{
  "best_metric": 71.37643973087012,
  "best_model_checkpoint": "./whisper-small-finetune_levantine/checkpoint-4000",
  "epoch": 3.5087719298245617,
  "eval_steps": 1000,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "grad_norm": 24.49530792236328,
      "learning_rate": 5.000000000000001e-07,
      "loss": 1.6584,
      "step": 25
    },
    {
      "epoch": 0.04,
      "grad_norm": 26.777080535888672,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.5114,
      "step": 50
    },
    {
      "epoch": 0.05,
      "grad_norm": 20.427642822265625,
      "learning_rate": 1.5e-06,
      "loss": 1.4767,
      "step": 75
    },
    {
      "epoch": 0.07,
      "grad_norm": 21.08566665649414,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.411,
      "step": 100
    },
    {
      "epoch": 0.09,
      "grad_norm": 19.165359497070312,
      "learning_rate": 2.5e-06,
      "loss": 1.415,
      "step": 125
    },
    {
      "epoch": 0.11,
      "grad_norm": 21.769067764282227,
      "learning_rate": 3e-06,
      "loss": 1.2559,
      "step": 150
    },
    {
      "epoch": 0.12,
      "grad_norm": 24.020814895629883,
      "learning_rate": 3.5e-06,
      "loss": 1.3334,
      "step": 175
    },
    {
      "epoch": 0.14,
      "grad_norm": 18.758386611938477,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.1788,
      "step": 200
    },
    {
      "epoch": 0.16,
      "grad_norm": 24.74569320678711,
      "learning_rate": 4.5e-06,
      "loss": 1.3336,
      "step": 225
    },
    {
      "epoch": 0.18,
      "grad_norm": 20.71479034423828,
      "learning_rate": 5e-06,
      "loss": 1.266,
      "step": 250
    },
    {
      "epoch": 0.19,
      "grad_norm": 22.093847274780273,
      "learning_rate": 5.500000000000001e-06,
      "loss": 1.2522,
      "step": 275
    },
    {
      "epoch": 0.21,
      "grad_norm": 20.75282859802246,
      "learning_rate": 6e-06,
      "loss": 1.364,
      "step": 300
    },
    {
      "epoch": 0.23,
      "grad_norm": 19.298858642578125,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 1.1445,
      "step": 325
    },
    {
      "epoch": 0.25,
      "grad_norm": 19.872005462646484,
      "learning_rate": 7e-06,
      "loss": 1.2237,
      "step": 350
    },
    {
      "epoch": 0.26,
      "grad_norm": 24.68609619140625,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.1531,
      "step": 375
    },
    {
      "epoch": 0.28,
      "grad_norm": 19.604598999023438,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.265,
      "step": 400
    },
    {
      "epoch": 0.3,
      "grad_norm": 20.639904022216797,
      "learning_rate": 8.5e-06,
      "loss": 1.2187,
      "step": 425
    },
    {
      "epoch": 0.32,
      "grad_norm": 20.088167190551758,
      "learning_rate": 9e-06,
      "loss": 1.2555,
      "step": 450
    },
    {
      "epoch": 0.33,
      "grad_norm": 18.40792465209961,
      "learning_rate": 9.5e-06,
      "loss": 1.2229,
      "step": 475
    },
    {
      "epoch": 0.35,
      "grad_norm": 18.657146453857422,
      "learning_rate": 1e-05,
      "loss": 1.2368,
      "step": 500
    },
    {
      "epoch": 0.37,
      "grad_norm": 18.283592224121094,
      "learning_rate": 9.944444444444445e-06,
      "loss": 1.2768,
      "step": 525
    },
    {
      "epoch": 0.39,
      "grad_norm": 20.509708404541016,
      "learning_rate": 9.88888888888889e-06,
      "loss": 1.2746,
      "step": 550
    },
    {
      "epoch": 0.4,
      "grad_norm": 29.214723587036133,
      "learning_rate": 9.833333333333333e-06,
      "loss": 1.2714,
      "step": 575
    },
    {
      "epoch": 0.42,
      "grad_norm": 22.353967666625977,
      "learning_rate": 9.777777777777779e-06,
      "loss": 1.2722,
      "step": 600
    },
    {
      "epoch": 0.44,
      "grad_norm": 18.456220626831055,
      "learning_rate": 9.722222222222223e-06,
      "loss": 1.1404,
      "step": 625
    },
    {
      "epoch": 0.46,
      "grad_norm": 13.63076114654541,
      "learning_rate": 9.666666666666667e-06,
      "loss": 1.1851,
      "step": 650
