{
  "best_metric": 0.4799608558338152,
  "best_model_checkpoint": "./whisper-small-ar_tsize_0.6/checkpoint-5000",
  "epoch": 2.1468441391155,
  "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": 59.05270767211914,
      "learning_rate": 5.000000000000001e-07,
      "loss": 3.2982,
      "step": 25
    },
    {
      "epoch": 0.02,
      "grad_norm": 46.39448165893555,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.5837,
      "step": 50
    },
    {
      "epoch": 0.03,
      "grad_norm": 25.321762084960938,
      "learning_rate": 1.5e-06,
      "loss": 1.8822,
      "step": 75
    },
    {
      "epoch": 0.04,
      "grad_norm": 15.204469680786133,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.2958,
      "step": 100
    },
    {
      "epoch": 0.05,
      "grad_norm": 21.59715461730957,
      "learning_rate": 2.5e-06,
      "loss": 1.0842,
      "step": 125
    },
    {
      "epoch": 0.06,
      "grad_norm": 15.583703994750977,
      "learning_rate": 3e-06,
      "loss": 1.0754,
      "step": 150
    },
    {
      "epoch": 0.08,
      "grad_norm": 15.472356796264648,
      "learning_rate": 3.5e-06,
      "loss": 0.977,
      "step": 175
    },
    {
      "epoch": 0.09,
      "grad_norm": 14.069989204406738,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.8513,
      "step": 200
    },
    {
      "epoch": 0.1,
      "grad_norm": 17.11322784423828,
      "learning_rate": 4.5e-06,
      "loss": 0.8297,
      "step": 225
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.451148986816406,
      "learning_rate": 5e-06,
      "loss": 0.7162,
      "step": 250
    },
    {
      "epoch": 0.12,
      "grad_norm": 16.34424591064453,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.731,
      "step": 275
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.942704200744629,
      "learning_rate": 6e-06,
      "loss": 0.6081,
      "step": 300
    },
    {
      "epoch": 0.14,
      "grad_norm": 18.032196044921875,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.4776,
      "step": 325
    },
    {
      "epoch": 0.15,
      "grad_norm": 18.75937843322754,
      "learning_rate": 7e-06,
      "loss": 0.4826,
      "step": 350
    },
    {
      "epoch": 0.16,
      "grad_norm": 15.238700866699219,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.4241,
      "step": 375
    },
    {
      "epoch": 0.17,
      "grad_norm": 15.744266510009766,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.4345,
      "step": 400
    },
    {
      "epoch": 0.18,
      "grad_norm": 20.355653762817383,
      "learning_rate": 8.5e-06,
      "loss": 0.4603,
      "step": 425
    },
    {
      "epoch": 0.19,
      "grad_norm": 13.174769401550293,
      "learning_rate": 9e-06,
      "loss": 0.4287,
      "step": 450
    },
    {
      "epoch": 0.2,
      "grad_norm": 11.769058227539062,
      "learning_rate": 9.5e-06,
      "loss": 0.3736,
      "step": 475
    },
    {
      "epoch": 0.21,
      "grad_norm": 14.914719581604004,
      "learning_rate": 1e-05,
      "loss": 0.407,
      "step": 500
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.257582664489746,
      "learning_rate": 9.944444444444445e-06,
      "loss": 0.4238,
      "step": 525
    },
    {
      "epoch": 0.24,
      "grad_norm": 15.175993919372559,
      "learning_rate": 9.88888888888889e-06,
      "loss": 0.3907,
      "step": 550
    },
    {
      "epoch": 0.25,
      "grad_norm": 11.015135765075684,
      "learning_rate": 9.833333333333333e-06,
      "loss": 0.3915,
      "step": 575
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.840285301208496,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.3861,
      "step": 600
    },
    {
      "epoch": 0.27,
      "grad_norm": 14.146044731140137,
      "learning_rate": 9.722222222222223e-06,
      "loss": 0.4046,
      "step": 625
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.024092674255371,
      "learning_rate": 9.666666666666667e-06,
      "loss": 0.3832,
      "step": 650
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.27486801147461,
      "learning_rate": 9.611111111111112e-06,
      "loss": 0.3641,
      "step": 675
    },
    {
      "epoch": 0.3,
      "grad_norm": 17.2608642578125,
      "learning_rate": 9.555555555555556e-06,
      "loss": 0.4064,
      "step": 700
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.510050773620605,
      "learning_rate": 9.5e-06,
      "loss": 0.3866,
      "step": 725
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.728179931640625,
      "learning_rate": 9.444444444444445e-06,
