{
  "best_metric": 53.662828506943114,
  "best_model_checkpoint": "./checkpoint-3000",
  "epoch": 357.0004,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 2.5e-08,
      "loss": 0.6457,
      "step": 25
    },
    {
      "epoch": 3.0,
      "learning_rate": 5e-08,
      "loss": 0.6364,
      "step": 50
    },
    {
      "epoch": 5.0,
      "learning_rate": 7.5e-08,
      "loss": 0.6418,
      "step": 75
    },
    {
      "epoch": 7.0,
      "learning_rate": 1e-07,
      "loss": 0.6378,
      "step": 100
    },
    {
      "epoch": 8.0,
      "learning_rate": 1.25e-07,
      "loss": 0.6318,
      "step": 125
    },
    {
      "epoch": 10.0,
      "learning_rate": 1.5e-07,
      "loss": 0.6215,
      "step": 150
    },
    {
      "epoch": 12.0,
      "learning_rate": 1.75e-07,
      "loss": 0.6273,
      "step": 175
    },
    {
      "epoch": 14.0,
      "learning_rate": 2e-07,
      "loss": 0.6081,
      "step": 200
    },
    {
      "epoch": 16.0,
      "learning_rate": 2.25e-07,
      "loss": 0.6053,
      "step": 225
    },
    {
      "epoch": 17.0,
      "learning_rate": 2.5e-07,
      "loss": 0.5989,
      "step": 250
    },
    {
      "epoch": 19.0,
      "learning_rate": 2.75e-07,
      "loss": 0.5812,
      "step": 275
    },
    {
      "epoch": 21.0,
      "learning_rate": 3e-07,
      "loss": 0.5828,
      "step": 300
    },
    {
      "epoch": 23.0,
      "learning_rate": 3.25e-07,
      "loss": 0.5657,
      "step": 325
    },
    {
      "epoch": 24.0,
      "learning_rate": 3.5e-07,
      "loss": 0.5588,
      "step": 350
    },
    {
      "epoch": 26.0,
      "learning_rate": 3.75e-07,
      "loss": 0.5519,
      "step": 375
    },
    {
      "epoch": 28.0,
      "learning_rate": 4e-07,
      "loss": 0.545,
      "step": 400
    },
    {
      "epoch": 30.0,
      "learning_rate": 4.2499999999999995e-07,
      "loss": 0.533,
      "step": 425
    },
    {
      "epoch": 32.0,
      "learning_rate": 4.5e-07,
      "loss": 0.5229,
      "step": 450
    },
    {
      "epoch": 33.0,
      "learning_rate": 4.7499999999999995e-07,
      "loss": 0.5166,
      "step": 475
    },
    {
      "epoch": 35.0,
      "learning_rate": 5e-07,
      "loss": 0.5011,
      "step": 500
    },
    {
      "epoch": 35.0,
      "eval_loss": 0.5963125228881836,
      "eval_runtime": 242.2717,
      "eval_samples_per_second": 4.214,
      "eval_steps_per_second": 0.132,
      "eval_wer": 59.88676791247466,
      "step": 500
    },
    {
      "epoch": 37.0,
      "learning_rate": 4.972222222222222e-07,
      "loss": 0.4965,
      "step": 525
    },
    {
      "epoch": 39.0,
      "learning_rate": 4.944444444444445e-07,
      "loss": 0.4828,
      "step": 550
    },
    {
      "epoch": 41.0,
      "learning_rate": 4.916666666666666e-07,
      "loss": 0.4732,
      "step": 575
    },
    {
      "epoch": 42.0,
      "learning_rate": 4.888888888888889e-07,
      "loss": 0.4708,
      "step": 600
    },
    {
      "epoch": 44.0,
      "learning_rate": 4.861111111111111e-07,
      "loss": 0.4586,
      "step": 625
    },
    {
      "epoch": 46.0,
      "learning_rate": 4.833333333333333e-07,
      "loss": 0.4551,
      "step": 650
    },
    {
      "epoch": 48.0,
      "learning_rate": 4.805555555555555e-07,
      "loss": 0.4483,
      "step": 675
    },
    {