    },
    {
      "epoch": 0.47,
      "grad_norm": 18.397729873657227,
      "learning_rate": 9.611111111111112e-06,
      "loss": 1.2517,
      "step": 675
    },
    {
      "epoch": 0.49,
      "grad_norm": 21.700220108032227,
      "learning_rate": 9.555555555555556e-06,
      "loss": 1.1377,
      "step": 700
    },
    {
      "epoch": 0.51,
      "grad_norm": 21.407974243164062,
      "learning_rate": 9.5e-06,
      "loss": 1.167,
      "step": 725
    },
    {
      "epoch": 0.53,
      "grad_norm": 23.072011947631836,
      "learning_rate": 9.444444444444445e-06,
      "loss": 1.1425,
      "step": 750
    },
    {
      "epoch": 0.54,
      "grad_norm": 18.679073333740234,
      "learning_rate": 9.38888888888889e-06,
      "loss": 1.3557,
      "step": 775
    },
    {
      "epoch": 0.56,
      "grad_norm": 17.260936737060547,
      "learning_rate": 9.333333333333334e-06,
      "loss": 1.1378,
      "step": 800
    },
    {
      "epoch": 0.58,
      "grad_norm": 21.285743713378906,
      "learning_rate": 9.277777777777778e-06,
      "loss": 1.1549,
      "step": 825
    },
    {
      "epoch": 0.6,
      "grad_norm": 15.386796951293945,
      "learning_rate": 9.222222222222224e-06,
      "loss": 1.1067,
      "step": 850
    },
    {
      "epoch": 0.61,
      "grad_norm": 16.645769119262695,
      "learning_rate": 9.166666666666666e-06,
      "loss": 1.0842,
      "step": 875
    },
    {
      "epoch": 0.63,
      "grad_norm": 20.210805892944336,
      "learning_rate": 9.111111111111112e-06,
      "loss": 1.2224,
      "step": 900
    },
    {
      "epoch": 0.65,
      "grad_norm": 16.632762908935547,
      "learning_rate": 9.055555555555556e-06,
      "loss": 1.1574,
      "step": 925
    },
    {
      "epoch": 0.67,
      "grad_norm": 18.083099365234375,
      "learning_rate": 9e-06,
      "loss": 1.0843,
      "step": 950
    },
    {
      "epoch": 0.68,
      "grad_norm": 17.896181106567383,
      "learning_rate": 8.944444444444446e-06,
      "loss": 1.1495,
      "step": 975
    },
    {
      "epoch": 0.7,
      "grad_norm": 17.45163917541504,
      "learning_rate": 8.888888888888888e-06,
      "loss": 1.1664,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "eval_cer": 49.3922912710928,
      "eval_loss": 1.1463403701782227,
      "eval_runtime": 691.8556,
      "eval_samples_per_second": 4.119,
      "eval_steps_per_second": 0.516,
      "eval_wer": 74.55810240620367,
      "step": 1000
    },
    {
      "epoch": 0.72,
      "grad_norm": 15.581072807312012,
      "learning_rate": 8.833333333333334e-06,
      "loss": 1.103,
      "step": 1025
    },
    {
      "epoch": 0.74,
      "grad_norm": 24.887859344482422,
      "learning_rate": 8.777777777777778e-06,
      "loss": 1.1038,
      "step": 1050
    },
    {
      "epoch": 0.75,
      "grad_norm": 18.880041122436523,
      "learning_rate": 8.722222222222224e-06,
      "loss": 1.1027,
      "step": 1075
    },
    {
      "epoch": 0.77,
      "grad_norm": 15.563136100769043,
      "learning_rate": 8.666666666666668e-06,
      "loss": 1.1111,
      "step": 1100
    },
    {
      "epoch": 0.79,
      "grad_norm": 15.4989595413208,
      "learning_rate": 8.611111111111112e-06,
      "loss": 1.1109,
      "step": 1125
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.751100540161133,
      "learning_rate": 8.555555555555556e-06,
      "loss": 1.1971,
      "step": 1150
    },
    {
      "epoch": 0.82,
      "grad_norm": 14.376218795776367,
      "learning_rate": 8.5e-06,
      "loss": 1.1702,
      "step": 1175
    },
    {
      "epoch": 0.84,
      "grad_norm": 23.62881851196289,
      "learning_rate": 8.444444444444446e-06,
      "loss": 1.162,
      "step": 1200
    },
    {
      "epoch": 0.86,
      "grad_norm": 18.038660049438477,
      "learning_rate": 8.38888888888889e-06,
      "loss": 1.1106,
      "step": 1225
    },
    {
      "epoch": 0.88,
      "grad_norm": 18.63202667236328,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.0821,
      "step": 1250
    },
    {
      "epoch": 0.89,
