{
  "best_metric": 75.78173858661663,
  "best_model_checkpoint": "./whisper-small-finetune_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": 26.755067825317383,
      "learning_rate": 5.000000000000001e-07,
      "loss": 2.3039,
      "step": 25
    },
    {
      "epoch": 0.03,
      "grad_norm": 34.66984558105469,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.1708,
      "step": 50
    },
    {
      "epoch": 0.04,
      "grad_norm": 26.377073287963867,
      "learning_rate": 1.5e-06,
      "loss": 1.8591,
      "step": 75
    },
    {
      "epoch": 0.06,
      "grad_norm": 30.776378631591797,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.8478,
      "step": 100
    },
    {
      "epoch": 0.07,
      "grad_norm": 30.643896102905273,
      "learning_rate": 2.5e-06,
      "loss": 1.8421,
      "step": 125
    },
    {
      "epoch": 0.09,
      "grad_norm": 34.42327880859375,
      "learning_rate": 3e-06,
      "loss": 1.9252,
      "step": 150
    },
    {
      "epoch": 0.1,
      "grad_norm": 24.890520095825195,
      "learning_rate": 3.5e-06,
      "loss": 1.7428,
      "step": 175
    },
    {
      "epoch": 0.12,
      "grad_norm": 30.156938552856445,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.5537,
      "step": 200
    },
    {
      "epoch": 0.13,
      "grad_norm": 27.96819496154785,
      "learning_rate": 4.5e-06,
      "loss": 1.6786,
      "step": 225
    },
    {
      "epoch": 0.15,
      "grad_norm": 28.565181732177734,
      "learning_rate": 5e-06,
      "loss": 1.6344,
      "step": 250
    },
    {
      "epoch": 0.16,
      "grad_norm": 24.439979553222656,
      "learning_rate": 5.500000000000001e-06,
      "loss": 1.7475,
      "step": 275
    },
    {
      "epoch": 0.18,
      "grad_norm": 19.079715728759766,
      "learning_rate": 6e-06,
      "loss": 1.5635,
      "step": 300
    },
    {
      "epoch": 0.19,
      "grad_norm": 31.042789459228516,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 1.6371,
      "step": 325
    },
    {
      "epoch": 0.21,
      "grad_norm": 21.58063316345215,
      "learning_rate": 7e-06,
      "loss": 1.5827,
      "step": 350
    },
    {
      "epoch": 0.22,
      "grad_norm": 20.27756118774414,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.6198,
      "step": 375
    },
    {
      "epoch": 0.24,
      "grad_norm": 29.146913528442383,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.462,
      "step": 400
    },
    {
      "epoch": 0.25,
      "grad_norm": 27.70607566833496,
      "learning_rate": 8.5e-06,
      "loss": 1.6481,
      "step": 425
    },
    {
      "epoch": 0.27,
      "grad_norm": 24.769601821899414,
      "learning_rate": 9e-06,
      "loss": 1.5338,
      "step": 450
    },
    {
      "epoch": 0.28,
      "grad_norm": 28.691078186035156,
      "learning_rate": 9.5e-06,
      "loss": 1.469,
      "step": 475
    },
    {
      "epoch": 0.3,
      "grad_norm": 27.30621337890625,
      "learning_rate": 1e-05,
      "loss": 1.5568,
      "step": 500
    },
    {
      "epoch": 0.31,
      "grad_norm": 23.094369888305664,
      "learning_rate": 9.944444444444445e-06,
      "loss": 1.4752,
      "step": 525
    },
    {
      "epoch": 0.33,
      "grad_norm": 25.179765701293945,
      "learning_rate": 9.88888888888889e-06,
      "loss": 1.564,
      "step": 550
    },
    {
      "epoch": 0.34,
      "grad_norm": 23.27145004272461,
      "learning_rate": 9.833333333333333e-06,
      "loss": 1.3967,
      "step": 575
    },
    {
      "epoch": 0.36,
      "grad_norm": 25.917579650878906,
      "learning_rate": 9.777777777777779e-06,
      "loss": 1.4498,
      "step": 600
    },
    {
      "epoch": 0.37,
      "grad_norm": 17.447675704956055,
      "learning_rate": 9.722222222222223e-06,
      "loss": 1.4077,
      "step": 625
    },
    {
      "epoch": 0.39,
      "grad_norm": 23.504993438720703,
      "learning_rate": 9.666666666666667e-06,
      "loss": 1.4868,
      "step": 650
    },
    {
      "epoch": 0.4,
      "grad_norm": 20.719358444213867,