      "epoch": 49.0,
      "learning_rate": 4.777777777777778e-07,
      "loss": 0.437,
      "step": 700
    },
    {
      "epoch": 51.0,
      "learning_rate": 4.7499999999999995e-07,
      "loss": 0.4322,
      "step": 725
    },
    {
      "epoch": 53.0,
      "learning_rate": 4.722222222222222e-07,
      "loss": 0.4207,
      "step": 750
    },
    {
      "epoch": 55.0,
      "learning_rate": 4.694444444444444e-07,
      "loss": 0.4165,
      "step": 775
    },
    {
      "epoch": 57.0,
      "learning_rate": 4.6666666666666666e-07,
      "loss": 0.4121,
      "step": 800
    },
    {
      "epoch": 58.0,
      "learning_rate": 4.6388888888888886e-07,
      "loss": 0.4014,
      "step": 825
    },
    {
      "epoch": 60.0,
      "learning_rate": 4.611111111111111e-07,
      "loss": 0.3975,
      "step": 850
    },
    {
      "epoch": 62.0,
      "learning_rate": 4.5833333333333327e-07,
      "loss": 0.3938,
      "step": 875
    },
    {
      "epoch": 64.0,
      "learning_rate": 4.555555555555555e-07,
      "loss": 0.3806,
      "step": 900
    },
    {
      "epoch": 66.0,
      "learning_rate": 4.527777777777778e-07,
      "loss": 0.3835,
      "step": 925
    },
    {
      "epoch": 67.0,
      "learning_rate": 4.5e-07,
      "loss": 0.372,
      "step": 950
    },
    {
      "epoch": 69.0,
      "learning_rate": 4.4722222222222223e-07,
      "loss": 0.3652,
      "step": 975
    },
    {
      "epoch": 71.0,
      "learning_rate": 4.444444444444444e-07,
      "loss": 0.3648,
      "step": 1000
    },
    {
      "epoch": 71.0,
      "eval_loss": 0.5612689256668091,
      "eval_runtime": 219.2597,
      "eval_samples_per_second": 4.657,
      "eval_steps_per_second": 0.146,
      "eval_wer": 55.95424811598638,
      "step": 1000
    },
    {
      "epoch": 73.0,
      "learning_rate": 4.4166666666666664e-07,
      "loss": 0.3552,
      "step": 1025
    },
    {
      "epoch": 74.0,
      "learning_rate": 4.3888888888888884e-07,
      "loss": 0.3504,
      "step": 1050
    },
    {
      "epoch": 76.0,
      "learning_rate": 4.361111111111111e-07,
      "loss": 0.3452,
      "step": 1075
    },
    {
      "epoch": 78.0,
      "learning_rate": 4.3333333333333335e-07,
      "loss": 0.3425,
      "step": 1100
    },
    {
      "epoch": 80.0,
      "learning_rate": 4.3055555555555555e-07,
      "loss": 0.3314,
      "step": 1125
    },
    {
      "epoch": 82.0,
      "learning_rate": 4.2777777777777775e-07,
      "loss": 0.3347,
      "step": 1150
    },
    {
      "epoch": 83.0,
      "learning_rate": 4.2499999999999995e-07,
      "loss": 0.3247,
      "step": 1175
    },
    {
      "epoch": 85.0,
      "learning_rate": 4.222222222222222e-07,
      "loss": 0.3229,
      "step": 1200
    },
    {
      "epoch": 87.0,
      "learning_rate": 4.194444444444444e-07,
      "loss": 0.3165,
      "step": 1225
    },
    {
      "epoch": 89.0,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 0.3168,
      "step": 1250
    },
    {
      "epoch": 91.0,
      "learning_rate": 4.1388888888888887e-07,
      "loss": 0.3051,
      "step": 1275
    },
    {
      "epoch": 92.0,
      "learning_rate": 4.1111111111111107e-07,
      "loss": 0.3035,
      "step": 1300
    },
    {
      "epoch": 94.0,
      "learning_rate": 4.083333333333333e-07,
      "loss": 0.3009,
      "step": 1325