      "grad_norm": 21.305559158325195,
      "learning_rate": 8.277777777777778e-06,
      "loss": 1.1619,
      "step": 1275
    },
    {
      "epoch": 0.91,
      "grad_norm": 23.507896423339844,
      "learning_rate": 8.222222222222222e-06,
      "loss": 1.1462,
      "step": 1300
    },
    {
      "epoch": 0.93,
      "grad_norm": 24.86905288696289,
      "learning_rate": 8.166666666666668e-06,
      "loss": 1.2142,
      "step": 1325
    },
    {
      "epoch": 0.95,
      "grad_norm": 19.211929321289062,
      "learning_rate": 8.111111111111112e-06,
      "loss": 1.153,
      "step": 1350
    },
    {
      "epoch": 0.96,
      "grad_norm": 16.071701049804688,
      "learning_rate": 8.055555555555557e-06,
      "loss": 1.1604,
      "step": 1375
    },
    {
      "epoch": 0.98,
      "grad_norm": 20.383394241333008,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.2233,
      "step": 1400
    },
    {
      "epoch": 1.0,
      "grad_norm": 25.448518753051758,
      "learning_rate": 7.944444444444445e-06,
      "loss": 1.0069,
      "step": 1425
    },
    {
      "epoch": 1.02,
      "grad_norm": 11.47100830078125,
      "learning_rate": 7.88888888888889e-06,
      "loss": 0.8217,
      "step": 1450
    },
    {
      "epoch": 1.04,
      "grad_norm": 12.604474067687988,
      "learning_rate": 7.833333333333333e-06,
      "loss": 0.8894,
      "step": 1475
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.27602481842041,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.8474,
      "step": 1500
    },
    {
      "epoch": 1.07,
      "grad_norm": 17.379961013793945,
      "learning_rate": 7.722222222222223e-06,
      "loss": 0.8747,
      "step": 1525
    },
    {
      "epoch": 1.09,
      "grad_norm": 15.369179725646973,
      "learning_rate": 7.666666666666667e-06,
      "loss": 0.826,
      "step": 1550
    },
    {
      "epoch": 1.11,
      "grad_norm": 14.735352516174316,
      "learning_rate": 7.611111111111111e-06,
      "loss": 0.8508,
      "step": 1575
    },
    {
      "epoch": 1.12,
      "grad_norm": 17.533004760742188,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.8139,
      "step": 1600
    },
    {
      "epoch": 1.14,
      "grad_norm": 18.504226684570312,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7383,
      "step": 1625
    },
    {
      "epoch": 1.16,
      "grad_norm": 14.30873966217041,
      "learning_rate": 7.444444444444445e-06,
      "loss": 0.8406,
      "step": 1650
    },
    {
      "epoch": 1.18,
      "grad_norm": 16.69165802001953,
      "learning_rate": 7.38888888888889e-06,
      "loss": 0.8421,
      "step": 1675
    },
    {
      "epoch": 1.19,
      "grad_norm": 19.381628036499023,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.8858,
      "step": 1700
    },
    {
      "epoch": 1.21,
      "grad_norm": 13.489551544189453,
      "learning_rate": 7.277777777777778e-06,
      "loss": 0.9216,
      "step": 1725
    },
    {
      "epoch": 1.23,
      "grad_norm": 12.954667091369629,
      "learning_rate": 7.222222222222223e-06,
      "loss": 0.8289,
      "step": 1750
    },
    {
      "epoch": 1.25,
      "grad_norm": 16.877836227416992,
      "learning_rate": 7.166666666666667e-06,
      "loss": 0.9608,
      "step": 1775
    },
    {
      "epoch": 1.26,
      "grad_norm": 22.067745208740234,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.8079,
      "step": 1800
    },
    {
      "epoch": 1.28,
      "grad_norm": 14.56628131866455,
      "learning_rate": 7.055555555555557e-06,
      "loss": 0.8918,
      "step": 1825
    },
    {
      "epoch": 1.3,
      "grad_norm": 18.21314811706543,
      "learning_rate": 7e-06,
      "loss": 0.8302,
      "step": 1850
    },
    {
      "epoch": 1.32,
      "grad_norm": 18.223928451538086,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.8743,
      "step": 1875
    },
    {
      "epoch": 1.33,
      "grad_norm": 9.887601852416992,
      "learning_rate": 6.88888888888889e-06,
      "loss": 0.8838,
      "step": 1900
    },
    {
      "epoch": 1.35,