      "loss": 0.3714,
      "step": 750
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.029936790466309,
      "learning_rate": 9.38888888888889e-06,
      "loss": 0.3995,
      "step": 775
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.523062705993652,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.3521,
      "step": 800
    },
    {
      "epoch": 0.35,
      "grad_norm": 16.931310653686523,
      "learning_rate": 9.277777777777778e-06,
      "loss": 0.4273,
      "step": 825
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.761600494384766,
      "learning_rate": 9.222222222222224e-06,
      "loss": 0.359,
      "step": 850
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.472681999206543,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.3641,
      "step": 875
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.588469505310059,
      "learning_rate": 9.111111111111112e-06,
      "loss": 0.3008,
      "step": 900
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.993001937866211,
      "learning_rate": 9.055555555555556e-06,
      "loss": 0.3504,
      "step": 925
    },
    {
      "epoch": 0.41,
      "grad_norm": 15.114642143249512,
      "learning_rate": 9e-06,
      "loss": 0.3608,
      "step": 950
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.145492553710938,
      "learning_rate": 8.944444444444446e-06,
      "loss": 0.3693,
      "step": 975
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.101957321166992,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.3396,
      "step": 1000
    },
    {
      "epoch": 0.43,
      "eval_loss": 0.34926944971084595,
      "eval_runtime": 2524.5618,
      "eval_samples_per_second": 3.074,
      "eval_steps_per_second": 0.385,
      "eval_wer": 0.5713268982696499,
      "step": 1000
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.783793449401855,
      "learning_rate": 8.833333333333334e-06,
      "loss": 0.3821,
      "step": 1025
    },
    {
      "epoch": 0.45,
      "grad_norm": 12.823246002197266,
      "learning_rate": 8.777777777777778e-06,
      "loss": 0.3537,
      "step": 1050
    },
    {
      "epoch": 0.46,
      "grad_norm": 14.500955581665039,
      "learning_rate": 8.722222222222224e-06,
      "loss": 0.3171,
      "step": 1075
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.166177749633789,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.3132,
      "step": 1100
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.919294357299805,
      "learning_rate": 8.611111111111112e-06,
      "loss": 0.3347,
      "step": 1125
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.559989929199219,
      "learning_rate": 8.555555555555556e-06,
      "loss": 0.2997,
      "step": 1150
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.830954551696777,
      "learning_rate": 8.5e-06,
      "loss": 0.3592,
      "step": 1175
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.9480619430542,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.309,
      "step": 1200
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.179547309875488,
      "learning_rate": 8.38888888888889e-06,
      "loss": 0.3348,
      "step": 1225
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.940916061401367,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.3464,
      "step": 1250
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.2194242477417,
      "learning_rate": 8.277777777777778e-06,
      "loss": 0.3515,
      "step": 1275
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.347893714904785,
      "learning_rate": 8.222222222222222e-06,
      "loss": 0.3375,
      "step": 1300
    },
    {
      "epoch": 0.57,
      "grad_norm": 7.932102203369141,
      "learning_rate": 8.166666666666668e-06,
      "loss": 0.3159,
      "step": 1325
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.098414421081543,
      "learning_rate": 8.111111111111112e-06,
      "loss": 0.31,
      "step": 1350
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.054365158081055,
      "learning_rate": 8.055555555555557e-06,
      "loss": 0.3196,
      "step": 1375
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.636041641235352,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.298,
      "step": 1400
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.017091751098633,
      "learning_rate": 7.944444444444445e-06,
      "loss": 0.3391,
      "step": 1425
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.708351135253906,
      "learning_rate": 7.88888888888889e-06,
