{
  "best_metric": 75.64947360448383,
  "best_model_checkpoint": "./whisper-small-finetune_gulf/checkpoint-4000",
  "epoch": 2.311604253351826,
  "eval_steps": 1000,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 23.472209930419922,
      "learning_rate": 5.000000000000001e-07,
      "loss": 1.8733,
      "step": 25
    },
    {
      "epoch": 0.02,
      "grad_norm": 25.33260154724121,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.6853,
      "step": 50
    },
    {
      "epoch": 0.03,
      "grad_norm": 28.451927185058594,
      "learning_rate": 1.5e-06,
      "loss": 1.6101,
      "step": 75
    },
    {
      "epoch": 0.05,
      "grad_norm": 29.835905075073242,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.7625,
      "step": 100
    },
    {
      "epoch": 0.06,
      "grad_norm": 28.589879989624023,
      "learning_rate": 2.5e-06,
      "loss": 1.6508,
      "step": 125
    },
    {
      "epoch": 0.07,
      "grad_norm": 28.195240020751953,
      "learning_rate": 3e-06,
      "loss": 1.6032,
      "step": 150
    },
    {
      "epoch": 0.08,
      "grad_norm": 23.07231903076172,
      "learning_rate": 3.5e-06,
      "loss": 1.5002,
      "step": 175
    },
    {
      "epoch": 0.09,
      "grad_norm": 26.471607208251953,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.3828,
      "step": 200
    },
    {
      "epoch": 0.1,
      "grad_norm": 20.687870025634766,
      "learning_rate": 4.5e-06,
      "loss": 1.52,
      "step": 225
    },
    {
      "epoch": 0.12,
      "grad_norm": 20.147708892822266,
      "learning_rate": 5e-06,
      "loss": 1.5193,
      "step": 250
    },
    {
      "epoch": 0.13,
      "grad_norm": 21.233428955078125,
      "learning_rate": 5.500000000000001e-06,
      "loss": 1.5131,
      "step": 275
    },
    {
      "epoch": 0.14,
      "grad_norm": 21.05452537536621,
      "learning_rate": 6e-06,
      "loss": 1.4526,
      "step": 300
    },
    {
      "epoch": 0.15,
      "grad_norm": 24.340858459472656,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 1.5262,
      "step": 325
    },
    {
      "epoch": 0.16,
      "grad_norm": 24.064132690429688,
      "learning_rate": 7e-06,
      "loss": 1.397,
      "step": 350
    },
    {
      "epoch": 0.17,
      "grad_norm": 21.471921920776367,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.5244,
      "step": 375
    },
    {
      "epoch": 0.18,
      "grad_norm": 22.753755569458008,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.3962,
      "step": 400
    },
    {
      "epoch": 0.2,
      "grad_norm": 18.438940048217773,
      "learning_rate": 8.5e-06,
      "loss": 1.3372,
      "step": 425
    },
    {
      "epoch": 0.21,
      "grad_norm": 19.2254581451416,
      "learning_rate": 9e-06,
      "loss": 1.426,
      "step": 450
    },
    {
      "epoch": 0.22,
      "grad_norm": 18.967300415039062,
      "learning_rate": 9.5e-06,
      "loss": 1.3301,
      "step": 475
    },
    {
      "epoch": 0.23,
      "grad_norm": 20.847196578979492,
      "learning_rate": 1e-05,
      "loss": 1.2433,
      "step": 500
    },
    {
      "epoch": 0.24,
      "grad_norm": 18.459182739257812,
      "learning_rate": 9.944444444444445e-06,
      "loss": 1.2761,
      "step": 525
    },
    {
      "epoch": 0.25,
      "grad_norm": 22.492895126342773,
      "learning_rate": 9.88888888888889e-06,
      "loss": 1.4599,
      "step": 550
    },
    {
      "epoch": 0.27,
      "grad_norm": 25.510766983032227,
      "learning_rate": 9.833333333333333e-06,
      "loss": 1.2511,
      "step": 575
    },
    {
      "epoch": 0.28,
      "grad_norm": 20.4246768951416,
      "learning_rate": 9.777777777777779e-06,
      "loss": 1.2121,
      "step": 600
    },
    {
      "epoch": 0.29,
      "grad_norm": 25.91058349609375,
      "learning_rate": 9.722222222222223e-06,
      "loss": 1.4637,
      "step": 625
    },
    {
      "epoch": 0.3,
      "grad_norm": 25.087705612182617,
      "learning_rate": 9.666666666666667e-06,
      "loss": 1.2607,
      "step": 650
    },
    {
