{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.7169,
      "step": 25
    },
    {
      "epoch": 0.01,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.7164,
      "step": 50
    },
    {
      "epoch": 0.01,
      "learning_rate": 7.5e-07,
      "loss": 0.7121,
      "step": 75
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.7051,
      "step": 100
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.2400000000000002e-06,
      "loss": 0.6987,
      "step": 125
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.48e-06,
      "loss": 0.6951,
      "step": 150
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.73e-06,
      "loss": 0.6815,
      "step": 175
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.98e-06,
      "loss": 0.659,
      "step": 200
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.2300000000000002e-06,
      "loss": 0.6204,
      "step": 225
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.4800000000000004e-06,
      "loss": 0.586,
      "step": 250
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.7300000000000005e-06,
      "loss": 0.5613,
      "step": 275
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.97e-06,
      "loss": 0.5296,
      "step": 300
    },
    {
      "epoch": 0.03,
      "learning_rate": 3.2200000000000005e-06,
      "loss": 0.4957,
      "step": 325
    },
    {
      "epoch": 0.04,
      "learning_rate": 3.4700000000000002e-06,
      "loss": 0.4583,
      "step": 350
    },
    {
      "epoch": 0.04,
      "learning_rate": 3.7200000000000004e-06,
      "loss": 0.4508,
      "step": 375
    },
    {
      "epoch": 0.04,
      "learning_rate": 3.97e-06,
      "loss": 0.3925,
      "step": 400
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.22e-06,
      "loss": 0.3554,
      "step": 425
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.47e-06,
      "loss": 0.3238,
      "step": 450
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.7200000000000005e-06,
      "loss": 0.2907,
      "step": 475
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.970000000000001e-06,
      "loss": 0.2741,
      "step": 500
    },
    {
      "epoch": 0.05,
      "eval_accuracy": 0.9106,
      "eval_loss": 0.2429841309785843,
      "eval_runtime": 157.4737,
      "eval_samples_per_second": 63.503,
      "eval_steps_per_second": 0.349,
      "step": 500
    },
    {
      "epoch": 0.05,
      "learning_rate": 5.220000000000001e-06,
      "loss": 0.2663,
      "step": 525
    },
    {
      "epoch": 0.06,
      "learning_rate": 5.470000000000001e-06,
      "loss": 0.247,
      "step": 550
    },
    {
      "epoch": 0.06,
      "learning_rate": 5.72e-06,
      "loss": 0.21,
      "step": 575
    },
    {
      "epoch": 0.06,
      "learning_rate": 5.9700000000000004e-06,
      "loss": 0.2073,
      "step": 600
    },
    {
      "epoch": 0.06,
      "learning_rate": 6.220000000000001e-06,
      "loss": 0.2088,
      "step": 625
    },
    {
      "epoch": 0.07,
      "learning_rate": 6.470000000000001e-06,
      "loss": 0.2218,
      "step": 650
    },
    {
      "epoch": 0.07,
      "learning_rate": 6.720000000000001e-06,
      "loss": 0.2045,
      "step": 675
    },
    {
      "epoch": 0.07,
      "learning_rate": 6.97e-06,
      "loss": 0.1942,
      "step": 700
    },
    {
      "epoch": 0.07,
      "learning_rate": 7.22e-06,
      "loss": 0.1776,
      "step": 725
    },
    {
      "epoch": 0.07,
      "learning_rate": 7.4700000000000005e-06,
      "loss": 0.1816,
      "step": 750
    },
    {
      "epoch": 0.08,
      "learning_rate": 7.72e-06,
      "loss": 0.177,
      "step": 775
    },
    {
      "epoch": 0.08,
      "learning_rate": 7.970000000000002e-06,
      "loss": 0.1678,
      "step": 800
    },
    {
      "epoch": 0.08,
      "learning_rate": 8.220000000000001e-06,
      "loss": 0.1682,
      "step": 825
    },
    {
      "epoch": 0.09,
      "learning_rate": 8.47e-06,
      "loss": 0.1619,
      "step": 850
    },
    {
      "epoch": 0.09,
      "learning_rate": 8.720000000000001e-06,
      "loss": 0.1631,
      "step": 875
    },
    {
      "epoch": 0.09,
      "learning_rate": 8.97e-06,
      "loss": 0.1629,
      "step": 900
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.220000000000002e-06,
      "loss": 0.1671,
      "step": 925
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.47e-06,
      "loss": 0.1479,
      "step": 950
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.72e-06,
      "loss": 0.1494,
      "step": 975
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.970000000000001e-06,
      "loss": 0.1624,
      "step": 1000
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.9495,
      "eval_loss": 0.14469695091247559,
      "eval_runtime": 146.152,
      "eval_samples_per_second": 68.422,
      "eval_steps_per_second": 0.376,
      "step": 1000
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.975555555555556e-06,
      "loss": 0.149,
      "step": 1025
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.947777777777779e-06,
      "loss": 0.1482,
      "step": 1050
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.920000000000002e-06,
      "loss": 0.1335,
      "step": 1075
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.892222222222223e-06,
      "loss": 0.1405,
      "step": 1100
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.864444444444446e-06,
      "loss": 0.1226,
      "step": 1125
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.836666666666668e-06,
      "loss": 0.1263,
      "step": 1150
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.80888888888889e-06,
      "loss": 0.1383,
      "step": 1175
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.781111111111113e-06,
      "loss": 0.1338,
      "step": 1200
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.753333333333335e-06,
      "loss": 0.1226,
      "step": 1225
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.725555555555557e-06,
      "loss": 0.1342,
      "step": 1250
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.697777777777778e-06,
      "loss": 0.137,
      "step": 1275
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.67e-06,
      "loss": 0.1345,
      "step": 1300
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.642222222222223e-06,
      "loss": 0.1304,
      "step": 1325
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.614444444444445e-06,