      "learning_rate": 9.611111111111112e-06,
      "loss": 1.3703,
      "step": 675
    },
    {
      "epoch": 0.42,
      "grad_norm": 19.470386505126953,
      "learning_rate": 9.555555555555556e-06,
      "loss": 1.4705,
      "step": 700
    },
    {
      "epoch": 0.43,
      "grad_norm": 21.750486373901367,
      "learning_rate": 9.5e-06,
      "loss": 1.5277,
      "step": 725
    },
    {
      "epoch": 0.45,
      "grad_norm": 23.778789520263672,
      "learning_rate": 9.444444444444445e-06,
      "loss": 1.5371,
      "step": 750
    },
    {
      "epoch": 0.46,
      "grad_norm": 23.14359474182129,
      "learning_rate": 9.38888888888889e-06,
      "loss": 1.5101,
      "step": 775
    },
    {
      "epoch": 0.48,
      "grad_norm": 25.576568603515625,
      "learning_rate": 9.333333333333334e-06,
      "loss": 1.4284,
      "step": 800
    },
    {
      "epoch": 0.49,
      "grad_norm": 26.62574005126953,
      "learning_rate": 9.277777777777778e-06,
      "loss": 1.4605,
      "step": 825
    },
    {
      "epoch": 0.51,
      "grad_norm": 18.61768341064453,
      "learning_rate": 9.222222222222224e-06,
      "loss": 1.4202,
      "step": 850
    },
    {
      "epoch": 0.52,
      "grad_norm": 26.829648971557617,
      "learning_rate": 9.166666666666666e-06,
      "loss": 1.3163,
      "step": 875
    },
    {
      "epoch": 0.54,
      "grad_norm": 21.73237419128418,
      "learning_rate": 9.111111111111112e-06,
      "loss": 1.4229,
      "step": 900
    },
    {
      "epoch": 0.55,
      "grad_norm": 26.230100631713867,
      "learning_rate": 9.055555555555556e-06,
      "loss": 1.4828,
      "step": 925
    },
    {
      "epoch": 0.57,
      "grad_norm": 20.698078155517578,
      "learning_rate": 9e-06,
      "loss": 1.5082,
      "step": 950
    },
    {
      "epoch": 0.58,
      "grad_norm": 24.733755111694336,
      "learning_rate": 8.944444444444446e-06,
      "loss": 1.291,
      "step": 975
    },
    {
      "epoch": 0.6,
      "grad_norm": 22.547061920166016,
      "learning_rate": 8.888888888888888e-06,
      "loss": 1.3906,
      "step": 1000
    },
    {
      "epoch": 0.6,
      "eval_cer": 57.395050172609885,
      "eval_loss": 1.3834500312805176,
      "eval_runtime": 827.1285,
      "eval_samples_per_second": 4.036,
      "eval_steps_per_second": 0.505,
      "eval_wer": 87.96643735668125,
      "step": 1000
    },
    {
      "epoch": 0.61,
      "grad_norm": 24.679819107055664,
      "learning_rate": 8.833333333333334e-06,
      "loss": 1.3797,
      "step": 1025
    },
    {
      "epoch": 0.63,
      "grad_norm": 22.122909545898438,
      "learning_rate": 8.777777777777778e-06,
      "loss": 1.4223,
      "step": 1050
    },
    {
      "epoch": 0.64,
      "grad_norm": 17.617143630981445,
      "learning_rate": 8.722222222222224e-06,
      "loss": 1.3753,
      "step": 1075
    },
    {
      "epoch": 0.66,
      "grad_norm": 18.17972755432129,
      "learning_rate": 8.666666666666668e-06,
      "loss": 1.2559,
      "step": 1100
    },
    {
      "epoch": 0.67,
      "grad_norm": 26.944639205932617,
      "learning_rate": 8.611111111111112e-06,
      "loss": 1.3918,
      "step": 1125
    },
    {
      "epoch": 0.69,
      "grad_norm": 20.75006675720215,
      "learning_rate": 8.555555555555556e-06,
      "loss": 1.4469,
      "step": 1150
    },
    {
      "epoch": 0.7,
      "grad_norm": 25.37113380432129,
      "learning_rate": 8.5e-06,
      "loss": 1.3639,
      "step": 1175
    },
    {
      "epoch": 0.72,
      "grad_norm": 16.128446578979492,
      "learning_rate": 8.444444444444446e-06,
      "loss": 1.3684,
      "step": 1200
    },
    {
      "epoch": 0.73,
      "grad_norm": 22.034616470336914,
      "learning_rate": 8.38888888888889e-06,
      "loss": 1.2884,
      "step": 1225
    },
    {
      "epoch": 0.75,
      "grad_norm": 18.15462875366211,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.5307,
      "step": 1250
    },
    {
      "epoch": 0.76,
      "grad_norm": 27.089126586914062,
