      "loss": 1.1217,
      "step": 920
    },
    {
      "epoch": 4.6,
      "learning_rate": 4.276677667766777e-06,
      "loss": 1.1633,
      "step": 930
    },
    {
      "epoch": 4.65,
      "learning_rate": 4.262926292629263e-06,
      "loss": 1.1553,
      "step": 940
    },
    {
      "epoch": 4.7,
      "learning_rate": 4.2491749174917495e-06,
      "loss": 1.1764,
      "step": 950
    },
    {
      "epoch": 4.75,
      "learning_rate": 4.235423542354235e-06,
      "loss": 1.1058,
      "step": 960
    },
    {
      "epoch": 4.8,
      "learning_rate": 4.221672167216722e-06,
      "loss": 1.1587,
      "step": 970
    },
    {
      "epoch": 4.85,
      "learning_rate": 4.207920792079208e-06,
      "loss": 1.1889,
      "step": 980
    },
    {
      "epoch": 4.9,
      "learning_rate": 4.194169416941694e-06,
      "loss": 1.1176,
      "step": 990
    },
    {
      "epoch": 4.95,
      "learning_rate": 4.180418041804181e-06,
      "loss": 1.1319,
      "step": 1000
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.145,
      "step": 1010
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6325322187391152,
      "eval_loss": 0.9768889546394348,
      "eval_runtime": 38.6786,
      "eval_samples_per_second": 74.227,
      "eval_steps_per_second": 2.327,
      "step": 1010
    },
    {
      "epoch": 5.05,
      "learning_rate": 4.1529152915291535e-06,
      "loss": 1.1321,
      "step": 1020
    },
    {
      "epoch": 5.1,
      "learning_rate": 4.139163916391639e-06,
      "loss": 1.1268,
      "step": 1030
    },
    {
      "epoch": 5.15,
      "learning_rate": 4.125412541254126e-06,
      "loss": 1.1032,
      "step": 1040
    },
    {
      "epoch": 5.2,
      "learning_rate": 4.111661166116613e-06,
      "loss": 1.1327,
      "step": 1050
    },
    {
      "epoch": 5.25,
      "learning_rate": 4.0979097909790984e-06,
      "loss": 1.141,
      "step": 1060
    },
    {
      "epoch": 5.3,
      "learning_rate": 4.084158415841584e-06,
      "loss": 1.1025,
      "step": 1070
    },
    {
      "epoch": 5.35,
      "learning_rate": 4.070407040704071e-06,
      "loss": 1.1138,
      "step": 1080
    },
    {
      "epoch": 5.4,
      "learning_rate": 4.056655665566557e-06,
      "loss": 1.122,
      "step": 1090
    },
    {
      "epoch": 5.45,
      "learning_rate": 4.042904290429043e-06,
      "loss": 1.0642,
      "step": 1100
    },
    {
      "epoch": 5.5,
      "learning_rate": 4.029152915291529e-06,
      "loss": 1.1162,
      "step": 1110
    },
    {
      "epoch": 5.54,
      "learning_rate": 4.015401540154016e-06,
      "loss": 1.0873,
      "step": 1120
    },
    {
      "epoch": 5.59,
      "learning_rate": 4.001650165016502e-06,
      "loss": 1.1188,
      "step": 1130
    },
    {
      "epoch": 5.64,
      "learning_rate": 3.987898789878988e-06,
      "loss": 1.0996,
      "step": 1140
    },
    {
      "epoch": 5.69,
      "learning_rate": 3.974147414741474e-06,
      "loss": 1.1377,
      "step": 1150
    },
    {
      "epoch": 5.74,
      "learning_rate": 3.960396039603961e-06,
      "loss": 1.0857,
      "step": 1160
    },
    {
      "epoch": 5.79,
      "learning_rate": 3.9466446644664465e-06,
      "loss": 1.0875,
      "step": 1170
    },
    {
      "epoch": 5.84,
      "learning_rate": 3.932893289328933e-06,
      "loss": 1.0532,
      "step": 1180
    },
    {
      "epoch": 5.89,
      "learning_rate": 3.91914191419142e-06,
      "loss": 1.1319,
      "step": 1190
    },
    {
      "epoch": 5.94,
      "learning_rate": 3.905390539053906e-06,
      "loss": 1.1125,
      "step": 1200
    },
    {
      "epoch": 5.99,
      "learning_rate": 3.891639163916392e-06,
      "loss": 1.1589,
      "step": 1210
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6405433646812957,
      "eval_loss": 0.9515116214752197,
      "eval_runtime": 38.8551,
      "eval_samples_per_second": 73.89,
      "eval_steps_per_second": 2.316,
      "step": 1212
    },
    {
      "epoch": 6.04,
      "learning_rate": 3.877887788778878e-06,
      "loss": 1.1213,
      "step": 1220
    },
    {
      "epoch": 6.09,
      "learning_rate": 3.864136413641365e-06,
      "loss": 1.0874,
      "step": 1230
    },
    {
      "epoch": 6.14,
      "learning_rate": 3.8503850385038505e-06,
      "loss": 1.0901,
      "step": 1240
    },
    {
      "epoch": 6.19,
      "learning_rate": 3.836633663366337e-06,
      "loss": 1.0979,
      "step": 1250
    },
    {
      "epoch": 6.24,
      "learning_rate": 3.822882288228823e-06,
      "loss": 1.074,
      "step": 1260
    },
    {
      "epoch": 6.29,
      "learning_rate": 3.8091309130913096e-06,
      "loss": 1.0936,
      "step": 1270
    },
    {
      "epoch": 6.34,
      "learning_rate": 3.7953795379537954e-06,
      "loss": 1.071,
      "step": 1280
    },
    {
      "epoch": 6.39,
      "learning_rate": 3.781628162816282e-06,
      "loss": 1.1341,
      "step": 1290
    },
    {
      "epoch": 6.44,
      "learning_rate": 3.767876787678768e-06,
      "loss": 1.0803,
      "step": 1300
    },
    {
      "epoch": 6.49,
      "learning_rate": 3.7541254125412545e-06,
      "loss": 1.1026,
      "step": 1310
    },
    {
      "epoch": 6.53,
      "learning_rate": 3.7403740374037408e-06,
      "loss": 1.0736,
