{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 22.52252252252252,
  "eval_steps": 1000,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.11,
      "grad_norm": 29.43235206604004,
      "learning_rate": 4.000000000000001e-06,
      "loss": 11.9112,
      "step": 25
    },
    {
      "epoch": 0.23,
      "grad_norm": 12.577169418334961,
      "learning_rate": 9e-06,
      "loss": 5.9608,
      "step": 50
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.2483415603637695,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 2.7899,
      "step": 75
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.498162746429443,
      "learning_rate": 1.9e-05,
      "loss": 1.9344,
      "step": 100
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.876449584960938,
      "learning_rate": 2.4e-05,
      "loss": 1.1848,
      "step": 125
    },
    {
      "epoch": 0.68,
      "grad_norm": 6.835580825805664,
      "learning_rate": 2.9e-05,
      "loss": 0.7884,
      "step": 150
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.131637573242188,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.6148,
      "step": 175
    },
    {
      "epoch": 0.9,
      "grad_norm": 3.9987149238586426,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.5234,
      "step": 200
    },
    {
      "epoch": 1.01,
      "grad_norm": 3.6429152488708496,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.453,
      "step": 225
    },
    {
      "epoch": 1.13,
      "grad_norm": 4.594738483428955,
      "learning_rate": 4.9e-05,
      "loss": 0.3913,
      "step": 250
    },
    {
      "epoch": 1.24,
      "grad_norm": 4.02340841293335,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.3729,
      "step": 275
    },
    {
      "epoch": 1.35,
      "grad_norm": 4.27403450012207,
      "learning_rate": 5.9e-05,
      "loss": 0.3544,
      "step": 300
    },
    {
      "epoch": 1.46,
      "grad_norm": 3.860103130340576,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.3225,
      "step": 325
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.948971748352051,
      "learning_rate": 6.9e-05,
      "loss": 0.3358,
      "step": 350
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.851400852203369,
      "learning_rate": 7.4e-05,
      "loss": 0.3143,
      "step": 375
    },
    {
      "epoch": 1.8,
      "grad_norm": 3.7811708450317383,
      "learning_rate": 7.900000000000001e-05,
      "loss": 0.2908,
      "step": 400
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.640065908432007,
      "learning_rate": 8.4e-05,
      "loss": 0.3066,
      "step": 425
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.669788122177124,
      "learning_rate": 8.900000000000001e-05,
      "loss": 0.2659,
      "step": 450
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.3060202598571777,
      "learning_rate": 9.4e-05,
      "loss": 0.2312,
      "step": 475
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.385368824005127,
      "learning_rate": 9.900000000000001e-05,
      "loss": 0.2275,
      "step": 500
    },
    {
      "epoch": 2.36,
      "grad_norm": 2.557762861251831,
      "learning_rate": 9.955555555555556e-05,
      "loss": 0.2424,
      "step": 525
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.552363872528076,
      "learning_rate": 9.900000000000001e-05,
      "loss": 0.2194,
      "step": 550
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.041868209838867,
      "learning_rate": 9.844444444444444e-05,
      "loss": 0.235,
      "step": 575
    },
    {
      "epoch": 2.7,
      "grad_norm": 2.3052070140838623,
      "learning_rate": 9.78888888888889e-05,
      "loss": 0.227,
      "step": 600
    },
    {
      "epoch": 2.82,
      "grad_norm": 2.061685562133789,
      "learning_rate": 9.733333333333335e-05,
      "loss": 0.2054,
      "step": 625
    },
    {
      "epoch": 2.93,
      "grad_norm": 2.4055511951446533,
      "learning_rate": 9.677777777777778e-05,
      "loss": 0.218,
      "step": 650
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.5952019691467285,
      "learning_rate": 9.622222222222222e-05,
      "loss": 0.1879,
      "step": 675