      "learning_rate": 8.277777777777778e-06,
      "loss": 1.2426,
      "step": 1275
    },
    {
      "epoch": 0.78,
      "grad_norm": 17.21516990661621,
      "learning_rate": 8.222222222222222e-06,
      "loss": 1.2479,
      "step": 1300
    },
    {
      "epoch": 0.79,
      "grad_norm": 21.62128257751465,
      "learning_rate": 8.166666666666668e-06,
      "loss": 1.3416,
      "step": 1325
    },
    {
      "epoch": 0.81,
      "grad_norm": 18.543970108032227,
      "learning_rate": 8.111111111111112e-06,
      "loss": 1.2417,
      "step": 1350
    },
    {
      "epoch": 0.82,
      "grad_norm": 25.98469352722168,
      "learning_rate": 8.055555555555557e-06,
      "loss": 1.4074,
      "step": 1375
    },
    {
      "epoch": 0.84,
      "grad_norm": 18.895614624023438,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.378,
      "step": 1400
    },
    {
      "epoch": 0.85,
      "grad_norm": 21.452943801879883,
      "learning_rate": 7.944444444444445e-06,
      "loss": 1.3844,
      "step": 1425
    },
    {
      "epoch": 0.87,
      "grad_norm": 21.81563949584961,
      "learning_rate": 7.88888888888889e-06,
      "loss": 1.3518,
      "step": 1450
    },
    {
      "epoch": 0.88,
      "grad_norm": 20.642499923706055,
      "learning_rate": 7.833333333333333e-06,
      "loss": 1.2924,
      "step": 1475
    },
    {
      "epoch": 0.9,
      "grad_norm": 33.33131408691406,
      "learning_rate": 7.77777777777778e-06,
      "loss": 1.2999,
      "step": 1500
    },
    {
      "epoch": 0.91,
      "grad_norm": 16.468210220336914,
      "learning_rate": 7.722222222222223e-06,
      "loss": 1.2118,
      "step": 1525
    },
    {
      "epoch": 0.93,
      "grad_norm": 19.231273651123047,
      "learning_rate": 7.666666666666667e-06,
      "loss": 1.3415,
      "step": 1550
    },
    {
      "epoch": 0.94,
      "grad_norm": 24.304243087768555,
      "learning_rate": 7.611111111111111e-06,
      "loss": 1.3811,
      "step": 1575
    },
    {
      "epoch": 0.96,
      "grad_norm": 20.244470596313477,
      "learning_rate": 7.555555555555556e-06,
      "loss": 1.2991,
      "step": 1600
    },
    {
      "epoch": 0.97,
      "grad_norm": 19.40199851989746,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.3042,
      "step": 1625
    },
    {
      "epoch": 0.99,
      "grad_norm": 18.921825408935547,
      "learning_rate": 7.444444444444445e-06,
      "loss": 1.2901,
      "step": 1650
    },
    {
      "epoch": 1.0,
      "grad_norm": 16.50575065612793,
      "learning_rate": 7.38888888888889e-06,
      "loss": 1.2142,
      "step": 1675
    },
    {
      "epoch": 1.02,
      "grad_norm": 14.586038589477539,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.9015,
      "step": 1700
    },
    {
      "epoch": 1.03,
      "grad_norm": 20.73207664489746,
      "learning_rate": 7.277777777777778e-06,
      "loss": 1.0864,
      "step": 1725
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.946000099182129,
      "learning_rate": 7.222222222222223e-06,
      "loss": 0.9158,
      "step": 1750
    },
    {
      "epoch": 1.06,
      "grad_norm": 15.355812072753906,
      "learning_rate": 7.166666666666667e-06,
      "loss": 0.8193,
      "step": 1775
    },
    {
      "epoch": 1.08,
      "grad_norm": 15.703187942504883,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.8603,
      "step": 1800
    },
    {
      "epoch": 1.09,
      "grad_norm": 17.89627456665039,
      "learning_rate": 7.055555555555557e-06,
      "loss": 0.9686,
      "step": 1825
    },
    {
      "epoch": 1.11,
      "grad_norm": 19.415363311767578,
      "learning_rate": 7e-06,
      "loss": 1.0957,
      "step": 1850
    },
    {
      "epoch": 1.12,
      "grad_norm": 15.288445472717285,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.9709,
      "step": 1875
    },
    {
      "epoch": 1.14,
      "grad_norm": 18.04073143005371,
      "learning_rate": 6.88888888888889e-06,
      "loss": 0.9127,
      "step": 1900
    },
    {
      "epoch": 1.15,