      "loss": 0.1351,
      "step": 1350
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.586666666666667e-06,
      "loss": 0.1232,
      "step": 1375
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.55888888888889e-06,
      "loss": 0.1275,
      "step": 1400
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.531111111111112e-06,
      "loss": 0.1218,
      "step": 1425
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.503333333333334e-06,
      "loss": 0.119,
      "step": 1450
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.475555555555557e-06,
      "loss": 0.1188,
      "step": 1475
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.447777777777778e-06,
      "loss": 0.1196,
      "step": 1500
    },
    {
      "epoch": 0.15,
      "eval_accuracy": 0.9481,
      "eval_loss": 0.13628898561000824,
      "eval_runtime": 151.5738,
      "eval_samples_per_second": 65.974,
      "eval_steps_per_second": 0.363,
      "step": 1500
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.42e-06,
      "loss": 0.1205,
      "step": 1525
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.392222222222222e-06,
      "loss": 0.1105,
      "step": 1550
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.364444444444445e-06,
      "loss": 0.122,
      "step": 1575
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.336666666666666e-06,
      "loss": 0.1313,
      "step": 1600
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.30888888888889e-06,
      "loss": 0.1181,
      "step": 1625
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.281111111111112e-06,
      "loss": 0.1216,
      "step": 1650
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.253333333333333e-06,
      "loss": 0.1193,
      "step": 1675
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.225555555555556e-06,
      "loss": 0.1128,
      "step": 1700
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.197777777777779e-06,
      "loss": 0.1098,
      "step": 1725
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.17e-06,
      "loss": 0.1003,
      "step": 1750
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.142222222222223e-06,
      "loss": 0.121,
      "step": 1775
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.114444444444444e-06,
      "loss": 0.1176,
      "step": 1800
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.086666666666667e-06,
      "loss": 0.1112,
      "step": 1825
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.05888888888889e-06,
      "loss": 0.1078,
      "step": 1850
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.031111111111111e-06,
      "loss": 0.1121,
      "step": 1875
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.003333333333334e-06,
      "loss": 0.1034,
      "step": 1900
    },
    {
      "epoch": 0.19,
      "learning_rate": 8.975555555555557e-06,
      "loss": 0.1108,
      "step": 1925
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.947777777777778e-06,
      "loss": 0.1052,
      "step": 1950
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.920000000000001e-06,
      "loss": 0.0919,
      "step": 1975
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.892222222222224e-06,
      "loss": 0.1195,
      "step": 2000
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.9547,
      "eval_loss": 0.1344193071126938,
      "eval_runtime": 151.3523,
      "eval_samples_per_second": 66.071,
      "eval_steps_per_second": 0.363,
      "step": 2000
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.864444444444445e-06,
      "loss": 0.0975,
      "step": 2025
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.836666666666668e-06,
      "loss": 0.115,
      "step": 2050
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.80888888888889e-06,
      "loss": 0.1068,
      "step": 2075
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.781111111111112e-06,
      "loss": 0.0911,
      "step": 2100
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.753333333333333e-06,
      "loss": 0.1056,
      "step": 2125
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.725555555555556e-06,
      "loss": 0.1117,
      "step": 2150
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.697777777777779e-06,
      "loss": 0.1086,
      "step": 2175
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.67e-06,
      "loss": 0.1084,
      "step": 2200
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.642222222222223e-06,
      "loss": 0.116,
      "step": 2225
    },
    {
      "epoch": 0.23,
      "learning_rate": 8.614444444444446e-06,
      "loss": 0.0924,
      "step": 2250
    },
    {
      "epoch": 0.23,
      "learning_rate": 8.586666666666667e-06,
      "loss": 0.1115,
      "step": 2275
    },
    {
      "epoch": 0.23,
      "learning_rate": 8.55888888888889e-06,
      "loss": 0.0984,
      "step": 2300
    },
    {
      "epoch": 0.23,
      "learning_rate": 8.531111111111113e-06,
      "loss": 0.101,
      "step": 2325
    },
    {
      "epoch": 0.23,
      "learning_rate": 8.503333333333334e-06,
      "loss": 0.0994,
      "step": 2350
    },
    {
      "epoch": 0.24,
      "learning_rate": 8.475555555555555e-06,
      "loss": 0.102,
      "step": 2375
    },
    {
      "epoch": 0.24,
      "learning_rate": 8.447777777777778e-06,
      "loss": 0.1065,
      "step": 2400
    },
    {
      "epoch": 0.24,
      "learning_rate": 8.42e-06,
      "loss": 0.105,
      "step": 2425
    },
    {
      "epoch": 0.24,
      "learning_rate": 8.392222222222222e-06,
      "loss": 0.104,
      "step": 2450
    },
    {
      "epoch": 0.25,
      "learning_rate": 8.364444444444445e-06,
      "loss": 0.1191,
      "step": 2475
    },
    {
      "epoch": 0.25,
      "learning_rate": 8.337777777777777e-06,
      "loss": 0.0973,
      "step": 2500
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 0.9553,
      "eval_loss": 0.130845844745636,
      "eval_runtime": 152.4264,
      "eval_samples_per_second": 65.605,
      "eval_steps_per_second": 0.361,
      "step": 2500
    },
    {
      "epoch": 0.25,
      "learning_rate": 8.31e-06,
      "loss": 0.1062,
      "step": 2525
    },
    {
      "epoch": 0.26,
      "learning_rate": 8.282222222222223e-06,
      "loss": 0.1026,
      "step": 2550
    },
    {
      "epoch": 0.26,
      "learning_rate": 8.254444444444444e-06,