      "epoch": 0.31,
      "grad_norm": 20.08827781677246,
      "learning_rate": 9.611111111111112e-06,
      "loss": 1.4525,
      "step": 675
    },
    {
      "epoch": 0.32,
      "grad_norm": 16.310239791870117,
      "learning_rate": 9.555555555555556e-06,
      "loss": 1.2905,
      "step": 700
    },
    {
      "epoch": 0.34,
      "grad_norm": 18.671165466308594,
      "learning_rate": 9.5e-06,
      "loss": 1.3568,
      "step": 725
    },
    {
      "epoch": 0.35,
      "grad_norm": 25.499460220336914,
      "learning_rate": 9.444444444444445e-06,
      "loss": 1.2893,
      "step": 750
    },
    {
      "epoch": 0.36,
      "grad_norm": 20.31481170654297,
      "learning_rate": 9.38888888888889e-06,
      "loss": 1.268,
      "step": 775
    },
    {
      "epoch": 0.37,
      "grad_norm": 22.260953903198242,
      "learning_rate": 9.333333333333334e-06,
      "loss": 1.2924,
      "step": 800
    },
    {
      "epoch": 0.38,
      "grad_norm": 19.559682846069336,
      "learning_rate": 9.277777777777778e-06,
      "loss": 1.3534,
      "step": 825
    },
    {
      "epoch": 0.39,
      "grad_norm": 21.260534286499023,
      "learning_rate": 9.222222222222224e-06,
      "loss": 1.265,
      "step": 850
    },
    {
      "epoch": 0.4,
      "grad_norm": 22.76539421081543,
      "learning_rate": 9.166666666666666e-06,
      "loss": 1.302,
      "step": 875
    },
    {
      "epoch": 0.42,
      "grad_norm": 23.064783096313477,
      "learning_rate": 9.111111111111112e-06,
      "loss": 1.25,
      "step": 900
    },
    {
      "epoch": 0.43,
      "grad_norm": 19.941423416137695,
      "learning_rate": 9.055555555555556e-06,
      "loss": 1.2915,
      "step": 925
    },
    {
      "epoch": 0.44,
      "grad_norm": 17.240646362304688,
      "learning_rate": 9e-06,
      "loss": 1.2868,
      "step": 950
    },
    {
      "epoch": 0.45,
      "grad_norm": 18.2373046875,
      "learning_rate": 8.944444444444446e-06,
      "loss": 1.2259,
      "step": 975
    },
    {
      "epoch": 0.46,
      "grad_norm": 22.34012794494629,
      "learning_rate": 8.888888888888888e-06,
      "loss": 1.3947,
      "step": 1000
    },
    {
      "epoch": 0.46,
      "eval_cer": 65.38537817965462,
      "eval_loss": 1.2471961975097656,
      "eval_runtime": 1151.935,
      "eval_samples_per_second": 3.755,
      "eval_steps_per_second": 0.47,
      "eval_wer": 84.87843671892752,
      "step": 1000
    },
    {
      "epoch": 0.47,
      "grad_norm": 18.88918685913086,
      "learning_rate": 8.833333333333334e-06,
      "loss": 1.2824,
      "step": 1025
    },
    {
      "epoch": 0.49,
      "grad_norm": 20.2314395904541,
      "learning_rate": 8.777777777777778e-06,
      "loss": 1.3135,
      "step": 1050
    },
    {
      "epoch": 0.5,
      "grad_norm": 21.175655364990234,
      "learning_rate": 8.722222222222224e-06,
      "loss": 1.2259,
      "step": 1075
    },
    {
      "epoch": 0.51,
      "grad_norm": 16.999879837036133,
      "learning_rate": 8.666666666666668e-06,
      "loss": 1.2408,
      "step": 1100
    },
    {
      "epoch": 0.52,
      "grad_norm": 19.178285598754883,
      "learning_rate": 8.611111111111112e-06,
      "loss": 1.2721,
      "step": 1125
    },
    {
      "epoch": 0.53,
      "grad_norm": 20.64499855041504,
      "learning_rate": 8.555555555555556e-06,
      "loss": 1.3392,
      "step": 1150
    },
    {
      "epoch": 0.54,
      "grad_norm": 19.396726608276367,
      "learning_rate": 8.5e-06,
      "loss": 1.2863,
      "step": 1175
    },
    {
      "epoch": 0.55,
      "grad_norm": 19.87407875061035,
      "learning_rate": 8.444444444444446e-06,
      "loss": 1.3248,
      "step": 1200
    },
    {
      "epoch": 0.57,
      "grad_norm": 16.245336532592773,
      "learning_rate": 8.38888888888889e-06,
      "loss": 1.3406,
      "step": 1225
    },
    {
      "epoch": 0.58,
      "grad_norm": 17.97079086303711,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.211,
      "step": 1250
    },
    {
      "epoch": 0.59,