    },
    {
      "epoch": 96.0,
      "learning_rate": 4.055555555555555e-07,
      "loss": 0.2981,
      "step": 1350
    },
    {
      "epoch": 98.0,
      "learning_rate": 4.027777777777778e-07,
      "loss": 0.29,
      "step": 1375
    },
    {
      "epoch": 99.0,
      "learning_rate": 4e-07,
      "loss": 0.2854,
      "step": 1400
    },
    {
      "epoch": 101.0,
      "learning_rate": 3.972222222222222e-07,
      "loss": 0.284,
      "step": 1425
    },
    {
      "epoch": 103.0,
      "learning_rate": 3.9444444444444444e-07,
      "loss": 0.2805,
      "step": 1450
    },
    {
      "epoch": 105.0,
      "learning_rate": 3.9166666666666664e-07,
      "loss": 0.2743,
      "step": 1475
    },
    {
      "epoch": 107.0,
      "learning_rate": 3.888888888888889e-07,
      "loss": 0.2732,
      "step": 1500
    },
    {
      "epoch": 107.0,
      "eval_loss": 0.5504247546195984,
      "eval_runtime": 221.6747,
      "eval_samples_per_second": 4.606,
      "eval_steps_per_second": 0.144,
      "eval_wer": 54.4585134463104,
      "step": 1500
    },
    {
      "epoch": 108.0,
      "learning_rate": 3.861111111111111e-07,
      "loss": 0.2677,
      "step": 1525
    },
    {
      "epoch": 110.0,
      "learning_rate": 3.8333333333333335e-07,
      "loss": 0.2633,
      "step": 1550
    },
    {
      "epoch": 112.0,
      "learning_rate": 3.805555555555555e-07,
      "loss": 0.2601,
      "step": 1575
    },
    {
      "epoch": 114.0,
      "learning_rate": 3.7777777777777775e-07,
      "loss": 0.2608,
      "step": 1600
    },
    {
      "epoch": 116.0,
      "learning_rate": 3.75e-07,
      "loss": 0.2539,
      "step": 1625
    },
    {
      "epoch": 117.0,
      "learning_rate": 3.722222222222222e-07,
      "loss": 0.2513,
      "step": 1650
    },
    {
      "epoch": 119.0,
      "learning_rate": 3.6944444444444447e-07,
      "loss": 0.2483,
      "step": 1675
    },
    {
      "epoch": 121.0,
      "learning_rate": 3.666666666666666e-07,
      "loss": 0.2436,
      "step": 1700
    },
    {
      "epoch": 123.0,
      "learning_rate": 3.6388888888888887e-07,
      "loss": 0.2394,
      "step": 1725
    },
    {
      "epoch": 124.0,
      "learning_rate": 3.6111111111111107e-07,
      "loss": 0.2372,
      "step": 1750
    },
    {
      "epoch": 126.0,
      "learning_rate": 3.583333333333333e-07,
      "loss": 0.2347,
      "step": 1775
    },
    {
      "epoch": 128.0,
      "learning_rate": 3.5555555555555553e-07,
      "loss": 0.2312,
      "step": 1800
    },
    {
      "epoch": 130.0,
      "learning_rate": 3.527777777777778e-07,
      "loss": 0.2294,
      "step": 1825
    },
    {
      "epoch": 132.0,
      "learning_rate": 3.5e-07,
      "loss": 0.2241,
      "step": 1850
    },
    {
      "epoch": 133.0,
      "learning_rate": 3.472222222222222e-07,
      "loss": 0.2249,
      "step": 1875
    },
    {
      "epoch": 135.0,
      "learning_rate": 3.4444444444444444e-07,
      "loss": 0.2187,
      "step": 1900
    },
    {
      "epoch": 137.0,
      "learning_rate": 3.4166666666666664e-07,
      "loss": 0.2171,
      "step": 1925
    },
    {
      "epoch": 139.0,
      "learning_rate": 3.388888888888889e-07,
      "loss": 0.2139,
      "step": 1950
    },
    {
      "epoch": 141.0,
      "learning_rate": 3.361111111111111e-07,
