{
  "best_metric": 11.732222816934506,
  "best_model_checkpoint": "./checkpoint-3000",
  "epoch": 4.09,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.4e-07,
      "loss": 1.3874,
      "step": 25
    },
    {
      "epoch": 0.01,
      "learning_rate": 9.200000000000001e-07,
      "loss": 1.224,
      "step": 50
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.42e-06,
      "loss": 0.7884,
      "step": 75
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.9200000000000003e-06,
      "loss": 0.466,
      "step": 100
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.42e-06,
      "loss": 0.2788,
      "step": 125
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.92e-06,
      "loss": 0.2233,
      "step": 150
    },
    {
      "epoch": 0.04,
      "learning_rate": 3.4200000000000007e-06,
      "loss": 0.2411,
      "step": 175
    },
    {
      "epoch": 0.05,
      "learning_rate": 3.920000000000001e-06,
      "loss": 0.2389,
      "step": 200
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.42e-06,
      "loss": 0.2536,
      "step": 225
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.92e-06,
      "loss": 0.2358,
      "step": 250
    },
    {
      "epoch": 0.07,
      "learning_rate": 5.420000000000001e-06,
      "loss": 0.2,
      "step": 275
    },
    {
      "epoch": 0.07,
      "learning_rate": 5.92e-06,
      "loss": 0.1946,
      "step": 300
    },
    {
      "epoch": 0.08,
      "learning_rate": 6.42e-06,
      "loss": 0.215,
      "step": 325
    },
    {
      "epoch": 0.09,
      "learning_rate": 6.92e-06,
      "loss": 0.1802,
      "step": 350
    },
    {
      "epoch": 0.09,
      "learning_rate": 7.420000000000001e-06,
      "loss": 0.2226,
      "step": 375
    },
    {
      "epoch": 0.1,
      "learning_rate": 7.92e-06,
      "loss": 0.2374,
      "step": 400
    },
    {
      "epoch": 0.11,
      "learning_rate": 8.42e-06,
      "loss": 0.2663,
      "step": 425
    },
    {
      "epoch": 0.11,
      "learning_rate": 8.920000000000001e-06,
      "loss": 0.3244,
      "step": 450
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.42e-06,
      "loss": 0.4516,
      "step": 475
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.920000000000002e-06,
      "loss": 0.5582,
      "step": 500
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.3659963607788086,
      "eval_runtime": 1037.0778,
      "eval_samples_per_second": 2.468,
      "eval_steps_per_second": 0.077,
      "eval_wer": 14.817038942050365,
      "step": 500
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.940000000000001e-06,
      "loss": 0.6142,
      "step": 525
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.871428571428572e-06,
      "loss": 0.5451,
      "step": 550
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.3885,
      "step": 575
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.72857142857143e-06,
      "loss": 0.2781,
      "step": 600
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.657142857142859e-06,
      "loss": 0.2405,
      "step": 625
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.585714285714286e-06,
      "loss": 0.2455,
      "step": 650
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.514285714285715e-06,
      "loss": 0.2063,
      "step": 675
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.442857142857144e-06,
      "loss": 0.2097,
      "step": 700
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.371428571428572e-06,
      "loss": 0.2479,
      "step": 725
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.3e-06,
      "loss": 0.2491,
      "step": 750
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.22857142857143e-06,
      "loss": 0.2182,
      "step": 775
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.157142857142857e-06,
      "loss": 0.2302,
      "step": 800
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.085714285714286e-06,
      "loss": 0.2773,
      "step": 825
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.014285714285715e-06,
      "loss": 0.2446,
      "step": 850
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.942857142857142e-06,
      "loss": 0.2213,
      "step": 875
    },
    {
      "epoch": 0.23,