      "grad_norm": 25.904541015625,
      "learning_rate": 8.277777777777778e-06,
      "loss": 1.334,
      "step": 1275
    },
    {
      "epoch": 0.6,
      "grad_norm": 21.328983306884766,
      "learning_rate": 8.222222222222222e-06,
      "loss": 1.2982,
      "step": 1300
    },
    {
      "epoch": 0.61,
      "grad_norm": 19.949731826782227,
      "learning_rate": 8.166666666666668e-06,
      "loss": 1.2365,
      "step": 1325
    },
    {
      "epoch": 0.62,
      "grad_norm": 15.301751136779785,
      "learning_rate": 8.111111111111112e-06,
      "loss": 1.2033,
      "step": 1350
    },
    {
      "epoch": 0.64,
      "grad_norm": 22.96265411376953,
      "learning_rate": 8.055555555555557e-06,
      "loss": 1.2319,
      "step": 1375
    },
    {
      "epoch": 0.65,
      "grad_norm": 17.26740074157715,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.1544,
      "step": 1400
    },
    {
      "epoch": 0.66,
      "grad_norm": 27.209585189819336,
      "learning_rate": 7.944444444444445e-06,
      "loss": 1.3496,
      "step": 1425
    },
    {
      "epoch": 0.67,
      "grad_norm": 20.707279205322266,
      "learning_rate": 7.88888888888889e-06,
      "loss": 1.2727,
      "step": 1450
    },
    {
      "epoch": 0.68,
      "grad_norm": 20.74648666381836,
      "learning_rate": 7.833333333333333e-06,
      "loss": 1.2371,
      "step": 1475
    },
    {
      "epoch": 0.69,
      "grad_norm": 20.16001319885254,
      "learning_rate": 7.77777777777778e-06,
      "loss": 1.2456,
      "step": 1500
    },
    {
      "epoch": 0.71,
      "grad_norm": 20.684553146362305,
      "learning_rate": 7.722222222222223e-06,
      "loss": 1.1862,
      "step": 1525
    },
    {
      "epoch": 0.72,
      "grad_norm": 17.660783767700195,
      "learning_rate": 7.666666666666667e-06,
      "loss": 1.1802,
      "step": 1550
    },
    {
      "epoch": 0.73,
      "grad_norm": 27.349889755249023,
      "learning_rate": 7.611111111111111e-06,
      "loss": 1.1869,
      "step": 1575
    },
    {
      "epoch": 0.74,
      "grad_norm": 20.43325424194336,
      "learning_rate": 7.555555555555556e-06,
      "loss": 1.2752,
      "step": 1600
    },
    {
      "epoch": 0.75,
      "grad_norm": 18.21103286743164,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.083,
      "step": 1625
    },
    {
      "epoch": 0.76,
      "grad_norm": 17.000732421875,
      "learning_rate": 7.444444444444445e-06,
      "loss": 1.1785,
      "step": 1650
    },
    {
      "epoch": 0.77,
      "grad_norm": 16.118410110473633,
      "learning_rate": 7.38888888888889e-06,
      "loss": 1.2115,
      "step": 1675
    },
    {
      "epoch": 0.79,
      "grad_norm": 16.376646041870117,
      "learning_rate": 7.333333333333333e-06,
      "loss": 1.1399,
      "step": 1700
    },
    {
      "epoch": 0.8,
      "grad_norm": 23.872587203979492,
      "learning_rate": 7.277777777777778e-06,
      "loss": 1.2943,
      "step": 1725
    },
    {
      "epoch": 0.81,
      "grad_norm": 17.694063186645508,
      "learning_rate": 7.222222222222223e-06,
      "loss": 1.1768,
      "step": 1750
    },
    {
      "epoch": 0.82,
      "grad_norm": 15.69645881652832,
      "learning_rate": 7.166666666666667e-06,
      "loss": 1.1779,
      "step": 1775
    },
    {
      "epoch": 0.83,
      "grad_norm": 17.946985244750977,
      "learning_rate": 7.111111111111112e-06,
      "loss": 1.1575,
      "step": 1800
    },
    {
      "epoch": 0.84,
      "grad_norm": 24.68529510498047,
      "learning_rate": 7.055555555555557e-06,
      "loss": 1.2993,
      "step": 1825
    },
    {
      "epoch": 0.86,
      "grad_norm": 20.466711044311523,
      "learning_rate": 7e-06,
      "loss": 1.1476,
      "step": 1850
    },
    {
      "epoch": 0.87,
      "grad_norm": 20.040111541748047,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.1855,
      "step": 1875
    },
    {
      "epoch": 0.88,
      "grad_norm": 20.88412857055664,
      "learning_rate": 6.88888888888889e-06,
      "loss": 1.2041,
      "step": 1900
    },
    {