      "loss": 0.2121,
      "step": 1975
    },
    {
      "epoch": 142.0,
      "learning_rate": 3.333333333333333e-07,
      "loss": 0.2081,
      "step": 2000
    },
    {
      "epoch": 142.0,
      "eval_loss": 0.5501875281333923,
      "eval_runtime": 220.7799,
      "eval_samples_per_second": 4.625,
      "eval_steps_per_second": 0.145,
      "eval_wer": 53.6704793236678,
      "step": 2000
    },
    {
      "epoch": 144.0,
      "learning_rate": 3.3055555555555556e-07,
      "loss": 0.2082,
      "step": 2025
    },
    {
      "epoch": 146.0,
      "learning_rate": 3.2777777777777776e-07,
      "loss": 0.2039,
      "step": 2050
    },
    {
      "epoch": 148.0,
      "learning_rate": 3.25e-07,
      "loss": 0.2022,
      "step": 2075
    },
    {
      "epoch": 149.0,
      "learning_rate": 3.222222222222222e-07,
      "loss": 0.1984,
      "step": 2100
    },
    {
      "epoch": 151.0,
      "learning_rate": 3.194444444444444e-07,
      "loss": 0.1969,
      "step": 2125
    },
    {
      "epoch": 153.0,
      "learning_rate": 3.166666666666666e-07,
      "loss": 0.1938,
      "step": 2150
    },
    {
      "epoch": 155.0,
      "learning_rate": 3.1388888888888887e-07,
      "loss": 0.193,
      "step": 2175
    },
    {
      "epoch": 157.0,
      "learning_rate": 3.111111111111111e-07,
      "loss": 0.1908,
      "step": 2200
    },
    {
      "epoch": 158.0,
      "learning_rate": 3.0833333333333333e-07,
      "loss": 0.1864,
      "step": 2225
    },
    {
      "epoch": 160.0,
      "learning_rate": 3.055555555555556e-07,
      "loss": 0.1846,
      "step": 2250
    },
    {
      "epoch": 162.0,
      "learning_rate": 3.0277777777777773e-07,
      "loss": 0.1834,
      "step": 2275
    },
    {
      "epoch": 164.0,
      "learning_rate": 3e-07,
      "loss": 0.1808,
      "step": 2300
    },
    {
      "epoch": 166.0,
      "learning_rate": 2.972222222222222e-07,
      "loss": 0.1787,
      "step": 2325
    },
    {
      "epoch": 167.0,
      "learning_rate": 2.9444444444444444e-07,
      "loss": 0.1769,
      "step": 2350
    },
    {
      "epoch": 169.0,
      "learning_rate": 2.916666666666667e-07,
      "loss": 0.174,
      "step": 2375
    },
    {
      "epoch": 171.0,
      "learning_rate": 2.8888888888888885e-07,
      "loss": 0.1718,
      "step": 2400
    },
    {
      "epoch": 173.0,
      "learning_rate": 2.861111111111111e-07,
      "loss": 0.1707,
      "step": 2425
    },
    {
      "epoch": 174.0,
      "learning_rate": 2.833333333333333e-07,
      "loss": 0.1684,
      "step": 2450
    },
    {
      "epoch": 176.0,
      "learning_rate": 2.8055555555555556e-07,
      "loss": 0.1671,
      "step": 2475
    },
    {
      "epoch": 178.0,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.1627,
      "step": 2500
    },
    {
      "epoch": 178.0,
      "eval_loss": 0.5558388233184814,
      "eval_runtime": 219.4948,
      "eval_samples_per_second": 4.652,
      "eval_steps_per_second": 0.146,
      "eval_wer": 53.82732106652385,
      "step": 2500
    },
    {
      "epoch": 180.0,
      "learning_rate": 2.75e-07,
      "loss": 0.1628,
      "step": 2525
    },
    {
      "epoch": 182.0,
      "learning_rate": 2.7222222222222216e-07,
      "loss": 0.1587,
      "step": 2550
    },
    {
      "epoch": 183.0,
      "learning_rate": 2.694444444444444e-07,
