{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 99.34640522875817,
  "eval_steps": 500,
  "global_step": 11400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.87,
      "learning_rate": 4.950000000000001e-06,
      "loss": 14.1952,
      "step": 100
    },
    {
      "epoch": 1.74,
      "learning_rate": 9.950000000000001e-06,
      "loss": 6.4908,
      "step": 200
    },
    {
      "epoch": 2.61,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 4.5428,
      "step": 300
    },
    {
      "epoch": 3.49,
      "learning_rate": 1.995e-05,
      "loss": 3.8598,
      "step": 400
    },
    {
      "epoch": 4.36,
      "learning_rate": 2.495e-05,
      "loss": 3.3958,
      "step": 500
    },
    {
      "epoch": 4.36,
      "eval_loss": 3.3137385845184326,
      "eval_runtime": 268.8924,
      "eval_samples_per_second": 12.161,
      "eval_steps_per_second": 1.521,
      "eval_wer": 1.0,
      "step": 500
    },
    {
      "epoch": 5.23,
      "learning_rate": 2.995e-05,
      "loss": 3.2375,
      "step": 600
    },
    {
      "epoch": 6.1,
      "learning_rate": 3.495e-05,
      "loss": 3.1969,
      "step": 700
    },
    {
      "epoch": 6.97,
      "learning_rate": 3.995e-05,
      "loss": 3.151,
      "step": 800
    },
    {
      "epoch": 7.84,
      "learning_rate": 4.495e-05,
      "loss": 3.1114,
      "step": 900
    },
    {
      "epoch": 8.71,
      "learning_rate": 4.995e-05,
      "loss": 3.032,
      "step": 1000
    },
    {
      "epoch": 8.71,
      "eval_loss": 2.858614921569824,
      "eval_runtime": 245.2843,
      "eval_samples_per_second": 13.331,
      "eval_steps_per_second": 1.667,
      "eval_wer": 0.9992608441937366,
      "step": 1000
    },
    {
      "epoch": 9.59,
      "learning_rate": 5.495e-05,
      "loss": 2.5279,
      "step": 1100
    },
    {
      "epoch": 10.46,
      "learning_rate": 5.995000000000001e-05,
      "loss": 1.8013,
      "step": 1200
    },
    {
      "epoch": 11.33,
      "learning_rate": 6.494999999999999e-05,
      "loss": 1.5785,
      "step": 1300
    },
    {
      "epoch": 12.2,
      "learning_rate": 6.995e-05,
      "loss": 1.468,
      "step": 1400
    },
    {
      "epoch": 13.07,
      "learning_rate": 7.495e-05,
      "loss": 1.3977,
      "step": 1500
    },
    {
      "epoch": 13.07,
      "eval_loss": 0.4785650372505188,
      "eval_runtime": 252.0133,
      "eval_samples_per_second": 12.976,
      "eval_steps_per_second": 1.623,
      "eval_wer": 0.637541334370745,
      "step": 1500
    },
    {
      "epoch": 13.94,
      "learning_rate": 7.99e-05,
      "loss": 1.3681,
      "step": 1600
    },
    {
      "epoch": 14.81,
      "learning_rate": 8.49e-05,
      "loss": 1.336,
      "step": 1700
    },
    {
      "epoch": 15.69,
      "learning_rate": 8.99e-05,
      "loss": 1.307,
      "step": 1800
    },
    {
      "epoch": 16.56,
      "learning_rate": 9.49e-05,
      "loss": 1.2934,
      "step": 1900
    },
    {
      "epoch": 17.43,
      "learning_rate": 9.99e-05,
      "loss": 1.2751,
      "step": 2000
    },
    {
      "epoch": 17.43,
      "eval_loss": 0.3816487491130829,
      "eval_runtime": 244.8161,
      "eval_samples_per_second": 13.357,
      "eval_steps_per_second": 1.671,