      "epoch": 0.89,
      "grad_norm": 16.89543342590332,
      "learning_rate": 6.833333333333334e-06,
      "loss": 1.2062,
      "step": 1925
    },
    {
      "epoch": 0.9,
      "grad_norm": 20.681123733520508,
      "learning_rate": 6.777777777777779e-06,
      "loss": 1.0671,
      "step": 1950
    },
    {
      "epoch": 0.91,
      "grad_norm": 20.805986404418945,
      "learning_rate": 6.7222222222222235e-06,
      "loss": 1.197,
      "step": 1975
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.39628791809082,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.333,
      "step": 2000
    },
    {
      "epoch": 0.92,
      "eval_cer": 72.85421517713513,
      "eval_loss": 1.1428595781326294,
      "eval_runtime": 1198.4252,
      "eval_samples_per_second": 3.61,
      "eval_steps_per_second": 0.451,
      "eval_wer": 92.31235325304856,
      "step": 2000
    },
    {
      "epoch": 0.94,
      "grad_norm": 18.98177719116211,
      "learning_rate": 6.6111111111111115e-06,
      "loss": 1.1586,
      "step": 2025
    },
    {
      "epoch": 0.95,
      "grad_norm": 26.32819366455078,
      "learning_rate": 6.555555555555556e-06,
      "loss": 1.2463,
      "step": 2050
    },
    {
      "epoch": 0.96,
      "grad_norm": 17.621980667114258,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 1.1764,
      "step": 2075
    },
    {
      "epoch": 0.97,
      "grad_norm": 19.834068298339844,
      "learning_rate": 6.444444444444445e-06,
      "loss": 1.2383,
      "step": 2100
    },
    {
      "epoch": 0.98,
      "grad_norm": 18.443336486816406,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 1.1966,
      "step": 2125
    },
    {
      "epoch": 0.99,
      "grad_norm": 23.796865463256836,
      "learning_rate": 6.333333333333333e-06,
      "loss": 1.1506,
      "step": 2150
    },
    {
      "epoch": 1.01,
      "grad_norm": 13.802796363830566,
      "learning_rate": 6.277777777777778e-06,
      "loss": 1.004,
      "step": 2175
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.86549186706543,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.8599,
      "step": 2200
    },
    {
      "epoch": 1.03,
      "grad_norm": 20.885765075683594,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.9144,
      "step": 2225
    },
    {
      "epoch": 1.04,
      "grad_norm": 15.317635536193848,
      "learning_rate": 6.111111111111112e-06,
      "loss": 0.863,
      "step": 2250
    },
    {
      "epoch": 1.05,
      "grad_norm": 16.995641708374023,
      "learning_rate": 6.055555555555555e-06,
      "loss": 0.8261,
      "step": 2275
    },
    {
      "epoch": 1.06,
      "grad_norm": 16.974023818969727,
      "learning_rate": 6e-06,
      "loss": 0.7936,
      "step": 2300
    },
    {
      "epoch": 1.07,
      "grad_norm": 15.467360496520996,
      "learning_rate": 5.944444444444445e-06,
      "loss": 0.8796,
      "step": 2325
    },
    {
      "epoch": 1.09,
      "grad_norm": 14.19994068145752,
      "learning_rate": 5.88888888888889e-06,
      "loss": 0.9499,
      "step": 2350
    },
    {
      "epoch": 1.1,
      "grad_norm": 18.083833694458008,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.8659,
      "step": 2375
    },
    {
      "epoch": 1.11,
      "grad_norm": 13.523449897766113,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.7468,
      "step": 2400
    },
    {
      "epoch": 1.12,
      "grad_norm": 22.28549575805664,
      "learning_rate": 5.722222222222222e-06,
      "loss": 0.9936,
      "step": 2425
    },
    {
      "epoch": 1.13,
      "grad_norm": 12.301285743713379,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.9272,
      "step": 2450
    },
    {
      "epoch": 1.14,
      "grad_norm": 15.811965942382812,
      "learning_rate": 5.611111111111112e-06,
      "loss": 0.8392,
      "step": 2475
    },
    {
      "epoch": 1.16,
      "grad_norm": 10.93188762664795,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.7856,
      "step": 2500
    },
    {
      "epoch": 1.17,