      "loss": 0.3254,
      "step": 1450
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.196100234985352,
      "learning_rate": 7.833333333333333e-06,
      "loss": 0.3192,
      "step": 1475
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.520899772644043,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.3749,
      "step": 1500
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.588493347167969,
      "learning_rate": 7.722222222222223e-06,
      "loss": 0.3031,
      "step": 1525
    },
    {
      "epoch": 0.67,
      "grad_norm": 7.777310848236084,
      "learning_rate": 7.666666666666667e-06,
      "loss": 0.3237,
      "step": 1550
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.77524471282959,
      "learning_rate": 7.611111111111111e-06,
      "loss": 0.3241,
      "step": 1575
    },
    {
      "epoch": 0.69,
      "grad_norm": 12.974262237548828,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.3137,
      "step": 1600
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.752847671508789,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3036,
      "step": 1625
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.93649673461914,
      "learning_rate": 7.444444444444445e-06,
      "loss": 0.3242,
      "step": 1650
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.082414627075195,
      "learning_rate": 7.38888888888889e-06,
      "loss": 0.3305,
      "step": 1675
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.735708236694336,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.2919,
      "step": 1700
    },
    {
      "epoch": 0.74,
      "grad_norm": 6.739984512329102,
      "learning_rate": 7.277777777777778e-06,
      "loss": 0.2963,
      "step": 1725
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.7431058883667,
      "learning_rate": 7.222222222222223e-06,
      "loss": 0.2995,
      "step": 1750
    },
    {
      "epoch": 0.76,
      "grad_norm": 7.32223653793335,
      "learning_rate": 7.166666666666667e-06,
      "loss": 0.297,
      "step": 1775
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.857797622680664,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.3037,
      "step": 1800
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.739462852478027,
      "learning_rate": 7.055555555555557e-06,
      "loss": 0.2904,
      "step": 1825
    },
    {
      "epoch": 0.79,
      "grad_norm": 6.398651599884033,
      "learning_rate": 7e-06,
      "loss": 0.2922,
      "step": 1850
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.660301208496094,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.2988,
      "step": 1875
    },
    {
      "epoch": 0.82,
      "grad_norm": 7.902390003204346,
      "learning_rate": 6.88888888888889e-06,
      "loss": 0.2855,
      "step": 1900
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.831525802612305,
      "learning_rate": 6.833333333333334e-06,
      "loss": 0.3228,
      "step": 1925
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.282029151916504,
      "learning_rate": 6.777777777777779e-06,
      "loss": 0.325,
      "step": 1950
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.988739967346191,
      "learning_rate": 6.7222222222222235e-06,
      "loss": 0.2969,
      "step": 1975
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.352680206298828,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2874,
      "step": 2000
    },
    {
      "epoch": 0.86,
      "eval_loss": 0.287245512008667,
      "eval_runtime": 2341.0722,
      "eval_samples_per_second": 3.315,
      "eval_steps_per_second": 0.415,
      "eval_wer": 0.5246875139006272,
      "step": 2000
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.70847225189209,
      "learning_rate": 6.6111111111111115e-06,
      "loss": 0.2838,
      "step": 2025
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.944375991821289,
      "learning_rate": 6.555555555555556e-06,
      "loss": 0.2956,
      "step": 2050
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.928349494934082,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.298,
      "step": 2075
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.797341346740723,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.2609,
      "step": 2100
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.493168830871582,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 0.2661,
      "step": 2125
    },
    {
      "epoch": 0.92,
      "grad_norm": 6.650107383728027,
      "learning_rate": 6.333333333333333e-06,
      "loss": 0.2848,