      "learning_rate": 8.871428571428571e-06,
      "loss": 0.1809,
      "step": 900
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.8e-06,
      "loss": 0.2946,
      "step": 925
    },
    {
      "epoch": 1.01,
      "learning_rate": 8.72857142857143e-06,
      "loss": 0.3836,
      "step": 950
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.657142857142858e-06,
      "loss": 0.2838,
      "step": 975
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.585714285714286e-06,
      "loss": 0.2285,
      "step": 1000
    },
    {
      "epoch": 1.02,
      "eval_loss": 0.2918657660484314,
      "eval_runtime": 1060.8322,
      "eval_samples_per_second": 2.412,
      "eval_steps_per_second": 0.075,
      "eval_wer": 12.630383415360114,
      "step": 1000
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.514285714285715e-06,
      "loss": 0.1869,
      "step": 1025
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.442857142857144e-06,
      "loss": 0.1679,
      "step": 1050
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.371428571428573e-06,
      "loss": 0.1406,
      "step": 1075
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.3e-06,
      "loss": 0.1489,
      "step": 1100
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.22857142857143e-06,
      "loss": 0.1214,
      "step": 1125
    },
    {
      "epoch": 1.06,
      "learning_rate": 8.157142857142858e-06,
      "loss": 0.12,
      "step": 1150
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.085714285714287e-06,
      "loss": 0.1122,
      "step": 1175
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.014285714285715e-06,
      "loss": 0.1117,
      "step": 1200
    },
    {
      "epoch": 1.08,
      "learning_rate": 7.942857142857144e-06,
      "loss": 0.1437,
      "step": 1225
    },
    {
      "epoch": 1.08,
      "learning_rate": 7.871428571428573e-06,
      "loss": 0.177,
      "step": 1250
    },
    {
      "epoch": 1.09,
      "learning_rate": 7.800000000000002e-06,
      "loss": 0.1541,
      "step": 1275
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.72857142857143e-06,
      "loss": 0.1613,
      "step": 1300
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.657142857142858e-06,
      "loss": 0.1439,
      "step": 1325
    },
    {
      "epoch": 1.11,
      "learning_rate": 7.585714285714286e-06,
      "loss": 0.144,
      "step": 1350
    },
    {
      "epoch": 1.12,
      "learning_rate": 7.514285714285715e-06,
      "loss": 0.1429,
      "step": 1375
    },
    {
      "epoch": 1.12,
      "learning_rate": 7.442857142857144e-06,
      "loss": 0.145,
      "step": 1400
    },
    {
      "epoch": 1.13,
      "learning_rate": 7.371428571428571e-06,
      "loss": 0.2797,
      "step": 1425
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.3e-06,
      "loss": 0.3132,
      "step": 1450
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.2285714285714294e-06,
      "loss": 0.2924,
      "step": 1475
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.1571428571428584e-06,
      "loss": 0.2038,
      "step": 1500
    },
    {
      "epoch": 1.15,
      "eval_loss": 0.2794933021068573,
      "eval_runtime": 1001.5942,
      "eval_samples_per_second": 2.555,
      "eval_steps_per_second": 0.08,
      "eval_wer": 11.384968864636232,
      "step": 1500
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.085714285714286e-06,
      "loss": 0.137,
      "step": 1525
    },
    {
      "epoch": 1.16,
      "learning_rate": 7.014285714285715e-06,
      "loss": 0.1311,
      "step": 1550
    },
    {
      "epoch": 1.17,
      "learning_rate": 6.942857142857144e-06,
      "loss": 0.112,
      "step": 1575
    },
    {
      "epoch": 1.17,
      "learning_rate": 6.871428571428572e-06,
      "loss": 0.1416,
      "step": 1600
    },
    {
      "epoch": 1.18,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.1226,
      "step": 1625
    },
    {
      "epoch": 1.19,
      "learning_rate": 6.7314285714285724e-06,
      "loss": 0.12,
      "step": 1650
    },
    {
      "epoch": 1.19,
      "learning_rate": 6.660000000000001e-06,
      "loss": 0.1247,
      "step": 1675
    },
    {
      "epoch": 1.2,