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.5567409992218018,
      "learning_rate": 9.566666666666667e-05,
      "loss": 0.1524,
      "step": 700
    },
    {
      "epoch": 3.27,
      "grad_norm": 2.305778741836548,
      "learning_rate": 9.511111111111112e-05,
      "loss": 0.1553,
      "step": 725
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.9750478267669678,
      "learning_rate": 9.455555555555556e-05,
      "loss": 0.1538,
      "step": 750
    },
    {
      "epoch": 3.49,
      "grad_norm": 2.064730405807495,
      "learning_rate": 9.4e-05,
      "loss": 0.158,
      "step": 775
    },
    {
      "epoch": 3.6,
      "grad_norm": 2.288181781768799,
      "learning_rate": 9.344444444444444e-05,
      "loss": 0.1488,
      "step": 800
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.7010929584503174,
      "learning_rate": 9.28888888888889e-05,
      "loss": 0.1491,
      "step": 825
    },
    {
      "epoch": 3.83,
      "grad_norm": 1.653943657875061,
      "learning_rate": 9.233333333333333e-05,
      "loss": 0.149,
      "step": 850
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.655405044555664,
      "learning_rate": 9.177777777777778e-05,
      "loss": 0.1505,
      "step": 875
    },
    {
      "epoch": 4.05,
      "grad_norm": 1.766348958015442,
      "learning_rate": 9.122222222222223e-05,
      "loss": 0.122,
      "step": 900
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.1792188882827759,
      "learning_rate": 9.066666666666667e-05,
      "loss": 0.0985,
      "step": 925
    },
    {
      "epoch": 4.28,
      "grad_norm": 1.3778753280639648,
      "learning_rate": 9.011111111111111e-05,
      "loss": 0.0994,
      "step": 950
    },
    {
      "epoch": 4.39,
      "grad_norm": 1.5239909887313843,
      "learning_rate": 8.955555555555556e-05,
      "loss": 0.1067,
      "step": 975
    },
    {
      "epoch": 4.5,
      "grad_norm": 1.2479132413864136,
      "learning_rate": 8.900000000000001e-05,
      "loss": 0.1035,
      "step": 1000
    },
    {
      "epoch": 4.5,
      "eval_loss": 0.30151915550231934,
      "eval_runtime": 1360.3971,
      "eval_samples_per_second": 2.296,
      "eval_steps_per_second": 0.072,
      "eval_wer": 0.3249633006258209,
      "step": 1000
    },
    {
      "epoch": 4.62,
      "grad_norm": 1.8175023794174194,
      "learning_rate": 8.844444444444445e-05,
      "loss": 0.1114,
      "step": 1025
    },
    {
      "epoch": 4.73,
      "grad_norm": 1.7170511484146118,
      "learning_rate": 8.78888888888889e-05,
      "loss": 0.1059,
      "step": 1050
    },
    {
      "epoch": 4.84,
      "grad_norm": 1.6984485387802124,
      "learning_rate": 8.733333333333333e-05,
      "loss": 0.1057,
      "step": 1075
    },
    {
      "epoch": 4.95,
      "grad_norm": 2.2961227893829346,
      "learning_rate": 8.677777777777778e-05,
      "loss": 0.1008,
      "step": 1100
    },
    {
      "epoch": 5.07,
      "grad_norm": 1.120532751083374,
      "learning_rate": 8.622222222222222e-05,
      "loss": 0.0785,
      "step": 1125
    },
    {
      "epoch": 5.18,
      "grad_norm": 2.4177873134613037,
      "learning_rate": 8.566666666666667e-05,
      "loss": 0.0587,
      "step": 1150
    },
    {
      "epoch": 5.29,
      "grad_norm": 1.216162085533142,
      "learning_rate": 8.511111111111112e-05,
      "loss": 0.0645,
      "step": 1175
    },
    {
      "epoch": 5.41,
      "grad_norm": 1.3237918615341187,
      "learning_rate": 8.455555555555556e-05,
      "loss": 0.0667,
      "step": 1200
    },
    {
      "epoch": 5.52,
      "grad_norm": 1.7567414045333862,
      "learning_rate": 8.4e-05,
      "loss": 0.0681,
      "step": 1225
    },
    {
      "epoch": 5.63,
      "grad_norm": 1.6587105989456177,
      "learning_rate": 8.344444444444445e-05,
      "loss": 0.0687,
      "step": 1250
    },
    {
      "epoch": 5.74,
      "grad_norm": 1.6716225147247314,
      "learning_rate": 8.28888888888889e-05,
      "loss": 0.0697,
      "step": 1275
    },
    {
      "epoch": 5.86,
      "grad_norm": 1.4438738822937012,
      "learning_rate": 8.233333333333333e-05,
      "loss": 0.0714,