      "eval_wer": 0.53931141801206,
      "step": 2000
    },
    {
      "epoch": 18.3,
      "learning_rate": 9.895744680851065e-05,
      "loss": 1.2545,
      "step": 2100
    },
    {
      "epoch": 19.17,
      "learning_rate": 9.78936170212766e-05,
      "loss": 1.2488,
      "step": 2200
    },
    {
      "epoch": 20.04,
      "learning_rate": 9.682978723404255e-05,
      "loss": 1.2315,
      "step": 2300
    },
    {
      "epoch": 20.92,
      "learning_rate": 9.576595744680852e-05,
      "loss": 1.2237,
      "step": 2400
    },
    {
      "epoch": 21.79,
      "learning_rate": 9.470212765957447e-05,
      "loss": 1.2113,
      "step": 2500
    },
    {
      "epoch": 21.79,
      "eval_loss": 0.3450602889060974,
      "eval_runtime": 260.0099,
      "eval_samples_per_second": 12.576,
      "eval_steps_per_second": 1.573,
      "eval_wer": 0.509900797510212,
      "step": 2500
    },
    {
      "epoch": 22.66,
      "learning_rate": 9.363829787234043e-05,
      "loss": 1.204,
      "step": 2600
    },
    {
      "epoch": 23.53,
      "learning_rate": 9.257446808510639e-05,
      "loss": 1.1834,
      "step": 2700
    },
    {
      "epoch": 24.4,
      "learning_rate": 9.151063829787234e-05,
      "loss": 1.1782,
      "step": 2800
    },
    {
      "epoch": 25.27,
      "learning_rate": 9.04468085106383e-05,
      "loss": 1.1616,
      "step": 2900
    },
    {
      "epoch": 26.14,
      "learning_rate": 8.938297872340426e-05,
      "loss": 1.156,
      "step": 3000
    },
    {
      "epoch": 26.14,
      "eval_loss": 0.3244517743587494,
      "eval_runtime": 254.1562,
      "eval_samples_per_second": 12.866,
      "eval_steps_per_second": 1.609,
      "eval_wer": 0.49192764053686056,
      "step": 3000
    },
    {
      "epoch": 27.02,
      "learning_rate": 8.831914893617022e-05,
      "loss": 1.165,
      "step": 3100
    },
    {
      "epoch": 27.89,
      "learning_rate": 8.725531914893618e-05,
      "loss": 1.1559,
      "step": 3200
    },
    {
      "epoch": 28.76,
      "learning_rate": 8.619148936170212e-05,
      "loss": 1.1239,
      "step": 3300
    },
    {
      "epoch": 29.63,
      "learning_rate": 8.51276595744681e-05,
      "loss": 1.1219,
      "step": 3400
    },
    {
      "epoch": 30.5,
      "learning_rate": 8.406382978723405e-05,
      "loss": 1.1226,
      "step": 3500
    },
    {
      "epoch": 30.5,
      "eval_loss": 0.2992089092731476,
      "eval_runtime": 263.7476,
      "eval_samples_per_second": 12.398,
      "eval_steps_per_second": 1.551,
      "eval_wer": 0.44411593075277184,
      "step": 3500
    },
    {
      "epoch": 31.37,
      "learning_rate": 8.3e-05,
      "loss": 1.1115,
      "step": 3600
    },
    {
      "epoch": 32.24,
      "learning_rate": 8.194680851063831e-05,
      "loss": 1.0983,
      "step": 3700
    },
    {
      "epoch": 33.12,
      "learning_rate": 8.088297872340426e-05,
      "loss": 1.0957,
      "step": 3800
    },
    {
      "epoch": 33.99,
      "learning_rate": 7.981914893617022e-05,
      "loss": 1.0874,
      "step": 3900
    },
    {
      "epoch": 34.86,
      "learning_rate": 7.875531914893617e-05,
      "loss": 1.0913,
      "step": 4000
    },
    {
      "epoch": 34.86,
      "eval_loss": 0.2831476330757141,
