{
  "best_metric": 18.301895430821354,
  "best_model_checkpoint": "./checkpoint-2000",
  "epoch": 19.559902200489,
  "eval_steps": 1000,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.24,
      "grad_norm": 39.86157989501953,
      "learning_rate": 5.000000000000001e-07,
      "loss": 2.0555,
      "step": 25
    },
    {
      "epoch": 0.49,
      "grad_norm": Infinity,
      "learning_rate": 9.800000000000001e-07,
      "loss": 1.5219,
      "step": 50
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.197519779205322,
      "learning_rate": 1.48e-06,
      "loss": 1.0167,
      "step": 75
    },
    {
      "epoch": 0.98,
      "grad_norm": 5.485505104064941,
      "learning_rate": 1.98e-06,
      "loss": 0.7299,
      "step": 100
    },
    {
      "epoch": 1.22,
      "grad_norm": 5.534335613250732,
      "learning_rate": 2.4800000000000004e-06,
      "loss": 0.6317,
      "step": 125
    },
    {
      "epoch": 1.47,
      "grad_norm": 4.898209095001221,
      "learning_rate": 2.9800000000000003e-06,
      "loss": 0.5503,
      "step": 150
    },
    {
      "epoch": 1.71,
      "grad_norm": 5.0602946281433105,
      "learning_rate": 3.48e-06,
      "loss": 0.4998,
      "step": 175
    },
    {
      "epoch": 1.96,
      "grad_norm": 4.8069305419921875,
      "learning_rate": 3.980000000000001e-06,
      "loss": 0.4457,
      "step": 200
    },
    {
      "epoch": 2.2,
      "grad_norm": 4.571457386016846,
      "learning_rate": 4.48e-06,
      "loss": 0.3778,
      "step": 225
    },
    {
      "epoch": 2.44,
      "grad_norm": 4.707397937774658,
      "learning_rate": 4.980000000000001e-06,
      "loss": 0.333,
      "step": 250
    },
    {
      "epoch": 2.69,
      "grad_norm": 4.317914009094238,
      "learning_rate": 5.480000000000001e-06,
      "loss": 0.2859,
      "step": 275
    },
    {
      "epoch": 2.93,
      "grad_norm": 2.6433582305908203,
      "learning_rate": 5.98e-06,
      "loss": 0.2229,
      "step": 300
    },
    {
      "epoch": 3.18,
      "grad_norm": 2.5579586029052734,
      "learning_rate": 6.480000000000001e-06,
      "loss": 0.1767,
      "step": 325
    },
    {
      "epoch": 3.42,
      "grad_norm": 2.2651941776275635,
      "learning_rate": 6.98e-06,
      "loss": 0.1474,
      "step": 350
    },
    {
      "epoch": 3.67,
      "grad_norm": 2.528773546218872,
      "learning_rate": 7.48e-06,
      "loss": 0.1493,
      "step": 375
    },
    {
      "epoch": 3.91,
      "grad_norm": 2.422232151031494,
      "learning_rate": 7.980000000000002e-06,
      "loss": 0.142,
      "step": 400
    },
    {
      "epoch": 4.16,
      "grad_norm": 2.209630012512207,
      "learning_rate": 8.48e-06,
      "loss": 0.1126,
      "step": 425
    },
    {
      "epoch": 4.4,
      "grad_norm": 1.9945831298828125,
      "learning_rate": 8.98e-06,
      "loss": 0.0909,
      "step": 450
    },
    {
      "epoch": 4.65,
      "grad_norm": 1.8972020149230957,
      "learning_rate": 9.48e-06,
      "loss": 0.0892,
      "step": 475
    },
    {
      "epoch": 4.89,
      "grad_norm": 2.345607042312622,
      "learning_rate": 9.980000000000001e-06,
      "loss": 0.0909,
      "step": 500
    },
    {
      "epoch": 5.13,
      "grad_norm": 1.3913277387619019,
      "learning_rate": 9.946666666666667e-06,
      "loss": 0.0713,
      "step": 525
    },
    {
      "epoch": 5.38,
      "grad_norm": 2.5306901931762695,
      "learning_rate": 9.891111111111113e-06,
      "loss": 0.0535,
      "step": 550
    },
    {
      "epoch": 5.62,
      "grad_norm": 1.3689407110214233,
      "learning_rate": 9.835555555555556e-06,
      "loss": 0.051,
      "step": 575
    },
    {
      "epoch": 5.87,
      "grad_norm": 2.091756820678711,
      "learning_rate": 9.780000000000001e-06,
      "loss": 0.0573,
      "step": 600
    },