      "step": 2150
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.574175834655762,
      "learning_rate": 6.277777777777778e-06,
      "loss": 0.2805,
      "step": 2175
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.462020874023438,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.2796,
      "step": 2200
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.76837158203125,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.2658,
      "step": 2225
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.173978805541992,
      "learning_rate": 6.111111111111112e-06,
      "loss": 0.2736,
      "step": 2250
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.383520126342773,
      "learning_rate": 6.055555555555555e-06,
      "loss": 0.2787,
      "step": 2275
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.080275535583496,
      "learning_rate": 6e-06,
      "loss": 0.2724,
      "step": 2300
    },
    {
      "epoch": 1.0,
      "grad_norm": 7.687675476074219,
      "learning_rate": 5.944444444444445e-06,
      "loss": 0.2918,
      "step": 2325
    },
    {
      "epoch": 1.01,
      "grad_norm": 5.718274116516113,
      "learning_rate": 5.88888888888889e-06,
      "loss": 0.197,
      "step": 2350
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.179292678833008,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.1709,
      "step": 2375
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.117734909057617,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.1831,
      "step": 2400
    },
    {
      "epoch": 1.04,
      "grad_norm": 5.605225563049316,
      "learning_rate": 5.722222222222222e-06,
      "loss": 0.1668,
      "step": 2425
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.287034034729004,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.1822,
      "step": 2450
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.633007049560547,
      "learning_rate": 5.611111111111112e-06,
      "loss": 0.1563,
      "step": 2475
    },
    {
      "epoch": 1.07,
      "grad_norm": 6.662952423095703,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.1814,
      "step": 2500
    },
    {
      "epoch": 1.08,
      "grad_norm": 6.421918869018555,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.1562,
      "step": 2525
    },
    {
      "epoch": 1.09,
      "grad_norm": 5.427768230438232,
      "learning_rate": 5.444444444444445e-06,
      "loss": 0.1579,
      "step": 2550
    },
    {
      "epoch": 1.11,
      "grad_norm": 7.273324012756348,
      "learning_rate": 5.388888888888889e-06,
      "loss": 0.1773,
      "step": 2575
    },
    {
      "epoch": 1.12,
      "grad_norm": 10.406336784362793,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.1839,
      "step": 2600
    },
    {
      "epoch": 1.13,
      "grad_norm": 9.072065353393555,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 0.1649,
      "step": 2625
    },
    {
      "epoch": 1.14,
      "grad_norm": 5.964203357696533,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 0.1792,
      "step": 2650
    },
    {
      "epoch": 1.15,
      "grad_norm": 11.519746780395508,
      "learning_rate": 5.1666666666666675e-06,
      "loss": 0.1725,
      "step": 2675
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.105682373046875,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.1568,
      "step": 2700
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.028508186340332,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 0.1991,
      "step": 2725
    },
    {
      "epoch": 1.18,
      "grad_norm": 7.226799488067627,
      "learning_rate": 5e-06,
      "loss": 0.1893,
      "step": 2750
    },
    {
      "epoch": 1.19,
      "grad_norm": 7.659578323364258,
      "learning_rate": 4.944444444444445e-06,
      "loss": 0.1775,
      "step": 2775
    },
    {
      "epoch": 1.2,
      "grad_norm": 8.088948249816895,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.174,
      "step": 2800
    },
    {
      "epoch": 1.21,
      "grad_norm": 5.826450824737549,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.1696,
      "step": 2825
    },
    {
      "epoch": 1.22,
      "grad_norm": 6.254305839538574,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.1635,
      "step": 2850
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.753539562225342,
      "learning_rate": 4.722222222222222e-06,
      "loss": 0.2123,
      "step": 2875
    },
    {
      "epoch": 1.25,
      "grad_norm": 7.371118545532227,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.1625,