      "step": 1300
    },
    {
      "epoch": 5.97,
      "grad_norm": 1.4982130527496338,
      "learning_rate": 8.177777777777778e-05,
      "loss": 0.0713,
      "step": 1325
    },
    {
      "epoch": 6.08,
      "grad_norm": 1.739702820777893,
      "learning_rate": 8.122222222222222e-05,
      "loss": 0.0457,
      "step": 1350
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.8361156582832336,
      "learning_rate": 8.066666666666667e-05,
      "loss": 0.0379,
      "step": 1375
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.9192413091659546,
      "learning_rate": 8.011111111111111e-05,
      "loss": 0.0437,
      "step": 1400
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.7922126054763794,
      "learning_rate": 7.955555555555556e-05,
      "loss": 0.0393,
      "step": 1425
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.7307619452476501,
      "learning_rate": 7.900000000000001e-05,
      "loss": 0.0398,
      "step": 1450
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.776818037033081,
      "learning_rate": 7.844444444444446e-05,
      "loss": 0.0417,
      "step": 1475
    },
    {
      "epoch": 6.76,
      "grad_norm": 1.0701009035110474,
      "learning_rate": 7.788888888888888e-05,
      "loss": 0.044,
      "step": 1500
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.9106244444847107,
      "learning_rate": 7.733333333333333e-05,
      "loss": 0.0449,
      "step": 1525
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.8038358688354492,
      "learning_rate": 7.677777777777778e-05,
      "loss": 0.0443,
      "step": 1550
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.8888857364654541,
      "learning_rate": 7.622222222222223e-05,
      "loss": 0.0295,
      "step": 1575
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.7159153819084167,
      "learning_rate": 7.566666666666667e-05,
      "loss": 0.0235,
      "step": 1600
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.9652548432350159,
      "learning_rate": 7.511111111111111e-05,
      "loss": 0.0276,
      "step": 1625
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.9230145215988159,
      "learning_rate": 7.455555555555556e-05,
      "loss": 0.0253,
      "step": 1650
    },
    {
      "epoch": 7.55,
      "grad_norm": 1.6814730167388916,
      "learning_rate": 7.4e-05,
      "loss": 0.029,
      "step": 1675
    },
    {
      "epoch": 7.66,
      "grad_norm": 1.036941409111023,
      "learning_rate": 7.344444444444445e-05,
      "loss": 0.0262,
      "step": 1700
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.6998699307441711,
      "learning_rate": 7.28888888888889e-05,
      "loss": 0.027,
      "step": 1725
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.7357670068740845,
      "learning_rate": 7.233333333333335e-05,
      "loss": 0.0254,
      "step": 1750
    },
    {
      "epoch": 8.0,
      "grad_norm": 2.0645382404327393,
      "learning_rate": 7.177777777777777e-05,
      "loss": 0.0299,
      "step": 1775
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.7657965421676636,
      "learning_rate": 7.122222222222222e-05,
      "loss": 0.0169,
      "step": 1800
    },
    {
      "epoch": 8.22,
      "grad_norm": 1.1174191236495972,
      "learning_rate": 7.066666666666667e-05,
      "loss": 0.0142,
      "step": 1825
    },
    {
      "epoch": 8.33,
      "grad_norm": 1.5552918910980225,
      "learning_rate": 7.011111111111112e-05,
      "loss": 0.0169,
      "step": 1850
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.6741360425949097,
      "learning_rate": 6.955555555555556e-05,
      "loss": 0.0171,
      "step": 1875
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.6866123080253601,
      "learning_rate": 6.9e-05,
      "loss": 0.017,
      "step": 1900
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.5000187754631042,
      "learning_rate": 6.844444444444445e-05,
      "loss": 0.0158,
      "step": 1925
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.7297791838645935,
      "learning_rate": 6.788888888888888e-05,
      "loss": 0.0175,
      "step": 1950
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.7372362017631531,