      "grad_norm": 22.714948654174805,
      "learning_rate": 6.833333333333334e-06,
      "loss": 0.8611,
      "step": 1925
    },
    {
      "epoch": 1.37,
      "grad_norm": 17.669092178344727,
      "learning_rate": 6.777777777777779e-06,
      "loss": 0.8815,
      "step": 1950
    },
    {
      "epoch": 1.39,
      "grad_norm": 17.025781631469727,
      "learning_rate": 6.7222222222222235e-06,
      "loss": 0.8293,
      "step": 1975
    },
    {
      "epoch": 1.4,
      "grad_norm": 15.673608779907227,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8629,
      "step": 2000
    },
    {
      "epoch": 1.4,
      "eval_cer": 52.48608116371126,
      "eval_loss": 1.1019113063812256,
      "eval_runtime": 711.271,
      "eval_samples_per_second": 4.007,
      "eval_steps_per_second": 0.502,
      "eval_wer": 77.20948796898163,
      "step": 2000
    },
    {
      "epoch": 1.42,
      "grad_norm": 18.508005142211914,
      "learning_rate": 6.6111111111111115e-06,
      "loss": 0.8398,
      "step": 2025
    },
    {
      "epoch": 1.44,
      "grad_norm": 18.747060775756836,
      "learning_rate": 6.555555555555556e-06,
      "loss": 0.9232,
      "step": 2050
    },
    {
      "epoch": 1.46,
      "grad_norm": 20.83848762512207,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.8977,
      "step": 2075
    },
    {
      "epoch": 1.47,
      "grad_norm": 19.12614631652832,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.84,
      "step": 2100
    },
    {
      "epoch": 1.49,
      "grad_norm": 15.908050537109375,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 0.9378,
      "step": 2125
    },
    {
      "epoch": 1.51,
      "grad_norm": 12.05090045928955,
      "learning_rate": 6.333333333333333e-06,
      "loss": 0.7247,
      "step": 2150
    },
    {
      "epoch": 1.53,
      "grad_norm": 16.157573699951172,
      "learning_rate": 6.277777777777778e-06,
      "loss": 0.8945,
      "step": 2175
    },
    {
      "epoch": 1.54,
      "grad_norm": 20.49693489074707,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.8605,
      "step": 2200
    },
    {
      "epoch": 1.56,
      "grad_norm": 15.67389965057373,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.8405,
      "step": 2225
    },
    {
      "epoch": 1.58,
      "grad_norm": 14.528761863708496,
      "learning_rate": 6.111111111111112e-06,
      "loss": 0.8315,
      "step": 2250
    },
    {
      "epoch": 1.6,
      "grad_norm": 21.247663497924805,
      "learning_rate": 6.055555555555555e-06,
      "loss": 0.8511,
      "step": 2275
    },
    {
      "epoch": 1.61,
      "grad_norm": 19.025577545166016,
      "learning_rate": 6e-06,
      "loss": 0.8066,
      "step": 2300
    },
    {
      "epoch": 1.63,
      "grad_norm": 20.215118408203125,
      "learning_rate": 5.944444444444445e-06,
      "loss": 0.8249,
      "step": 2325
    },
    {
      "epoch": 1.65,
      "grad_norm": 14.20770263671875,
      "learning_rate": 5.88888888888889e-06,
      "loss": 0.8892,
      "step": 2350
    },
    {
      "epoch": 1.67,
      "grad_norm": 12.155647277832031,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.8731,
      "step": 2375
    },
    {
      "epoch": 1.68,
      "grad_norm": 16.259695053100586,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.7707,
      "step": 2400
    },
    {
      "epoch": 1.7,
      "grad_norm": 18.88398551940918,
      "learning_rate": 5.722222222222222e-06,
      "loss": 0.8606,
      "step": 2425
    },
    {
      "epoch": 1.72,
      "grad_norm": 14.643707275390625,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.8134,
      "step": 2450
    },
    {
      "epoch": 1.74,
      "grad_norm": 14.3607177734375,
      "learning_rate": 5.611111111111112e-06,
      "loss": 0.8725,
      "step": 2475
    },
    {
      "epoch": 1.75,
      "grad_norm": 23.07048225402832,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.9089,
      "step": 2500
    },
    {
      "epoch": 1.77,
      "grad_norm": 13.498936653137207,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.882,