      "loss": 0.1598,
      "step": 2575
    },
    {
      "epoch": 185.0,
      "learning_rate": 2.6666666666666667e-07,
      "loss": 0.1575,
      "step": 2600
    },
    {
      "epoch": 187.0,
      "learning_rate": 2.638888888888889e-07,
      "loss": 0.1547,
      "step": 2625
    },
    {
      "epoch": 189.0,
      "learning_rate": 2.6111111111111113e-07,
      "loss": 0.1524,
      "step": 2650
    },
    {
      "epoch": 191.0,
      "learning_rate": 2.5833333333333333e-07,
      "loss": 0.1523,
      "step": 2675
    },
    {
      "epoch": 192.0,
      "learning_rate": 2.5555555555555553e-07,
      "loss": 0.1507,
      "step": 2700
    },
    {
      "epoch": 194.0,
      "learning_rate": 2.5277777777777773e-07,
      "loss": 0.1484,
      "step": 2725
    },
    {
      "epoch": 196.0,
      "learning_rate": 2.5e-07,
      "loss": 0.1468,
      "step": 2750
    },
    {
      "epoch": 198.0,
      "learning_rate": 2.4722222222222224e-07,
      "loss": 0.144,
      "step": 2775
    },
    {
      "epoch": 199.0,
      "learning_rate": 2.4444444444444445e-07,
      "loss": 0.1452,
      "step": 2800
    },
    {
      "epoch": 201.0,
      "learning_rate": 2.4166666666666665e-07,
      "loss": 0.1427,
      "step": 2825
    },
    {
      "epoch": 203.0,
      "learning_rate": 2.388888888888889e-07,
      "loss": 0.14,
      "step": 2850
    },
    {
      "epoch": 205.0,
      "learning_rate": 2.361111111111111e-07,
      "loss": 0.1399,
      "step": 2875
    },
    {
      "epoch": 207.0,
      "learning_rate": 2.3333333333333333e-07,
      "loss": 0.1384,
      "step": 2900
    },
    {
      "epoch": 208.0,
      "learning_rate": 2.3055555555555556e-07,
      "loss": 0.1366,
      "step": 2925
    },
    {
      "epoch": 210.0,
      "learning_rate": 2.2777777777777776e-07,
      "loss": 0.1363,
      "step": 2950
    },
    {
      "epoch": 212.0,
      "learning_rate": 2.25e-07,
      "loss": 0.1331,
      "step": 2975
    },
    {
      "epoch": 214.0,
      "learning_rate": 2.222222222222222e-07,
      "loss": 0.133,
      "step": 3000
    },
    {
      "epoch": 214.0,
      "eval_loss": 0.5628067255020142,
      "eval_runtime": 220.2103,
      "eval_samples_per_second": 4.636,
      "eval_steps_per_second": 0.145,
      "eval_wer": 53.662828506943114,
      "step": 3000
    },
    {
      "epoch": 216.0,
      "learning_rate": 2.1944444444444442e-07,
      "loss": 0.1324,
      "step": 3025
    },
    {
      "epoch": 217.0,
      "learning_rate": 2.1666666666666667e-07,
      "loss": 0.1301,
      "step": 3050
    },
    {
      "epoch": 219.0,
      "learning_rate": 2.1388888888888888e-07,
      "loss": 0.1288,
      "step": 3075
    },
    {
      "epoch": 221.0,
      "learning_rate": 2.111111111111111e-07,
      "loss": 0.128,
      "step": 3100
    },
    {
      "epoch": 223.0,
      "learning_rate": 2.0833333333333333e-07,
      "loss": 0.1266,
      "step": 3125
    },
    {
      "epoch": 224.0,
      "learning_rate": 2.0555555555555553e-07,
      "loss": 0.1258,
      "step": 3150
    },
    {
      "epoch": 226.0,
      "learning_rate": 2.0277777777777776e-07,
      "loss": 0.1251,
      "step": 3175
    },
    {
      "epoch": 228.0,
      "learning_rate": 2e-07,
      "loss": 0.1219,
      "step": 3200
    },
    {
      "epoch": 230.0,