      "grad_norm": 17.967145919799805,
      "learning_rate": 6.833333333333334e-06,
      "loss": 0.8762,
      "step": 1925
    },
    {
      "epoch": 1.17,
      "grad_norm": 22.560022354125977,
      "learning_rate": 6.777777777777779e-06,
      "loss": 1.0106,
      "step": 1950
    },
    {
      "epoch": 1.18,
      "grad_norm": 16.264883041381836,
      "learning_rate": 6.7222222222222235e-06,
      "loss": 0.919,
      "step": 1975
    },
    {
      "epoch": 1.2,
      "grad_norm": 15.793697357177734,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9716,
      "step": 2000
    },
    {
      "epoch": 1.2,
      "eval_cer": 54.57994907253641,
      "eval_loss": 1.2673263549804688,
      "eval_runtime": 825.2901,
      "eval_samples_per_second": 4.045,
      "eval_steps_per_second": 0.506,
      "eval_wer": 81.23827392120076,
      "step": 2000
    },
    {
      "epoch": 1.21,
      "grad_norm": 17.40117645263672,
      "learning_rate": 6.6111111111111115e-06,
      "loss": 0.9739,
      "step": 2025
    },
    {
      "epoch": 1.23,
      "grad_norm": 22.668455123901367,
      "learning_rate": 6.555555555555556e-06,
      "loss": 1.0057,
      "step": 2050
    },
    {
      "epoch": 1.24,
      "grad_norm": 21.214128494262695,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.9888,
      "step": 2075
    },
    {
      "epoch": 1.26,
      "grad_norm": 19.409332275390625,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.9321,
      "step": 2100
    },
    {
      "epoch": 1.27,
      "grad_norm": 19.678197860717773,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 0.973,
      "step": 2125
    },
    {
      "epoch": 1.29,
      "grad_norm": 16.86955451965332,
      "learning_rate": 6.333333333333333e-06,
      "loss": 1.0285,
      "step": 2150
    },
    {
      "epoch": 1.3,
      "grad_norm": 15.891237258911133,
      "learning_rate": 6.277777777777778e-06,
      "loss": 0.9538,
      "step": 2175
    },
    {
      "epoch": 1.32,
      "grad_norm": 19.17290687561035,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.9328,
      "step": 2200
    },
    {
      "epoch": 1.33,
      "grad_norm": 16.7691593170166,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.8896,
      "step": 2225
    },
    {
      "epoch": 1.35,
      "grad_norm": 18.572256088256836,
      "learning_rate": 6.111111111111112e-06,
      "loss": 1.0089,
      "step": 2250
    },
    {
      "epoch": 1.36,
      "grad_norm": 17.122636795043945,
      "learning_rate": 6.055555555555555e-06,
      "loss": 1.0513,
      "step": 2275
    },
    {
      "epoch": 1.38,
      "grad_norm": 16.53435707092285,
      "learning_rate": 6e-06,
      "loss": 0.9443,
      "step": 2300
    },
    {
      "epoch": 1.39,
      "grad_norm": 20.76995086669922,
      "learning_rate": 5.944444444444445e-06,
      "loss": 1.0072,
      "step": 2325
    },
    {
      "epoch": 1.41,
      "grad_norm": 19.78141975402832,
      "learning_rate": 5.88888888888889e-06,
      "loss": 0.9116,
      "step": 2350
    },
    {
      "epoch": 1.42,
      "grad_norm": 18.738779067993164,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9244,
      "step": 2375
    },
    {
      "epoch": 1.44,
      "grad_norm": 16.18290901184082,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.8844,
      "step": 2400
    },
    {
      "epoch": 1.45,
      "grad_norm": 23.43638038635254,
      "learning_rate": 5.722222222222222e-06,
      "loss": 0.9252,
      "step": 2425
    },
    {
      "epoch": 1.47,
      "grad_norm": 18.08245849609375,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.9689,
      "step": 2450
    },
    {
      "epoch": 1.48,
      "grad_norm": 20.146072387695312,
      "learning_rate": 5.611111111111112e-06,
      "loss": 0.9982,
      "step": 2475
    },
    {
      "epoch": 1.5,
      "grad_norm": 19.40115737915039,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.8793,
      "step": 2500
    },
    {
      "epoch": 1.51,