      "eval_runtime": 264.6688,
      "eval_samples_per_second": 12.355,
      "eval_steps_per_second": 1.545,
      "eval_wer": 0.431472476171951,
      "step": 4000
    },
    {
      "epoch": 35.73,
      "learning_rate": 7.769148936170213e-05,
      "loss": 1.0896,
      "step": 4100
    },
    {
      "epoch": 36.6,
      "learning_rate": 7.663829787234042e-05,
      "loss": 1.071,
      "step": 4200
    },
    {
      "epoch": 37.47,
      "learning_rate": 7.557446808510639e-05,
      "loss": 1.0635,
      "step": 4300
    },
    {
      "epoch": 38.34,
      "learning_rate": 7.451063829787235e-05,
      "loss": 1.0689,
      "step": 4400
    },
    {
      "epoch": 39.22,
      "learning_rate": 7.34468085106383e-05,
      "loss": 1.0615,
      "step": 4500
    },
    {
      "epoch": 39.22,
      "eval_loss": 0.2807982563972473,
      "eval_runtime": 253.9269,
      "eval_samples_per_second": 12.878,
      "eval_steps_per_second": 1.611,
      "eval_wer": 0.4339622641509434,
      "step": 4500
    },
    {
      "epoch": 40.09,
      "learning_rate": 7.238297872340425e-05,
      "loss": 1.0553,
      "step": 4600
    },
    {
      "epoch": 40.96,
      "learning_rate": 7.131914893617023e-05,
      "loss": 1.0533,
      "step": 4700
    },
    {
      "epoch": 41.83,
      "learning_rate": 7.025531914893617e-05,
      "loss": 1.0546,
      "step": 4800
    },
    {
      "epoch": 42.7,
      "learning_rate": 6.919148936170213e-05,
      "loss": 1.0517,
      "step": 4900
    },
    {
      "epoch": 43.57,
      "learning_rate": 6.812765957446809e-05,
      "loss": 1.0455,
      "step": 5000
    },
    {
      "epoch": 43.57,
      "eval_loss": 0.27128756046295166,
      "eval_runtime": 251.6872,
      "eval_samples_per_second": 12.992,
      "eval_steps_per_second": 1.625,
      "eval_wer": 0.4087531608636452,
      "step": 5000
    },
    {
      "epoch": 44.44,
      "learning_rate": 6.706382978723405e-05,
      "loss": 1.0386,
      "step": 5100
    },
    {
      "epoch": 45.32,
      "learning_rate": 6.6e-05,
      "loss": 1.0287,
      "step": 5200
    },
    {
      "epoch": 46.19,
      "learning_rate": 6.493617021276595e-05,
      "loss": 1.0251,
      "step": 5300
    },
    {
      "epoch": 47.06,
      "learning_rate": 6.387234042553192e-05,
      "loss": 1.0335,
      "step": 5400
    },
    {
      "epoch": 47.93,
      "learning_rate": 6.280851063829788e-05,
      "loss": 1.0228,
      "step": 5500
    },
    {
      "epoch": 47.93,
      "eval_loss": 0.2622440457344055,
      "eval_runtime": 249.5189,
      "eval_samples_per_second": 13.105,
      "eval_steps_per_second": 1.639,
      "eval_wer": 0.39599299747130906,
      "step": 5500
    },
    {
      "epoch": 48.8,
      "learning_rate": 6.174468085106383e-05,
      "loss": 1.0184,
      "step": 5600
    },
    {
      "epoch": 49.67,
      "learning_rate": 6.068085106382979e-05,
      "loss": 1.0174,
      "step": 5700
    },
    {
      "epoch": 50.54,
      "learning_rate": 5.9617021276595755e-05,
      "loss": 1.0133,
      "step": 5800
    },
    {
      "epoch": 51.42,
      "learning_rate": 5.85531914893617e-05,
      "loss": 1.0108,
      "step": 5900
    },
    {
      "epoch": 52.29,
      "learning_rate": 5.7489361702127666e-05,