      "loss": 0.1072,
      "step": 2575
    },
    {
      "epoch": 0.26,
      "learning_rate": 8.226666666666667e-06,
      "loss": 0.1015,
      "step": 2600
    },
    {
      "epoch": 0.26,
      "learning_rate": 8.19888888888889e-06,
      "loss": 0.0997,
      "step": 2625
    },
    {
      "epoch": 0.27,
      "learning_rate": 8.171111111111113e-06,
      "loss": 0.1102,
      "step": 2650
    },
    {
      "epoch": 0.27,
      "learning_rate": 8.143333333333334e-06,
      "loss": 0.1022,
      "step": 2675
    },
    {
      "epoch": 0.27,
      "learning_rate": 8.115555555555557e-06,
      "loss": 0.0958,
      "step": 2700
    },
    {
      "epoch": 0.27,
      "learning_rate": 8.08777777777778e-06,
      "loss": 0.102,
      "step": 2725
    },
    {
      "epoch": 0.28,
      "learning_rate": 8.06e-06,
      "loss": 0.1028,
      "step": 2750
    },
    {
      "epoch": 0.28,
      "learning_rate": 8.032222222222222e-06,
      "loss": 0.1041,
      "step": 2775
    },
    {
      "epoch": 0.28,
      "learning_rate": 8.004444444444445e-06,
      "loss": 0.1026,
      "step": 2800
    },
    {
      "epoch": 0.28,
      "learning_rate": 7.976666666666668e-06,
      "loss": 0.1158,
      "step": 2825
    },
    {
      "epoch": 0.28,
      "learning_rate": 7.948888888888889e-06,
      "loss": 0.1059,
      "step": 2850
    },
    {
      "epoch": 0.29,
      "learning_rate": 7.921111111111112e-06,
      "loss": 0.0931,
      "step": 2875
    },
    {
      "epoch": 0.29,
      "learning_rate": 7.893333333333335e-06,
      "loss": 0.0988,
      "step": 2900
    },
    {
      "epoch": 0.29,
      "learning_rate": 7.865555555555556e-06,
      "loss": 0.1071,
      "step": 2925
    },
    {
      "epoch": 0.29,
      "learning_rate": 7.837777777777779e-06,
      "loss": 0.1026,
      "step": 2950
    },
    {
      "epoch": 0.3,
      "learning_rate": 7.810000000000001e-06,
      "loss": 0.108,
      "step": 2975
    },
    {
      "epoch": 0.3,
      "learning_rate": 7.782222222222223e-06,
      "loss": 0.1039,
      "step": 3000
    },
    {
      "epoch": 0.3,
      "eval_accuracy": 0.9479,
      "eval_loss": 0.15836530923843384,
      "eval_runtime": 144.1908,
      "eval_samples_per_second": 69.353,
      "eval_steps_per_second": 0.381,
      "step": 3000
    },
    {
      "epoch": 0.3,
      "learning_rate": 7.754444444444445e-06,
      "loss": 0.1099,
      "step": 3025
    },
    {
      "epoch": 0.3,
      "learning_rate": 7.726666666666667e-06,
      "loss": 0.1021,
      "step": 3050
    },
    {
      "epoch": 0.31,
      "learning_rate": 7.69888888888889e-06,
      "loss": 0.0981,
      "step": 3075
    },
    {
      "epoch": 0.31,
      "learning_rate": 7.67111111111111e-06,
      "loss": 0.1085,
      "step": 3100
    },
    {
      "epoch": 0.31,
      "learning_rate": 7.643333333333334e-06,
      "loss": 0.0973,
      "step": 3125
    },
    {
      "epoch": 0.32,
      "learning_rate": 7.6155555555555564e-06,
      "loss": 0.107,
      "step": 3150
    },
    {
      "epoch": 0.32,
      "learning_rate": 7.5877777777777784e-06,
      "loss": 0.1035,
      "step": 3175
    },
    {
      "epoch": 0.32,
      "learning_rate": 7.5600000000000005e-06,
      "loss": 0.098,
      "step": 3200
    },
    {
      "epoch": 0.32,
      "learning_rate": 7.532222222222223e-06,
      "loss": 0.1066,
      "step": 3225
    },
    {
      "epoch": 0.33,
      "learning_rate": 7.504444444444445e-06,
      "loss": 0.0945,
      "step": 3250
    },
    {
      "epoch": 0.33,
      "learning_rate": 7.476666666666667e-06,
      "loss": 0.1128,
      "step": 3275
    },
    {
      "epoch": 0.33,
      "learning_rate": 7.44888888888889e-06,
      "loss": 0.1143,
      "step": 3300
    },
    {
      "epoch": 0.33,
      "learning_rate": 7.421111111111112e-06,
      "loss": 0.1122,
      "step": 3325
    },
    {
      "epoch": 0.34,
      "learning_rate": 7.393333333333333e-06,
      "loss": 0.1083,
      "step": 3350
    },
    {
      "epoch": 0.34,
      "learning_rate": 7.3655555555555554e-06,
      "loss": 0.108,
      "step": 3375
    },
    {
      "epoch": 0.34,
      "learning_rate": 7.337777777777778e-06,
      "loss": 0.1222,
      "step": 3400
    },
    {
      "epoch": 0.34,
      "learning_rate": 7.31e-06,
      "loss": 0.1238,
      "step": 3425
    },
    {
      "epoch": 0.34,
      "learning_rate": 7.282222222222222e-06,
      "loss": 0.1102,
      "step": 3450
    },
    {
      "epoch": 0.35,
      "learning_rate": 7.254444444444445e-06,
      "loss": 0.1293,
      "step": 3475
    },
    {
      "epoch": 0.35,
      "learning_rate": 7.226666666666667e-06,
      "loss": 0.1183,
      "step": 3500
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 0.952,
      "eval_loss": 0.141157865524292,
      "eval_runtime": 146.5068,
      "eval_samples_per_second": 68.256,
      "eval_steps_per_second": 0.375,
      "step": 3500
    },
    {
      "epoch": 0.35,
      "learning_rate": 7.19888888888889e-06,
      "loss": 0.1272,
      "step": 3525
    },
    {
      "epoch": 0.35,
      "learning_rate": 7.171111111111112e-06,
      "loss": 0.1219,
      "step": 3550
    },
    {
      "epoch": 0.36,
      "learning_rate": 7.143333333333334e-06,
      "loss": 0.1157,
      "step": 3575
    },
    {
      "epoch": 0.36,
      "learning_rate": 7.115555555555557e-06,
      "loss": 0.1077,
      "step": 3600
    },
    {
      "epoch": 0.36,
      "learning_rate": 7.087777777777778e-06,
      "loss": 0.1179,
      "step": 3625
    },
    {
      "epoch": 0.36,
      "learning_rate": 7.06e-06,
      "loss": 0.1219,
      "step": 3650
    },
    {
      "epoch": 0.37,
      "learning_rate": 7.032222222222222e-06,
      "loss": 0.1155,
      "step": 3675
    },
    {
      "epoch": 0.37,
      "learning_rate": 7.004444444444445e-06,
      "loss": 0.1293,
      "step": 3700
    },
    {
      "epoch": 0.37,
      "learning_rate": 6.976666666666667e-06,
      "loss": 0.1277,
      "step": 3725
    },
    {
      "epoch": 0.38,
      "learning_rate": 6.948888888888889e-06,
      "loss": 0.1191,
      "step": 3750
    },
    {
      "epoch": 0.38,
      "learning_rate": 6.921111111111112e-06,
      "loss": 0.1328,
      "step": 3775
    },
    {
      "epoch": 0.38,
      "learning_rate": 6.893333333333334e-06,
      "loss": 0.1188,
      "step": 3800
    },
    {
      "epoch": 0.38,