      "learning_rate": 6.733333333333333e-05,
      "loss": 0.0156,
      "step": 1975
    },
    {
      "epoch": 9.01,
      "grad_norm": 0.7431686520576477,
      "learning_rate": 6.677777777777779e-05,
      "loss": 0.0165,
      "step": 2000
    },
    {
      "epoch": 9.01,
      "eval_loss": 0.3495735228061676,
      "eval_runtime": 1282.4445,
      "eval_samples_per_second": 2.435,
      "eval_steps_per_second": 0.076,
      "eval_wer": 0.30066445182724255,
      "step": 2000
    },
    {
      "epoch": 9.12,
      "grad_norm": 0.7414669990539551,
      "learning_rate": 6.622222222222224e-05,
      "loss": 0.0124,
      "step": 2025
    },
    {
      "epoch": 9.23,
      "grad_norm": 0.7123928666114807,
      "learning_rate": 6.566666666666666e-05,
      "loss": 0.0111,
      "step": 2050
    },
    {
      "epoch": 9.35,
      "grad_norm": 0.5316759943962097,
      "learning_rate": 6.511111111111111e-05,
      "loss": 0.0111,
      "step": 2075
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.2563645839691162,
      "learning_rate": 6.455555555555556e-05,
      "loss": 0.012,
      "step": 2100
    },
    {
      "epoch": 9.57,
      "grad_norm": 1.2440812587738037,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.0106,
      "step": 2125
    },
    {
      "epoch": 9.68,
      "grad_norm": 1.0652848482131958,
      "learning_rate": 6.344444444444445e-05,
      "loss": 0.0082,
      "step": 2150
    },
    {
      "epoch": 9.8,
      "grad_norm": 1.4494587182998657,
      "learning_rate": 6.28888888888889e-05,
      "loss": 0.0089,
      "step": 2175
    },
    {
      "epoch": 9.91,
      "grad_norm": 1.4650260210037231,
      "learning_rate": 6.233333333333334e-05,
      "loss": 0.0087,
      "step": 2200
    },
    {
      "epoch": 10.02,
      "grad_norm": 0.8734236359596252,
      "learning_rate": 6.177777777777779e-05,
      "loss": 0.0081,
      "step": 2225
    },
    {
      "epoch": 10.14,
      "grad_norm": 0.3450920283794403,
      "learning_rate": 6.122222222222222e-05,
      "loss": 0.0083,
      "step": 2250
    },
    {
      "epoch": 10.25,
      "grad_norm": 0.21007080376148224,
      "learning_rate": 6.066666666666667e-05,
      "loss": 0.0084,
      "step": 2275
    },
    {
      "epoch": 10.36,
      "grad_norm": 0.646001935005188,
      "learning_rate": 6.011111111111112e-05,
      "loss": 0.0093,
      "step": 2300
    },
    {
      "epoch": 10.47,
      "grad_norm": 0.2686345875263214,
      "learning_rate": 5.9555555555555554e-05,
      "loss": 0.009,
      "step": 2325
    },
    {
      "epoch": 10.59,
      "grad_norm": 0.4910290241241455,
      "learning_rate": 5.9e-05,
      "loss": 0.0076,
      "step": 2350
    },
    {
      "epoch": 10.7,
      "grad_norm": 0.2715873718261719,
      "learning_rate": 5.844444444444445e-05,
      "loss": 0.0068,
      "step": 2375
    },
    {
      "epoch": 10.81,
      "grad_norm": 0.4012056589126587,
      "learning_rate": 5.788888888888889e-05,
      "loss": 0.009,
      "step": 2400
    },
    {
      "epoch": 10.92,
      "grad_norm": 0.7254907488822937,
      "learning_rate": 5.7333333333333336e-05,
      "loss": 0.0083,
      "step": 2425
    },
    {
      "epoch": 11.04,
      "grad_norm": 0.8048310279846191,
      "learning_rate": 5.6777777777777786e-05,
      "loss": 0.007,
      "step": 2450
    },
    {
      "epoch": 11.15,
      "grad_norm": 0.4055149555206299,
      "learning_rate": 5.622222222222222e-05,
      "loss": 0.0063,
      "step": 2475
    },
    {
      "epoch": 11.26,
      "grad_norm": 0.4453238248825073,
      "learning_rate": 5.566666666666667e-05,
      "loss": 0.0061,
      "step": 2500
    },
    {
      "epoch": 11.37,
      "grad_norm": 0.7719666361808777,
      "learning_rate": 5.511111111111111e-05,
      "loss": 0.006,
      "step": 2525
    },
    {
      "epoch": 11.49,
      "grad_norm": 0.6845406889915466,
      "learning_rate": 5.455555555555556e-05,
      "loss": 0.006,
      "step": 2550
    },
    {
      "epoch": 11.6,
      "grad_norm": 1.624002456665039,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.006,
      "step": 2575
    },
    {