      "step": 1320
    },
    {
      "epoch": 6.58,
      "learning_rate": 3.726622662266227e-06,
      "loss": 1.1181,
      "step": 1330
    },
    {
      "epoch": 6.63,
      "learning_rate": 3.712871287128713e-06,
      "loss": 1.0721,
      "step": 1340
    },
    {
      "epoch": 6.68,
      "learning_rate": 3.6991199119912e-06,
      "loss": 1.0743,
      "step": 1350
    },
    {
      "epoch": 6.73,
      "learning_rate": 3.6853685368536857e-06,
      "loss": 1.1077,
      "step": 1360
    },
    {
      "epoch": 6.78,
      "learning_rate": 3.6716171617161715e-06,
      "loss": 1.0815,
      "step": 1370
    },
    {
      "epoch": 6.83,
      "learning_rate": 3.657865786578658e-06,
      "loss": 1.0834,
      "step": 1380
    },
    {
      "epoch": 6.88,
      "learning_rate": 3.6441144114411443e-06,
      "loss": 1.1362,
      "step": 1390
    },
    {
      "epoch": 6.93,
      "learning_rate": 3.6303630363036306e-06,
      "loss": 1.0713,
      "step": 1400
    },
    {
      "epoch": 6.98,
      "learning_rate": 3.616611661166117e-06,
      "loss": 1.0752,
      "step": 1410
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.64576802507837,
      "eval_loss": 0.9395281076431274,
      "eval_runtime": 38.3561,
      "eval_samples_per_second": 74.851,
      "eval_steps_per_second": 2.346,
      "step": 1414
    },
    {
      "epoch": 7.03,
      "learning_rate": 3.6028602860286034e-06,
      "loss": 1.0504,
      "step": 1420
    },
    {
      "epoch": 7.08,
      "learning_rate": 3.5891089108910892e-06,
      "loss": 1.1083,
      "step": 1430
    },
    {
      "epoch": 7.13,
      "learning_rate": 3.575357535753576e-06,
      "loss": 1.0908,
      "step": 1440
    },
    {
      "epoch": 7.18,
      "learning_rate": 3.5616061606160617e-06,
      "loss": 1.1186,
      "step": 1450
    },
    {
      "epoch": 7.23,
      "learning_rate": 3.5478547854785484e-06,
      "loss": 1.096,
      "step": 1460
    },
    {
      "epoch": 7.28,
      "learning_rate": 3.534103410341034e-06,
      "loss": 1.1121,
      "step": 1470
    },
    {
      "epoch": 7.33,
      "learning_rate": 3.520352035203521e-06,
      "loss": 1.0535,
      "step": 1480
    },
    {
      "epoch": 7.38,
      "learning_rate": 3.506600660066007e-06,
      "loss": 1.0403,
      "step": 1490
    },
    {
      "epoch": 7.43,
      "learning_rate": 3.4928492849284933e-06,
      "loss": 1.0904,
      "step": 1500
    },
    {
      "epoch": 7.48,
      "learning_rate": 3.4790979097909795e-06,
      "loss": 1.0987,
      "step": 1510
    },
    {
      "epoch": 7.52,
      "learning_rate": 3.4653465346534653e-06,
      "loss": 1.0786,
      "step": 1520
    },
    {
      "epoch": 7.57,
      "learning_rate": 3.451595159515952e-06,
      "loss": 1.1133,
      "step": 1530
    },
    {
      "epoch": 7.62,
      "learning_rate": 3.4378437843784377e-06,
      "loss": 1.0804,
      "step": 1540
    },
    {
      "epoch": 7.67,
      "learning_rate": 3.4240924092409244e-06,
      "loss": 1.0237,
      "step": 1550
    },
    {
      "epoch": 7.72,
      "learning_rate": 3.4103410341034106e-06,
      "loss": 1.0276,
      "step": 1560
    },
    {
      "epoch": 7.77,
      "learning_rate": 3.396589658965897e-06,
      "loss": 1.0769,
      "step": 1570
    },
    {
      "epoch": 7.82,
      "learning_rate": 3.382838283828383e-06,
      "loss": 1.0962,
      "step": 1580
    },
    {
      "epoch": 7.87,
      "learning_rate": 3.3690869086908697e-06,
      "loss": 1.0586,
      "step": 1590
    },
    {
      "epoch": 7.92,
      "learning_rate": 3.3553355335533555e-06,
      "loss": 1.0786,
      "step": 1600
    },
    {
      "epoch": 7.97,
      "learning_rate": 3.341584158415842e-06,
      "loss": 1.0524,
      "step": 1610
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.64576802507837,
      "eval_loss": 0.9330951571464539,
      "eval_runtime": 38.8204,
      "eval_samples_per_second": 73.956,
      "eval_steps_per_second": 2.318,
      "step": 1616
    },
    {
      "epoch": 8.02,
      "learning_rate": 3.327832783278328e-06,
      "loss": 1.0425,
      "step": 1620
    },
    {
      "epoch": 8.07,
      "learning_rate": 3.3140814081408146e-06,
      "loss": 1.0729,
      "step": 1630
    },
    {
      "epoch": 8.12,
      "learning_rate": 3.3003300330033004e-06,
      "loss": 1.0748,
      "step": 1640
    },
    {
      "epoch": 8.17,
      "learning_rate": 3.286578657865787e-06,
      "loss": 1.0766,
      "step": 1650
    },
    {
      "epoch": 8.22,
      "learning_rate": 3.272827282728273e-06,
      "loss": 1.0779,
      "step": 1660
    },
    {
      "epoch": 8.27,
      "learning_rate": 3.259075907590759e-06,
      "loss": 1.0591,
      "step": 1670
    },
    {
      "epoch": 8.32,
      "learning_rate": 3.2453245324532458e-06,
      "loss": 1.0551,
      "step": 1680
    },
    {
      "epoch": 8.37,
      "learning_rate": 3.2315731573157316e-06,
      "loss": 1.031,
      "step": 1690
    },
    {
      "epoch": 8.42,
      "learning_rate": 3.2178217821782182e-06,
      "loss": 1.0348,
      "step": 1700
    },
    {
      "epoch": 8.47,
      "learning_rate": 3.204070407040704e-06,
      "loss": 1.064,
      "step": 1710
    },
    {
      "epoch": 8.51,