      "step": 2525
    },
    {
      "epoch": 1.79,
      "grad_norm": 16.4595947265625,
      "learning_rate": 5.444444444444445e-06,
      "loss": 0.8559,
      "step": 2550
    },
    {
      "epoch": 1.81,
      "grad_norm": 14.928861618041992,
      "learning_rate": 5.388888888888889e-06,
      "loss": 0.7985,
      "step": 2575
    },
    {
      "epoch": 1.82,
      "grad_norm": 12.721780776977539,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.8446,
      "step": 2600
    },
    {
      "epoch": 1.84,
      "grad_norm": 15.892865180969238,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 0.8533,
      "step": 2625
    },
    {
      "epoch": 1.86,
      "grad_norm": 15.438478469848633,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 0.8228,
      "step": 2650
    },
    {
      "epoch": 1.88,
      "grad_norm": 15.4628324508667,
      "learning_rate": 5.1666666666666675e-06,
      "loss": 0.8099,
      "step": 2675
    },
    {
      "epoch": 1.89,
      "grad_norm": 22.47260856628418,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.7983,
      "step": 2700
    },
    {
      "epoch": 1.91,
      "grad_norm": 14.621177673339844,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 0.7938,
      "step": 2725
    },
    {
      "epoch": 1.93,
      "grad_norm": 17.526037216186523,
      "learning_rate": 5e-06,
      "loss": 0.7927,
      "step": 2750
    },
    {
      "epoch": 1.95,
      "grad_norm": 18.350086212158203,
      "learning_rate": 4.944444444444445e-06,
      "loss": 0.7404,
      "step": 2775
    },
    {
      "epoch": 1.96,
      "grad_norm": 13.313730239868164,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.8781,
      "step": 2800
    },
    {
      "epoch": 1.98,
      "grad_norm": 13.593079566955566,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.7972,
      "step": 2825
    },
    {
      "epoch": 2.0,
      "grad_norm": 21.296812057495117,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.7721,
      "step": 2850
    },
    {
      "epoch": 2.02,
      "grad_norm": 15.535265922546387,
      "learning_rate": 4.722222222222222e-06,
      "loss": 0.6832,
      "step": 2875
    },
    {
      "epoch": 2.04,
      "grad_norm": 14.036383628845215,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.6262,
      "step": 2900
    },
    {
      "epoch": 2.05,
      "grad_norm": 12.284743309020996,
      "learning_rate": 4.611111111111112e-06,
      "loss": 0.5659,
      "step": 2925
    },
    {
      "epoch": 2.07,
      "grad_norm": 13.954039573669434,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.5496,
      "step": 2950
    },
    {
      "epoch": 2.09,
      "grad_norm": 16.266000747680664,
      "learning_rate": 4.5e-06,
      "loss": 0.5937,
      "step": 2975
    },
    {
      "epoch": 2.11,
      "grad_norm": 17.434391021728516,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.5846,
      "step": 3000
    },
    {
      "epoch": 2.11,
      "eval_cer": 47.78639547731686,
      "eval_loss": 1.1020084619522095,
      "eval_runtime": 697.2267,
      "eval_samples_per_second": 4.088,
      "eval_steps_per_second": 0.512,
      "eval_wer": 72.4027825293648,
      "step": 3000
    },
    {
      "epoch": 2.12,
      "grad_norm": 17.443408966064453,
      "learning_rate": 4.388888888888889e-06,
      "loss": 0.5241,
      "step": 3025
    },
    {
      "epoch": 2.14,
      "grad_norm": 10.447083473205566,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.5926,
      "step": 3050
    },
    {
      "epoch": 2.16,
      "grad_norm": 19.269094467163086,
      "learning_rate": 4.277777777777778e-06,
      "loss": 0.5737,
      "step": 3075
    },
    {
      "epoch": 2.18,
      "grad_norm": 16.834049224853516,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.5023,
      "step": 3100
    },
    {
      "epoch": 2.19,
      "grad_norm": 18.942148208618164,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.6057,
      "step": 3125
    },
    {
      "epoch": 2.21,