      "learning_rate": 6.588571428571429e-06,
      "loss": 0.2006,
      "step": 1700
    },
    {
      "epoch": 1.2,
      "learning_rate": 6.517142857142858e-06,
      "loss": 0.3048,
      "step": 1725
    },
    {
      "epoch": 1.21,
      "learning_rate": 6.445714285714286e-06,
      "loss": 0.2532,
      "step": 1750
    },
    {
      "epoch": 1.22,
      "learning_rate": 6.374285714285715e-06,
      "loss": 0.1975,
      "step": 1775
    },
    {
      "epoch": 1.22,
      "learning_rate": 6.302857142857144e-06,
      "loss": 0.1865,
      "step": 1800
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.231428571428571e-06,
      "loss": 0.1657,
      "step": 1825
    },
    {
      "epoch": 2.01,
      "learning_rate": 6.16e-06,
      "loss": 0.0563,
      "step": 1850
    },
    {
      "epoch": 2.01,
      "learning_rate": 6.088571428571429e-06,
      "loss": 0.0617,
      "step": 1875
    },
    {
      "epoch": 2.02,
      "learning_rate": 6.017142857142858e-06,
      "loss": 0.0556,
      "step": 1900
    },
    {
      "epoch": 2.03,
      "learning_rate": 5.945714285714286e-06,
      "loss": 0.0645,
      "step": 1925
    },
    {
      "epoch": 2.03,
      "learning_rate": 5.874285714285715e-06,
      "loss": 0.0642,
      "step": 1950
    },
    {
      "epoch": 2.04,
      "learning_rate": 5.802857142857144e-06,
      "loss": 0.0786,
      "step": 1975
    },
    {
      "epoch": 2.04,
      "learning_rate": 5.731428571428572e-06,
      "loss": 0.074,
      "step": 2000
    },
    {
      "epoch": 2.04,
      "eval_loss": 0.3149697780609131,
      "eval_runtime": 1006.1929,
      "eval_samples_per_second": 2.543,
      "eval_steps_per_second": 0.08,
      "eval_wer": 12.10428062296837,
      "step": 2000
    },
    {
      "epoch": 2.05,
      "learning_rate": 5.66e-06,
      "loss": 0.0721,
      "step": 2025
    },
    {
      "epoch": 2.06,
      "learning_rate": 5.588571428571429e-06,
      "loss": 0.0649,
      "step": 2050
    },
    {
      "epoch": 2.06,
      "learning_rate": 5.517142857142857e-06,
      "loss": 0.0571,
      "step": 2075
    },
    {
      "epoch": 2.07,
      "learning_rate": 5.445714285714286e-06,
      "loss": 0.0728,
      "step": 2100
    },
    {
      "epoch": 2.08,
      "learning_rate": 5.374285714285715e-06,
      "loss": 0.0524,
      "step": 2125
    },
    {
      "epoch": 2.08,
      "learning_rate": 5.3028571428571425e-06,
      "loss": 0.077,
      "step": 2150
    },
    {
      "epoch": 2.09,
      "learning_rate": 5.2314285714285716e-06,
      "loss": 0.0717,
      "step": 2175
    },
    {
      "epoch": 2.1,
      "learning_rate": 5.1600000000000006e-06,
      "loss": 0.0977,
      "step": 2200
    },
    {
      "epoch": 2.1,
      "learning_rate": 5.08857142857143e-06,
      "loss": 0.1385,
      "step": 2225
    },
    {
      "epoch": 2.11,
      "learning_rate": 5.017142857142857e-06,
      "loss": 0.1941,
      "step": 2250
    },
    {
      "epoch": 2.11,
      "learning_rate": 4.945714285714286e-06,
      "loss": 0.1323,
      "step": 2275
    },
    {
      "epoch": 2.12,
      "learning_rate": 4.874285714285715e-06,
      "loss": 0.1016,
      "step": 2300
    },
    {
      "epoch": 2.13,
      "learning_rate": 4.802857142857143e-06,
      "loss": 0.1343,
      "step": 2325
    },
    {
      "epoch": 2.13,
      "learning_rate": 4.731428571428572e-06,
      "loss": 0.2062,
      "step": 2350
    },
    {
      "epoch": 2.14,
      "learning_rate": 4.66e-06,
      "loss": 0.1978,
      "step": 2375
    },
    {
      "epoch": 2.15,
      "learning_rate": 4.588571428571429e-06,
      "loss": 0.1342,
      "step": 2400
    },
    {
      "epoch": 2.15,
      "learning_rate": 4.5171428571428575e-06,
      "loss": 0.1188,
      "step": 2425
    },
    {
      "epoch": 2.16,
      "learning_rate": 4.445714285714286e-06,
      "loss": 0.1035,
      "step": 2450
    },
    {
      "epoch": 2.16,
      "learning_rate": 4.374285714285715e-06,
      "loss": 0.1606,
      "step": 2475
    },
    {
      "epoch": 2.17,
      "learning_rate": 4.302857142857143e-06,
      "loss": 0.2165,
      "step": 2500
    },
    {
      "epoch": 2.17,
      "eval_loss": 0.2978155016899109,
      "eval_runtime": 939.5906,