      "epoch": 11.71,
      "grad_norm": 0.28533199429512024,
      "learning_rate": 5.3444444444444455e-05,
      "loss": 0.006,
      "step": 2600
    },
    {
      "epoch": 11.82,
      "grad_norm": 0.6341890096664429,
      "learning_rate": 5.2888888888888885e-05,
      "loss": 0.0058,
      "step": 2625
    },
    {
      "epoch": 11.94,
      "grad_norm": 0.8105676770210266,
      "learning_rate": 5.2333333333333336e-05,
      "loss": 0.0048,
      "step": 2650
    },
    {
      "epoch": 12.05,
      "grad_norm": 0.4783516824245453,
      "learning_rate": 5.177777777777778e-05,
      "loss": 0.006,
      "step": 2675
    },
    {
      "epoch": 12.16,
      "grad_norm": 0.8321937322616577,
      "learning_rate": 5.122222222222223e-05,
      "loss": 0.0042,
      "step": 2700
    },
    {
      "epoch": 12.27,
      "grad_norm": 0.1419239044189453,
      "learning_rate": 5.0666666666666674e-05,
      "loss": 0.0045,
      "step": 2725
    },
    {
      "epoch": 12.39,
      "grad_norm": 0.19317950308322906,
      "learning_rate": 5.011111111111111e-05,
      "loss": 0.0036,
      "step": 2750
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.13593174517154694,
      "learning_rate": 4.955555555555556e-05,
      "loss": 0.004,
      "step": 2775
    },
    {
      "epoch": 12.61,
      "grad_norm": 0.3188939094543457,
      "learning_rate": 4.9e-05,
      "loss": 0.0034,
      "step": 2800
    },
    {
      "epoch": 12.73,
      "grad_norm": 0.31336820125579834,
      "learning_rate": 4.844444444444445e-05,
      "loss": 0.0035,
      "step": 2825
    },
    {
      "epoch": 12.84,
      "grad_norm": 0.25262606143951416,
      "learning_rate": 4.7888888888888886e-05,
      "loss": 0.0034,
      "step": 2850
    },
    {
      "epoch": 12.95,
      "grad_norm": 0.4232746362686157,
      "learning_rate": 4.7333333333333336e-05,
      "loss": 0.0036,
      "step": 2875
    },
    {
      "epoch": 13.06,
      "grad_norm": 0.37615010142326355,
      "learning_rate": 4.677777777777778e-05,
      "loss": 0.0034,
      "step": 2900
    },
    {
      "epoch": 13.18,
      "grad_norm": 0.13130201399326324,
      "learning_rate": 4.6222222222222224e-05,
      "loss": 0.0024,
      "step": 2925
    },
    {
      "epoch": 13.29,
      "grad_norm": 0.31216856837272644,
      "learning_rate": 4.566666666666667e-05,
      "loss": 0.0034,
      "step": 2950
    },
    {
      "epoch": 13.4,
      "grad_norm": 0.06603355705738068,
      "learning_rate": 4.511111111111112e-05,
      "loss": 0.0019,
      "step": 2975
    },
    {
      "epoch": 13.51,
      "grad_norm": 0.2945907711982727,
      "learning_rate": 4.4555555555555555e-05,
      "loss": 0.0022,
      "step": 3000
    },
    {
      "epoch": 13.51,
      "eval_loss": 0.36493760347366333,
      "eval_runtime": 1357.8705,
      "eval_samples_per_second": 2.3,
      "eval_steps_per_second": 0.072,
      "eval_wer": 0.2786061963995982,
      "step": 3000
    },
    {
      "epoch": 13.63,
      "grad_norm": 0.2547701597213745,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.002,
      "step": 3025
    },
    {
      "epoch": 13.74,
      "grad_norm": 0.7029439806938171,
      "learning_rate": 4.344444444444445e-05,
      "loss": 0.002,
      "step": 3050
    },
    {
      "epoch": 13.85,
      "grad_norm": 0.821772038936615,
      "learning_rate": 4.2888888888888886e-05,
      "loss": 0.0025,
      "step": 3075
    },
    {
      "epoch": 13.96,
      "grad_norm": 0.17252065241336823,
      "learning_rate": 4.233333333333334e-05,
      "loss": 0.0027,
      "step": 3100
    },
    {
      "epoch": 14.08,
      "grad_norm": 0.18165266513824463,
      "learning_rate": 4.177777777777778e-05,
      "loss": 0.0022,
      "step": 3125
    },
    {
      "epoch": 14.19,
      "grad_norm": 0.11963178962469101,
      "learning_rate": 4.1222222222222224e-05,
      "loss": 0.0023,
      "step": 3150
    },
    {
      "epoch": 14.3,
      "grad_norm": 1.3975796699523926,
      "learning_rate": 4.066666666666667e-05,
      "loss": 0.0024,
      "step": 3175
    },
    {
      "epoch": 14.41,
      "grad_norm": 0.3654703199863434,