      "loss": 0.9936,
      "step": 6000
    },
    {
      "epoch": 52.29,
      "eval_loss": 0.25247836112976074,
      "eval_runtime": 249.2066,
      "eval_samples_per_second": 13.122,
      "eval_steps_per_second": 1.641,
      "eval_wer": 0.37957595798482785,
      "step": 6000
    },
    {
      "epoch": 53.16,
      "learning_rate": 5.642553191489362e-05,
      "loss": 1.0015,
      "step": 6100
    },
    {
      "epoch": 54.03,
      "learning_rate": 5.5361702127659576e-05,
      "loss": 1.0003,
      "step": 6200
    },
    {
      "epoch": 54.9,
      "learning_rate": 5.429787234042554e-05,
      "loss": 0.988,
      "step": 6300
    },
    {
      "epoch": 55.77,
      "learning_rate": 5.3234042553191486e-05,
      "loss": 0.9845,
      "step": 6400
    },
    {
      "epoch": 56.64,
      "learning_rate": 5.217021276595745e-05,
      "loss": 0.968,
      "step": 6500
    },
    {
      "epoch": 56.64,
      "eval_loss": 0.2505591809749603,
      "eval_runtime": 253.8437,
      "eval_samples_per_second": 12.882,
      "eval_steps_per_second": 1.611,
      "eval_wer": 0.3798093756078584,
      "step": 6500
    },
    {
      "epoch": 57.52,
      "learning_rate": 5.110638297872341e-05,
      "loss": 0.9852,
      "step": 6600
    },
    {
      "epoch": 58.39,
      "learning_rate": 5.004255319148936e-05,
      "loss": 0.9737,
      "step": 6700
    },
    {
      "epoch": 59.26,
      "learning_rate": 4.897872340425532e-05,
      "loss": 0.9752,
      "step": 6800
    },
    {
      "epoch": 60.13,
      "learning_rate": 4.792553191489362e-05,
      "loss": 0.9717,
      "step": 6900
    },
    {
      "epoch": 61.0,
      "learning_rate": 4.686170212765958e-05,
      "loss": 0.9704,
      "step": 7000
    },
    {
      "epoch": 61.0,
      "eval_loss": 0.24809595942497253,
      "eval_runtime": 257.0988,
      "eval_samples_per_second": 12.719,
      "eval_steps_per_second": 1.591,
      "eval_wer": 0.37346819684886207,
      "step": 7000
    },
    {
      "epoch": 61.87,
      "learning_rate": 4.579787234042554e-05,
      "loss": 0.9603,
      "step": 7100
    },
    {
      "epoch": 62.75,
      "learning_rate": 4.474468085106383e-05,
      "loss": 0.967,
      "step": 7200
    },
    {
      "epoch": 63.62,
      "learning_rate": 4.368085106382979e-05,
      "loss": 0.9623,
      "step": 7300
    },
    {
      "epoch": 64.49,
      "learning_rate": 4.261702127659575e-05,
      "loss": 0.9506,
      "step": 7400
    },
    {
      "epoch": 65.36,
      "learning_rate": 4.15531914893617e-05,
      "loss": 0.9552,
      "step": 7500
    },
    {
      "epoch": 65.36,
      "eval_loss": 0.23935528099536896,
      "eval_runtime": 264.2571,
      "eval_samples_per_second": 12.374,
      "eval_steps_per_second": 1.548,
      "eval_wer": 0.36432600661349934,
      "step": 7500
    },
    {
      "epoch": 66.23,
      "learning_rate": 4.048936170212766e-05,
      "loss": 0.9449,
      "step": 7600
    },
    {
      "epoch": 67.1,
      "learning_rate": 3.9425531914893624e-05,
      "loss": 0.9411,
      "step": 7700
    },
    {
      "epoch": 67.97,
      "learning_rate": 3.8361702127659576e-05,
      "loss": 0.9468,
      "step": 7800
    },
    {
      "epoch": 68.85,