      "grad_norm": 15.368990898132324,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.991,
      "step": 2525
    },
    {
      "epoch": 1.18,
      "grad_norm": 23.429412841796875,
      "learning_rate": 5.444444444444445e-06,
      "loss": 0.956,
      "step": 2550
    },
    {
      "epoch": 1.19,
      "grad_norm": 15.563962936401367,
      "learning_rate": 5.388888888888889e-06,
      "loss": 0.9221,
      "step": 2575
    },
    {
      "epoch": 1.2,
      "grad_norm": 15.646391868591309,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.8138,
      "step": 2600
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.61239242553711,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 0.9199,
      "step": 2625
    },
    {
      "epoch": 1.23,
      "grad_norm": 16.09264373779297,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 0.9146,
      "step": 2650
    },
    {
      "epoch": 1.24,
      "grad_norm": 18.509414672851562,
      "learning_rate": 5.1666666666666675e-06,
      "loss": 0.8323,
      "step": 2675
    },
    {
      "epoch": 1.25,
      "grad_norm": 12.82259750366211,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.8451,
      "step": 2700
    },
    {
      "epoch": 1.26,
      "grad_norm": 13.008484840393066,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 0.8415,
      "step": 2725
    },
    {
      "epoch": 1.27,
      "grad_norm": 15.046307563781738,
      "learning_rate": 5e-06,
      "loss": 0.8555,
      "step": 2750
    },
    {
      "epoch": 1.28,
      "grad_norm": 16.91812515258789,
      "learning_rate": 4.944444444444445e-06,
      "loss": 1.015,
      "step": 2775
    },
    {
      "epoch": 1.29,
      "grad_norm": 16.76761245727539,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.8488,
      "step": 2800
    },
    {
      "epoch": 1.31,
      "grad_norm": 20.7960262298584,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.9098,
      "step": 2825
    },
    {
      "epoch": 1.32,
      "grad_norm": 15.695616722106934,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.8866,
      "step": 2850
    },
    {
      "epoch": 1.33,
      "grad_norm": 18.41292953491211,
      "learning_rate": 4.722222222222222e-06,
      "loss": 0.8872,
      "step": 2875
    },
    {
      "epoch": 1.34,
      "grad_norm": 22.192108154296875,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.8102,
      "step": 2900
    },
    {
      "epoch": 1.35,
      "grad_norm": 21.19410514831543,
      "learning_rate": 4.611111111111112e-06,
      "loss": 0.9069,
      "step": 2925
    },
    {
      "epoch": 1.36,
      "grad_norm": 18.010290145874023,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.8646,
      "step": 2950
    },
    {
      "epoch": 1.38,
      "grad_norm": 18.863052368164062,
      "learning_rate": 4.5e-06,
      "loss": 0.9674,
      "step": 2975
    },
    {
      "epoch": 1.39,
      "grad_norm": 13.40828800201416,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.918,
      "step": 3000
    },
    {
      "epoch": 1.39,
      "eval_cer": 64.30927214097356,
      "eval_loss": 1.1143155097961426,
      "eval_runtime": 1160.055,
      "eval_samples_per_second": 3.729,
      "eval_steps_per_second": 0.466,
      "eval_wer": 82.85238203438612,
      "step": 3000
    },
    {
      "epoch": 1.4,
      "grad_norm": 16.100902557373047,
      "learning_rate": 4.388888888888889e-06,
      "loss": 0.8247,
      "step": 3025
    },
    {
      "epoch": 1.41,
      "grad_norm": 14.138301849365234,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.8077,
      "step": 3050
    },
    {
      "epoch": 1.42,
      "grad_norm": 21.747058868408203,
      "learning_rate": 4.277777777777778e-06,
      "loss": 0.9274,
      "step": 3075
    },
    {
      "epoch": 1.43,
      "grad_norm": 12.049752235412598,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.9581,
      "step": 3100
    },
    {
      "epoch": 1.44,
      "grad_norm": 13.890820503234863,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.8945,
      "step": 3125
    },
    {
      "epoch": 1.46,