      "learning_rate": 1.9722222222222222e-07,
      "loss": 0.1226,
      "step": 3225
    },
    {
      "epoch": 232.0,
      "learning_rate": 1.9444444444444445e-07,
      "loss": 0.1202,
      "step": 3250
    },
    {
      "epoch": 233.0,
      "learning_rate": 1.9166666666666668e-07,
      "loss": 0.1199,
      "step": 3275
    },
    {
      "epoch": 235.0,
      "learning_rate": 1.8888888888888888e-07,
      "loss": 0.1186,
      "step": 3300
    },
    {
      "epoch": 237.0,
      "learning_rate": 1.861111111111111e-07,
      "loss": 0.1173,
      "step": 3325
    },
    {
      "epoch": 239.0,
      "learning_rate": 1.833333333333333e-07,
      "loss": 0.1172,
      "step": 3350
    },
    {
      "epoch": 241.0,
      "learning_rate": 1.8055555555555554e-07,
      "loss": 0.115,
      "step": 3375
    },
    {
      "epoch": 242.0,
      "learning_rate": 1.7777777777777776e-07,
      "loss": 0.1151,
      "step": 3400
    },
    {
      "epoch": 244.0,
      "learning_rate": 1.75e-07,
      "loss": 0.1141,
      "step": 3425
    },
    {
      "epoch": 246.0,
      "learning_rate": 1.7222222222222222e-07,
      "loss": 0.1137,
      "step": 3450
    },
    {
      "epoch": 248.0,
      "learning_rate": 1.6944444444444445e-07,
      "loss": 0.1109,
      "step": 3475
    },
    {
      "epoch": 249.0,
      "learning_rate": 1.6666666666666665e-07,
      "loss": 0.1112,
      "step": 3500
    },
    {
      "epoch": 249.0,
      "eval_loss": 0.5695652365684509,
      "eval_runtime": 216.96,
      "eval_samples_per_second": 4.706,
      "eval_steps_per_second": 0.147,
      "eval_wer": 54.07979801843846,
      "step": 3500
    },
    {
      "epoch": 251.0,
      "learning_rate": 1.6388888888888888e-07,
      "loss": 0.1103,
      "step": 3525
    },
    {
      "epoch": 253.0,
      "learning_rate": 1.611111111111111e-07,
      "loss": 0.1112,
      "step": 3550
    },
    {
      "epoch": 255.0,
      "learning_rate": 1.583333333333333e-07,
      "loss": 0.1081,
      "step": 3575
    },
    {
      "epoch": 257.0,
      "learning_rate": 1.5555555555555556e-07,
      "loss": 0.108,
      "step": 3600
    },
    {
      "epoch": 258.0,
      "learning_rate": 1.527777777777778e-07,
      "loss": 0.107,
      "step": 3625
    },
    {
      "epoch": 260.0,
      "learning_rate": 1.5e-07,
      "loss": 0.1057,
      "step": 3650
    },
    {
      "epoch": 262.0,
      "learning_rate": 1.4722222222222222e-07,
      "loss": 0.1053,
      "step": 3675
    },
    {
      "epoch": 264.0,
      "learning_rate": 1.4444444444444442e-07,
      "loss": 0.1055,
      "step": 3700
    },
    {
      "epoch": 266.0,
      "learning_rate": 1.4166666666666665e-07,
      "loss": 0.1053,
      "step": 3725
    },
    {
      "epoch": 267.0,
      "learning_rate": 1.3888888888888888e-07,
      "loss": 0.1043,
      "step": 3750
    },
    {
      "epoch": 269.0,
      "learning_rate": 1.3611111111111108e-07,
      "loss": 0.1031,
      "step": 3775
    },
    {
      "epoch": 271.0,
      "learning_rate": 1.3333333333333334e-07,
      "loss": 0.1008,
      "step": 3800
    },
    {
      "epoch": 273.0,
      "learning_rate": 1.3055555555555556e-07,
      "loss": 0.1027,
      "step": 3825
    },
    {
      "epoch": 274.0,
      "learning_rate": 1.2777777777777777e-07,