      "step": 2900
    },
    {
      "epoch": 1.26,
      "grad_norm": 8.816264152526855,
      "learning_rate": 4.611111111111112e-06,
      "loss": 0.1647,
      "step": 2925
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.907712936401367,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.1584,
      "step": 2950
    },
    {
      "epoch": 1.28,
      "grad_norm": 10.592573165893555,
      "learning_rate": 4.5e-06,
      "loss": 0.1695,
      "step": 2975
    },
    {
      "epoch": 1.29,
      "grad_norm": 6.8986992835998535,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.1626,
      "step": 3000
    },
    {
      "epoch": 1.29,
      "eval_loss": 0.2664027810096741,
      "eval_runtime": 2325.9517,
      "eval_samples_per_second": 3.337,
      "eval_steps_per_second": 0.417,
      "eval_wer": 0.4912815266224812,
      "step": 3000
    },
    {
      "epoch": 1.3,
      "grad_norm": 6.402446269989014,
      "learning_rate": 4.388888888888889e-06,
      "loss": 0.1473,
      "step": 3025
    },
    {
      "epoch": 1.31,
      "grad_norm": 7.3218183517456055,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.1639,
      "step": 3050
    },
    {
      "epoch": 1.32,
      "grad_norm": 6.958987712860107,
      "learning_rate": 4.277777777777778e-06,
      "loss": 0.1513,
      "step": 3075
    },
    {
      "epoch": 1.33,
      "grad_norm": 7.87843132019043,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.1599,
      "step": 3100
    },
    {
      "epoch": 1.34,
      "grad_norm": 5.358362674713135,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.1528,
      "step": 3125
    },
    {
      "epoch": 1.35,
      "grad_norm": 7.011563301086426,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.1543,
      "step": 3150
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.992841720581055,
      "learning_rate": 4.055555555555556e-06,
      "loss": 0.1857,
      "step": 3175
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.283531665802002,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.1849,
      "step": 3200
    },
    {
      "epoch": 1.38,
      "grad_norm": 7.267604827880859,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.1735,
      "step": 3225
    },
    {
      "epoch": 1.4,
      "grad_norm": 7.122161865234375,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.1692,
      "step": 3250
    },
    {
      "epoch": 1.41,
      "grad_norm": 4.604694843292236,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.1821,
      "step": 3275
    },
    {
      "epoch": 1.42,
      "grad_norm": 9.58035945892334,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.1744,
      "step": 3300
    },
    {
      "epoch": 1.43,
      "grad_norm": 7.21837043762207,
      "learning_rate": 3.7222222222222225e-06,
      "loss": 0.1511,
      "step": 3325
    },
    {
      "epoch": 1.44,
      "grad_norm": 4.654594898223877,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.1492,
      "step": 3350
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.780755996704102,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 0.1634,
      "step": 3375
    },
    {
      "epoch": 1.46,
      "grad_norm": 7.447520732879639,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.1668,
      "step": 3400
    },
    {
      "epoch": 1.47,
      "grad_norm": 5.384232521057129,
      "learning_rate": 3.5e-06,
      "loss": 0.1616,
      "step": 3425
    },
    {
      "epoch": 1.48,
      "grad_norm": 8.949358940124512,
      "learning_rate": 3.444444444444445e-06,
      "loss": 0.1618,
      "step": 3450
    },
    {
      "epoch": 1.49,
      "grad_norm": 7.167361736297607,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.139,
      "step": 3475
    },
    {
      "epoch": 1.5,
      "grad_norm": 4.681729316711426,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.1557,
      "step": 3500
    },
    {
      "epoch": 1.51,
      "grad_norm": 4.427718639373779,
      "learning_rate": 3.277777777777778e-06,
      "loss": 0.1623,
      "step": 3525
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.309398651123047,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 0.1633,
      "step": 3550
    },
    {
      "epoch": 1.53,
      "grad_norm": 7.46133279800415,
      "learning_rate": 3.1666666666666667e-06,
      "loss": 0.1855,
      "step": 3575
    },
    {
      "epoch": 1.55,
      "grad_norm": 6.646737098693848,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.15,
      "step": 3600
    },
    {
      "epoch": 1.56,