      "grad_norm": 16.04409408569336,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.8455,
      "step": 3150
    },
    {
      "epoch": 1.47,
      "grad_norm": 14.720856666564941,
      "learning_rate": 4.055555555555556e-06,
      "loss": 0.8707,
      "step": 3175
    },
    {
      "epoch": 1.48,
      "grad_norm": 23.8120174407959,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.9138,
      "step": 3200
    },
    {
      "epoch": 1.49,
      "grad_norm": 19.527795791625977,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.8293,
      "step": 3225
    },
    {
      "epoch": 1.5,
      "grad_norm": 16.076868057250977,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.8586,
      "step": 3250
    },
    {
      "epoch": 1.51,
      "grad_norm": 18.909570693969727,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.9239,
      "step": 3275
    },
    {
      "epoch": 1.53,
      "grad_norm": 20.29988670349121,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.8682,
      "step": 3300
    },
    {
      "epoch": 1.54,
      "grad_norm": 15.01752758026123,
      "learning_rate": 3.7222222222222225e-06,
      "loss": 0.8631,
      "step": 3325
    },
    {
      "epoch": 1.55,
      "grad_norm": 14.974254608154297,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.8882,
      "step": 3350
    },
    {
      "epoch": 1.56,
      "grad_norm": 15.673378944396973,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 0.958,
      "step": 3375
    },
    {
      "epoch": 1.57,
      "grad_norm": 20.833036422729492,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.8278,
      "step": 3400
    },
    {
      "epoch": 1.58,
      "grad_norm": 15.56125545501709,
      "learning_rate": 3.5e-06,
      "loss": 0.8853,
      "step": 3425
    },
    {
      "epoch": 1.6,
      "grad_norm": 13.465311050415039,
      "learning_rate": 3.444444444444445e-06,
      "loss": 0.8877,
      "step": 3450
    },
    {
      "epoch": 1.61,
      "grad_norm": 19.25458335876465,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.8935,
      "step": 3475
    },
    {
      "epoch": 1.62,
      "grad_norm": 17.35223960876465,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.8421,
      "step": 3500
    },
    {
      "epoch": 1.63,
      "grad_norm": 17.846412658691406,
      "learning_rate": 3.277777777777778e-06,
      "loss": 0.791,
      "step": 3525
    },
    {
      "epoch": 1.64,
      "grad_norm": 12.937151908874512,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 0.811,
      "step": 3550
    },
    {
      "epoch": 1.65,
      "grad_norm": 19.468202590942383,
      "learning_rate": 3.1666666666666667e-06,
      "loss": 0.8527,
      "step": 3575
    },
    {
      "epoch": 1.66,
      "grad_norm": 14.417373657226562,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.8507,
      "step": 3600
    },
    {
      "epoch": 1.68,
      "grad_norm": 19.41051483154297,
      "learning_rate": 3.055555555555556e-06,
      "loss": 0.8945,
      "step": 3625
    },
    {
      "epoch": 1.69,
      "grad_norm": 17.270503997802734,
      "learning_rate": 3e-06,
      "loss": 0.8866,
      "step": 3650
    },
    {
      "epoch": 1.7,
      "grad_norm": 15.31128215789795,
      "learning_rate": 2.944444444444445e-06,
      "loss": 0.8862,
      "step": 3675
    },
    {
      "epoch": 1.71,
      "grad_norm": 16.625324249267578,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.8744,
      "step": 3700
    },
    {
      "epoch": 1.72,
      "grad_norm": 23.434606552124023,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.927,
      "step": 3725
    },
    {
      "epoch": 1.73,
      "grad_norm": 14.484862327575684,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 0.8662,
      "step": 3750
    },
    {
      "epoch": 1.75,
      "grad_norm": 15.611286163330078,
      "learning_rate": 2.7222222222222224e-06,
      "loss": 0.8296,
      "step": 3775
    },
    {
      "epoch": 1.76,
      "grad_norm": 14.025354385375977,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.8567,
