{
  "best_metric": 0.7646445695226183,
  "best_model_checkpoint": "./whisper-small-dialect_maghrebi/checkpoint-5000",
  "epoch": 2.995805871779509,
  "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": 116.86954498291016,
      "learning_rate": 5.000000000000001e-07,
      "loss": 5.1978,
      "step": 25
    },
    {
      "epoch": 0.03,
      "grad_norm": 53.237552642822266,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 4.3824,
      "step": 50
    },
    {
      "epoch": 0.04,
      "grad_norm": 35.781192779541016,
      "learning_rate": 1.5e-06,
      "loss": 3.4928,
      "step": 75
    },
    {
      "epoch": 0.06,
      "grad_norm": 33.04690933227539,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.8692,
      "step": 100
    },
    {
      "epoch": 0.07,
      "grad_norm": 34.77001190185547,
      "learning_rate": 2.5e-06,
      "loss": 2.6458,
      "step": 125
    },
    {
      "epoch": 0.09,
      "grad_norm": 32.21883010864258,
      "learning_rate": 3e-06,
      "loss": 2.6372,
      "step": 150
    },
    {
      "epoch": 0.1,
      "grad_norm": 27.46969985961914,
      "learning_rate": 3.5e-06,
      "loss": 2.3463,
      "step": 175
    },
    {
      "epoch": 0.12,
      "grad_norm": 31.40074348449707,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.1219,
      "step": 200
    },
    {
      "epoch": 0.13,
      "grad_norm": 29.314828872680664,
      "learning_rate": 4.5e-06,
      "loss": 2.2193,
      "step": 225
    },
    {
      "epoch": 0.15,
      "grad_norm": 31.154987335205078,
      "learning_rate": 5e-06,
      "loss": 2.0789,
      "step": 250
    },
    {
      "epoch": 0.16,
      "grad_norm": 26.621301651000977,
      "learning_rate": 5.500000000000001e-06,
      "loss": 2.1032,
      "step": 275
    },
    {
      "epoch": 0.18,
      "grad_norm": 22.231800079345703,
      "learning_rate": 6e-06,
      "loss": 1.8349,
      "step": 300
    },
    {
      "epoch": 0.19,
      "grad_norm": 34.36363983154297,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 1.8653,
      "step": 325
    },
    {
      "epoch": 0.21,
      "grad_norm": 24.102861404418945,
      "learning_rate": 7e-06,
      "loss": 1.7016,
      "step": 350
    },
    {
      "epoch": 0.22,
      "grad_norm": 23.33128547668457,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.6353,
      "step": 375
    },
    {
      "epoch": 0.24,
      "grad_norm": 24.99433708190918,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.4623,
      "step": 400
    },
    {
      "epoch": 0.25,
      "grad_norm": 30.544538497924805,
      "learning_rate": 8.5e-06,
      "loss": 1.665,
      "step": 425
    },
    {
      "epoch": 0.27,
      "grad_norm": 27.5089168548584,
      "learning_rate": 9e-06,
      "loss": 1.5468,
      "step": 450
    },
    {
      "epoch": 0.28,
      "grad_norm": 30.92209243774414,
      "learning_rate": 9.5e-06,
      "loss": 1.4819,
      "step": 475
    },
    {
      "epoch": 0.3,
      "grad_norm": 27.407058715820312,
      "learning_rate": 1e-05,
      "loss": 1.5511,
      "step": 500
    },
    {
      "epoch": 0.31,
      "grad_norm": 25.37053108215332,
      "learning_rate": 9.944444444444445e-06,
      "loss": 1.4731,
      "step": 525
    },
    {
      "epoch": 0.33,
      "grad_norm": 27.973140716552734,
      "learning_rate": 9.88888888888889e-06,
      "loss": 1.5525,
      "step": 550
    },
    {
      "epoch": 0.34,
      "grad_norm": 24.72431182861328,
      "learning_rate": 9.833333333333333e-06,
      "loss": 1.3978,
      "step": 575
    },
    {
      "epoch": 0.36,
      "grad_norm": 28.63240623474121,
      "learning_rate": 9.777777777777779e-06,
      "loss": 1.4535,
      "step": 600
    },
    {
      "epoch": 0.37,
      "grad_norm": 19.769922256469727,
      "learning_rate": 9.722222222222223e-06,
      "loss": 1.4124,
      "step": 625
    },
    {
      "epoch": 0.39,
      "grad_norm": 25.757123947143555,
      "learning_rate": 9.666666666666667e-06,
      "loss": 1.4984,
      "step": 650