      "grad_norm": 7.056100845336914,
      "learning_rate": 3.055555555555556e-06,
      "loss": 0.1651,
      "step": 3625
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.165300369262695,
      "learning_rate": 3e-06,
      "loss": 0.1484,
      "step": 3650
    },
    {
      "epoch": 1.58,
      "grad_norm": 6.445197582244873,
      "learning_rate": 2.944444444444445e-06,
      "loss": 0.1826,
      "step": 3675
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.774392127990723,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.1485,
      "step": 3700
    },
    {
      "epoch": 1.6,
      "grad_norm": 7.94968318939209,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.1479,
      "step": 3725
    },
    {
      "epoch": 1.61,
      "grad_norm": 10.860284805297852,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 0.189,
      "step": 3750
    },
    {
      "epoch": 1.62,
      "grad_norm": 6.700891494750977,
      "learning_rate": 2.7222222222222224e-06,
      "loss": 0.1619,
      "step": 3775
    },
    {
      "epoch": 1.63,
      "grad_norm": 7.410924911499023,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.1653,
      "step": 3800
    },
    {
      "epoch": 1.64,
      "grad_norm": 8.572102546691895,
      "learning_rate": 2.6111111111111113e-06,
      "loss": 0.1796,
      "step": 3825
    },
    {
      "epoch": 1.65,
      "grad_norm": 6.531749725341797,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 0.2131,
      "step": 3850
    },
    {
      "epoch": 1.66,
      "grad_norm": 10.115751266479492,
      "learning_rate": 2.5e-06,
      "loss": 0.176,
      "step": 3875
    },
    {
      "epoch": 1.67,
      "grad_norm": 8.534689903259277,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.1776,
      "step": 3900
    },
    {
      "epoch": 1.69,
      "grad_norm": 6.934342861175537,
      "learning_rate": 2.388888888888889e-06,
      "loss": 0.1721,
      "step": 3925
    },
    {
      "epoch": 1.7,
      "grad_norm": 9.225700378417969,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.1574,
      "step": 3950
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.1533966064453125,
      "learning_rate": 2.277777777777778e-06,
      "loss": 0.1818,
      "step": 3975
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.937667846679688,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.1896,
      "step": 4000
    },
    {
      "epoch": 1.72,
      "eval_loss": 0.2509143352508545,
      "eval_runtime": 2071.836,
      "eval_samples_per_second": 3.746,
      "eval_steps_per_second": 0.469,
      "eval_wer": 0.4858102397580179,
      "step": 4000
    },
    {
      "epoch": 1.73,
      "grad_norm": 6.895401954650879,
      "learning_rate": 2.166666666666667e-06,
      "loss": 0.1406,
      "step": 4025
    },
    {
      "epoch": 1.74,
      "grad_norm": 6.403042793273926,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 0.1682,
      "step": 4050
    },
    {
      "epoch": 1.75,
      "grad_norm": 9.697944641113281,
      "learning_rate": 2.0555555555555555e-06,
      "loss": 0.1784,
      "step": 4075
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.50242805480957,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.1618,
      "step": 4100
    },
    {
      "epoch": 1.77,
      "grad_norm": 8.654674530029297,
      "learning_rate": 1.944444444444445e-06,
      "loss": 0.1654,
      "step": 4125
    },
    {
      "epoch": 1.78,
      "grad_norm": 8.766165733337402,
      "learning_rate": 1.888888888888889e-06,
      "loss": 0.1439,
      "step": 4150
    },
    {
      "epoch": 1.79,
      "grad_norm": 5.462009906768799,
      "learning_rate": 1.8333333333333333e-06,
      "loss": 0.1604,
      "step": 4175
    },
    {
      "epoch": 1.8,
      "grad_norm": 9.384346961975098,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.1316,
      "step": 4200
    },
    {
      "epoch": 1.81,
      "grad_norm": 6.890562057495117,
      "learning_rate": 1.7222222222222224e-06,
      "loss": 0.1548,
      "step": 4225
    },
    {
      "epoch": 1.82,
      "grad_norm": 7.512774467468262,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.1419,
      "step": 4250
    },
    {
      "epoch": 1.84,
      "grad_norm": 9.803231239318848,
      "learning_rate": 1.6111111111111113e-06,
      "loss": 0.1555,
      "step": 4275
    },
    {
      "epoch": 1.85,
      "grad_norm": 6.2830376625061035,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.1463,
      "step": 4300
    },
    {
      "epoch": 1.86,
      "grad_norm": 7.151432037353516,