      "eval_samples_per_second": 2.724,
      "eval_steps_per_second": 0.085,
      "eval_wer": 12.851007167008262,
      "step": 2500
    },
    {
      "epoch": 2.18,
      "learning_rate": 4.231428571428572e-06,
      "loss": 0.204,
      "step": 2525
    },
    {
      "epoch": 2.18,
      "learning_rate": 4.16e-06,
      "loss": 0.1276,
      "step": 2550
    },
    {
      "epoch": 2.19,
      "learning_rate": 4.088571428571429e-06,
      "loss": 0.1049,
      "step": 2575
    },
    {
      "epoch": 2.19,
      "learning_rate": 4.017142857142857e-06,
      "loss": 0.1006,
      "step": 2600
    },
    {
      "epoch": 2.2,
      "learning_rate": 3.945714285714286e-06,
      "loss": 0.0767,
      "step": 2625
    },
    {
      "epoch": 2.21,
      "learning_rate": 3.874285714285715e-06,
      "loss": 0.083,
      "step": 2650
    },
    {
      "epoch": 2.21,
      "learning_rate": 3.802857142857143e-06,
      "loss": 0.0746,
      "step": 2675
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.731428571428572e-06,
      "loss": 0.0573,
      "step": 2700
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.66e-06,
      "loss": 0.0699,
      "step": 2725
    },
    {
      "epoch": 3.0,
      "learning_rate": 3.588571428571429e-06,
      "loss": 0.0473,
      "step": 2750
    },
    {
      "epoch": 3.01,
      "learning_rate": 3.5171428571428573e-06,
      "loss": 0.0565,
      "step": 2775
    },
    {
      "epoch": 3.02,
      "learning_rate": 3.4457142857142863e-06,
      "loss": 0.1134,
      "step": 2800
    },
    {
      "epoch": 3.02,
      "learning_rate": 3.3742857142857145e-06,
      "loss": 0.1051,
      "step": 2825
    },
    {
      "epoch": 3.03,
      "learning_rate": 3.302857142857143e-06,
      "loss": 0.0684,
      "step": 2850
    },
    {
      "epoch": 3.04,
      "learning_rate": 3.2314285714285716e-06,
      "loss": 0.0543,
      "step": 2875
    },
    {
      "epoch": 3.04,
      "learning_rate": 3.1600000000000002e-06,
      "loss": 0.039,
      "step": 2900
    },
    {
      "epoch": 3.05,
      "learning_rate": 3.0885714285714284e-06,
      "loss": 0.0344,
      "step": 2925
    },
    {
      "epoch": 3.06,
      "learning_rate": 3.0171428571428574e-06,
      "loss": 0.0298,
      "step": 2950
    },
    {
      "epoch": 3.06,
      "learning_rate": 2.945714285714286e-06,
      "loss": 0.0299,
      "step": 2975
    },
    {
      "epoch": 3.07,
      "learning_rate": 2.8742857142857146e-06,
      "loss": 0.0399,
      "step": 3000
    },
    {
      "epoch": 3.07,
      "eval_loss": 0.34670406579971313,
      "eval_runtime": 923.9331,
      "eval_samples_per_second": 2.77,
      "eval_steps_per_second": 0.087,
      "eval_wer": 11.732222816934506,
      "step": 3000
    },
    {
      "epoch": 3.07,
      "learning_rate": 2.802857142857143e-06,
      "loss": 0.0472,
      "step": 3025
    },
    {
      "epoch": 3.08,
      "learning_rate": 2.7314285714285714e-06,
      "loss": 0.0461,
      "step": 3050
    },
    {
      "epoch": 3.09,
      "learning_rate": 2.6600000000000004e-06,
      "loss": 0.0263,
      "step": 3075
    },
    {
      "epoch": 3.09,
      "learning_rate": 2.5885714285714285e-06,
      "loss": 0.0246,
      "step": 3100
    },
    {
      "epoch": 3.1,
      "learning_rate": 2.5171428571428575e-06,
      "loss": 0.0296,
      "step": 3125
    },
    {
      "epoch": 3.1,
      "learning_rate": 2.445714285714286e-06,
      "loss": 0.0223,
      "step": 3150
    },
    {
      "epoch": 3.11,
      "learning_rate": 2.3742857142857147e-06,
      "loss": 0.0278,
      "step": 3175
    },
    {
      "epoch": 3.12,
      "learning_rate": 2.302857142857143e-06,
      "loss": 0.0387,
      "step": 3200
    },
    {
      "epoch": 3.12,
      "learning_rate": 2.2314285714285715e-06,
      "loss": 0.0411,
      "step": 3225
    },
    {
      "epoch": 3.13,
      "learning_rate": 2.16e-06,
      "loss": 0.0396,
      "step": 3250
    },
    {
      "epoch": 3.14,
      "learning_rate": 2.0885714285714287e-06,
      "loss": 0.0308,
      "step": 3275
    },
    {
      "epoch": 3.14,
      "learning_rate": 2.0171428571428573e-06,
      "loss": 0.0369,
      "step": 3300