    },
    {
      "epoch": 0.4,
      "grad_norm": 22.6812801361084,
      "learning_rate": 9.611111111111112e-06,
      "loss": 1.3713,
      "step": 675
    },
    {
      "epoch": 0.42,
      "grad_norm": 20.56770896911621,
      "learning_rate": 9.555555555555556e-06,
      "loss": 1.4864,
      "step": 700
    },
    {
      "epoch": 0.43,
      "grad_norm": 23.555574417114258,
      "learning_rate": 9.5e-06,
      "loss": 1.5274,
      "step": 725
    },
    {
      "epoch": 0.45,
      "grad_norm": 25.852689743041992,
      "learning_rate": 9.444444444444445e-06,
      "loss": 1.5334,
      "step": 750
    },
    {
      "epoch": 0.46,
      "grad_norm": 25.466737747192383,
      "learning_rate": 9.38888888888889e-06,
      "loss": 1.5196,
      "step": 775
    },
    {
      "epoch": 0.48,
      "grad_norm": 28.5491943359375,
      "learning_rate": 9.333333333333334e-06,
      "loss": 1.4496,
      "step": 800
    },
    {
      "epoch": 0.49,
      "grad_norm": 30.337696075439453,
      "learning_rate": 9.277777777777778e-06,
      "loss": 1.4737,
      "step": 825
    },
    {
      "epoch": 0.51,
      "grad_norm": 21.093345642089844,
      "learning_rate": 9.222222222222224e-06,
      "loss": 1.4252,
      "step": 850
    },
    {
      "epoch": 0.52,
      "grad_norm": 29.34189224243164,
      "learning_rate": 9.166666666666666e-06,
      "loss": 1.3419,
      "step": 875
    },
    {
      "epoch": 0.54,
      "grad_norm": 23.60466766357422,
      "learning_rate": 9.111111111111112e-06,
      "loss": 1.4271,
      "step": 900
    },
    {
      "epoch": 0.55,
      "grad_norm": 28.1879940032959,
      "learning_rate": 9.055555555555556e-06,
      "loss": 1.4775,
      "step": 925
    },
    {
      "epoch": 0.57,
      "grad_norm": 21.133268356323242,
      "learning_rate": 9e-06,
      "loss": 1.5143,
      "step": 950
    },
    {
      "epoch": 0.58,
      "grad_norm": 26.829559326171875,
      "learning_rate": 8.944444444444446e-06,
      "loss": 1.2894,
      "step": 975
    },
    {
      "epoch": 0.6,
      "grad_norm": 25.009366989135742,
      "learning_rate": 8.888888888888888e-06,
      "loss": 1.4033,
      "step": 1000
    },
    {
      "epoch": 0.6,
      "eval_loss": 1.3855018615722656,
      "eval_runtime": 880.9008,
      "eval_samples_per_second": 3.789,
      "eval_steps_per_second": 0.475,
      "eval_wer": 0.8908171774025433,
      "step": 1000
    },
    {
      "epoch": 0.61,
      "grad_norm": 26.44243049621582,
      "learning_rate": 8.833333333333334e-06,
      "loss": 1.4051,
      "step": 1025
    },
    {
      "epoch": 0.63,
      "grad_norm": 26.72956085205078,
      "learning_rate": 8.777777777777778e-06,
      "loss": 1.4225,
      "step": 1050
    },
    {
      "epoch": 0.64,
      "grad_norm": 18.200281143188477,
      "learning_rate": 8.722222222222224e-06,
      "loss": 1.3932,
      "step": 1075
    },
    {
      "epoch": 0.66,
      "grad_norm": 20.901416778564453,
      "learning_rate": 8.666666666666668e-06,
      "loss": 1.2544,
      "step": 1100
    },
    {
      "epoch": 0.67,
      "grad_norm": 30.09878158569336,
      "learning_rate": 8.611111111111112e-06,
      "loss": 1.4005,
      "step": 1125
    },
    {
      "epoch": 0.69,
      "grad_norm": 24.29834747314453,
      "learning_rate": 8.555555555555556e-06,
      "loss": 1.4653,
      "step": 1150
    },
    {
      "epoch": 0.7,
      "grad_norm": 28.41468620300293,
      "learning_rate": 8.5e-06,
      "loss": 1.3723,
      "step": 1175
    },
    {
      "epoch": 0.72,
      "grad_norm": 15.405891418457031,
      "learning_rate": 8.444444444444446e-06,
      "loss": 1.3921,
      "step": 1200
    },
    {
      "epoch": 0.73,
      "grad_norm": 24.989072799682617,
      "learning_rate": 8.38888888888889e-06,
      "loss": 1.2972,
      "step": 1225
    },
    {
      "epoch": 0.75,
      "grad_norm": 20.0754451751709,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.529,
      "step": 1250
    },
    {
      "epoch": 0.76,