      "grad_norm": 19.48564910888672,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.9001,
      "step": 2525
    },
    {
      "epoch": 1.53,
      "grad_norm": 19.529861450195312,
      "learning_rate": 5.444444444444445e-06,
      "loss": 1.0249,
      "step": 2550
    },
    {
      "epoch": 1.54,
      "grad_norm": 19.736713409423828,
      "learning_rate": 5.388888888888889e-06,
      "loss": 0.9182,
      "step": 2575
    },
    {
      "epoch": 1.56,
      "grad_norm": 16.709518432617188,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.9904,
      "step": 2600
    },
    {
      "epoch": 1.57,
      "grad_norm": 15.45124340057373,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 1.0358,
      "step": 2625
    },
    {
      "epoch": 1.59,
      "grad_norm": 11.870710372924805,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 0.9011,
      "step": 2650
    },
    {
      "epoch": 1.6,
      "grad_norm": 17.464656829833984,
      "learning_rate": 5.1666666666666675e-06,
      "loss": 0.918,
      "step": 2675
    },
    {
      "epoch": 1.62,
      "grad_norm": 24.91024398803711,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.927,
      "step": 2700
    },
    {
      "epoch": 1.63,
      "grad_norm": 18.0244140625,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 0.9138,
      "step": 2725
    },
    {
      "epoch": 1.65,
      "grad_norm": 18.960786819458008,
      "learning_rate": 5e-06,
      "loss": 0.9982,
      "step": 2750
    },
    {
      "epoch": 1.66,
      "grad_norm": 17.35990333557129,
      "learning_rate": 4.944444444444445e-06,
      "loss": 0.9706,
      "step": 2775
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.708779335021973,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.9477,
      "step": 2800
    },
    {
      "epoch": 1.69,
      "grad_norm": 19.692068099975586,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.9244,
      "step": 2825
    },
    {
      "epoch": 1.71,
      "grad_norm": 15.64211654663086,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.9989,
      "step": 2850
    },
    {
      "epoch": 1.72,
      "grad_norm": 24.154022216796875,
      "learning_rate": 4.722222222222222e-06,
      "loss": 0.9411,
      "step": 2875
    },
    {
      "epoch": 1.74,
      "grad_norm": 26.374160766601562,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.9849,
      "step": 2900
    },
    {
      "epoch": 1.75,
      "grad_norm": 16.004297256469727,
      "learning_rate": 4.611111111111112e-06,
      "loss": 1.031,
      "step": 2925
    },
    {
      "epoch": 1.77,
      "grad_norm": 19.26837730407715,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.9211,
      "step": 2950
    },
    {
      "epoch": 1.78,
      "grad_norm": 15.54520320892334,
      "learning_rate": 4.5e-06,
      "loss": 0.9234,
      "step": 2975
    },
    {
      "epoch": 1.8,
      "grad_norm": 21.567298889160156,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.9518,
      "step": 3000
    },
    {
      "epoch": 1.8,
      "eval_cer": 63.4530028080558,
      "eval_loss": 1.204471230506897,
      "eval_runtime": 857.8987,
      "eval_samples_per_second": 3.891,
      "eval_steps_per_second": 0.487,
      "eval_wer": 85.61079841567646,
      "step": 3000
    },
    {
      "epoch": 1.81,
      "grad_norm": 15.287321090698242,
      "learning_rate": 4.388888888888889e-06,
      "loss": 0.9597,
      "step": 3025
    },
    {
      "epoch": 1.83,
      "grad_norm": 19.927249908447266,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.9353,
      "step": 3050
    },
    {
      "epoch": 1.84,
      "grad_norm": 23.143543243408203,
      "learning_rate": 4.277777777777778e-06,
      "loss": 1.0893,
      "step": 3075
    },
    {
      "epoch": 1.86,
      "grad_norm": 16.31475257873535,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.8705,
      "step": 3100
    },
    {
      "epoch": 1.87,
      "grad_norm": 11.8856840133667,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.8493,
      "step": 3125
    },
    {