      "loss": 0.0995,
      "step": 3850
    },
    {
      "epoch": 276.0,
      "learning_rate": 1.25e-07,
      "loss": 0.0997,
      "step": 3875
    },
    {
      "epoch": 278.0,
      "learning_rate": 1.2222222222222222e-07,
      "loss": 0.101,
      "step": 3900
    },
    {
      "epoch": 280.0,
      "learning_rate": 1.1944444444444445e-07,
      "loss": 0.0985,
      "step": 3925
    },
    {
      "epoch": 282.0,
      "learning_rate": 1.1666666666666667e-07,
      "loss": 0.0983,
      "step": 3950
    },
    {
      "epoch": 283.0,
      "learning_rate": 1.1388888888888888e-07,
      "loss": 0.0979,
      "step": 3975
    },
    {
      "epoch": 285.0,
      "learning_rate": 1.111111111111111e-07,
      "loss": 0.0973,
      "step": 4000
    },
    {
      "epoch": 285.0,
      "eval_loss": 0.5748891830444336,
      "eval_runtime": 217.6279,
      "eval_samples_per_second": 4.691,
      "eval_steps_per_second": 0.147,
      "eval_wer": 53.99946444282927,
      "step": 4000
    },
    {
      "epoch": 287.0,
      "learning_rate": 1.0833333333333334e-07,
      "loss": 0.0967,
      "step": 4025
    },
    {
      "epoch": 289.0,
      "learning_rate": 1.0555555555555555e-07,
      "loss": 0.0958,
      "step": 4050
    },
    {
      "epoch": 291.0,
      "learning_rate": 1.0277777777777777e-07,
      "loss": 0.0965,
      "step": 4075
    },
    {
      "epoch": 292.0,
      "learning_rate": 1e-07,
      "loss": 0.0965,
      "step": 4100
    },
    {
      "epoch": 294.0,
      "learning_rate": 9.722222222222222e-08,
      "loss": 0.0948,
      "step": 4125
    },
    {
      "epoch": 296.0,
      "learning_rate": 9.444444444444444e-08,
      "loss": 0.0941,
      "step": 4150
    },
    {
      "epoch": 298.0,
      "learning_rate": 9.166666666666665e-08,
      "loss": 0.0935,
      "step": 4175
    },
    {
      "epoch": 299.0,
      "learning_rate": 8.888888888888888e-08,
      "loss": 0.0937,
      "step": 4200
    },
    {
      "epoch": 301.0,
      "learning_rate": 8.611111111111111e-08,
      "loss": 0.0928,
      "step": 4225
    },
    {
      "epoch": 303.0,
      "learning_rate": 8.333333333333333e-08,
      "loss": 0.0924,
      "step": 4250
    },
    {
      "epoch": 305.0,
      "learning_rate": 8.055555555555555e-08,
      "loss": 0.0924,
      "step": 4275
    },
    {
      "epoch": 307.0,
      "learning_rate": 7.777777777777778e-08,
      "loss": 0.091,
      "step": 4300
    },
    {
      "epoch": 308.0,
      "learning_rate": 7.5e-08,
      "loss": 0.0916,
      "step": 4325
    },
    {
      "epoch": 310.0,
      "learning_rate": 7.222222222222221e-08,
      "loss": 0.0915,
      "step": 4350
    },
    {
      "epoch": 312.0,
      "learning_rate": 6.944444444444444e-08,
      "loss": 0.0907,
      "step": 4375
    },
    {
      "epoch": 314.0,
      "learning_rate": 6.666666666666667e-08,
      "loss": 0.0894,
      "step": 4400
    },
    {
      "epoch": 316.0,
      "learning_rate": 6.388888888888888e-08,
      "loss": 0.0911,
      "step": 4425
    },
    {
      "epoch": 317.0,
      "learning_rate": 6.111111111111111e-08,
      "loss": 0.0899,
      "step": 4450
    },
    {
      "epoch": 319.0,
      "learning_rate": 5.833333333333333e-08,
      "loss": 0.0892,
      "step": 4475
    },
    {
      "epoch": 321.0,