      "learning_rate": 3.1903190319031907e-06,
      "loss": 1.0142,
      "step": 1720
    },
    {
      "epoch": 8.56,
      "learning_rate": 3.1765676567656765e-06,
      "loss": 1.0871,
      "step": 1730
    },
    {
      "epoch": 8.61,
      "learning_rate": 3.162816281628163e-06,
      "loss": 1.0586,
      "step": 1740
    },
    {
      "epoch": 8.66,
      "learning_rate": 3.1490649064906494e-06,
      "loss": 1.05,
      "step": 1750
    },
    {
      "epoch": 8.71,
      "learning_rate": 3.1353135313531356e-06,
      "loss": 1.0503,
      "step": 1760
    },
    {
      "epoch": 8.76,
      "learning_rate": 3.121562156215622e-06,
      "loss": 1.0318,
      "step": 1770
    },
    {
      "epoch": 8.81,
      "learning_rate": 3.1078107810781085e-06,
      "loss": 1.087,
      "step": 1780
    },
    {
      "epoch": 8.86,
      "learning_rate": 3.0940594059405943e-06,
      "loss": 1.0563,
      "step": 1790
    },
    {
      "epoch": 8.91,
      "learning_rate": 3.080308030803081e-06,
      "loss": 1.0416,
      "step": 1800
    },
    {
      "epoch": 8.96,
      "learning_rate": 3.0665566556655667e-06,
      "loss": 1.0829,
      "step": 1810
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6572622779519331,
      "eval_loss": 0.9173280000686646,
      "eval_runtime": 38.5578,
      "eval_samples_per_second": 74.46,
      "eval_steps_per_second": 2.334,
      "step": 1818
    },
    {
      "epoch": 9.01,
      "learning_rate": 3.052805280528053e-06,
      "loss": 1.0785,
      "step": 1820
    },
    {
      "epoch": 9.06,
      "learning_rate": 3.039053905390539e-06,
      "loss": 1.0458,
      "step": 1830
    },
    {
      "epoch": 9.11,
      "learning_rate": 3.0253025302530254e-06,
      "loss": 1.0499,
      "step": 1840
    },
    {
      "epoch": 9.16,
      "learning_rate": 3.011551155115512e-06,
      "loss": 1.0528,
      "step": 1850
    },
    {
      "epoch": 9.21,
      "learning_rate": 2.997799779977998e-06,
      "loss": 1.0891,
      "step": 1860
    },
    {
      "epoch": 9.26,
      "learning_rate": 2.9840484048404845e-06,
      "loss": 1.0229,
      "step": 1870
    },
    {
      "epoch": 9.31,
      "learning_rate": 2.9702970297029703e-06,
      "loss": 1.0544,
      "step": 1880
    },
    {
      "epoch": 9.36,
      "learning_rate": 2.956545654565457e-06,
      "loss": 1.0305,
      "step": 1890
    },
    {
      "epoch": 9.41,
      "learning_rate": 2.9427942794279428e-06,
      "loss": 1.036,
      "step": 1900
    },
    {
      "epoch": 9.46,
      "learning_rate": 2.9290429042904294e-06,
      "loss": 1.0191,
      "step": 1910
    },
    {
      "epoch": 9.5,
      "learning_rate": 2.9152915291529156e-06,
      "loss": 1.0612,
      "step": 1920
    },
    {
      "epoch": 9.55,
      "learning_rate": 2.901540154015402e-06,
      "loss": 1.0258,
      "step": 1930
    },
    {
      "epoch": 9.6,
      "learning_rate": 2.887788778877888e-06,
      "loss": 0.9895,
      "step": 1940
    },
    {
      "epoch": 9.65,
      "learning_rate": 2.8740374037403747e-06,
      "loss": 1.0716,
      "step": 1950
    },
    {
      "epoch": 9.7,
      "learning_rate": 2.8602860286028605e-06,
      "loss": 1.0605,
      "step": 1960
    },
    {
      "epoch": 9.75,
      "learning_rate": 2.8465346534653464e-06,
      "loss": 1.0462,
      "step": 1970
    },
    {
      "epoch": 9.8,
      "learning_rate": 2.832783278327833e-06,
      "loss": 1.0582,
      "step": 1980
    },
    {
      "epoch": 9.85,
      "learning_rate": 2.8190319031903192e-06,
      "loss": 1.0484,
      "step": 1990
    },
    {
      "epoch": 9.9,
      "learning_rate": 2.8052805280528055e-06,
      "loss": 0.9998,
      "step": 2000
    },
    {
      "epoch": 9.95,
      "learning_rate": 2.7915291529152917e-06,
      "loss": 1.0437,
      "step": 2010
    },
    {
      "epoch": 10.0,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 1.0219,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6597004528039011,
      "eval_loss": 0.9113705158233643,
      "eval_runtime": 38.5342,
      "eval_samples_per_second": 74.505,
      "eval_steps_per_second": 2.336,
      "step": 2020
    },
    {
      "epoch": 10.05,
      "learning_rate": 2.764026402640264e-06,
      "loss": 1.0749,
      "step": 2030
    },
    {
      "epoch": 10.1,
      "learning_rate": 2.750275027502751e-06,
      "loss": 1.0419,
      "step": 2040
    },
    {
      "epoch": 10.15,
      "learning_rate": 2.7365236523652366e-06,
      "loss": 1.0407,
      "step": 2050
    },
    {
      "epoch": 10.2,
      "learning_rate": 2.7227722772277232e-06,
      "loss": 0.9984,
      "step": 2060
    },
    {
      "epoch": 10.25,
      "learning_rate": 2.709020902090209e-06,
      "loss": 0.9575,
      "step": 2070
    },
    {
      "epoch": 10.3,
      "learning_rate": 2.6952695269526957e-06,
      "loss": 1.021,
      "step": 2080
    },
    {
      "epoch": 10.35,
      "learning_rate": 2.681518151815182e-06,
      "loss": 1.0191,
      "step": 2090
    },
    {
      "epoch": 10.4,
      "learning_rate": 2.667766776677668e-06,
      "loss": 1.0502,
      "step": 2100
    },
    {
      "epoch": 10.45,
      "learning_rate": 2.6540154015401544e-06,
      "loss": 1.0292,
      "step": 2110
    },