      "step": 3800
    },
    {
      "epoch": 1.77,
      "grad_norm": 17.058929443359375,
      "learning_rate": 2.6111111111111113e-06,
      "loss": 0.8872,
      "step": 3825
    },
    {
      "epoch": 1.78,
      "grad_norm": 18.206918716430664,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 0.9316,
      "step": 3850
    },
    {
      "epoch": 1.79,
      "grad_norm": 14.466046333312988,
      "learning_rate": 2.5e-06,
      "loss": 0.8486,
      "step": 3875
    },
    {
      "epoch": 1.8,
      "grad_norm": 16.97430992126465,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.8618,
      "step": 3900
    },
    {
      "epoch": 1.81,
      "grad_norm": 11.907506942749023,
      "learning_rate": 2.388888888888889e-06,
      "loss": 0.7407,
      "step": 3925
    },
    {
      "epoch": 1.83,
      "grad_norm": 17.779897689819336,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.762,
      "step": 3950
    },
    {
      "epoch": 1.84,
      "grad_norm": 14.020506858825684,
      "learning_rate": 2.277777777777778e-06,
      "loss": 0.8425,
      "step": 3975
    },
    {
      "epoch": 1.85,
      "grad_norm": 15.239115715026855,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.8402,
      "step": 4000
    },
    {
      "epoch": 1.85,
      "eval_cer": 53.700996775354604,
      "eval_loss": 1.0890547037124634,
      "eval_runtime": 1101.5884,
      "eval_samples_per_second": 3.927,
      "eval_steps_per_second": 0.491,
      "eval_wer": 75.64947360448383,
      "step": 4000
    },
    {
      "epoch": 1.86,
      "grad_norm": 16.369224548339844,
      "learning_rate": 2.166666666666667e-06,
      "loss": 0.9486,
      "step": 4025
    },
    {
      "epoch": 1.87,
      "grad_norm": 13.583636283874512,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 0.8256,
      "step": 4050
    },
    {
      "epoch": 1.88,
      "grad_norm": 11.406463623046875,
      "learning_rate": 2.0555555555555555e-06,
      "loss": 0.7751,
      "step": 4075
    },
    {
      "epoch": 1.9,
      "grad_norm": 17.16242790222168,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.777,
      "step": 4100
    },
    {
      "epoch": 1.91,
      "grad_norm": 10.725821495056152,
      "learning_rate": 1.944444444444445e-06,
      "loss": 0.7842,
      "step": 4125
    },
    {
      "epoch": 1.92,
      "grad_norm": 18.373470306396484,
      "learning_rate": 1.888888888888889e-06,
      "loss": 0.7934,
      "step": 4150
    },
    {
      "epoch": 1.93,
      "grad_norm": 19.23195457458496,
      "learning_rate": 1.8333333333333333e-06,
      "loss": 0.9092,
      "step": 4175
    },
    {
      "epoch": 1.94,
      "grad_norm": 16.823266983032227,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.8693,
      "step": 4200
    },
    {
      "epoch": 1.95,
      "grad_norm": 17.290674209594727,
      "learning_rate": 1.7222222222222224e-06,
      "loss": 0.8973,
      "step": 4225
    },
    {
      "epoch": 1.96,
      "grad_norm": 17.71505355834961,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.8385,
      "step": 4250
    },
    {
      "epoch": 1.98,
      "grad_norm": 16.774585723876953,
      "learning_rate": 1.6111111111111113e-06,
      "loss": 0.7899,
      "step": 4275
    },
    {
      "epoch": 1.99,
      "grad_norm": 19.144384384155273,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.9822,
      "step": 4300
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.863649368286133,
      "learning_rate": 1.5e-06,
      "loss": 0.7994,
      "step": 4325
    },
    {
      "epoch": 2.01,
      "grad_norm": 13.130481719970703,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 0.5804,
      "step": 4350
    },
    {
      "epoch": 2.02,
      "grad_norm": 19.954849243164062,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 0.7005,
      "step": 4375
    },
    {
      "epoch": 2.03,
      "grad_norm": 16.30689811706543,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.6735,
      "step": 4400
    },
    {
      "epoch": 2.05,