      "grad_norm": 13.578940391540527,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.6517,
      "step": 3150
    },
    {
      "epoch": 2.23,
      "grad_norm": 13.73069953918457,
      "learning_rate": 4.055555555555556e-06,
      "loss": 0.6041,
      "step": 3175
    },
    {
      "epoch": 2.25,
      "grad_norm": 14.027520179748535,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.5526,
      "step": 3200
    },
    {
      "epoch": 2.26,
      "grad_norm": 8.801058769226074,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.5499,
      "step": 3225
    },
    {
      "epoch": 2.28,
      "grad_norm": 15.136823654174805,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.639,
      "step": 3250
    },
    {
      "epoch": 2.3,
      "grad_norm": 12.88554859161377,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.5577,
      "step": 3275
    },
    {
      "epoch": 2.32,
      "grad_norm": 14.318121910095215,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.6627,
      "step": 3300
    },
    {
      "epoch": 2.33,
      "grad_norm": 13.512727737426758,
      "learning_rate": 3.7222222222222225e-06,
      "loss": 0.6483,
      "step": 3325
    },
    {
      "epoch": 2.35,
      "grad_norm": 17.256370544433594,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.5252,
      "step": 3350
    },
    {
      "epoch": 2.37,
      "grad_norm": 15.82744026184082,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 0.5955,
      "step": 3375
    },
    {
      "epoch": 2.39,
      "grad_norm": 15.740201950073242,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.594,
      "step": 3400
    },
    {
      "epoch": 2.4,
      "grad_norm": 11.461688041687012,
      "learning_rate": 3.5e-06,
      "loss": 0.5577,
      "step": 3425
    },
    {
      "epoch": 2.42,
      "grad_norm": 11.742351531982422,
      "learning_rate": 3.444444444444445e-06,
      "loss": 0.5899,
      "step": 3450
    },
    {
      "epoch": 2.44,
      "grad_norm": 12.598740577697754,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.6051,
      "step": 3475
    },
    {
      "epoch": 2.46,
      "grad_norm": 13.464423179626465,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.5789,
      "step": 3500
    },
    {
      "epoch": 2.47,
      "grad_norm": 18.55891990661621,
      "learning_rate": 3.277777777777778e-06,
      "loss": 0.627,
      "step": 3525
    },
    {
      "epoch": 2.49,
      "grad_norm": 15.414612770080566,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 0.5117,
      "step": 3550
    },
    {
      "epoch": 2.51,
      "grad_norm": 12.111812591552734,
      "learning_rate": 3.1666666666666667e-06,
      "loss": 0.5448,
      "step": 3575
    },
    {
      "epoch": 2.53,
      "grad_norm": 13.01075267791748,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.5588,
      "step": 3600
    },
    {
      "epoch": 2.54,
      "grad_norm": 10.141874313354492,
      "learning_rate": 3.055555555555556e-06,
      "loss": 0.6379,
      "step": 3625
    },
    {
      "epoch": 2.56,
      "grad_norm": 16.219833374023438,
      "learning_rate": 3e-06,
      "loss": 0.5916,
      "step": 3650
    },
    {
      "epoch": 2.58,
      "grad_norm": 15.017302513122559,
      "learning_rate": 2.944444444444445e-06,
      "loss": 0.5663,
      "step": 3675
    },
    {
      "epoch": 2.6,
      "grad_norm": 13.934078216552734,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.5667,
      "step": 3700
    },
    {
      "epoch": 2.61,
      "grad_norm": 21.70482635498047,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.5694,
      "step": 3725
    },
    {
      "epoch": 2.63,
      "grad_norm": 13.521171569824219,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 0.5453,
      "step": 3750
    },
    {
      "epoch": 2.65,
      "grad_norm": 14.25871753692627,
      "learning_rate": 2.7222222222222224e-06,
      "loss": 0.568,
      "step": 3775
    },
    {
      "epoch": 2.67,
      "grad_norm": 15.246468544006348,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.6255,