      "learning_rate": 4.011111111111111e-05,
      "loss": 0.0026,
      "step": 3200
    },
    {
      "epoch": 14.53,
      "grad_norm": 0.22537653148174286,
      "learning_rate": 3.9555555555555556e-05,
      "loss": 0.003,
      "step": 3225
    },
    {
      "epoch": 14.64,
      "grad_norm": 0.531537652015686,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.0025,
      "step": 3250
    },
    {
      "epoch": 14.75,
      "grad_norm": 0.09146568179130554,
      "learning_rate": 3.844444444444444e-05,
      "loss": 0.0023,
      "step": 3275
    },
    {
      "epoch": 14.86,
      "grad_norm": 0.3482789993286133,
      "learning_rate": 3.7888888888888894e-05,
      "loss": 0.0028,
      "step": 3300
    },
    {
      "epoch": 14.98,
      "grad_norm": 0.15211211144924164,
      "learning_rate": 3.733333333333334e-05,
      "loss": 0.0032,
      "step": 3325
    },
    {
      "epoch": 15.09,
      "grad_norm": 0.02796722762286663,
      "learning_rate": 3.677777777777778e-05,
      "loss": 0.002,
      "step": 3350
    },
    {
      "epoch": 15.2,
      "grad_norm": 0.022246429696679115,
      "learning_rate": 3.6222222222222225e-05,
      "loss": 0.0019,
      "step": 3375
    },
    {
      "epoch": 15.32,
      "grad_norm": 0.2861407995223999,
      "learning_rate": 3.566666666666667e-05,
      "loss": 0.0029,
      "step": 3400
    },
    {
      "epoch": 15.43,
      "grad_norm": 0.24615710973739624,
      "learning_rate": 3.511111111111111e-05,
      "loss": 0.003,
      "step": 3425
    },
    {
      "epoch": 15.54,
      "grad_norm": 0.20990662276744843,
      "learning_rate": 3.4555555555555556e-05,
      "loss": 0.003,
      "step": 3450
    },
    {
      "epoch": 15.65,
      "grad_norm": 0.21401448547840118,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.0017,
      "step": 3475
    },
    {
      "epoch": 15.77,
      "grad_norm": 0.04294591024518013,
      "learning_rate": 3.3444444444444443e-05,
      "loss": 0.0019,
      "step": 3500
    },
    {
      "epoch": 15.88,
      "grad_norm": 0.7296904921531677,
      "learning_rate": 3.2888888888888894e-05,
      "loss": 0.002,
      "step": 3525
    },
    {
      "epoch": 15.99,
      "grad_norm": 0.23458455502986908,
      "learning_rate": 3.233333333333333e-05,
      "loss": 0.0016,
      "step": 3550
    },
    {
      "epoch": 16.1,
      "grad_norm": 0.02373860962688923,
      "learning_rate": 3.177777777777778e-05,
      "loss": 0.0007,
      "step": 3575
    },
    {
      "epoch": 16.22,
      "grad_norm": 0.028949666768312454,
      "learning_rate": 3.1222222222222225e-05,
      "loss": 0.0017,
      "step": 3600
    },
    {
      "epoch": 16.33,
      "grad_norm": 0.01981484517455101,
      "learning_rate": 3.066666666666667e-05,
      "loss": 0.0013,
      "step": 3625
    },
    {
      "epoch": 16.44,
      "grad_norm": 0.017177563160657883,
      "learning_rate": 3.0111111111111113e-05,
      "loss": 0.0014,
      "step": 3650
    },
    {
      "epoch": 16.55,
      "grad_norm": 0.021722471341490746,
      "learning_rate": 2.955555555555556e-05,
      "loss": 0.0008,
      "step": 3675
    },
    {
      "epoch": 16.67,
      "grad_norm": 0.10685452073812485,
      "learning_rate": 2.9e-05,
      "loss": 0.0013,
      "step": 3700
    },
    {
      "epoch": 16.78,
      "grad_norm": 0.01891660876572132,
      "learning_rate": 2.8444444444444447e-05,
      "loss": 0.0007,
      "step": 3725
    },
    {
      "epoch": 16.89,
      "grad_norm": 0.021835455670952797,
      "learning_rate": 2.788888888888889e-05,
      "loss": 0.0011,
      "step": 3750
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.017706584185361862,
      "learning_rate": 2.733333333333333e-05,
      "loss": 0.0006,
      "step": 3775
    },
    {
      "epoch": 17.12,
      "grad_norm": 0.2573525011539459,
      "learning_rate": 2.677777777777778e-05,
      "loss": 0.0013,
      "step": 3800
    },
    {
      "epoch": 17.23,
      "grad_norm": 0.015161894261837006,
      "learning_rate": 2.6222222222222226e-05,
      "loss": 0.0004,
      "step": 3825
    },
    {
      "epoch": 17.34,
      "grad_norm": 0.017167283222079277,