      "learning_rate": 6.865555555555557e-06,
      "loss": 0.1124,
      "step": 3825
    },
    {
      "epoch": 0.39,
      "learning_rate": 6.837777777777779e-06,
      "loss": 0.1272,
      "step": 3850
    },
    {
      "epoch": 0.39,
      "learning_rate": 6.810000000000001e-06,
      "loss": 0.1149,
      "step": 3875
    },
    {
      "epoch": 0.39,
      "learning_rate": 6.782222222222222e-06,
      "loss": 0.1173,
      "step": 3900
    },
    {
      "epoch": 0.39,
      "learning_rate": 6.754444444444445e-06,
      "loss": 0.1283,
      "step": 3925
    },
    {
      "epoch": 0.4,
      "learning_rate": 6.726666666666667e-06,
      "loss": 0.1104,
      "step": 3950
    },
    {
      "epoch": 0.4,
      "learning_rate": 6.698888888888889e-06,
      "loss": 0.1336,
      "step": 3975
    },
    {
      "epoch": 0.4,
      "learning_rate": 6.671111111111112e-06,
      "loss": 0.1103,
      "step": 4000
    },
    {
      "epoch": 0.4,
      "eval_accuracy": 0.9528,
      "eval_loss": 0.1352531760931015,
      "eval_runtime": 148.8877,
      "eval_samples_per_second": 67.165,
      "eval_steps_per_second": 0.369,
      "step": 4000
    },
    {
      "epoch": 0.4,
      "learning_rate": 6.643333333333334e-06,
      "loss": 0.1158,
      "step": 4025
    },
    {
      "epoch": 0.41,
      "learning_rate": 6.615555555555556e-06,
      "loss": 0.1001,
      "step": 4050
    },
    {
      "epoch": 0.41,
      "learning_rate": 6.587777777777779e-06,
      "loss": 0.1147,
      "step": 4075
    },
    {
      "epoch": 0.41,
      "learning_rate": 6.560000000000001e-06,
      "loss": 0.1122,
      "step": 4100
    },
    {
      "epoch": 0.41,
      "learning_rate": 6.532222222222223e-06,
      "loss": 0.1053,
      "step": 4125
    },
    {
      "epoch": 0.41,
      "learning_rate": 6.504444444444446e-06,
      "loss": 0.1126,
      "step": 4150
    },
    {
      "epoch": 0.42,
      "learning_rate": 6.476666666666667e-06,
      "loss": 0.1198,
      "step": 4175
    },
    {
      "epoch": 0.42,
      "learning_rate": 6.448888888888889e-06,
      "loss": 0.1138,
      "step": 4200
    },
    {
      "epoch": 0.42,
      "learning_rate": 6.421111111111112e-06,
      "loss": 0.1145,
      "step": 4225
    },
    {
      "epoch": 0.42,
      "learning_rate": 6.393333333333334e-06,
      "loss": 0.1186,
      "step": 4250
    },
    {
      "epoch": 0.43,
      "learning_rate": 6.365555555555556e-06,
      "loss": 0.1152,
      "step": 4275
    },
    {
      "epoch": 0.43,
      "learning_rate": 6.3377777777777786e-06,
      "loss": 0.108,
      "step": 4300
    },
    {
      "epoch": 0.43,
      "learning_rate": 6.3100000000000006e-06,
      "loss": 0.122,
      "step": 4325
    },
    {
      "epoch": 0.43,
      "learning_rate": 6.282222222222223e-06,
      "loss": 0.1036,
      "step": 4350
    },
    {
      "epoch": 0.44,
      "learning_rate": 6.2544444444444455e-06,
      "loss": 0.1058,
      "step": 4375
    },
    {
      "epoch": 0.44,
      "learning_rate": 6.2266666666666675e-06,
      "loss": 0.1083,
      "step": 4400
    },
    {
      "epoch": 0.44,
      "learning_rate": 6.1988888888888895e-06,
      "loss": 0.1252,
      "step": 4425
    },
    {
      "epoch": 0.45,
      "learning_rate": 6.171111111111112e-06,
      "loss": 0.1181,
      "step": 4450
    },
    {
      "epoch": 0.45,
      "learning_rate": 6.1433333333333335e-06,
      "loss": 0.1092,
      "step": 4475
    },
    {
      "epoch": 0.45,
      "learning_rate": 6.1155555555555555e-06,
      "loss": 0.1084,
      "step": 4500
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.9578,
      "eval_loss": 0.11692265421152115,
      "eval_runtime": 147.2838,
      "eval_samples_per_second": 67.896,
      "eval_steps_per_second": 0.373,
      "step": 4500
    },
    {
      "epoch": 0.45,
      "learning_rate": 6.087777777777778e-06,
      "loss": 0.0946,
      "step": 4525
    },
    {
      "epoch": 0.46,
      "learning_rate": 6.0600000000000004e-06,
      "loss": 0.1256,
      "step": 4550
    },
    {
      "epoch": 0.46,
      "learning_rate": 6.0322222222222224e-06,
      "loss": 0.1077,
      "step": 4575
    },
    {
      "epoch": 0.46,
      "learning_rate": 6.004444444444445e-06,
      "loss": 0.1159,
      "step": 4600
    },
    {
      "epoch": 0.46,
      "learning_rate": 5.976666666666667e-06,
      "loss": 0.0976,
      "step": 4625
    },
    {
      "epoch": 0.47,
      "learning_rate": 5.948888888888889e-06,
      "loss": 0.1128,
      "step": 4650
    },
    {
      "epoch": 0.47,
      "learning_rate": 5.921111111111112e-06,
      "loss": 0.1148,
      "step": 4675
    },
    {
      "epoch": 0.47,
      "learning_rate": 5.893333333333334e-06,
      "loss": 0.0886,
      "step": 4700
    },
    {
      "epoch": 0.47,
      "learning_rate": 5.8666666666666675e-06,
      "loss": 0.1125,
      "step": 4725
    },
    {
      "epoch": 0.47,
      "learning_rate": 5.8388888888888895e-06,
      "loss": 0.116,
      "step": 4750
    },
    {
      "epoch": 0.48,
      "learning_rate": 5.8111111111111116e-06,
      "loss": 0.1091,
      "step": 4775
    },
    {
      "epoch": 0.48,
      "learning_rate": 5.7833333333333344e-06,
      "loss": 0.1155,
      "step": 4800
    },
    {
      "epoch": 0.48,
      "learning_rate": 5.7555555555555564e-06,
      "loss": 0.1104,
      "step": 4825
    },
    {
      "epoch": 0.48,
      "learning_rate": 5.7277777777777785e-06,
      "loss": 0.1037,
      "step": 4850
    },
    {
      "epoch": 0.49,
      "learning_rate": 5.7e-06,
      "loss": 0.1101,
      "step": 4875
    },
    {
      "epoch": 0.49,
      "learning_rate": 5.6722222222222225e-06,
      "loss": 0.1039,
      "step": 4900
    },
    {
      "epoch": 0.49,
      "learning_rate": 5.6444444444444445e-06,
      "loss": 0.109,
      "step": 4925
    },
    {
      "epoch": 0.49,
      "learning_rate": 5.6166666666666665e-06,
      "loss": 0.1196,
      "step": 4950
    },
    {
      "epoch": 0.5,
      "learning_rate": 5.588888888888889e-06,
      "loss": 0.1116,
      "step": 4975
    },
    {
      "epoch": 0.5,
      "learning_rate": 5.561111111111111e-06,
      "loss": 0.1041,
      "step": 5000
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.952,
      "eval_loss": 0.1560685634613037,
      "eval_runtime": 155.1342,
      "eval_samples_per_second": 64.46,
      "eval_steps_per_second": 0.355,