      "grad_norm": 29.132904052734375,
      "learning_rate": 8.277777777777778e-06,
      "loss": 1.2448,
      "step": 1275
    },
    {
      "epoch": 0.78,
      "grad_norm": 20.090288162231445,
      "learning_rate": 8.222222222222222e-06,
      "loss": 1.2525,
      "step": 1300
    },
    {
      "epoch": 0.79,
      "grad_norm": 22.652437210083008,
      "learning_rate": 8.166666666666668e-06,
      "loss": 1.3458,
      "step": 1325
    },
    {
      "epoch": 0.81,
      "grad_norm": 19.910520553588867,
      "learning_rate": 8.111111111111112e-06,
      "loss": 1.2503,
      "step": 1350
    },
    {
      "epoch": 0.82,
      "grad_norm": 19.039749145507812,
      "learning_rate": 8.055555555555557e-06,
      "loss": 1.4319,
      "step": 1375
    },
    {
      "epoch": 0.84,
      "grad_norm": 21.227113723754883,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.388,
      "step": 1400
    },
    {
      "epoch": 0.85,
      "grad_norm": 23.50383949279785,
      "learning_rate": 7.944444444444445e-06,
      "loss": 1.3927,
      "step": 1425
    },
    {
      "epoch": 0.87,
      "grad_norm": 25.4440860748291,
      "learning_rate": 7.88888888888889e-06,
      "loss": 1.3669,
      "step": 1450
    },
    {
      "epoch": 0.88,
      "grad_norm": 23.226131439208984,
      "learning_rate": 7.833333333333333e-06,
      "loss": 1.3087,
      "step": 1475
    },
    {
      "epoch": 0.9,
      "grad_norm": 23.878108978271484,
      "learning_rate": 7.77777777777778e-06,
      "loss": 1.3217,
      "step": 1500
    },
    {
      "epoch": 0.91,
      "grad_norm": 18.915708541870117,
      "learning_rate": 7.722222222222223e-06,
      "loss": 1.2178,
      "step": 1525
    },
    {
      "epoch": 0.93,
      "grad_norm": 21.104673385620117,
      "learning_rate": 7.666666666666667e-06,
      "loss": 1.3529,
      "step": 1550
    },
    {
      "epoch": 0.94,
      "grad_norm": 27.815702438354492,
      "learning_rate": 7.611111111111111e-06,
      "loss": 1.3759,
      "step": 1575
    },
    {
      "epoch": 0.96,
      "grad_norm": 21.6750431060791,
      "learning_rate": 7.555555555555556e-06,
      "loss": 1.3114,
      "step": 1600
    },
    {
      "epoch": 0.97,
      "grad_norm": 21.750614166259766,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.3145,
      "step": 1625
    },
    {
      "epoch": 0.99,
      "grad_norm": 20.031118392944336,
      "learning_rate": 7.444444444444445e-06,
      "loss": 1.2907,
      "step": 1650
    },
    {
      "epoch": 1.0,
      "grad_norm": 20.69341468811035,
      "learning_rate": 7.38888888888889e-06,
      "loss": 1.2293,
      "step": 1675
    },
    {
      "epoch": 1.02,
      "grad_norm": 17.094104766845703,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.9131,
      "step": 1700
    },
    {
      "epoch": 1.03,
      "grad_norm": 22.092281341552734,
      "learning_rate": 7.277777777777778e-06,
      "loss": 1.0874,
      "step": 1725
    },
    {
      "epoch": 1.05,
      "grad_norm": 13.262706756591797,
      "learning_rate": 7.222222222222223e-06,
      "loss": 0.9151,
      "step": 1750
    },
    {
      "epoch": 1.06,
      "grad_norm": 16.924537658691406,
      "learning_rate": 7.166666666666667e-06,
      "loss": 0.8105,
      "step": 1775
    },
    {
      "epoch": 1.08,
      "grad_norm": 17.41377830505371,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.8549,
      "step": 1800
    },
    {
      "epoch": 1.09,
      "grad_norm": 20.281648635864258,
      "learning_rate": 7.055555555555557e-06,
      "loss": 0.9726,
      "step": 1825
    },
    {
      "epoch": 1.11,
      "grad_norm": 21.788726806640625,
      "learning_rate": 7e-06,
      "loss": 1.0932,
      "step": 1850
    },
    {
      "epoch": 1.12,
      "grad_norm": 16.48975372314453,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.984,
      "step": 1875
    },
    {
      "epoch": 1.14,
      "grad_norm": 18.869504928588867,
      "learning_rate": 6.88888888888889e-06,
      "loss": 0.9154,
      "step": 1900