      "epoch": 1.89,
      "grad_norm": 14.792439460754395,
      "learning_rate": 4.111111111111111e-06,
      "loss": 0.9663,
      "step": 3150
    },
    {
      "epoch": 1.9,
      "grad_norm": 13.951410293579102,
      "learning_rate": 4.055555555555556e-06,
      "loss": 0.9282,
      "step": 3175
    },
    {
      "epoch": 1.92,
      "grad_norm": 19.062644958496094,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.8443,
      "step": 3200
    },
    {
      "epoch": 1.93,
      "grad_norm": 18.890697479248047,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.97,
      "step": 3225
    },
    {
      "epoch": 1.95,
      "grad_norm": 16.668237686157227,
      "learning_rate": 3.88888888888889e-06,
      "loss": 0.8924,
      "step": 3250
    },
    {
      "epoch": 1.96,
      "grad_norm": 20.245948791503906,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.983,
      "step": 3275
    },
    {
      "epoch": 1.98,
      "grad_norm": 17.0203800201416,
      "learning_rate": 3.777777777777778e-06,
      "loss": 1.0049,
      "step": 3300
    },
    {
      "epoch": 1.99,
      "grad_norm": 23.726200103759766,
      "learning_rate": 3.7222222222222225e-06,
      "loss": 0.8701,
      "step": 3325
    },
    {
      "epoch": 2.01,
      "grad_norm": 9.839333534240723,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.7549,
      "step": 3350
    },
    {
      "epoch": 2.02,
      "grad_norm": 16.390169143676758,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 0.603,
      "step": 3375
    },
    {
      "epoch": 2.04,
      "grad_norm": 16.594236373901367,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.6646,
      "step": 3400
    },
    {
      "epoch": 2.05,
      "grad_norm": 18.335918426513672,
      "learning_rate": 3.5e-06,
      "loss": 0.7228,
      "step": 3425
    },
    {
      "epoch": 2.07,
      "grad_norm": 14.09089183807373,
      "learning_rate": 3.444444444444445e-06,
      "loss": 0.6843,
      "step": 3450
    },
    {
      "epoch": 2.08,
      "grad_norm": 19.89991569519043,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.6836,
      "step": 3475
    },
    {
      "epoch": 2.1,
      "grad_norm": 16.895814895629883,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.628,
      "step": 3500
    },
    {
      "epoch": 2.11,
      "grad_norm": 17.576400756835938,
      "learning_rate": 3.277777777777778e-06,
      "loss": 0.6666,
      "step": 3525
    },
    {
      "epoch": 2.13,
      "grad_norm": 15.989606857299805,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 0.6256,
      "step": 3550
    },
    {
      "epoch": 2.14,
      "grad_norm": 16.086654663085938,
      "learning_rate": 3.1666666666666667e-06,
      "loss": 0.6641,
      "step": 3575
    },
    {
      "epoch": 2.16,
      "grad_norm": 14.707154273986816,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.62,
      "step": 3600
    },
    {
      "epoch": 2.17,
      "grad_norm": 14.809082984924316,
      "learning_rate": 3.055555555555556e-06,
      "loss": 0.6564,
      "step": 3625
    },
    {
      "epoch": 2.19,
      "grad_norm": 15.021686553955078,
      "learning_rate": 3e-06,
      "loss": 0.669,
      "step": 3650
    },
    {
      "epoch": 2.2,
      "grad_norm": 17.62252426147461,
      "learning_rate": 2.944444444444445e-06,
      "loss": 0.6823,
      "step": 3675
    },
    {
      "epoch": 2.22,
      "grad_norm": 15.664830207824707,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.695,
      "step": 3700
    },
    {
      "epoch": 2.23,
      "grad_norm": 8.63792610168457,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.6279,
      "step": 3725
    },
    {
      "epoch": 2.25,
      "grad_norm": 15.918222427368164,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 0.6861,
      "step": 3750
    },
    {
      "epoch": 2.26,
      "grad_norm": 14.93791389465332,
      "learning_rate": 2.7222222222222224e-06,
      "loss": 0.6068,
      "step": 3775
    },
    {
      "epoch": 2.28,