      "step": 5000
    },
    {
      "epoch": 0.5,
      "learning_rate": 5.533333333333334e-06,
      "loss": 0.1097,
      "step": 5025
    },
    {
      "epoch": 0.51,
      "learning_rate": 5.505555555555556e-06,
      "loss": 0.1029,
      "step": 5050
    },
    {
      "epoch": 0.51,
      "learning_rate": 5.477777777777778e-06,
      "loss": 0.1015,
      "step": 5075
    },
    {
      "epoch": 0.51,
      "learning_rate": 5.450000000000001e-06,
      "loss": 0.1203,
      "step": 5100
    },
    {
      "epoch": 0.51,
      "learning_rate": 5.422222222222223e-06,
      "loss": 0.0967,
      "step": 5125
    },
    {
      "epoch": 0.52,
      "learning_rate": 5.394444444444444e-06,
      "loss": 0.1082,
      "step": 5150
    },
    {
      "epoch": 0.52,
      "learning_rate": 5.366666666666666e-06,
      "loss": 0.1175,
      "step": 5175
    },
    {
      "epoch": 0.52,
      "learning_rate": 5.338888888888889e-06,
      "loss": 0.1043,
      "step": 5200
    },
    {
      "epoch": 0.52,
      "learning_rate": 5.311111111111111e-06,
      "loss": 0.1035,
      "step": 5225
    },
    {
      "epoch": 0.53,
      "learning_rate": 5.283333333333333e-06,
      "loss": 0.1043,
      "step": 5250
    },
    {
      "epoch": 0.53,
      "learning_rate": 5.255555555555556e-06,
      "loss": 0.1002,
      "step": 5275
    },
    {
      "epoch": 0.53,
      "learning_rate": 5.227777777777778e-06,
      "loss": 0.0964,
      "step": 5300
    },
    {
      "epoch": 0.53,
      "learning_rate": 5.2e-06,
      "loss": 0.1069,
      "step": 5325
    },
    {
      "epoch": 0.54,
      "learning_rate": 5.172222222222223e-06,
      "loss": 0.1089,
      "step": 5350
    },
    {
      "epoch": 0.54,
      "learning_rate": 5.144444444444445e-06,
      "loss": 0.1092,
      "step": 5375
    },
    {
      "epoch": 0.54,
      "learning_rate": 5.116666666666668e-06,
      "loss": 0.1026,
      "step": 5400
    },
    {
      "epoch": 0.54,
      "learning_rate": 5.088888888888889e-06,
      "loss": 0.0982,
      "step": 5425
    },
    {
      "epoch": 0.55,
      "learning_rate": 5.061111111111111e-06,
      "loss": 0.0983,
      "step": 5450
    },
    {
      "epoch": 0.55,
      "learning_rate": 5.033333333333333e-06,
      "loss": 0.0978,
      "step": 5475
    },
    {
      "epoch": 0.55,
      "learning_rate": 5.005555555555556e-06,
      "loss": 0.0966,
      "step": 5500
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.9549,
      "eval_loss": 0.12704403698444366,
      "eval_runtime": 152.5968,
      "eval_samples_per_second": 65.532,
      "eval_steps_per_second": 0.36,
      "step": 5500
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.977777777777778e-06,
      "loss": 0.1094,
      "step": 5525
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.95e-06,
      "loss": 0.1042,
      "step": 5550
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.922222222222223e-06,
      "loss": 0.1,
      "step": 5575
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.894444444444445e-06,
      "loss": 0.1079,
      "step": 5600
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.866666666666667e-06,
      "loss": 0.1017,
      "step": 5625
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.838888888888889e-06,
      "loss": 0.0953,
      "step": 5650
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.811111111111111e-06,
      "loss": 0.094,
      "step": 5675
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.783333333333334e-06,
      "loss": 0.0966,
      "step": 5700
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.755555555555556e-06,
      "loss": 0.1056,
      "step": 5725
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.727777777777779e-06,
      "loss": 0.104,
      "step": 5750
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.7e-06,
      "loss": 0.0862,
      "step": 5775
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.672222222222223e-06,
      "loss": 0.1007,
      "step": 5800
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.644444444444445e-06,
      "loss": 0.1061,
      "step": 5825
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.616666666666667e-06,
      "loss": 0.0974,
      "step": 5850
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.58888888888889e-06,
      "loss": 0.097,
      "step": 5875
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.561111111111112e-06,
      "loss": 0.1002,
      "step": 5900
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.533333333333334e-06,
      "loss": 0.1069,
      "step": 5925
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.505555555555556e-06,
      "loss": 0.1128,
      "step": 5950
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.477777777777778e-06,
      "loss": 0.0957,
      "step": 5975
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.1,
      "step": 6000
    },
    {
      "epoch": 0.6,
      "eval_accuracy": 0.9594,
      "eval_loss": 0.11799036711454391,
      "eval_runtime": 145.5771,
      "eval_samples_per_second": 68.692,
      "eval_steps_per_second": 0.378,
      "step": 6000
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.422222222222223e-06,
      "loss": 0.098,
      "step": 6025
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.3944444444444455e-06,
      "loss": 0.098,
      "step": 6050
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.366666666666667e-06,
      "loss": 0.1047,
      "step": 6075
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.3388888888888895e-06,
      "loss": 0.0998,
      "step": 6100
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.3111111111111115e-06,
      "loss": 0.115,
      "step": 6125
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.2833333333333335e-06,
      "loss": 0.0965,
      "step": 6150
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.255555555555556e-06,
      "loss": 0.1031,
      "step": 6175
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.227777777777778e-06,
      "loss": 0.1105,
      "step": 6200
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.105,
      "step": 6225
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.1722222222222225e-06,
      "loss": 0.1022,
      "step": 6250
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.1444444444444445e-06,
      "loss": 0.0991,
      "step": 6275