    },
    {
      "epoch": 1.15,
      "grad_norm": 21.09754180908203,
      "learning_rate": 6.833333333333334e-06,
      "loss": 0.8897,
      "step": 1925
    },
    {
      "epoch": 1.17,
      "grad_norm": 26.01792335510254,
      "learning_rate": 6.777777777777779e-06,
      "loss": 1.0268,
      "step": 1950
    },
    {
      "epoch": 1.18,
      "grad_norm": 17.146648406982422,
      "learning_rate": 6.7222222222222235e-06,
      "loss": 0.9361,
      "step": 1975
    },
    {
      "epoch": 1.2,
      "grad_norm": 17.298044204711914,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9838,
      "step": 2000
    },
    {
      "epoch": 1.2,
      "eval_loss": 1.278802752494812,
      "eval_runtime": 865.4527,
      "eval_samples_per_second": 3.857,
      "eval_steps_per_second": 0.483,
      "eval_wer": 0.8505315822388994,
      "step": 2000
    },
    {
      "epoch": 1.21,
      "grad_norm": 19.29056739807129,
      "learning_rate": 6.6111111111111115e-06,
      "loss": 0.9819,
      "step": 2025
    },
    {
      "epoch": 1.23,
      "grad_norm": 26.150436401367188,
      "learning_rate": 6.555555555555556e-06,
      "loss": 1.005,
      "step": 2050
    },
    {
      "epoch": 1.24,
      "grad_norm": 23.540571212768555,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.9917,
      "step": 2075
    },
    {
      "epoch": 1.26,
      "grad_norm": 21.85613441467285,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.9446,
      "step": 2100
    },
    {
      "epoch": 1.27,
      "grad_norm": 22.249435424804688,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 0.9781,
      "step": 2125
    },
    {
      "epoch": 1.29,
      "grad_norm": 18.47481918334961,
      "learning_rate": 6.333333333333333e-06,
      "loss": 1.0354,
      "step": 2150
    },
    {
      "epoch": 1.3,
      "grad_norm": 17.81864356994629,
      "learning_rate": 6.277777777777778e-06,
      "loss": 0.961,
      "step": 2175
    },
    {
      "epoch": 1.32,
      "grad_norm": 21.85215950012207,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.9379,
      "step": 2200
    },
    {
      "epoch": 1.33,
      "grad_norm": 17.410579681396484,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.8977,
      "step": 2225
    },
    {
      "epoch": 1.35,
      "grad_norm": 18.829809188842773,
      "learning_rate": 6.111111111111112e-06,
      "loss": 1.0192,
      "step": 2250
    },
    {
      "epoch": 1.36,
      "grad_norm": 17.973054885864258,
      "learning_rate": 6.055555555555555e-06,
      "loss": 1.0612,
      "step": 2275
    },
    {
      "epoch": 1.38,
      "grad_norm": 19.412118911743164,
      "learning_rate": 6e-06,
      "loss": 0.9552,
      "step": 2300
    },
    {
      "epoch": 1.39,
      "grad_norm": 22.247821807861328,
      "learning_rate": 5.944444444444445e-06,
      "loss": 1.0233,
      "step": 2325
    },
    {
      "epoch": 1.41,
      "grad_norm": 20.862869262695312,
      "learning_rate": 5.88888888888889e-06,
      "loss": 0.912,
      "step": 2350
    },
    {
      "epoch": 1.42,
      "grad_norm": 19.814916610717773,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9388,
      "step": 2375
    },
    {
      "epoch": 1.44,
      "grad_norm": 16.922496795654297,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.8858,
      "step": 2400
    },
    {
      "epoch": 1.45,
      "grad_norm": 24.972301483154297,
      "learning_rate": 5.722222222222222e-06,
      "loss": 0.9323,
      "step": 2425
    },
    {
      "epoch": 1.47,
      "grad_norm": 20.58715057373047,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.9798,
      "step": 2450
    },
    {
      "epoch": 1.48,
      "grad_norm": 22.14284324645996,
      "learning_rate": 5.611111111111112e-06,
      "loss": 1.0126,
      "step": 2475
    },
    {
      "epoch": 1.5,
      "grad_norm": 21.637020111083984,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.8938,
      "step": 2500
    },
    {
      "epoch": 1.51,