    {
      "epoch": 6.11,
      "grad_norm": 1.5179609060287476,
      "learning_rate": 9.724444444444445e-06,
      "loss": 0.0454,
      "step": 625
    },
    {
      "epoch": 6.36,
      "grad_norm": 1.3439680337905884,
      "learning_rate": 9.66888888888889e-06,
      "loss": 0.029,
      "step": 650
    },
    {
      "epoch": 6.6,
      "grad_norm": 1.279569387435913,
      "learning_rate": 9.613333333333335e-06,
      "loss": 0.032,
      "step": 675
    },
    {
      "epoch": 6.85,
      "grad_norm": 1.6171939373016357,
      "learning_rate": 9.557777777777777e-06,
      "loss": 0.0314,
      "step": 700
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.8949939608573914,
      "learning_rate": 9.502222222222223e-06,
      "loss": 0.0245,
      "step": 725
    },
    {
      "epoch": 7.33,
      "grad_norm": 1.3610639572143555,
      "learning_rate": 9.446666666666667e-06,
      "loss": 0.0165,
      "step": 750
    },
    {
      "epoch": 7.58,
      "grad_norm": 1.5446442365646362,
      "learning_rate": 9.391111111111111e-06,
      "loss": 0.0169,
      "step": 775
    },
    {
      "epoch": 7.82,
      "grad_norm": 1.2825671434402466,
      "learning_rate": 9.335555555555557e-06,
      "loss": 0.0193,
      "step": 800
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.7063258290290833,
      "learning_rate": 9.280000000000001e-06,
      "loss": 0.0154,
      "step": 825
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.72287917137146,
      "learning_rate": 9.224444444444445e-06,
      "loss": 0.0102,
      "step": 850
    },
    {
      "epoch": 8.56,
      "grad_norm": 1.2877657413482666,
      "learning_rate": 9.168888888888889e-06,
      "loss": 0.0093,
      "step": 875
    },
    {
      "epoch": 8.8,
      "grad_norm": 1.6262348890304565,
      "learning_rate": 9.113333333333335e-06,
      "loss": 0.0117,
      "step": 900
    },
    {
      "epoch": 9.05,
      "grad_norm": 0.49374520778656006,
      "learning_rate": 9.057777777777779e-06,
      "loss": 0.0121,
      "step": 925
    },
    {
      "epoch": 9.29,
      "grad_norm": 0.8290354013442993,
      "learning_rate": 9.002222222222223e-06,
      "loss": 0.0063,
      "step": 950
    },
    {
      "epoch": 9.54,
      "grad_norm": 0.9974227547645569,
      "learning_rate": 8.946666666666669e-06,
      "loss": 0.006,
      "step": 975
    },
    {
      "epoch": 9.78,
      "grad_norm": 0.8566415309906006,
      "learning_rate": 8.891111111111111e-06,
      "loss": 0.007,
      "step": 1000
    },
    {
      "epoch": 9.78,
      "eval_loss": 0.41129249334335327,
      "eval_runtime": 1450.197,
      "eval_samples_per_second": 1.996,
      "eval_steps_per_second": 0.499,
      "eval_wer": 18.926838201629888,
      "step": 1000
    },
    {
      "epoch": 10.02,
      "grad_norm": 1.7889363765716553,
      "learning_rate": 8.835555555555557e-06,
      "loss": 0.0067,
      "step": 1025
    },
    {
      "epoch": 10.27,
      "grad_norm": 0.9269456267356873,
      "learning_rate": 8.78e-06,
      "loss": 0.0047,
      "step": 1050
    },
    {
      "epoch": 10.51,
      "grad_norm": 0.7966364622116089,
      "learning_rate": 8.724444444444445e-06,
      "loss": 0.0047,
      "step": 1075
    },
    {
      "epoch": 10.76,
      "grad_norm": 0.6406449675559998,
      "learning_rate": 8.66888888888889e-06,
      "loss": 0.0049,
      "step": 1100
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.4881477355957031,
      "learning_rate": 8.613333333333333e-06,
      "loss": 0.0049,
      "step": 1125
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.4843809902667999,
      "learning_rate": 8.557777777777778e-06,
      "loss": 0.0036,
      "step": 1150
    },
    {
      "epoch": 11.49,
      "grad_norm": 0.8177527189254761,
      "learning_rate": 8.502222222222223e-06,