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.116666666666667e-06,
      "loss": 0.1004,
      "step": 6300
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.088888888888889e-06,
      "loss": 0.1009,
      "step": 6325
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.061111111111111e-06,
      "loss": 0.1041,
      "step": 6350
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.033333333333333e-06,
      "loss": 0.091,
      "step": 6375
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.005555555555556e-06,
      "loss": 0.1042,
      "step": 6400
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.977777777777778e-06,
      "loss": 0.0923,
      "step": 6425
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.95e-06,
      "loss": 0.0957,
      "step": 6450
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.922222222222223e-06,
      "loss": 0.0927,
      "step": 6475
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.894444444444444e-06,
      "loss": 0.0965,
      "step": 6500
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.9551,
      "eval_loss": 0.12714883685112,
      "eval_runtime": 146.7045,
      "eval_samples_per_second": 68.164,
      "eval_steps_per_second": 0.375,
      "step": 6500
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.866666666666667e-06,
      "loss": 0.1084,
      "step": 6525
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.838888888888889e-06,
      "loss": 0.1,
      "step": 6550
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.8111111111111117e-06,
      "loss": 0.0976,
      "step": 6575
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.7833333333333337e-06,
      "loss": 0.0974,
      "step": 6600
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.7555555555555557e-06,
      "loss": 0.0927,
      "step": 6625
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.727777777777778e-06,
      "loss": 0.1042,
      "step": 6650
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.7e-06,
      "loss": 0.0954,
      "step": 6675
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.6722222222222226e-06,
      "loss": 0.0904,
      "step": 6700
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.644444444444445e-06,
      "loss": 0.0976,
      "step": 6725
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.616666666666667e-06,
      "loss": 0.0978,
      "step": 6750
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.588888888888889e-06,
      "loss": 0.0975,
      "step": 6775
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.561111111111111e-06,
      "loss": 0.1042,
      "step": 6800
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.5333333333333335e-06,
      "loss": 0.0997,
      "step": 6825
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.505555555555556e-06,
      "loss": 0.0996,
      "step": 6850
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.4777777777777784e-06,
      "loss": 0.1028,
      "step": 6875
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.45e-06,
      "loss": 0.0917,
      "step": 6900
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.4222222222222224e-06,
      "loss": 0.0857,
      "step": 6925
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.3944444444444445e-06,
      "loss": 0.0913,
      "step": 6950
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.366666666666667e-06,
      "loss": 0.0885,
      "step": 6975
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.3388888888888893e-06,
      "loss": 0.1008,
      "step": 7000
    },
    {
      "epoch": 0.7,
      "eval_accuracy": 0.9545,
      "eval_loss": 0.12300202995538712,
      "eval_runtime": 180.9152,
      "eval_samples_per_second": 55.275,
      "eval_steps_per_second": 0.304,
      "step": 7000
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.3111111111111118e-06,
      "loss": 0.1016,
      "step": 7025
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.2833333333333334e-06,
      "loss": 0.0993,
      "step": 7050
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.255555555555556e-06,
      "loss": 0.0938,
      "step": 7075
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.227777777777778e-06,
      "loss": 0.0941,
      "step": 7100
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0906,
      "step": 7125
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.1722222222222227e-06,
      "loss": 0.0857,
      "step": 7150
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.144444444444445e-06,
      "loss": 0.0876,
      "step": 7175
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.1166666666666668e-06,
      "loss": 0.089,
      "step": 7200
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.088888888888889e-06,
      "loss": 0.0919,
      "step": 7225
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.0611111111111112e-06,
      "loss": 0.0887,
      "step": 7250
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.0333333333333337e-06,
      "loss": 0.0882,
      "step": 7275
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.005555555555556e-06,
      "loss": 0.0915,
      "step": 7300
    },
    {
      "epoch": 0.73,
      "learning_rate": 2.9777777777777777e-06,
      "loss": 0.0835,
      "step": 7325
    },
    {
      "epoch": 0.73,
      "learning_rate": 2.95e-06,
      "loss": 0.0909,
      "step": 7350
    },
    {
      "epoch": 0.74,
      "learning_rate": 2.9222222222222226e-06,
      "loss": 0.0934,
      "step": 7375
    },
    {
      "epoch": 0.74,
      "learning_rate": 2.8944444444444446e-06,
      "loss": 0.0941,
      "step": 7400
    },
    {
      "epoch": 0.74,
      "learning_rate": 2.866666666666667e-06,
      "loss": 0.0885,
      "step": 7425
    },
    {
      "epoch": 0.74,
      "learning_rate": 2.8388888888888895e-06,
      "loss": 0.0843,
      "step": 7450
    },
    {
      "epoch": 0.75,
      "learning_rate": 2.811111111111111e-06,
      "loss": 0.0894,
      "step": 7475
    },
    {
      "epoch": 0.75,
      "learning_rate": 2.7833333333333335e-06,
      "loss": 0.1005,
      "step": 7500
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.9566,
      "eval_loss": 0.1161317378282547,
      "eval_runtime": 156.7623,
      "eval_samples_per_second": 63.791,
      "eval_steps_per_second": 0.351,
      "step": 7500
    },
    {
      "epoch": 0.75,
      "learning_rate": 2.755555555555556e-06,
      "loss": 0.0928,