      "learning_rate": 3.7297872340425534e-05,
      "loss": 0.9448,
      "step": 7900
    },
    {
      "epoch": 69.72,
      "learning_rate": 3.6234042553191486e-05,
      "loss": 0.9417,
      "step": 8000
    },
    {
      "epoch": 69.72,
      "eval_loss": 0.23499037325382233,
      "eval_runtime": 254.8702,
      "eval_samples_per_second": 12.83,
      "eval_steps_per_second": 1.605,
      "eval_wer": 0.35366660182843807,
      "step": 8000
    },
    {
      "epoch": 70.59,
      "learning_rate": 3.517021276595745e-05,
      "loss": 0.9342,
      "step": 8100
    },
    {
      "epoch": 71.46,
      "learning_rate": 3.41063829787234e-05,
      "loss": 0.9338,
      "step": 8200
    },
    {
      "epoch": 72.33,
      "learning_rate": 3.304255319148936e-05,
      "loss": 0.9298,
      "step": 8300
    },
    {
      "epoch": 73.2,
      "learning_rate": 3.1978723404255326e-05,
      "loss": 0.9239,
      "step": 8400
    },
    {
      "epoch": 74.07,
      "learning_rate": 3.091489361702128e-05,
      "loss": 0.9215,
      "step": 8500
    },
    {
      "epoch": 74.07,
      "eval_loss": 0.23256558179855347,
      "eval_runtime": 242.8597,
      "eval_samples_per_second": 13.465,
      "eval_steps_per_second": 1.684,
      "eval_wer": 0.35070997860338454,
      "step": 8500
    },
    {
      "epoch": 74.95,
      "learning_rate": 2.9851063829787236e-05,
      "loss": 0.9233,
      "step": 8600
    },
    {
      "epoch": 75.82,
      "learning_rate": 2.878723404255319e-05,
      "loss": 0.9193,
      "step": 8700
    },
    {
      "epoch": 76.69,
      "learning_rate": 2.7723404255319153e-05,
      "loss": 0.9186,
      "step": 8800
    },
    {
      "epoch": 77.56,
      "learning_rate": 2.6659574468085108e-05,
      "loss": 0.9064,
      "step": 8900
    },
    {
      "epoch": 78.43,
      "learning_rate": 2.5595744680851063e-05,
      "loss": 0.9097,
      "step": 9000
    },
    {
      "epoch": 78.43,
      "eval_loss": 0.227674201130867,
      "eval_runtime": 258.8243,
      "eval_samples_per_second": 12.634,
      "eval_steps_per_second": 1.58,
      "eval_wer": 0.348725928807625,
      "step": 9000
    },
    {
      "epoch": 79.3,
      "learning_rate": 2.453191489361702e-05,
      "loss": 0.9029,
      "step": 9100
    },
    {
      "epoch": 80.17,
      "learning_rate": 2.346808510638298e-05,
      "loss": 0.9137,
      "step": 9200
    },
    {
      "epoch": 81.05,
      "learning_rate": 2.2404255319148938e-05,
      "loss": 0.9045,
      "step": 9300
    },
    {
      "epoch": 81.92,
      "learning_rate": 2.1340425531914897e-05,
      "loss": 0.905,
      "step": 9400
    },
    {
      "epoch": 82.79,
      "learning_rate": 2.027659574468085e-05,
      "loss": 0.9003,
      "step": 9500
    },
    {
      "epoch": 82.79,
      "eval_loss": 0.2229508012533188,
      "eval_runtime": 256.366,
      "eval_samples_per_second": 12.755,
      "eval_steps_per_second": 1.595,
      "eval_wer": 0.33623808597549115,
      "step": 9500
    },
    {
      "epoch": 83.66,
      "learning_rate": 1.921276595744681e-05,
      "loss": 0.8983,
      "step": 9600
    },
    {
      "epoch": 84.53,
      "learning_rate": 1.8148936170212765e-05,
      "loss": 0.9026,
      "step": 9700