      "grad_norm": 20.220205307006836,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.8939,
      "step": 2525
    },
    {
      "epoch": 1.53,
      "grad_norm": 20.443422317504883,
      "learning_rate": 5.444444444444445e-06,
      "loss": 1.0328,
      "step": 2550
    },
    {
      "epoch": 1.54,
      "grad_norm": 21.949094772338867,
      "learning_rate": 5.388888888888889e-06,
      "loss": 0.931,
      "step": 2575
    },
    {
      "epoch": 1.56,
      "grad_norm": 17.825851440429688,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.9842,
      "step": 2600
    },
    {
      "epoch": 1.57,
      "grad_norm": 16.849262237548828,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 1.0379,
      "step": 2625
    },
    {
      "epoch": 1.59,
      "grad_norm": 13.391217231750488,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 0.9041,
      "step": 2650
    },
    {
      "epoch": 1.6,
      "grad_norm": 19.400392532348633,
      "learning_rate": 5.1666666666666675e-06,
      "loss": 0.9205,
      "step": 2675
    },
    {
      "epoch": 1.62,
      "grad_norm": 27.793262481689453,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.9372,
      "step": 2700
    },
    {
      "epoch": 1.63,
      "grad_norm": 18.764686584472656,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 0.9059,
      "step": 2725
    },
    {
      "epoch": 1.65,
      "grad_norm": 19.35423469543457,
      "learning_rate": 5e-06,
      "loss": 0.9997,
      "step": 2750
    },
    {
      "epoch": 1.66,
      "grad_norm": 18.365015029907227,
      "learning_rate": 4.944444444444445e-06,
      "loss": 0.9733,
      "step": 2775
    },
    {
      "epoch": 1.68,
      "grad_norm": 14.752123832702637,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.9581,
      "step": 2800
    },
    {
      "epoch": 1.69,
      "grad_norm": 21.784793853759766,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.9323,
      "step": 2825
    },
    {
      "epoch": 1.71,
      "grad_norm": 18.17945098876953,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.9977,
      "step": 2850
    },
    {
      "epoch": 1.72,
      "grad_norm": 26.37334442138672,
      "learning_rate": 4.722222222222222e-06,
      "loss": 0.9415,
      "step": 2875
    },
    {
      "epoch": 1.74,
      "grad_norm": 26.018369674682617,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.9731,
      "step": 2900
    },
    {
      "epoch": 1.75,
      "grad_norm": 17.425535202026367,
      "learning_rate": 4.611111111111112e-06,
      "loss": 1.0247,
      "step": 2925
    },
    {
      "epoch": 1.77,
      "grad_norm": 20.60196304321289,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.9286,
      "step": 2950
    },
    {
      "epoch": 1.78,
      "grad_norm": 15.926254272460938,
      "learning_rate": 4.5e-06,
      "loss": 0.9287,
      "step": 2975
    },
    {
      "epoch": 1.8,
      "grad_norm": 22.282596588134766,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.9488,
      "step": 3000
    },
    {
      "epoch": 1.8,
      "eval_loss": 1.2125581502914429,
      "eval_runtime": 939.7465,
      "eval_samples_per_second": 3.552,
      "eval_steps_per_second": 0.445,
      "eval_wer": 1.0568584531999166,
      "step": 3000
    },
    {
      "epoch": 1.81,
      "grad_norm": 16.973588943481445,
      "learning_rate": 4.388888888888889e-06,
      "loss": 0.9748,
      "step": 3025
    },
    {
      "epoch": 1.83,
      "grad_norm": 22.70237159729004,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.9352,
      "step": 3050
    },
    {
      "epoch": 1.84,
      "grad_norm": 23.420825958251953,
      "learning_rate": 4.277777777777778e-06,
      "loss": 1.0914,
      "step": 3075
    },
    {
      "epoch": 1.86,
      "grad_norm": 16.404287338256836,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.8749,
      "step": 3100
    },
    {
      "epoch": 1.87,
      "grad_norm": 12.428851127624512,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.8473,
      "step": 3125
    },
    {
      "epoch": 1.89,