      "step": 3800
    },
    {
      "epoch": 2.68,
      "grad_norm": 12.953627586364746,
      "learning_rate": 2.6111111111111113e-06,
      "loss": 0.597,
      "step": 3825
    },
    {
      "epoch": 2.7,
      "grad_norm": 13.217804908752441,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 0.5954,
      "step": 3850
    },
    {
      "epoch": 2.72,
      "grad_norm": 9.510717391967773,
      "learning_rate": 2.5e-06,
      "loss": 0.5497,
      "step": 3875
    },
    {
      "epoch": 2.74,
      "grad_norm": 14.46406078338623,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.6277,
      "step": 3900
    },
    {
      "epoch": 2.75,
      "grad_norm": 9.987096786499023,
      "learning_rate": 2.388888888888889e-06,
      "loss": 0.6343,
      "step": 3925
    },
    {
      "epoch": 2.77,
      "grad_norm": 20.157745361328125,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.5747,
      "step": 3950
    },
    {
      "epoch": 2.79,
      "grad_norm": 13.586224555969238,
      "learning_rate": 2.277777777777778e-06,
      "loss": 0.5702,
      "step": 3975
    },
    {
      "epoch": 2.81,
      "grad_norm": 12.312579154968262,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.5732,
      "step": 4000
    },
    {
      "epoch": 2.81,
      "eval_cer": 44.95220931355196,
      "eval_loss": 1.0969101190567017,
      "eval_runtime": 680.5084,
      "eval_samples_per_second": 4.188,
      "eval_steps_per_second": 0.525,
      "eval_wer": 71.37643973087012,
      "step": 4000
    },
    {
      "epoch": 2.82,
      "grad_norm": 13.381065368652344,
      "learning_rate": 2.166666666666667e-06,
      "loss": 0.6482,
      "step": 4025
    },
    {
      "epoch": 2.84,
      "grad_norm": 14.35582447052002,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 0.562,
      "step": 4050
    },
    {
      "epoch": 2.86,
      "grad_norm": 13.286849975585938,
      "learning_rate": 2.0555555555555555e-06,
      "loss": 0.5772,
      "step": 4075
    },
    {
      "epoch": 2.88,
      "grad_norm": 15.850845336914062,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.5406,
      "step": 4100
    },
    {
      "epoch": 2.89,
      "grad_norm": 14.659579277038574,
      "learning_rate": 1.944444444444445e-06,
      "loss": 0.5307,
      "step": 4125
    },
    {
      "epoch": 2.91,
      "grad_norm": 14.19031810760498,
      "learning_rate": 1.888888888888889e-06,
      "loss": 0.5941,
      "step": 4150
    },
    {
      "epoch": 2.93,
      "grad_norm": 15.247186660766602,
      "learning_rate": 1.8333333333333333e-06,
      "loss": 0.5901,
      "step": 4175
    },
    {
      "epoch": 2.95,
      "grad_norm": 20.510732650756836,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.5564,
      "step": 4200
    },
    {
      "epoch": 2.96,
      "grad_norm": 17.80103874206543,
      "learning_rate": 1.7222222222222224e-06,
      "loss": 0.595,
      "step": 4225
    },
    {
      "epoch": 2.98,
      "grad_norm": 17.83429718017578,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.6067,
      "step": 4250
    },
    {
      "epoch": 3.0,
      "grad_norm": 25.276424407958984,
      "learning_rate": 1.6111111111111113e-06,
      "loss": 0.5515,
      "step": 4275
    },
    {
      "epoch": 3.02,
      "grad_norm": 10.023052215576172,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.4601,
      "step": 4300
    },
    {
      "epoch": 3.04,
      "grad_norm": 13.198165893554688,
      "learning_rate": 1.5e-06,
      "loss": 0.4379,
      "step": 4325
    },
    {
      "epoch": 3.05,
      "grad_norm": 10.112595558166504,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 0.4504,
      "step": 4350
    },
    {
      "epoch": 3.07,
      "grad_norm": 11.835315704345703,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 0.4759,
      "step": 4375
    },
    {
      "epoch": 3.09,
      "grad_norm": 9.861777305603027,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.3991,
      "step": 4400
    },
    {
      "epoch": 3.11,
      "grad_norm": 10.269206047058105,