      "grad_norm": 12.148699760437012,
      "learning_rate": 1.2777777777777779e-06,
      "loss": 0.6317,
      "step": 4425
    },
    {
      "epoch": 2.06,
      "grad_norm": 9.524985313415527,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 0.5642,
      "step": 4450
    },
    {
      "epoch": 2.07,
      "grad_norm": 17.790037155151367,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 0.5801,
      "step": 4475
    },
    {
      "epoch": 2.08,
      "grad_norm": 11.228938102722168,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.7253,
      "step": 4500
    },
    {
      "epoch": 2.09,
      "grad_norm": 15.2631254196167,
      "learning_rate": 1.0555555555555557e-06,
      "loss": 0.6465,
      "step": 4525
    },
    {
      "epoch": 2.1,
      "grad_norm": 20.890735626220703,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.6849,
      "step": 4550
    },
    {
      "epoch": 2.12,
      "grad_norm": 15.065770149230957,
      "learning_rate": 9.444444444444445e-07,
      "loss": 0.5892,
      "step": 4575
    },
    {
      "epoch": 2.13,
      "grad_norm": 10.663036346435547,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.5681,
      "step": 4600
    },
    {
      "epoch": 2.14,
      "grad_norm": 17.5535831451416,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.6496,
      "step": 4625
    },
    {
      "epoch": 2.15,
      "grad_norm": 14.98912239074707,
      "learning_rate": 7.777777777777779e-07,
      "loss": 0.6773,
      "step": 4650
    },
    {
      "epoch": 2.16,
      "grad_norm": 14.315783500671387,
      "learning_rate": 7.222222222222222e-07,
      "loss": 0.6852,
      "step": 4675
    },
    {
      "epoch": 2.17,
      "grad_norm": 17.369815826416016,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.6886,
      "step": 4700
    },
    {
      "epoch": 2.18,
      "grad_norm": 9.153715133666992,
      "learning_rate": 6.111111111111112e-07,
      "loss": 0.6393,
      "step": 4725
    },
    {
      "epoch": 2.2,
      "grad_norm": 7.958740234375,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.6516,
      "step": 4750
    },
    {
      "epoch": 2.21,
      "grad_norm": 12.058722496032715,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.6594,
      "step": 4775
    },
    {
      "epoch": 2.22,
      "grad_norm": 10.650951385498047,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.6383,
      "step": 4800
    },
    {
      "epoch": 2.23,
      "grad_norm": 9.111865043640137,
      "learning_rate": 3.8888888888888895e-07,
      "loss": 0.5531,
      "step": 4825
    },
    {
      "epoch": 2.24,
      "grad_norm": 15.841039657592773,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.6738,
      "step": 4850
    },
    {
      "epoch": 2.25,
      "grad_norm": 16.60221290588379,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.643,
      "step": 4875
    },
    {
      "epoch": 2.27,
      "grad_norm": 17.007307052612305,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.6843,
      "step": 4900
    },
    {
      "epoch": 2.28,
      "grad_norm": 13.0649995803833,
      "learning_rate": 1.6666666666666668e-07,
      "loss": 0.6416,
      "step": 4925
    },
    {
      "epoch": 2.29,
      "grad_norm": 15.913739204406738,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 0.6124,
      "step": 4950
    },
    {
      "epoch": 2.3,
      "grad_norm": 17.32563018798828,
      "learning_rate": 5.555555555555556e-08,
      "loss": 0.6627,
      "step": 4975
    },
    {
      "epoch": 2.31,
      "grad_norm": 16.87995147705078,
      "learning_rate": 0.0,
      "loss": 0.6894,
      "step": 5000
    },
    {
      "epoch": 2.31,
      "eval_cer": 57.09752532338271,
      "eval_loss": 1.100616455078125,
      "eval_runtime": 1127.1011,
      "eval_samples_per_second": 3.838,
      "eval_steps_per_second": 0.48,
      "eval_wer": 79.2130576384155,
      "step": 5000
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "total_flos": 1.154283884199936e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}