      "grad_norm": 17.85788345336914,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.9619,
      "step": 3150
    },
    {
      "epoch": 1.9,
      "grad_norm": 15.915815353393555,
      "learning_rate": 4.055555555555556e-06,
      "loss": 0.9358,
      "step": 3175
    },
    {
      "epoch": 1.92,
      "grad_norm": 18.625694274902344,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.8463,
      "step": 3200
    },
    {
      "epoch": 1.93,
      "grad_norm": 20.894052505493164,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.9683,
      "step": 3225
    },
    {
      "epoch": 1.95,
      "grad_norm": 17.617666244506836,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.8803,
      "step": 3250
    },
    {
      "epoch": 1.96,
      "grad_norm": 22.090686798095703,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.9894,
      "step": 3275
    },
    {
      "epoch": 1.98,
      "grad_norm": 19.619741439819336,
      "learning_rate": 3.777777777777778e-06,
      "loss": 1.0189,
      "step": 3300
    },
    {
      "epoch": 1.99,
      "grad_norm": 25.635848999023438,
      "learning_rate": 3.7222222222222225e-06,
      "loss": 0.8735,
      "step": 3325
    },
    {
      "epoch": 2.01,
      "grad_norm": 10.868768692016602,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.755,
      "step": 3350
    },
    {
      "epoch": 2.02,
      "grad_norm": 18.025487899780273,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 0.6058,
      "step": 3375
    },
    {
      "epoch": 2.04,
      "grad_norm": 18.096664428710938,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.6674,
      "step": 3400
    },
    {
      "epoch": 2.05,
      "grad_norm": 19.073925018310547,
      "learning_rate": 3.5e-06,
      "loss": 0.725,
      "step": 3425
    },
    {
      "epoch": 2.07,
      "grad_norm": 15.58212661743164,
      "learning_rate": 3.444444444444445e-06,
      "loss": 0.6849,
      "step": 3450
    },
    {
      "epoch": 2.08,
      "grad_norm": 21.028173446655273,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.6903,
      "step": 3475
    },
    {
      "epoch": 2.1,
      "grad_norm": 15.783544540405273,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.6258,
      "step": 3500
    },
    {
      "epoch": 2.11,
      "grad_norm": 17.83692741394043,
      "learning_rate": 3.277777777777778e-06,
      "loss": 0.6701,
      "step": 3525
    },
    {
      "epoch": 2.13,
      "grad_norm": 18.08365821838379,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 0.6288,
      "step": 3550
    },
    {
      "epoch": 2.14,
      "grad_norm": 17.74977684020996,
      "learning_rate": 3.1666666666666667e-06,
      "loss": 0.6646,
      "step": 3575
    },
    {
      "epoch": 2.16,
      "grad_norm": 16.543344497680664,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.6207,
      "step": 3600
    },
    {
      "epoch": 2.17,
      "grad_norm": 15.900822639465332,
      "learning_rate": 3.055555555555556e-06,
      "loss": 0.6549,
      "step": 3625
    },
    {
      "epoch": 2.19,
      "grad_norm": 15.244269371032715,
      "learning_rate": 3e-06,
      "loss": 0.6731,
      "step": 3650
    },
    {
      "epoch": 2.2,
      "grad_norm": 18.52103614807129,
      "learning_rate": 2.944444444444445e-06,
      "loss": 0.6826,
      "step": 3675
    },
    {
      "epoch": 2.22,
      "grad_norm": 16.776268005371094,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.7041,
      "step": 3700
    },
    {
      "epoch": 2.23,
      "grad_norm": 9.410231590270996,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.6298,
      "step": 3725
    },
    {
      "epoch": 2.25,
      "grad_norm": 16.997234344482422,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 0.6884,
      "step": 3750
    },
    {
      "epoch": 2.26,
      "grad_norm": 16.043540954589844,
      "learning_rate": 2.7222222222222224e-06,
      "loss": 0.6071,
      "step": 3775
    },
    {
      "epoch": 2.28,
      "grad_norm": 17.286903381347656,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.6079,