    {
      "epoch": 10.5,
      "learning_rate": 2.64026402640264e-06,
      "loss": 0.9986,
      "step": 2120
    },
    {
      "epoch": 10.54,
      "learning_rate": 2.626512651265127e-06,
      "loss": 0.9604,
      "step": 2130
    },
    {
      "epoch": 10.59,
      "learning_rate": 2.6127612761276126e-06,
      "loss": 1.0127,
      "step": 2140
    },
    {
      "epoch": 10.64,
      "learning_rate": 2.5990099009900993e-06,
      "loss": 1.048,
      "step": 2150
    },
    {
      "epoch": 10.69,
      "learning_rate": 2.585258525852585e-06,
      "loss": 1.0475,
      "step": 2160
    },
    {
      "epoch": 10.74,
      "learning_rate": 2.5715071507150717e-06,
      "loss": 1.0552,
      "step": 2170
    },
    {
      "epoch": 10.79,
      "learning_rate": 2.557755775577558e-06,
      "loss": 1.0351,
      "step": 2180
    },
    {
      "epoch": 10.84,
      "learning_rate": 2.544004400440044e-06,
      "loss": 1.0465,
      "step": 2190
    },
    {
      "epoch": 10.89,
      "learning_rate": 2.5302530253025304e-06,
      "loss": 1.0807,
      "step": 2200
    },
    {
      "epoch": 10.94,
      "learning_rate": 2.516501650165017e-06,
      "loss": 1.0255,
      "step": 2210
    },
    {
      "epoch": 10.99,
      "learning_rate": 2.502750275027503e-06,
      "loss": 0.9986,
      "step": 2220
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6579588993382097,
      "eval_loss": 0.9033769369125366,
      "eval_runtime": 38.6552,
      "eval_samples_per_second": 74.272,
      "eval_steps_per_second": 2.328,
      "step": 2222
    },
    {
      "epoch": 11.04,
      "learning_rate": 2.488998899889989e-06,
      "loss": 1.0336,
      "step": 2230
    },
    {
      "epoch": 11.09,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 1.0238,
      "step": 2240
    },
    {
      "epoch": 11.14,
      "learning_rate": 2.4614961496149616e-06,
      "loss": 1.04,
      "step": 2250
    },
    {
      "epoch": 11.19,
      "learning_rate": 2.4477447744774478e-06,
      "loss": 1.0165,
      "step": 2260
    },
    {
      "epoch": 11.24,
      "learning_rate": 2.4339933993399344e-06,
      "loss": 1.0155,
      "step": 2270
    },
    {
      "epoch": 11.29,
      "learning_rate": 2.4202420242024207e-06,
      "loss": 1.0273,
      "step": 2280
    },
    {
      "epoch": 11.34,
      "learning_rate": 2.406490649064907e-06,
      "loss": 1.0048,
      "step": 2290
    },
    {
      "epoch": 11.39,
      "learning_rate": 2.392739273927393e-06,
      "loss": 0.988,
      "step": 2300
    },
    {
      "epoch": 11.44,
      "learning_rate": 2.3789878987898793e-06,
      "loss": 1.0388,
      "step": 2310
    },
    {
      "epoch": 11.49,
      "learning_rate": 2.3652365236523656e-06,
      "loss": 1.0389,
      "step": 2320
    },
    {
      "epoch": 11.53,
      "learning_rate": 2.3514851485148514e-06,
      "loss": 1.0651,
      "step": 2330
    },
    {
      "epoch": 11.58,
      "learning_rate": 2.337733773377338e-06,
      "loss": 1.0789,
      "step": 2340
    },
    {
      "epoch": 11.63,
      "learning_rate": 2.3239823982398242e-06,
      "loss": 0.9802,
      "step": 2350
    },
    {
      "epoch": 11.68,
      "learning_rate": 2.3102310231023105e-06,
      "loss": 1.054,
      "step": 2360
    },
    {
      "epoch": 11.73,
      "learning_rate": 2.2964796479647967e-06,
      "loss": 1.0351,
      "step": 2370
    },
    {
      "epoch": 11.78,
      "learning_rate": 2.282728272827283e-06,
      "loss": 1.0146,
      "step": 2380
    },
    {
      "epoch": 11.83,
      "learning_rate": 2.268976897689769e-06,
      "loss": 1.0293,
      "step": 2390
    },
    {
      "epoch": 11.88,
      "learning_rate": 2.2552255225522554e-06,
      "loss": 0.9827,
      "step": 2400
    },
    {
      "epoch": 11.93,
      "learning_rate": 2.2414741474147416e-06,
      "loss": 0.9807,
      "step": 2410
    },
    {
      "epoch": 11.98,
      "learning_rate": 2.227722772277228e-06,
      "loss": 1.013,
      "step": 2420
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6656217345872518,
      "eval_loss": 0.9003686904907227,
      "eval_runtime": 38.455,
      "eval_samples_per_second": 74.659,
      "eval_steps_per_second": 2.34,
      "step": 2424
    },
    {
      "epoch": 12.03,
      "learning_rate": 2.213971397139714e-06,
      "loss": 1.0289,
      "step": 2430
    },
    {
      "epoch": 12.08,
      "learning_rate": 2.2002200220022003e-06,
      "loss": 1.0303,
      "step": 2440
    },
    {
      "epoch": 12.13,
      "learning_rate": 2.186468646864687e-06,
      "loss": 1.0304,
      "step": 2450
    },
    {
      "epoch": 12.18,
      "learning_rate": 2.172717271727173e-06,
      "loss": 1.0318,
      "step": 2460
    },
    {
      "epoch": 12.23,
      "learning_rate": 2.1589658965896594e-06,
      "loss": 1.0566,
      "step": 2470
    },
    {
      "epoch": 12.28,
      "learning_rate": 2.145214521452145e-06,
      "loss": 1.0052,
      "step": 2480
    },
    {
      "epoch": 12.33,
      "learning_rate": 2.1314631463146314e-06,
      "loss": 0.9798,
      "step": 2490
    },
    {
      "epoch": 12.38,
      "learning_rate": 2.1177117711771177e-06,
      "loss": 1.0139,
      "step": 2500
    },
    {
      "epoch": 12.43,