      "learning_rate": 1.2777777777777779e-06,
      "loss": 0.4102,
      "step": 4425
    },
    {
      "epoch": 3.12,
      "grad_norm": 10.480530738830566,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 0.4027,
      "step": 4450
    },
    {
      "epoch": 3.14,
      "grad_norm": 8.999235153198242,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 0.4232,
      "step": 4475
    },
    {
      "epoch": 3.16,
      "grad_norm": 11.057938575744629,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.3943,
      "step": 4500
    },
    {
      "epoch": 3.18,
      "grad_norm": 11.833608627319336,
      "learning_rate": 1.0555555555555557e-06,
      "loss": 0.3893,
      "step": 4525
    },
    {
      "epoch": 3.19,
      "grad_norm": 12.620944023132324,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.435,
      "step": 4550
    },
    {
      "epoch": 3.21,
      "grad_norm": 11.93198299407959,
      "learning_rate": 9.444444444444445e-07,
      "loss": 0.4328,
      "step": 4575
    },
    {
      "epoch": 3.23,
      "grad_norm": 8.604935646057129,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.427,
      "step": 4600
    },
    {
      "epoch": 3.25,
      "grad_norm": 11.086129188537598,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.4351,
      "step": 4625
    },
    {
      "epoch": 3.26,
      "grad_norm": 17.6994571685791,
      "learning_rate": 7.777777777777779e-07,
      "loss": 0.4108,
      "step": 4650
    },
    {
      "epoch": 3.28,
      "grad_norm": 9.501194953918457,
      "learning_rate": 7.222222222222222e-07,
      "loss": 0.4369,
      "step": 4675
    },
    {
      "epoch": 3.3,
      "grad_norm": 12.139167785644531,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.4041,
      "step": 4700
    },
    {
      "epoch": 3.32,
      "grad_norm": 8.777289390563965,
      "learning_rate": 6.111111111111112e-07,
      "loss": 0.4543,
      "step": 4725
    },
    {
      "epoch": 3.33,
      "grad_norm": 8.665961265563965,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.4471,
      "step": 4750
    },
    {
      "epoch": 3.35,
      "grad_norm": 15.121393203735352,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.5016,
      "step": 4775
    },
    {
      "epoch": 3.37,
      "grad_norm": 14.372379302978516,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.4648,
      "step": 4800
    },
    {
      "epoch": 3.39,
      "grad_norm": 9.729578018188477,
      "learning_rate": 3.8888888888888895e-07,
      "loss": 0.4787,
      "step": 4825
    },
    {
      "epoch": 3.4,
      "grad_norm": 8.461750030517578,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.3607,
      "step": 4850
    },
    {
      "epoch": 3.42,
      "grad_norm": 13.118003845214844,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.4459,
      "step": 4875
    },
    {
      "epoch": 3.44,
      "grad_norm": 11.437525749206543,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.4861,
      "step": 4900
    },
    {
      "epoch": 3.46,
      "grad_norm": 13.136241912841797,
      "learning_rate": 1.6666666666666668e-07,
      "loss": 0.482,
      "step": 4925
    },
    {
      "epoch": 3.47,
      "grad_norm": 12.513566017150879,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 0.4167,
      "step": 4950
    },
    {
      "epoch": 3.49,
      "grad_norm": 10.880446434020996,
      "learning_rate": 5.555555555555556e-08,
      "loss": 0.3889,
      "step": 4975
    },
    {
      "epoch": 3.51,
      "grad_norm": 10.85265827178955,
      "learning_rate": 0.0,
      "loss": 0.3676,
      "step": 5000
    },
    {
      "epoch": 3.51,
      "eval_cer": 46.004569883821965,
      "eval_loss": 1.1264092922210693,
      "eval_runtime": 688.5775,
      "eval_samples_per_second": 4.139,
      "eval_steps_per_second": 0.518,
      "eval_wer": 72.96727106853689,
      "step": 5000
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 1000,
  "total_flos": 1.154081874419712e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}