      "step": 7525
    },
    {
      "epoch": 0.76,
      "learning_rate": 2.727777777777778e-06,
      "loss": 0.0872,
      "step": 7550
    },
    {
      "epoch": 0.76,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.0908,
      "step": 7575
    },
    {
      "epoch": 0.76,
      "learning_rate": 2.672222222222223e-06,
      "loss": 0.0889,
      "step": 7600
    },
    {
      "epoch": 0.76,
      "learning_rate": 2.6444444444444444e-06,
      "loss": 0.0902,
      "step": 7625
    },
    {
      "epoch": 0.77,
      "learning_rate": 2.616666666666667e-06,
      "loss": 0.0907,
      "step": 7650
    },
    {
      "epoch": 0.77,
      "learning_rate": 2.5888888888888893e-06,
      "loss": 0.0853,
      "step": 7675
    },
    {
      "epoch": 0.77,
      "learning_rate": 2.5611111111111113e-06,
      "loss": 0.1042,
      "step": 7700
    },
    {
      "epoch": 0.77,
      "learning_rate": 2.5333333333333338e-06,
      "loss": 0.0931,
      "step": 7725
    },
    {
      "epoch": 0.78,
      "learning_rate": 2.5055555555555554e-06,
      "loss": 0.0949,
      "step": 7750
    },
    {
      "epoch": 0.78,
      "learning_rate": 2.4777777777777782e-06,
      "loss": 0.0746,
      "step": 7775
    },
    {
      "epoch": 0.78,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.0874,
      "step": 7800
    },
    {
      "epoch": 0.78,
      "learning_rate": 2.4222222222222223e-06,
      "loss": 0.0856,
      "step": 7825
    },
    {
      "epoch": 0.79,
      "learning_rate": 2.3944444444444447e-06,
      "loss": 0.0802,
      "step": 7850
    },
    {
      "epoch": 0.79,
      "learning_rate": 2.3666666666666667e-06,
      "loss": 0.0711,
      "step": 7875
    },
    {
      "epoch": 0.79,
      "learning_rate": 2.338888888888889e-06,
      "loss": 0.0828,
      "step": 7900
    },
    {
      "epoch": 0.79,
      "learning_rate": 2.311111111111111e-06,
      "loss": 0.0852,
      "step": 7925
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.2833333333333336e-06,
      "loss": 0.0823,
      "step": 7950
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.2555555555555557e-06,
      "loss": 0.0853,
      "step": 7975
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.2277777777777777e-06,
      "loss": 0.0868,
      "step": 8000
    },
    {
      "epoch": 0.8,
      "eval_accuracy": 0.9543,
      "eval_loss": 0.12787339091300964,
      "eval_runtime": 146.1122,
      "eval_samples_per_second": 68.441,
      "eval_steps_per_second": 0.376,
      "step": 8000
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.2e-06,
      "loss": 0.0888,
      "step": 8025
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.1722222222222226e-06,
      "loss": 0.089,
      "step": 8050
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.1444444444444446e-06,
      "loss": 0.0798,
      "step": 8075
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.116666666666667e-06,
      "loss": 0.0798,
      "step": 8100
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.088888888888889e-06,
      "loss": 0.0783,
      "step": 8125
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.061111111111111e-06,
      "loss": 0.0872,
      "step": 8150
    },
    {
      "epoch": 0.82,
      "learning_rate": 2.0333333333333335e-06,
      "loss": 0.0756,
      "step": 8175
    },
    {
      "epoch": 0.82,
      "learning_rate": 2.0055555555555555e-06,
      "loss": 0.0891,
      "step": 8200
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.977777777777778e-06,
      "loss": 0.0878,
      "step": 8225
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.0965,
      "step": 8250
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.9222222222222224e-06,
      "loss": 0.0844,
      "step": 8275
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.8944444444444446e-06,
      "loss": 0.1039,
      "step": 8300
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.8666666666666669e-06,
      "loss": 0.0919,
      "step": 8325
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.8388888888888889e-06,
      "loss": 0.0914,
      "step": 8350
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.8111111111111113e-06,
      "loss": 0.1006,
      "step": 8375
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.7833333333333336e-06,
      "loss": 0.0891,
      "step": 8400
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.7555555555555556e-06,
      "loss": 0.0945,
      "step": 8425
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.727777777777778e-06,
      "loss": 0.0798,
      "step": 8450
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.0888,
      "step": 8475
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6722222222222223e-06,
      "loss": 0.0747,
      "step": 8500
    },
    {
      "epoch": 0.85,
      "eval_accuracy": 0.9578,
      "eval_loss": 0.12338610738515854,
      "eval_runtime": 146.8921,
      "eval_samples_per_second": 68.077,
      "eval_steps_per_second": 0.374,
      "step": 8500
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6444444444444447e-06,
      "loss": 0.084,
      "step": 8525
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6166666666666667e-06,
      "loss": 0.0764,
      "step": 8550
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.588888888888889e-06,
      "loss": 0.082,
      "step": 8575
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.5611111111111114e-06,
      "loss": 0.0892,
      "step": 8600
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.5333333333333334e-06,
      "loss": 0.0853,
      "step": 8625
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.5055555555555556e-06,
      "loss": 0.0833,
      "step": 8650
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.477777777777778e-06,
      "loss": 0.0917,
      "step": 8675
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.45e-06,
      "loss": 0.0783,
      "step": 8700
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.4222222222222223e-06,
      "loss": 0.0959,
      "step": 8725
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.3944444444444446e-06,
      "loss": 0.0758,
      "step": 8750
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.3666666666666668e-06,
      "loss": 0.0806,
      "step": 8775
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.338888888888889e-06,
      "loss": 0.0808,
      "step": 8800