      "learning_rate": 2.103960396039604e-06,
      "loss": 0.9932,
      "step": 2510
    },
    {
      "epoch": 12.48,
      "learning_rate": 2.0902090209020905e-06,
      "loss": 1.0196,
      "step": 2520
    },
    {
      "epoch": 12.52,
      "learning_rate": 2.0764576457645768e-06,
      "loss": 1.0261,
      "step": 2530
    },
    {
      "epoch": 12.57,
      "learning_rate": 2.062706270627063e-06,
      "loss": 1.0209,
      "step": 2540
    },
    {
      "epoch": 12.62,
      "learning_rate": 2.0489548954895492e-06,
      "loss": 0.9964,
      "step": 2550
    },
    {
      "epoch": 12.67,
      "learning_rate": 2.0352035203520354e-06,
      "loss": 1.0248,
      "step": 2560
    },
    {
      "epoch": 12.72,
      "learning_rate": 2.0214521452145217e-06,
      "loss": 0.9682,
      "step": 2570
    },
    {
      "epoch": 12.77,
      "learning_rate": 2.007700770077008e-06,
      "loss": 1.0485,
      "step": 2580
    },
    {
      "epoch": 12.82,
      "learning_rate": 1.993949394939494e-06,
      "loss": 0.996,
      "step": 2590
    },
    {
      "epoch": 12.87,
      "learning_rate": 1.9801980198019803e-06,
      "loss": 1.0432,
      "step": 2600
    },
    {
      "epoch": 12.92,
      "learning_rate": 1.9664466446644666e-06,
      "loss": 1.0554,
      "step": 2610
    },
    {
      "epoch": 12.97,
      "learning_rate": 1.952695269526953e-06,
      "loss": 1.0133,
      "step": 2620
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6628352490421456,
      "eval_loss": 0.8939908146858215,
      "eval_runtime": 38.8578,
      "eval_samples_per_second": 73.885,
      "eval_steps_per_second": 2.316,
      "step": 2626
    },
    {
      "epoch": 13.02,
      "learning_rate": 1.938943894389439e-06,
      "loss": 0.9908,
      "step": 2630
    },
    {
      "epoch": 13.07,
      "learning_rate": 1.9251925192519253e-06,
      "loss": 0.9901,
      "step": 2640
    },
    {
      "epoch": 13.12,
      "learning_rate": 1.9114411441144115e-06,
      "loss": 0.9948,
      "step": 2650
    },
    {
      "epoch": 13.17,
      "learning_rate": 1.8976897689768977e-06,
      "loss": 1.0064,
      "step": 2660
    },
    {
      "epoch": 13.22,
      "learning_rate": 1.883938393839384e-06,
      "loss": 1.0087,
      "step": 2670
    },
    {
      "epoch": 13.27,
      "learning_rate": 1.8701870187018704e-06,
      "loss": 0.9959,
      "step": 2680
    },
    {
      "epoch": 13.32,
      "learning_rate": 1.8564356435643566e-06,
      "loss": 0.9834,
      "step": 2690
    },
    {
      "epoch": 13.37,
      "learning_rate": 1.8426842684268428e-06,
      "loss": 1.0346,
      "step": 2700
    },
    {
      "epoch": 13.42,
      "learning_rate": 1.828932893289329e-06,
      "loss": 0.9487,
      "step": 2710
    },
    {
      "epoch": 13.47,
      "learning_rate": 1.8151815181518153e-06,
      "loss": 0.9952,
      "step": 2720
    },
    {
      "epoch": 13.51,
      "learning_rate": 1.8014301430143017e-06,
      "loss": 1.0243,
      "step": 2730
    },
    {
      "epoch": 13.56,
      "learning_rate": 1.787678767876788e-06,
      "loss": 0.9498,
      "step": 2740
    },
    {
      "epoch": 13.61,
      "learning_rate": 1.7739273927392742e-06,
      "loss": 0.9573,
      "step": 2750
    },
    {
      "epoch": 13.66,
      "learning_rate": 1.7601760176017604e-06,
      "loss": 0.9838,
      "step": 2760
    },
    {
      "epoch": 13.71,
      "learning_rate": 1.7464246424642466e-06,
      "loss": 1.0254,
      "step": 2770
    },
    {
      "epoch": 13.76,
      "learning_rate": 1.7326732673267326e-06,
      "loss": 1.0186,
      "step": 2780
    },
    {
      "epoch": 13.81,
      "learning_rate": 1.7189218921892189e-06,
      "loss": 1.032,
      "step": 2790
    },
    {
      "epoch": 13.86,
      "learning_rate": 1.7051705170517053e-06,
      "loss": 1.0525,
      "step": 2800
    },
    {
      "epoch": 13.91,
      "learning_rate": 1.6914191419141915e-06,
      "loss": 1.0422,
      "step": 2810
    },
    {
      "epoch": 13.96,
      "learning_rate": 1.6776677667766778e-06,
      "loss": 1.0064,
      "step": 2820
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.6649251132009752,
      "eval_loss": 0.8916007876396179,
      "eval_runtime": 38.5634,
      "eval_samples_per_second": 74.449,
      "eval_steps_per_second": 2.334,
      "step": 2828
    },
    {
      "epoch": 14.01,
      "learning_rate": 1.663916391639164e-06,
      "loss": 1.0109,
      "step": 2830
    },
    {
      "epoch": 14.06,
      "learning_rate": 1.6501650165016502e-06,
      "loss": 1.016,
      "step": 2840
    },
    {
      "epoch": 14.11,
      "learning_rate": 1.6364136413641364e-06,
      "loss": 1.0287,
      "step": 2850
    },
    {
      "epoch": 14.16,
      "learning_rate": 1.6226622662266229e-06,
      "loss": 1.0048,
      "step": 2860
    },
    {
      "epoch": 14.21,
      "learning_rate": 1.6089108910891091e-06,
      "loss": 1.0639,
      "step": 2870
    },
    {
      "epoch": 14.26,
      "learning_rate": 1.5951595159515953e-06,
      "loss": 0.9976,
      "step": 2880
    },
    {
      "epoch": 14.31,
      "learning_rate": 1.5814081408140816e-06,
      "loss": 0.9993,
      "step": 2890
    },
    {
      "epoch": 14.36,