      "grad_norm": 17.324039459228516,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.6098,
      "step": 3800
    },
    {
      "epoch": 2.29,
      "grad_norm": 17.186784744262695,
      "learning_rate": 2.6111111111111113e-06,
      "loss": 0.7561,
      "step": 3825
    },
    {
      "epoch": 2.31,
      "grad_norm": 16.359756469726562,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 0.5851,
      "step": 3850
    },
    {
      "epoch": 2.32,
      "grad_norm": 16.602548599243164,
      "learning_rate": 2.5e-06,
      "loss": 0.64,
      "step": 3875
    },
    {
      "epoch": 2.34,
      "grad_norm": 15.252476692199707,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.6555,
      "step": 3900
    },
    {
      "epoch": 2.35,
      "grad_norm": 15.010741233825684,
      "learning_rate": 2.388888888888889e-06,
      "loss": 0.5832,
      "step": 3925
    },
    {
      "epoch": 2.37,
      "grad_norm": 8.625840187072754,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.5749,
      "step": 3950
    },
    {
      "epoch": 2.38,
      "grad_norm": 11.733368873596191,
      "learning_rate": 2.277777777777778e-06,
      "loss": 0.6156,
      "step": 3975
    },
    {
      "epoch": 2.4,
      "grad_norm": 16.896711349487305,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.6477,
      "step": 4000
    },
    {
      "epoch": 2.4,
      "eval_cer": 55.91453043066921,
      "eval_loss": 1.2121813297271729,
      "eval_runtime": 829.4327,
      "eval_samples_per_second": 4.024,
      "eval_steps_per_second": 0.504,
      "eval_wer": 79.73733583489681,
      "step": 4000
    },
    {
      "epoch": 2.41,
      "grad_norm": 16.266002655029297,
      "learning_rate": 2.166666666666667e-06,
      "loss": 0.6686,
      "step": 4025
    },
    {
      "epoch": 2.43,
      "grad_norm": 14.70505428314209,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 0.6445,
      "step": 4050
    },
    {
      "epoch": 2.44,
      "grad_norm": 18.52985191345215,
      "learning_rate": 2.0555555555555555e-06,
      "loss": 0.6195,
      "step": 4075
    },
    {
      "epoch": 2.46,
      "grad_norm": 13.347681045532227,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.6454,
      "step": 4100
    },
    {
      "epoch": 2.47,
      "grad_norm": 19.63964080810547,
      "learning_rate": 1.944444444444445e-06,
      "loss": 0.6106,
      "step": 4125
    },
    {
      "epoch": 2.49,
      "grad_norm": 20.261411666870117,
      "learning_rate": 1.888888888888889e-06,
      "loss": 0.6703,
      "step": 4150
    },
    {
      "epoch": 2.5,
      "grad_norm": 14.775351524353027,
      "learning_rate": 1.8333333333333333e-06,
      "loss": 0.6232,
      "step": 4175
    },
    {
      "epoch": 2.52,
      "grad_norm": 15.170753479003906,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.6405,
      "step": 4200
    },
    {
      "epoch": 2.53,
      "grad_norm": 17.451885223388672,
      "learning_rate": 1.7222222222222224e-06,
      "loss": 0.668,
      "step": 4225
    },
    {
      "epoch": 2.55,
      "grad_norm": 16.877351760864258,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.7075,
      "step": 4250
    },
    {
      "epoch": 2.56,
      "grad_norm": 17.570926666259766,
      "learning_rate": 1.6111111111111113e-06,
      "loss": 0.7038,
      "step": 4275
    },
    {
      "epoch": 2.58,
      "grad_norm": 11.851263999938965,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.5578,
      "step": 4300
    },
    {
      "epoch": 2.59,
      "grad_norm": 15.071412086486816,
      "learning_rate": 1.5e-06,
      "loss": 0.6107,
      "step": 4325
    },
    {
      "epoch": 2.61,
      "grad_norm": 13.913065910339355,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 0.5909,
      "step": 4350
    },
    {
      "epoch": 2.62,
      "grad_norm": 19.536226272583008,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 0.6972,
      "step": 4375
    },
    {
      "epoch": 2.64,
      "grad_norm": 12.989717483520508,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.5659,