      "learning_rate": 1.5e-06,
      "loss": 0.1535,
      "step": 4325
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.497450828552246,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 0.1379,
      "step": 4350
    },
    {
      "epoch": 1.88,
      "grad_norm": 6.720860481262207,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 0.1765,
      "step": 4375
    },
    {
      "epoch": 1.89,
      "grad_norm": 6.73000955581665,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.1555,
      "step": 4400
    },
    {
      "epoch": 1.9,
      "grad_norm": 7.2651591300964355,
      "learning_rate": 1.2777777777777779e-06,
      "loss": 0.1409,
      "step": 4425
    },
    {
      "epoch": 1.91,
      "grad_norm": 5.468505859375,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 0.1449,
      "step": 4450
    },
    {
      "epoch": 1.92,
      "grad_norm": 7.236633777618408,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 0.1639,
      "step": 4475
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.03482437133789,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.1632,
      "step": 4500
    },
    {
      "epoch": 1.94,
      "grad_norm": 5.117453098297119,
      "learning_rate": 1.0555555555555557e-06,
      "loss": 0.1392,
      "step": 4525
    },
    {
      "epoch": 1.95,
      "grad_norm": 5.924282550811768,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.1448,
      "step": 4550
    },
    {
      "epoch": 1.96,
      "grad_norm": 11.092741012573242,
      "learning_rate": 9.444444444444445e-07,
      "loss": 0.1458,
      "step": 4575
    },
    {
      "epoch": 1.98,
      "grad_norm": 6.324784278869629,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.1509,
      "step": 4600
    },
    {
      "epoch": 1.99,
      "grad_norm": 7.572363376617432,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.1612,
      "step": 4625
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.5995659828186035,
      "learning_rate": 7.777777777777779e-07,
      "loss": 0.1412,
      "step": 4650
    },
    {
      "epoch": 2.01,
      "grad_norm": 4.05815315246582,
      "learning_rate": 7.222222222222222e-07,
      "loss": 0.1035,
      "step": 4675
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.89007568359375,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.0823,
      "step": 4700
    },
    {
      "epoch": 2.03,
      "grad_norm": 4.5809783935546875,
      "learning_rate": 6.111111111111112e-07,
      "loss": 0.1022,
      "step": 4725
    },
    {
      "epoch": 2.04,
      "grad_norm": 6.360114574432373,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.0845,
      "step": 4750
    },
    {
      "epoch": 2.05,
      "grad_norm": 4.002993583679199,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.0968,
      "step": 4775
    },
    {
      "epoch": 2.06,
      "grad_norm": 3.601863384246826,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.1006,
      "step": 4800
    },
    {
      "epoch": 2.07,
      "grad_norm": 4.200375556945801,
      "learning_rate": 3.8888888888888895e-07,
      "loss": 0.1003,
      "step": 4825
    },
    {
      "epoch": 2.08,
      "grad_norm": 5.6527180671691895,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.1016,
      "step": 4850
    },
    {
      "epoch": 2.09,
      "grad_norm": 6.326806545257568,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.0986,
      "step": 4875
    },
    {
      "epoch": 2.1,
      "grad_norm": 4.397176265716553,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.0925,
      "step": 4900
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.7298688888549805,
      "learning_rate": 1.6666666666666668e-07,
      "loss": 0.0757,
      "step": 4925
    },
    {
      "epoch": 2.13,
      "grad_norm": 4.760688304901123,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 0.0982,
      "step": 4950
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.136986255645752,
      "learning_rate": 5.555555555555556e-08,
      "loss": 0.08,
      "step": 4975
    },
    {
      "epoch": 2.15,
      "grad_norm": 3.902303695678711,
      "learning_rate": 0.0,
      "loss": 0.0933,
      "step": 5000
    },
    {
      "epoch": 2.15,
      "eval_loss": 0.24663741886615753,
      "eval_runtime": 2084.3454,
      "eval_samples_per_second": 3.723,
      "eval_steps_per_second": 0.466,
      "eval_wer": 0.4799608558338152,
      "step": 5000
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "total_flos": 1.153995298799616e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}