      "step": 3800
    },
    {
      "epoch": 2.29,
      "grad_norm": 19.805654525756836,
      "learning_rate": 2.6111111111111113e-06,
      "loss": 0.7621,
      "step": 3825
    },
    {
      "epoch": 2.31,
      "grad_norm": 17.18813705444336,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 0.5951,
      "step": 3850
    },
    {
      "epoch": 2.32,
      "grad_norm": 17.5855770111084,
      "learning_rate": 2.5e-06,
      "loss": 0.6432,
      "step": 3875
    },
    {
      "epoch": 2.34,
      "grad_norm": 16.48206901550293,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.6648,
      "step": 3900
    },
    {
      "epoch": 2.35,
      "grad_norm": 15.761092185974121,
      "learning_rate": 2.388888888888889e-06,
      "loss": 0.5789,
      "step": 3925
    },
    {
      "epoch": 2.37,
      "grad_norm": 10.336495399475098,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.5722,
      "step": 3950
    },
    {
      "epoch": 2.38,
      "grad_norm": 12.249388694763184,
      "learning_rate": 2.277777777777778e-06,
      "loss": 0.6276,
      "step": 3975
    },
    {
      "epoch": 2.4,
      "grad_norm": 17.24640655517578,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.654,
      "step": 4000
    },
    {
      "epoch": 2.4,
      "eval_loss": 1.2189723253250122,
      "eval_runtime": 880.2772,
      "eval_samples_per_second": 3.792,
      "eval_steps_per_second": 0.475,
      "eval_wer": 0.7822597456743798,
      "step": 4000
    },
    {
      "epoch": 2.41,
      "grad_norm": 17.82291030883789,
      "learning_rate": 2.166666666666667e-06,
      "loss": 0.6819,
      "step": 4025
    },
    {
      "epoch": 2.43,
      "grad_norm": 15.322093963623047,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 0.6462,
      "step": 4050
    },
    {
      "epoch": 2.44,
      "grad_norm": 18.799236297607422,
      "learning_rate": 2.0555555555555555e-06,
      "loss": 0.6248,
      "step": 4075
    },
    {
      "epoch": 2.46,
      "grad_norm": 13.401931762695312,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.6459,
      "step": 4100
    },
    {
      "epoch": 2.47,
      "grad_norm": 19.656177520751953,
      "learning_rate": 1.944444444444445e-06,
      "loss": 0.6104,
      "step": 4125
    },
    {
      "epoch": 2.49,
      "grad_norm": 20.60839080810547,
      "learning_rate": 1.888888888888889e-06,
      "loss": 0.6743,
      "step": 4150
    },
    {
      "epoch": 2.5,
      "grad_norm": 15.391945838928223,
      "learning_rate": 1.8333333333333333e-06,
      "loss": 0.6277,
      "step": 4175
    },
    {
      "epoch": 2.52,
      "grad_norm": 16.259719848632812,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.6386,
      "step": 4200
    },
    {
      "epoch": 2.53,
      "grad_norm": 14.335466384887695,
      "learning_rate": 1.7222222222222224e-06,
      "loss": 0.6621,
      "step": 4225
    },
    {
      "epoch": 2.55,
      "grad_norm": 17.98448371887207,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.7163,
      "step": 4250
    },
    {
      "epoch": 2.56,
      "grad_norm": 18.504575729370117,
      "learning_rate": 1.6111111111111113e-06,
      "loss": 0.7152,
      "step": 4275
    },
    {
      "epoch": 2.58,
      "grad_norm": 11.735873222351074,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.565,
      "step": 4300
    },
    {
      "epoch": 2.59,
      "grad_norm": 16.514970779418945,
      "learning_rate": 1.5e-06,
      "loss": 0.6267,
      "step": 4325
    },
    {
      "epoch": 2.61,
      "grad_norm": 14.953248023986816,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 0.5904,
      "step": 4350
    },
    {
      "epoch": 2.62,
      "grad_norm": 21.26978874206543,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 0.7009,
      "step": 4375
    },
    {
      "epoch": 2.64,
      "grad_norm": 13.684809684753418,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.5642,
      "step": 4400
    },
    {
      "epoch": 2.65,