      "step": 4400
    },
    {
      "epoch": 2.65,
      "grad_norm": 12.693081855773926,
      "learning_rate": 1.2777777777777779e-06,
      "loss": 0.5787,
      "step": 4425
    },
    {
      "epoch": 2.67,
      "grad_norm": 17.102140426635742,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 0.6361,
      "step": 4450
    },
    {
      "epoch": 2.68,
      "grad_norm": 13.896649360656738,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 0.6509,
      "step": 4475
    },
    {
      "epoch": 2.7,
      "grad_norm": 22.483936309814453,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.6536,
      "step": 4500
    },
    {
      "epoch": 2.71,
      "grad_norm": 15.584601402282715,
      "learning_rate": 1.0555555555555557e-06,
      "loss": 0.5859,
      "step": 4525
    },
    {
      "epoch": 2.73,
      "grad_norm": 18.10873794555664,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.6344,
      "step": 4550
    },
    {
      "epoch": 2.74,
      "grad_norm": 11.653306007385254,
      "learning_rate": 9.444444444444445e-07,
      "loss": 0.6182,
      "step": 4575
    },
    {
      "epoch": 2.76,
      "grad_norm": 10.643468856811523,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.6861,
      "step": 4600
    },
    {
      "epoch": 2.77,
      "grad_norm": 14.426728248596191,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.6564,
      "step": 4625
    },
    {
      "epoch": 2.79,
      "grad_norm": 17.03219223022461,
      "learning_rate": 7.777777777777779e-07,
      "loss": 0.6172,
      "step": 4650
    },
    {
      "epoch": 2.8,
      "grad_norm": 17.09849739074707,
      "learning_rate": 7.222222222222222e-07,
      "loss": 0.7349,
      "step": 4675
    },
    {
      "epoch": 2.82,
      "grad_norm": 12.162270545959473,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.5356,
      "step": 4700
    },
    {
      "epoch": 2.83,
      "grad_norm": 14.960759162902832,
      "learning_rate": 6.111111111111112e-07,
      "loss": 0.5448,
      "step": 4725
    },
    {
      "epoch": 2.85,
      "grad_norm": 16.547992706298828,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.6641,
      "step": 4750
    },
    {
      "epoch": 2.86,
      "grad_norm": 14.882378578186035,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.508,
      "step": 4775
    },
    {
      "epoch": 2.88,
      "grad_norm": 19.316282272338867,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.5793,
      "step": 4800
    },
    {
      "epoch": 2.89,
      "grad_norm": 15.735701560974121,
      "learning_rate": 3.8888888888888895e-07,
      "loss": 0.6396,
      "step": 4825
    },
    {
      "epoch": 2.91,
      "grad_norm": 16.93009376525879,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.5591,
      "step": 4850
    },
    {
      "epoch": 2.92,
      "grad_norm": 14.681782722473145,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.6488,
      "step": 4875
    },
    {
      "epoch": 2.94,
      "grad_norm": 16.71626091003418,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.633,
      "step": 4900
    },
    {
      "epoch": 2.95,
      "grad_norm": 20.87442398071289,
      "learning_rate": 1.6666666666666668e-07,
      "loss": 0.6695,
      "step": 4925
    },
    {
      "epoch": 2.97,
      "grad_norm": 17.04303741455078,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 0.5856,
      "step": 4950
    },
    {
      "epoch": 2.98,
      "grad_norm": 14.400847434997559,
      "learning_rate": 5.555555555555556e-08,
      "loss": 0.5595,
      "step": 4975
    },
    {
      "epoch": 3.0,
      "grad_norm": 12.768998146057129,
      "learning_rate": 0.0,
      "loss": 0.6857,
      "step": 5000
    },
    {
      "epoch": 3.0,
      "eval_cer": 48.94773392916453,
      "eval_loss": 1.1982322931289673,
      "eval_runtime": 797.5519,
      "eval_samples_per_second": 4.185,
      "eval_steps_per_second": 0.524,
      "eval_wer": 75.78173858661663,
      "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
}