      "learning_rate": 1.5676567656765678e-06,
      "loss": 1.045,
      "step": 2900
    },
    {
      "epoch": 14.41,
      "learning_rate": 1.5539053905390542e-06,
      "loss": 1.0057,
      "step": 2910
    },
    {
      "epoch": 14.46,
      "learning_rate": 1.5401540154015405e-06,
      "loss": 1.0304,
      "step": 2920
    },
    {
      "epoch": 14.5,
      "learning_rate": 1.5264026402640265e-06,
      "loss": 1.0412,
      "step": 2930
    },
    {
      "epoch": 14.55,
      "learning_rate": 1.5126512651265127e-06,
      "loss": 1.0259,
      "step": 2940
    },
    {
      "epoch": 14.6,
      "learning_rate": 1.498899889988999e-06,
      "loss": 1.0207,
      "step": 2950
    },
    {
      "epoch": 14.65,
      "learning_rate": 1.4851485148514852e-06,
      "loss": 0.9956,
      "step": 2960
    },
    {
      "epoch": 14.7,
      "learning_rate": 1.4713971397139714e-06,
      "loss": 0.9838,
      "step": 2970
    },
    {
      "epoch": 14.75,
      "learning_rate": 1.4576457645764578e-06,
      "loss": 1.0084,
      "step": 2980
    },
    {
      "epoch": 14.8,
      "learning_rate": 1.443894389438944e-06,
      "loss": 0.9956,
      "step": 2990
    },
    {
      "epoch": 14.85,
      "learning_rate": 1.4301430143014303e-06,
      "loss": 0.9595,
      "step": 3000
    },
    {
      "epoch": 14.9,
      "learning_rate": 1.4163916391639165e-06,
      "loss": 0.9391,
      "step": 3010
    },
    {
      "epoch": 14.95,
      "learning_rate": 1.4026402640264027e-06,
      "loss": 1.0213,
      "step": 3020
    },
    {
      "epoch": 15.0,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 0.9858,
      "step": 3030
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6732845698362939,
      "eval_loss": 0.8882250785827637,
      "eval_runtime": 38.5965,
      "eval_samples_per_second": 74.385,
      "eval_steps_per_second": 2.332,
      "step": 3030
    },
    {
      "epoch": 15.05,
      "learning_rate": 1.3751375137513754e-06,
      "loss": 0.9837,
      "step": 3040
    },
    {
      "epoch": 15.1,
      "learning_rate": 1.3613861386138616e-06,
      "loss": 0.9723,
      "step": 3050
    },
    {
      "epoch": 15.15,
      "learning_rate": 1.3476347634763478e-06,
      "loss": 1.018,
      "step": 3060
    },
    {
      "epoch": 15.2,
      "learning_rate": 1.333883388338834e-06,
      "loss": 1.0612,
      "step": 3070
    },
    {
      "epoch": 15.25,
      "learning_rate": 1.32013201320132e-06,
      "loss": 1.0081,
      "step": 3080
    },
    {
      "epoch": 15.3,
      "learning_rate": 1.3063806380638063e-06,
      "loss": 1.0096,
      "step": 3090
    },
    {
      "epoch": 15.35,
      "learning_rate": 1.2926292629262925e-06,
      "loss": 0.9442,
      "step": 3100
    },
    {
      "epoch": 15.4,
      "learning_rate": 1.278877887788779e-06,
      "loss": 1.0434,
      "step": 3110
    },
    {
      "epoch": 15.45,
      "learning_rate": 1.2651265126512652e-06,
      "loss": 1.025,
      "step": 3120
    },
    {
      "epoch": 15.5,
      "learning_rate": 1.2513751375137514e-06,
      "loss": 1.0102,
      "step": 3130
    },
    {
      "epoch": 15.54,
      "learning_rate": 1.2376237623762377e-06,
      "loss": 0.9965,
      "step": 3140
    },
    {
      "epoch": 15.59,
      "learning_rate": 1.2238723872387239e-06,
      "loss": 1.0295,
      "step": 3150
    },
    {
      "epoch": 15.64,
      "learning_rate": 1.2101210121012103e-06,
      "loss": 0.988,
      "step": 3160
    },
    {
      "epoch": 15.69,
      "learning_rate": 1.1963696369636966e-06,
      "loss": 1.001,
      "step": 3170
    },
    {
      "epoch": 15.74,
      "learning_rate": 1.1826182618261828e-06,
      "loss": 1.0232,
      "step": 3180
    },
    {
      "epoch": 15.79,
      "learning_rate": 1.168866886688669e-06,
      "loss": 1.0148,
      "step": 3190
    },
    {
      "epoch": 15.84,
      "learning_rate": 1.1551155115511552e-06,
      "loss": 0.9523,
      "step": 3200
    },
    {
      "epoch": 15.89,
      "learning_rate": 1.1413641364136415e-06,
      "loss": 1.0571,
      "step": 3210
    },
    {
      "epoch": 15.94,
      "learning_rate": 1.1276127612761277e-06,
      "loss": 0.973,
      "step": 3220
    },
    {
      "epoch": 15.99,
      "learning_rate": 1.113861386138614e-06,
      "loss": 0.9863,
      "step": 3230
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6739811912225705,
      "eval_loss": 0.8849846720695496,
      "eval_runtime": 38.9067,
      "eval_samples_per_second": 73.792,
      "eval_steps_per_second": 2.313,
      "step": 3232
    },
    {
      "epoch": 16.04,
      "learning_rate": 1.1001100110011001e-06,
      "loss": 0.9616,
      "step": 3240
    },
    {
      "epoch": 16.09,
      "learning_rate": 1.0863586358635866e-06,
      "loss": 1.0094,
      "step": 3250
    },
    {
      "epoch": 16.14,
      "learning_rate": 1.0726072607260726e-06,
      "loss": 0.9672,
      "step": 3260
    },
    {
      "epoch": 16.19,
      "learning_rate": 1.0588558855885588e-06,
      "loss": 1.0321,
      "step": 3270
    },
    {
      "epoch": 16.24,
      "learning_rate": 1.0451045104510453e-06,
      "loss": 1.0313,
      "step": 3280
    },
    {
      "epoch": 16.29,
      "learning_rate": 1.0313531353135315e-06,