      "grad_norm": 13.99679946899414,
      "learning_rate": 1.2777777777777779e-06,
      "loss": 0.5808,
      "step": 4425
    },
    {
      "epoch": 2.67,
      "grad_norm": 18.248409271240234,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 0.6285,
      "step": 4450
    },
    {
      "epoch": 2.68,
      "grad_norm": 13.999139785766602,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 0.6578,
      "step": 4475
    },
    {
      "epoch": 2.7,
      "grad_norm": 23.412248611450195,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.6514,
      "step": 4500
    },
    {
      "epoch": 2.71,
      "grad_norm": 16.740997314453125,
      "learning_rate": 1.0555555555555557e-06,
      "loss": 0.5876,
      "step": 4525
    },
    {
      "epoch": 2.73,
      "grad_norm": 19.925724029541016,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.6444,
      "step": 4550
    },
    {
      "epoch": 2.74,
      "grad_norm": 12.167197227478027,
      "learning_rate": 9.444444444444445e-07,
      "loss": 0.6231,
      "step": 4575
    },
    {
      "epoch": 2.76,
      "grad_norm": 12.373435974121094,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.691,
      "step": 4600
    },
    {
      "epoch": 2.77,
      "grad_norm": 14.85899543762207,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.6554,
      "step": 4625
    },
    {
      "epoch": 2.79,
      "grad_norm": 17.101654052734375,
      "learning_rate": 7.777777777777779e-07,
      "loss": 0.6281,
      "step": 4650
    },
    {
      "epoch": 2.8,
      "grad_norm": 18.191896438598633,
      "learning_rate": 7.222222222222222e-07,
      "loss": 0.735,
      "step": 4675
    },
    {
      "epoch": 2.82,
      "grad_norm": 13.210752487182617,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.5379,
      "step": 4700
    },
    {
      "epoch": 2.83,
      "grad_norm": 15.177491188049316,
      "learning_rate": 6.111111111111112e-07,
      "loss": 0.5441,
      "step": 4725
    },
    {
      "epoch": 2.85,
      "grad_norm": 17.19512367248535,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.6691,
      "step": 4750
    },
    {
      "epoch": 2.86,
      "grad_norm": 17.00133514404297,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.5082,
      "step": 4775
    },
    {
      "epoch": 2.88,
      "grad_norm": 21.2026424407959,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.583,
      "step": 4800
    },
    {
      "epoch": 2.89,
      "grad_norm": 16.55152130126953,
      "learning_rate": 3.8888888888888895e-07,
      "loss": 0.6382,
      "step": 4825
    },
    {
      "epoch": 2.91,
      "grad_norm": 17.66421127319336,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.5711,
      "step": 4850
    },
    {
      "epoch": 2.92,
      "grad_norm": 15.728326797485352,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.6598,
      "step": 4875
    },
    {
      "epoch": 2.94,
      "grad_norm": 17.651121139526367,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.6427,
      "step": 4900
    },
    {
      "epoch": 2.95,
      "grad_norm": 18.308921813964844,
      "learning_rate": 1.6666666666666668e-07,
      "loss": 0.6662,
      "step": 4925
    },
    {
      "epoch": 2.97,
      "grad_norm": 17.988201141357422,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 0.5818,
      "step": 4950
    },
    {
      "epoch": 2.98,
      "grad_norm": 14.850335121154785,
      "learning_rate": 5.555555555555556e-08,
      "loss": 0.5628,
      "step": 4975
    },
    {
      "epoch": 3.0,
      "grad_norm": 12.471959114074707,
      "learning_rate": 0.0,
      "loss": 0.6891,
      "step": 5000
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.2056573629379272,
      "eval_runtime": 966.5601,
      "eval_samples_per_second": 3.453,
      "eval_steps_per_second": 0.432,
      "eval_wer": 0.7646445695226183,
      "step": 5000
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "total_flos": 1.15434160128e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}