    },
    {
      "epoch": 85.4,
      "learning_rate": 1.7095744680851063e-05,
      "loss": 0.8921,
      "step": 9800
    },
    {
      "epoch": 86.27,
      "learning_rate": 1.603191489361702e-05,
      "loss": 0.895,
      "step": 9900
    },
    {
      "epoch": 87.15,
      "learning_rate": 1.4968085106382978e-05,
      "loss": 0.8857,
      "step": 10000
    },
    {
      "epoch": 87.15,
      "eval_loss": 0.2246018499135971,
      "eval_runtime": 255.6354,
      "eval_samples_per_second": 12.792,
      "eval_steps_per_second": 1.6,
      "eval_wer": 0.3361602801011476,
      "step": 10000
    },
    {
      "epoch": 88.02,
      "learning_rate": 1.3904255319148937e-05,
      "loss": 0.8916,
      "step": 10100
    },
    {
      "epoch": 88.89,
      "learning_rate": 1.2840425531914893e-05,
      "loss": 0.8808,
      "step": 10200
    },
    {
      "epoch": 89.76,
      "learning_rate": 1.1776595744680852e-05,
      "loss": 0.8817,
      "step": 10300
    },
    {
      "epoch": 90.63,
      "learning_rate": 1.0712765957446808e-05,
      "loss": 0.8805,
      "step": 10400
    },
    {
      "epoch": 91.5,
      "learning_rate": 9.648936170212767e-06,
      "loss": 0.882,
      "step": 10500
    },
    {
      "epoch": 91.5,
      "eval_loss": 0.2236073762178421,
      "eval_runtime": 260.0762,
      "eval_samples_per_second": 12.573,
      "eval_steps_per_second": 1.573,
      "eval_wer": 0.3315308305777086,
      "step": 10500
    },
    {
      "epoch": 92.37,
      "learning_rate": 8.585106382978724e-06,
      "loss": 0.8768,
      "step": 10600
    },
    {
      "epoch": 93.25,
      "learning_rate": 7.521276595744681e-06,
      "loss": 0.8711,
      "step": 10700
    },
    {
      "epoch": 94.12,
      "learning_rate": 6.457446808510638e-06,
      "loss": 0.8774,
      "step": 10800
    },
    {
      "epoch": 94.99,
      "learning_rate": 5.393617021276596e-06,
      "loss": 0.8781,
      "step": 10900
    },
    {
      "epoch": 95.86,
      "learning_rate": 4.329787234042553e-06,
      "loss": 0.8719,
      "step": 11000
    },
    {
      "epoch": 95.86,
      "eval_loss": 0.22029033303260803,
      "eval_runtime": 256.8466,
      "eval_samples_per_second": 12.731,
      "eval_steps_per_second": 1.592,
      "eval_wer": 0.32705699280295664,
      "step": 11000
    },
    {
      "epoch": 96.73,
      "learning_rate": 3.265957446808511e-06,
      "loss": 0.8687,
      "step": 11100
    },
    {
      "epoch": 97.6,
      "learning_rate": 2.2021276595744685e-06,
      "loss": 0.8771,
      "step": 11200
    },
    {
      "epoch": 98.47,
      "learning_rate": 1.1382978723404256e-06,
      "loss": 0.8633,
      "step": 11300
    },
    {
      "epoch": 99.35,
      "learning_rate": 7.446808510638299e-08,
      "loss": 0.8779,
      "step": 11400
    },
    {
      "epoch": 99.35,
      "step": 11400,
      "total_flos": 1.353536402091675e+20,
      "train_loss": 1.3810767638892458,
      "train_runtime": 73797.3063,
      "train_samples_per_second": 9.949,
      "train_steps_per_second": 0.154
    }
  ],
  "logging_steps": 100,
  "max_steps": 11400,
  "num_train_epochs": 100,
  "save_steps": 500,
  "total_flos": 1.353536402091675e+20,
  "trial_name": null,
  "trial_params": null
}