      "learning_rate": 2.5666666666666666e-05,
      "loss": 0.0005,
      "step": 3850
    },
    {
      "epoch": 17.45,
      "grad_norm": 0.019201019778847694,
      "learning_rate": 2.5111111111111113e-05,
      "loss": 0.0005,
      "step": 3875
    },
    {
      "epoch": 17.57,
      "grad_norm": 0.024687746539711952,
      "learning_rate": 2.4555555555555557e-05,
      "loss": 0.0006,
      "step": 3900
    },
    {
      "epoch": 17.68,
      "grad_norm": 0.016668912023305893,
      "learning_rate": 2.4e-05,
      "loss": 0.0006,
      "step": 3925
    },
    {
      "epoch": 17.79,
      "grad_norm": 0.013516202569007874,
      "learning_rate": 2.3444444444444448e-05,
      "loss": 0.0006,
      "step": 3950
    },
    {
      "epoch": 17.91,
      "grad_norm": 0.012521643191576004,
      "learning_rate": 2.288888888888889e-05,
      "loss": 0.0006,
      "step": 3975
    },
    {
      "epoch": 18.02,
      "grad_norm": 0.013049867004156113,
      "learning_rate": 2.2333333333333335e-05,
      "loss": 0.0011,
      "step": 4000
    },
    {
      "epoch": 18.02,
      "eval_loss": 0.3699657618999481,
      "eval_runtime": 1350.6575,
      "eval_samples_per_second": 2.312,
      "eval_steps_per_second": 0.073,
      "eval_wer": 0.2680985861083211,
      "step": 4000
    },
    {
      "epoch": 18.13,
      "grad_norm": 0.0116911381483078,
      "learning_rate": 2.177777777777778e-05,
      "loss": 0.0005,
      "step": 4025
    },
    {
      "epoch": 18.24,
      "grad_norm": 0.011522598564624786,
      "learning_rate": 2.1222222222222223e-05,
      "loss": 0.0007,
      "step": 4050
    },
    {
      "epoch": 18.36,
      "grad_norm": 0.013516987673938274,
      "learning_rate": 2.0666666666666666e-05,
      "loss": 0.0004,
      "step": 4075
    },
    {
      "epoch": 18.47,
      "grad_norm": 0.021075060591101646,
      "learning_rate": 2.011111111111111e-05,
      "loss": 0.0006,
      "step": 4100
    },
    {
      "epoch": 18.58,
      "grad_norm": 0.012502779252827168,
      "learning_rate": 1.9555555555555557e-05,
      "loss": 0.0004,
      "step": 4125
    },
    {
      "epoch": 18.69,
      "grad_norm": 0.011508314870297909,
      "learning_rate": 1.9e-05,
      "loss": 0.0003,
      "step": 4150
    },
    {
      "epoch": 18.81,
      "grad_norm": 0.013346145860850811,
      "learning_rate": 1.8444444444444445e-05,
      "loss": 0.0004,
      "step": 4175
    },
    {
      "epoch": 18.92,
      "grad_norm": 0.011840825900435448,
      "learning_rate": 1.788888888888889e-05,
      "loss": 0.0004,
      "step": 4200
    },
    {
      "epoch": 19.03,
      "grad_norm": 0.009515935555100441,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 0.0004,
      "step": 4225
    },
    {
      "epoch": 19.14,
      "grad_norm": 0.01005704328417778,
      "learning_rate": 1.677777777777778e-05,
      "loss": 0.0003,
      "step": 4250
    },
    {
      "epoch": 19.26,
      "grad_norm": 0.010899914428591728,
      "learning_rate": 1.6222222222222223e-05,
      "loss": 0.0003,
      "step": 4275
    },
    {
      "epoch": 19.37,
      "grad_norm": 0.007188358344137669,
      "learning_rate": 1.5666666666666667e-05,
      "loss": 0.0003,
      "step": 4300
    },
    {
      "epoch": 19.48,
      "grad_norm": 0.00955110415816307,
      "learning_rate": 1.5111111111111112e-05,
      "loss": 0.0003,
      "step": 4325
    },
    {
      "epoch": 19.59,
      "grad_norm": 0.007988874800503254,
      "learning_rate": 1.4555555555555556e-05,
      "loss": 0.0003,
      "step": 4350
    },
    {
      "epoch": 19.71,
      "grad_norm": 0.009029334411025047,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.0005,
      "step": 4375
    },
    {
      "epoch": 19.82,
      "grad_norm": 0.01156931184232235,
      "learning_rate": 1.3444444444444445e-05,
      "loss": 0.0003,
      "step": 4400
    },
    {
      "epoch": 19.93,
      "grad_norm": 0.01129342895001173,
      "learning_rate": 1.2888888888888889e-05,
      "loss": 0.0003,
      "step": 4425
    },
    {
      "epoch": 20.05,
      "grad_norm": 0.00786085519939661,
      "learning_rate": 1.2333333333333334e-05,
      "loss": 0.0004,