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.945714285714286e-06,
      "loss": 0.0376,
      "step": 3325
    },
    {
      "epoch": 3.15,
      "learning_rate": 1.8742857142857142e-06,
      "loss": 0.0439,
      "step": 3350
    },
    {
      "epoch": 3.16,
      "learning_rate": 1.8028571428571432e-06,
      "loss": 0.0563,
      "step": 3375
    },
    {
      "epoch": 3.17,
      "learning_rate": 1.7314285714285716e-06,
      "loss": 0.1056,
      "step": 3400
    },
    {
      "epoch": 3.17,
      "learning_rate": 1.6600000000000002e-06,
      "loss": 0.1284,
      "step": 3425
    },
    {
      "epoch": 3.18,
      "learning_rate": 1.5885714285714288e-06,
      "loss": 0.0671,
      "step": 3450
    },
    {
      "epoch": 3.19,
      "learning_rate": 1.5171428571428574e-06,
      "loss": 0.0579,
      "step": 3475
    },
    {
      "epoch": 3.19,
      "learning_rate": 1.4457142857142858e-06,
      "loss": 0.045,
      "step": 3500
    },
    {
      "epoch": 3.19,
      "eval_loss": 0.3500906527042389,
      "eval_runtime": 931.3693,
      "eval_samples_per_second": 2.748,
      "eval_steps_per_second": 0.086,
      "eval_wer": 11.721779089045835,
      "step": 3500
    },
    {
      "epoch": 3.2,
      "learning_rate": 1.3742857142857143e-06,
      "loss": 0.0313,
      "step": 3525
    },
    {
      "epoch": 3.21,
      "learning_rate": 1.302857142857143e-06,
      "loss": 0.0386,
      "step": 3550
    },
    {
      "epoch": 3.21,
      "learning_rate": 1.2314285714285715e-06,
      "loss": 0.0584,
      "step": 3575
    },
    {
      "epoch": 3.22,
      "learning_rate": 1.1600000000000001e-06,
      "loss": 0.0967,
      "step": 3600
    },
    {
      "epoch": 3.22,
      "learning_rate": 1.0885714285714287e-06,
      "loss": 0.1048,
      "step": 3625
    },
    {
      "epoch": 4.0,
      "learning_rate": 1.0171428571428573e-06,
      "loss": 0.0759,
      "step": 3650
    },
    {
      "epoch": 4.01,
      "learning_rate": 9.457142857142858e-07,
      "loss": 0.0125,
      "step": 3675
    },
    {
      "epoch": 4.01,
      "learning_rate": 8.742857142857144e-07,
      "loss": 0.0108,
      "step": 3700
    },
    {
      "epoch": 4.02,
      "learning_rate": 8.028571428571429e-07,
      "loss": 0.0093,
      "step": 3725
    },
    {
      "epoch": 4.03,
      "learning_rate": 7.314285714285715e-07,
      "loss": 0.0094,
      "step": 3750
    },
    {
      "epoch": 4.03,
      "learning_rate": 6.6e-07,
      "loss": 0.0164,
      "step": 3775
    },
    {
      "epoch": 4.04,
      "learning_rate": 5.885714285714286e-07,
      "loss": 0.02,
      "step": 3800
    },
    {
      "epoch": 4.05,
      "learning_rate": 5.171428571428572e-07,
      "loss": 0.0165,
      "step": 3825
    },
    {
      "epoch": 4.05,
      "learning_rate": 4.457142857142858e-07,
      "loss": 0.0171,
      "step": 3850
    },
    {
      "epoch": 4.06,
      "learning_rate": 3.7428571428571434e-07,
      "loss": 0.0227,
      "step": 3875
    },
    {
      "epoch": 4.07,
      "learning_rate": 3.028571428571429e-07,
      "loss": 0.0413,
      "step": 3900
    },
    {
      "epoch": 4.07,
      "learning_rate": 2.3142857142857144e-07,
      "loss": 0.0547,
      "step": 3925
    },
    {
      "epoch": 4.08,
      "learning_rate": 1.6e-07,
      "loss": 0.037,
      "step": 3950
    },
    {
      "epoch": 4.08,
      "learning_rate": 8.857142857142858e-08,
      "loss": 0.0281,
      "step": 3975
    },
    {
      "epoch": 4.09,
      "learning_rate": 1.7142857142857143e-08,
      "loss": 0.0187,
      "step": 4000
    },
    {
      "epoch": 4.09,
      "eval_loss": 0.38336294889450073,
      "eval_runtime": 948.6243,
      "eval_samples_per_second": 2.698,
      "eval_steps_per_second": 0.084,
      "eval_wer": 11.888878735264553,
      "step": 4000
    },
    {
      "epoch": 4.09,
      "step": 4000,
      "total_flos": 1.358659599630336e+20,
      "train_loss": 0.1602003634274006,
      "train_runtime": 36777.8553,
      "train_samples_per_second": 1.74,
      "train_steps_per_second": 0.109
    }
  ],
  "max_steps": 4000,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 1.358659599630336e+20,
  "trial_name": null,
  "trial_params": null
}