      "loss": 1.0116,
      "step": 3290
    },
    {
      "epoch": 16.34,
      "learning_rate": 1.0176017601760177e-06,
      "loss": 0.9632,
      "step": 3300
    },
    {
      "epoch": 16.39,
      "learning_rate": 1.003850385038504e-06,
      "loss": 1.0021,
      "step": 3310
    },
    {
      "epoch": 16.44,
      "learning_rate": 9.900990099009902e-07,
      "loss": 0.9827,
      "step": 3320
    },
    {
      "epoch": 16.49,
      "learning_rate": 9.763476347634764e-07,
      "loss": 0.9916,
      "step": 3330
    },
    {
      "epoch": 16.53,
      "learning_rate": 9.625962596259626e-07,
      "loss": 0.9863,
      "step": 3340
    },
    {
      "epoch": 16.58,
      "learning_rate": 9.488448844884489e-07,
      "loss": 0.9428,
      "step": 3350
    },
    {
      "epoch": 16.63,
      "learning_rate": 9.350935093509352e-07,
      "loss": 0.9882,
      "step": 3360
    },
    {
      "epoch": 16.68,
      "learning_rate": 9.213421342134214e-07,
      "loss": 1.0206,
      "step": 3370
    },
    {
      "epoch": 16.73,
      "learning_rate": 9.075907590759076e-07,
      "loss": 0.9906,
      "step": 3380
    },
    {
      "epoch": 16.78,
      "learning_rate": 8.93839383938394e-07,
      "loss": 0.9814,
      "step": 3390
    },
    {
      "epoch": 16.83,
      "learning_rate": 8.800880088008802e-07,
      "loss": 0.9964,
      "step": 3400
    },
    {
      "epoch": 16.88,
      "learning_rate": 8.663366336633663e-07,
      "loss": 0.9854,
      "step": 3410
    },
    {
      "epoch": 16.93,
      "learning_rate": 8.525852585258527e-07,
      "loss": 1.0075,
      "step": 3420
    },
    {
      "epoch": 16.98,
      "learning_rate": 8.388338833883389e-07,
      "loss": 1.0058,
      "step": 3430
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.6746778126088471,
      "eval_loss": 0.8856182098388672,
      "eval_runtime": 38.4783,
      "eval_samples_per_second": 74.614,
      "eval_steps_per_second": 2.339,
      "step": 3434
    },
    {
      "epoch": 17.03,
      "learning_rate": 8.250825082508251e-07,
      "loss": 1.0082,
      "step": 3440
    },
    {
      "epoch": 17.08,
      "learning_rate": 8.113311331133114e-07,
      "loss": 0.9692,
      "step": 3450
    },
    {
      "epoch": 17.13,
      "learning_rate": 7.975797579757977e-07,
      "loss": 0.9329,
      "step": 3460
    },
    {
      "epoch": 17.18,
      "learning_rate": 7.838283828382839e-07,
      "loss": 0.9938,
      "step": 3470
    },
    {
      "epoch": 17.23,
      "learning_rate": 7.700770077007702e-07,
      "loss": 0.9911,
      "step": 3480
    },
    {
      "epoch": 17.28,
      "learning_rate": 7.563256325632564e-07,
      "loss": 1.0087,
      "step": 3490
    },
    {
      "epoch": 17.33,
      "learning_rate": 7.425742574257426e-07,
      "loss": 0.9697,
      "step": 3500
    },
    {
      "epoch": 17.38,
      "learning_rate": 7.288228822882289e-07,
      "loss": 0.973,
      "step": 3510
    },
    {
      "epoch": 17.43,
      "learning_rate": 7.150715071507151e-07,
      "loss": 1.0056,
      "step": 3520
    },
    {
      "epoch": 17.48,
      "learning_rate": 7.013201320132014e-07,
      "loss": 0.9723,
      "step": 3530
    },
    {
      "epoch": 17.52,
      "learning_rate": 6.875687568756877e-07,
      "loss": 0.9905,
      "step": 3540
    },
    {
      "epoch": 17.57,
      "learning_rate": 6.738173817381739e-07,
      "loss": 1.0209,
      "step": 3550
    },
    {
      "epoch": 17.62,
      "learning_rate": 6.6006600660066e-07,
      "loss": 0.9991,
      "step": 3560
    },
    {
      "epoch": 17.67,
      "learning_rate": 6.463146314631463e-07,
      "loss": 1.0039,
      "step": 3570
    },
    {
      "epoch": 17.72,
      "learning_rate": 6.325632563256326e-07,
      "loss": 0.946,
      "step": 3580
    },
    {
      "epoch": 17.77,
      "learning_rate": 6.188118811881188e-07,
      "loss": 0.9709,
      "step": 3590
    },
    {
      "epoch": 17.82,
      "learning_rate": 6.050605060506052e-07,
      "loss": 1.0213,
      "step": 3600
    },
    {
      "epoch": 17.87,
      "learning_rate": 5.913091309130914e-07,
      "loss": 1.0016,
      "step": 3610
    },
    {
      "epoch": 17.92,
      "learning_rate": 5.775577557755776e-07,
      "loss": 0.9889,
      "step": 3620
    },
    {
      "epoch": 17.97,
      "learning_rate": 5.638063806380638e-07,
      "loss": 0.9637,
      "step": 3630
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.6722396377568791,
      "eval_loss": 0.8851516842842102,
      "eval_runtime": 38.5349,
      "eval_samples_per_second": 74.504,
      "eval_steps_per_second": 2.336,
      "step": 3636
    },
    {
      "epoch": 18.02,
      "learning_rate": 5.500550055005501e-07,
      "loss": 0.9832,
      "step": 3640
    },
    {
      "epoch": 18.07,
      "learning_rate": 5.363036303630363e-07,
      "loss": 0.9492,
      "step": 3650
    },
    {
      "epoch": 18.12,
      "learning_rate": 5.225522552255226e-07,
      "loss": 0.9676,
      "step": 3660
    },
    {
      "epoch": 18.17,
      "learning_rate": 5.088008800880089e-07,
      "loss": 0.991,
      "step": 3670
    },
    {
      "epoch": 18.22,
      "learning_rate": 4.950495049504951e-07,
      "loss": 1.0016,
      "step": 3680
    },