      "step": 4450
    },
    {
      "epoch": 20.16,
      "grad_norm": 0.009362996555864811,
      "learning_rate": 1.1777777777777778e-05,
      "loss": 0.0003,
      "step": 4475
    },
    {
      "epoch": 20.27,
      "grad_norm": 0.008518415503203869,
      "learning_rate": 1.1222222222222224e-05,
      "loss": 0.0003,
      "step": 4500
    },
    {
      "epoch": 20.38,
      "grad_norm": 0.007652095053344965,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 0.0003,
      "step": 4525
    },
    {
      "epoch": 20.5,
      "grad_norm": 0.008384721353650093,
      "learning_rate": 1.0111111111111111e-05,
      "loss": 0.0003,
      "step": 4550
    },
    {
      "epoch": 20.61,
      "grad_norm": 0.010271112434566021,
      "learning_rate": 9.555555555555556e-06,
      "loss": 0.0003,
      "step": 4575
    },
    {
      "epoch": 20.72,
      "grad_norm": 0.0075312405824661255,
      "learning_rate": 9e-06,
      "loss": 0.0004,
      "step": 4600
    },
    {
      "epoch": 20.83,
      "grad_norm": 0.009318512864410877,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.0003,
      "step": 4625
    },
    {
      "epoch": 20.95,
      "grad_norm": 0.0078095910139381886,
      "learning_rate": 7.88888888888889e-06,
      "loss": 0.0003,
      "step": 4650
    },
    {
      "epoch": 21.06,
      "grad_norm": 0.00864331517368555,
      "learning_rate": 7.333333333333334e-06,
      "loss": 0.0003,
      "step": 4675
    },
    {
      "epoch": 21.17,
      "grad_norm": 0.007982113398611546,
      "learning_rate": 6.777777777777779e-06,
      "loss": 0.0003,
      "step": 4700
    },
    {
      "epoch": 21.28,
      "grad_norm": 0.009959988296031952,
      "learning_rate": 6.222222222222222e-06,
      "loss": 0.0003,
      "step": 4725
    },
    {
      "epoch": 21.4,
      "grad_norm": 0.006970000918954611,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.0003,
      "step": 4750
    },
    {
      "epoch": 21.51,
      "grad_norm": 0.0091372299939394,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.0003,
      "step": 4775
    },
    {
      "epoch": 21.62,
      "grad_norm": 0.010007310658693314,
      "learning_rate": 4.555555555555556e-06,
      "loss": 0.0003,
      "step": 4800
    },
    {
      "epoch": 21.73,
      "grad_norm": 0.008494430221617222,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0003,
      "step": 4825
    },
    {
      "epoch": 21.85,
      "grad_norm": 0.0071678003296256065,
      "learning_rate": 3.4444444444444444e-06,
      "loss": 0.0003,
      "step": 4850
    },
    {
      "epoch": 21.96,
      "grad_norm": 0.00614485889673233,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.0003,
      "step": 4875
    },
    {
      "epoch": 22.07,
      "grad_norm": 0.008994187228381634,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.0003,
      "step": 4900
    },
    {
      "epoch": 22.18,
      "grad_norm": 0.008750267326831818,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.0003,
      "step": 4925
    },
    {
      "epoch": 22.3,
      "grad_norm": 0.007052999921143055,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 0.0002,
      "step": 4950
    },
    {
      "epoch": 22.41,
      "grad_norm": 0.007113702595233917,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.0002,
      "step": 4975
    },
    {
      "epoch": 22.52,
      "grad_norm": 0.008453252725303173,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 0.0003,
      "step": 5000
    },
    {
      "epoch": 22.52,
      "eval_loss": 0.37487614154815674,
      "eval_runtime": 1348.7243,
      "eval_samples_per_second": 2.316,
      "eval_steps_per_second": 0.073,
      "eval_wer": 0.26639882562002626,
      "step": 5000
    },
    {
      "epoch": 22.52,
      "step": 5000,
      "total_flos": 2.532745423355904e+20,
      "train_loss": 0.1750582966186106,
      "train_runtime": 15499.9794,
      "train_samples_per_second": 10.323,
      "train_steps_per_second": 0.323
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 23,
  "save_steps": 1000,
  "total_flos": 2.532745423355904e+20,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}