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.3111111111111112e-06,
      "loss": 0.0808,
      "step": 8825
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.2833333333333335e-06,
      "loss": 0.0805,
      "step": 8850
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.2555555555555557e-06,
      "loss": 0.0822,
      "step": 8875
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.227777777777778e-06,
      "loss": 0.0861,
      "step": 8900
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.094,
      "step": 8925
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.1722222222222224e-06,
      "loss": 0.0798,
      "step": 8950
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.1444444444444446e-06,
      "loss": 0.0881,
      "step": 8975
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.1166666666666666e-06,
      "loss": 0.1015,
      "step": 9000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.9578,
      "eval_loss": 0.11542944610118866,
      "eval_runtime": 146.1261,
      "eval_samples_per_second": 68.434,
      "eval_steps_per_second": 0.376,
      "step": 9000
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.0888888888888889e-06,
      "loss": 0.0892,
      "step": 9025
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.0611111111111113e-06,
      "loss": 0.0888,
      "step": 9050
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.0333333333333333e-06,
      "loss": 0.0877,
      "step": 9075
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.0055555555555556e-06,
      "loss": 0.1019,
      "step": 9100
    },
    {
      "epoch": 0.91,
      "learning_rate": 9.77777777777778e-07,
      "loss": 0.0959,
      "step": 9125
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.500000000000001e-07,
      "loss": 0.0919,
      "step": 9150
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.222222222222222e-07,
      "loss": 0.0957,
      "step": 9175
    },
    {
      "epoch": 0.92,
      "learning_rate": 8.944444444444445e-07,
      "loss": 0.0943,
      "step": 9200
    },
    {
      "epoch": 0.92,
      "learning_rate": 8.666666666666668e-07,
      "loss": 0.1008,
      "step": 9225
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.388888888888889e-07,
      "loss": 0.0911,
      "step": 9250
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.111111111111112e-07,
      "loss": 0.0967,
      "step": 9275
    },
    {
      "epoch": 0.93,
      "learning_rate": 7.833333333333335e-07,
      "loss": 0.0922,
      "step": 9300
    },
    {
      "epoch": 0.93,
      "learning_rate": 7.555555555555556e-07,
      "loss": 0.0969,
      "step": 9325
    },
    {
      "epoch": 0.94,
      "learning_rate": 7.277777777777778e-07,
      "loss": 0.1004,
      "step": 9350
    },
    {
      "epoch": 0.94,
      "learning_rate": 7.000000000000001e-07,
      "loss": 0.0875,
      "step": 9375
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.722222222222223e-07,
      "loss": 0.0918,
      "step": 9400
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.444444444444445e-07,
      "loss": 0.0946,
      "step": 9425
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.166666666666668e-07,
      "loss": 0.0878,
      "step": 9450
    },
    {
      "epoch": 0.95,
      "learning_rate": 5.888888888888889e-07,
      "loss": 0.0967,
      "step": 9475
    },
    {
      "epoch": 0.95,
      "learning_rate": 5.611111111111111e-07,
      "loss": 0.0973,
      "step": 9500
    },
    {
      "epoch": 0.95,
      "eval_accuracy": 0.9573,
      "eval_loss": 0.12337620556354523,
      "eval_runtime": 145.4242,
      "eval_samples_per_second": 68.764,
      "eval_steps_per_second": 0.378,
      "step": 9500
    },
    {
      "epoch": 0.95,
      "learning_rate": 5.333333333333335e-07,
      "loss": 0.0863,
      "step": 9525
    },
    {
      "epoch": 0.95,
      "learning_rate": 5.055555555555556e-07,
      "loss": 0.0944,
      "step": 9550
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.777777777777778e-07,
      "loss": 0.0927,
      "step": 9575
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 0.09,
      "step": 9600
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.2222222222222226e-07,
      "loss": 0.0919,
      "step": 9625
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.9444444444444444e-07,
      "loss": 0.1121,
      "step": 9650
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.666666666666667e-07,
      "loss": 0.0811,
      "step": 9675
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.4000000000000003e-07,
      "loss": 0.1077,
      "step": 9700
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.1222222222222226e-07,
      "loss": 0.093,
      "step": 9725
    },
    {
      "epoch": 0.97,
      "learning_rate": 2.844444444444445e-07,
      "loss": 0.0869,
      "step": 9750
    },
    {
      "epoch": 0.98,
      "learning_rate": 2.566666666666667e-07,
      "loss": 0.1002,
      "step": 9775
    },
    {
      "epoch": 0.98,
      "learning_rate": 2.2888888888888892e-07,
      "loss": 0.0925,
      "step": 9800
    },
    {
      "epoch": 0.98,
      "learning_rate": 2.0111111111111112e-07,
      "loss": 0.087,
      "step": 9825
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.7333333333333335e-07,
      "loss": 0.0909,
      "step": 9850
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.4555555555555558e-07,
      "loss": 0.1016,
      "step": 9875
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.1777777777777778e-07,
      "loss": 0.0943,
      "step": 9900
    },
    {
      "epoch": 0.99,
      "learning_rate": 9e-08,
      "loss": 0.1004,
      "step": 9925
    },
    {
      "epoch": 0.99,
      "learning_rate": 6.222222222222223e-08,
      "loss": 0.099,
      "step": 9950
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.4444444444444444e-08,
      "loss": 0.0976,
      "step": 9975
    },
    {
      "epoch": 1.0,
      "learning_rate": 6.666666666666667e-09,
      "loss": 0.0901,
      "step": 10000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9562,
      "eval_loss": 0.12256840616464615,
      "eval_runtime": 150.3828,
      "eval_samples_per_second": 66.497,
      "eval_steps_per_second": 0.366,
      "step": 10000
    }
  ],
  "max_steps": 10000,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 2.418844065244224e+17,
  "trial_name": null,
  "trial_params": null
}