    {
      "epoch": 18.27,
      "learning_rate": 4.812981298129813e-07,
      "loss": 1.0346,
      "step": 3690
    },
    {
      "epoch": 18.32,
      "learning_rate": 4.675467546754676e-07,
      "loss": 1.0101,
      "step": 3700
    },
    {
      "epoch": 18.37,
      "learning_rate": 4.537953795379538e-07,
      "loss": 1.0226,
      "step": 3710
    },
    {
      "epoch": 18.42,
      "learning_rate": 4.400440044004401e-07,
      "loss": 1.0054,
      "step": 3720
    },
    {
      "epoch": 18.47,
      "learning_rate": 4.2629262926292633e-07,
      "loss": 0.994,
      "step": 3730
    },
    {
      "epoch": 18.51,
      "learning_rate": 4.1254125412541255e-07,
      "loss": 1.0177,
      "step": 3740
    },
    {
      "epoch": 18.56,
      "learning_rate": 3.9878987898789883e-07,
      "loss": 0.97,
      "step": 3750
    },
    {
      "epoch": 18.61,
      "learning_rate": 3.850385038503851e-07,
      "loss": 0.9796,
      "step": 3760
    },
    {
      "epoch": 18.66,
      "learning_rate": 3.712871287128713e-07,
      "loss": 1.0034,
      "step": 3770
    },
    {
      "epoch": 18.71,
      "learning_rate": 3.5753575357535757e-07,
      "loss": 0.9773,
      "step": 3780
    },
    {
      "epoch": 18.76,
      "learning_rate": 3.4378437843784385e-07,
      "loss": 0.984,
      "step": 3790
    },
    {
      "epoch": 18.81,
      "learning_rate": 3.3003300330033e-07,
      "loss": 0.9513,
      "step": 3800
    },
    {
      "epoch": 18.86,
      "learning_rate": 3.162816281628163e-07,
      "loss": 0.9628,
      "step": 3810
    },
    {
      "epoch": 18.91,
      "learning_rate": 3.025302530253026e-07,
      "loss": 0.9573,
      "step": 3820
    },
    {
      "epoch": 18.96,
      "learning_rate": 2.887788778877888e-07,
      "loss": 0.9803,
      "step": 3830
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.6753744339951236,
      "eval_loss": 0.8829284906387329,
      "eval_runtime": 38.4857,
      "eval_samples_per_second": 74.599,
      "eval_steps_per_second": 2.339,
      "step": 3838
    },
    {
      "epoch": 19.01,
      "learning_rate": 2.7502750275027504e-07,
      "loss": 0.9945,
      "step": 3840
    },
    {
      "epoch": 19.06,
      "learning_rate": 2.612761276127613e-07,
      "loss": 1.0255,
      "step": 3850
    },
    {
      "epoch": 19.11,
      "learning_rate": 2.4752475247524754e-07,
      "loss": 0.966,
      "step": 3860
    },
    {
      "epoch": 19.16,
      "learning_rate": 2.337733773377338e-07,
      "loss": 0.9895,
      "step": 3870
    },
    {
      "epoch": 19.21,
      "learning_rate": 2.2002200220022005e-07,
      "loss": 0.9694,
      "step": 3880
    },
    {
      "epoch": 19.26,
      "learning_rate": 2.0627062706270628e-07,
      "loss": 0.9622,
      "step": 3890
    },
    {
      "epoch": 19.31,
      "learning_rate": 1.9251925192519256e-07,
      "loss": 0.9951,
      "step": 3900
    },
    {
      "epoch": 19.36,
      "learning_rate": 1.7876787678767878e-07,
      "loss": 0.9759,
      "step": 3910
    },
    {
      "epoch": 19.41,
      "learning_rate": 1.65016501650165e-07,
      "loss": 1.019,
      "step": 3920
    },
    {
      "epoch": 19.46,
      "learning_rate": 1.512651265126513e-07,
      "loss": 1.0533,
      "step": 3930
    },
    {
      "epoch": 19.5,
      "learning_rate": 1.3751375137513752e-07,
      "loss": 0.9828,
      "step": 3940
    },
    {
      "epoch": 19.55,
      "learning_rate": 1.2376237623762377e-07,
      "loss": 1.0073,
      "step": 3950
    },
    {
      "epoch": 19.6,
      "learning_rate": 1.1001100110011003e-07,
      "loss": 0.9639,
      "step": 3960
    },
    {
      "epoch": 19.65,
      "learning_rate": 9.625962596259628e-08,
      "loss": 0.9632,
      "step": 3970
    },
    {
      "epoch": 19.7,
      "learning_rate": 8.25082508250825e-08,
      "loss": 0.9732,
      "step": 3980
    },
    {
      "epoch": 19.75,
      "learning_rate": 6.875687568756876e-08,
      "loss": 0.9832,
      "step": 3990
    },
    {
      "epoch": 19.8,
      "learning_rate": 5.500550055005501e-08,
      "loss": 1.0138,
      "step": 4000
    },
    {
      "epoch": 19.85,
      "learning_rate": 4.125412541254125e-08,
      "loss": 0.9895,
      "step": 4010
    },
    {
      "epoch": 19.9,
      "learning_rate": 2.7502750275027506e-08,
      "loss": 0.9244,
      "step": 4020
    },
    {
      "epoch": 19.95,
      "learning_rate": 1.3751375137513753e-08,
      "loss": 0.9557,
      "step": 4030
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.0,
      "loss": 0.9356,
      "step": 4040
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.6788575409265064,
      "eval_loss": 0.8812496662139893,
      "eval_runtime": 38.3816,
      "eval_samples_per_second": 74.801,
      "eval_steps_per_second": 2.345,
      "step": 4040
    },
    {
      "epoch": 20.0,
      "step": 4040,
      "total_flos": 4.0029439499828675e+19,
      "train_loss": 1.115002474690428,
      "train_runtime": 19507.0594,
      "train_samples_per_second": 26.491,
      "train_steps_per_second": 0.207
    }
  ],
  "logging_steps": 10,
  "max_steps": 4040,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 4.0029439499828675e+19,
  "trial_name": null,
  "trial_params": null
}