      "learning_rate": 5.555555555555555e-08,
      "loss": 0.0906,
      "step": 4500
    },
    {
      "epoch": 321.0,
      "eval_loss": 0.5782893300056458,
      "eval_runtime": 236.4641,
      "eval_samples_per_second": 4.318,
      "eval_steps_per_second": 0.135,
      "eval_wer": 54.14865536896064,
      "step": 4500
    },
    {
      "epoch": 323.0,
      "learning_rate": 5.2777777777777776e-08,
      "loss": 0.089,
      "step": 4525
    },
    {
      "epoch": 324.0,
      "learning_rate": 5e-08,
      "loss": 0.0891,
      "step": 4550
    },
    {
      "epoch": 326.0,
      "learning_rate": 4.722222222222222e-08,
      "loss": 0.0891,
      "step": 4575
    },
    {
      "epoch": 328.0,
      "learning_rate": 4.444444444444444e-08,
      "loss": 0.0878,
      "step": 4600
    },
    {
      "epoch": 330.0,
      "learning_rate": 4.166666666666666e-08,
      "loss": 0.0886,
      "step": 4625
    },
    {
      "epoch": 332.0,
      "learning_rate": 3.888888888888889e-08,
      "loss": 0.0877,
      "step": 4650
    },
    {
      "epoch": 333.0,
      "learning_rate": 3.6111111111111106e-08,
      "loss": 0.0882,
      "step": 4675
    },
    {
      "epoch": 335.0,
      "learning_rate": 3.3333333333333334e-08,
      "loss": 0.0871,
      "step": 4700
    },
    {
      "epoch": 337.0,
      "learning_rate": 3.0555555555555556e-08,
      "loss": 0.0886,
      "step": 4725
    },
    {
      "epoch": 339.0,
      "learning_rate": 2.7777777777777774e-08,
      "loss": 0.0882,
      "step": 4750
    },
    {
      "epoch": 341.0,
      "learning_rate": 2.5e-08,
      "loss": 0.0872,
      "step": 4775
    },
    {
      "epoch": 342.0,
      "learning_rate": 2.222222222222222e-08,
      "loss": 0.0871,
      "step": 4800
    },
    {
      "epoch": 344.0,
      "learning_rate": 1.9444444444444445e-08,
      "loss": 0.0884,
      "step": 4825
    },
    {
      "epoch": 346.0,
      "learning_rate": 1.6666666666666667e-08,
      "loss": 0.0864,
      "step": 4850
    },
    {
      "epoch": 348.0,
      "learning_rate": 1.3888888888888887e-08,
      "loss": 0.0878,
      "step": 4875
    },
    {
      "epoch": 349.0,
      "learning_rate": 1.111111111111111e-08,
      "loss": 0.0865,
      "step": 4900
    },
    {
      "epoch": 351.0,
      "learning_rate": 8.333333333333334e-09,
      "loss": 0.0873,
      "step": 4925
    },
    {
      "epoch": 353.0,
      "learning_rate": 5.555555555555555e-09,
      "loss": 0.0868,
      "step": 4950
    },
    {
      "epoch": 355.0,
      "learning_rate": 2.7777777777777776e-09,
      "loss": 0.0865,
      "step": 4975
    },
    {
      "epoch": 357.0,
      "learning_rate": 0.0,
      "loss": 0.0874,
      "step": 5000
    },
    {
      "epoch": 357.0,
      "eval_loss": 0.5793120861053467,
      "eval_runtime": 218.3558,
      "eval_samples_per_second": 4.676,
      "eval_steps_per_second": 0.147,
      "eval_wer": 54.228988944569835,
      "step": 5000
    },
    {
      "epoch": 357.0,
      "step": 5000,
      "total_flos": 2.024578836135936e+19,
      "train_loss": 0.22935251636505127,
      "train_runtime": 22650.8672,
      "train_samples_per_second": 14.127,
      "train_steps_per_second": 0.221
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "total_flos": 2.024578836135936e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}