      "loss": 0.0028,
      "step": 1175
    },
    {
      "epoch": 11.74,
      "grad_norm": 0.676511287689209,
      "learning_rate": 8.446666666666668e-06,
      "loss": 0.0031,
      "step": 1200
    },
    {
      "epoch": 11.98,
      "grad_norm": 0.9199188351631165,
      "learning_rate": 8.391111111111112e-06,
      "loss": 0.003,
      "step": 1225
    },
    {
      "epoch": 12.22,
      "grad_norm": 0.248287171125412,
      "learning_rate": 8.335555555555556e-06,
      "loss": 0.0024,
      "step": 1250
    },
    {
      "epoch": 12.47,
      "grad_norm": 0.9288859963417053,
      "learning_rate": 8.28e-06,
      "loss": 0.0022,
      "step": 1275
    },
    {
      "epoch": 12.71,
      "grad_norm": 0.8308430910110474,
      "learning_rate": 8.224444444444444e-06,
      "loss": 0.0025,
      "step": 1300
    },
    {
      "epoch": 12.96,
      "grad_norm": 0.33064720034599304,
      "learning_rate": 8.16888888888889e-06,
      "loss": 0.0027,
      "step": 1325
    },
    {
      "epoch": 13.2,
      "grad_norm": 0.5254776477813721,
      "learning_rate": 8.113333333333334e-06,
      "loss": 0.0019,
      "step": 1350
    },
    {
      "epoch": 13.45,
      "grad_norm": 0.4681105315685272,
      "learning_rate": 8.057777777777778e-06,
      "loss": 0.0021,
      "step": 1375
    },
    {
      "epoch": 13.69,
      "grad_norm": 0.352115273475647,
      "learning_rate": 8.002222222222222e-06,
      "loss": 0.0026,
      "step": 1400
    },
    {
      "epoch": 13.94,
      "grad_norm": 1.5301597118377686,
      "learning_rate": 7.946666666666666e-06,
      "loss": 0.0027,
      "step": 1425
    },
    {
      "epoch": 14.18,
      "grad_norm": 0.28577202558517456,
      "learning_rate": 7.891111111111112e-06,
      "loss": 0.0018,
      "step": 1450
    },
    {
      "epoch": 14.43,
      "grad_norm": 0.7940084338188171,
      "learning_rate": 7.835555555555556e-06,
      "loss": 0.002,
      "step": 1475
    },
    {
      "epoch": 14.67,
      "grad_norm": 1.031543493270874,
      "learning_rate": 7.78e-06,
      "loss": 0.0021,
      "step": 1500
    },
    {
      "epoch": 14.91,
      "grad_norm": 0.5695396661758423,
      "learning_rate": 7.724444444444446e-06,
      "loss": 0.0016,
      "step": 1525
    },
    {
      "epoch": 15.16,
      "grad_norm": 0.2285182625055313,
      "learning_rate": 7.66888888888889e-06,
      "loss": 0.0014,
      "step": 1550
    },
    {
      "epoch": 15.4,
      "grad_norm": 0.40289613604545593,
      "learning_rate": 7.613333333333334e-06,
      "loss": 0.0017,
      "step": 1575
    },
    {
      "epoch": 15.65,
      "grad_norm": 0.5758986473083496,
      "learning_rate": 7.557777777777779e-06,
      "loss": 0.0012,
      "step": 1600
    },
    {
      "epoch": 15.89,
      "grad_norm": 0.5524174571037292,
      "learning_rate": 7.502222222222223e-06,
      "loss": 0.001,
      "step": 1625
    },
    {
      "epoch": 16.14,
      "grad_norm": 0.7031832933425903,
      "learning_rate": 7.446666666666668e-06,
      "loss": 0.0015,
      "step": 1650
    },
    {
      "epoch": 16.38,
      "grad_norm": 0.22649440169334412,
      "learning_rate": 7.3911111111111125e-06,
      "loss": 0.0012,
      "step": 1675
    },
    {
      "epoch": 16.63,
      "grad_norm": 0.46751469373703003,
      "learning_rate": 7.335555555555556e-06,
      "loss": 0.0012,
      "step": 1700
    },
    {
      "epoch": 16.87,
      "grad_norm": 0.3201611340045929,
      "learning_rate": 7.280000000000001e-06,
      "loss": 0.0009,
      "step": 1725
    },
    {
      "epoch": 17.11,
      "grad_norm": 0.05176452174782753,
      "learning_rate": 7.224444444444445e-06,
      "loss": 0.0007,
      "step": 1750
    },
    {
      "epoch": 17.36,
      "grad_norm": 0.5956466794013977,
      "learning_rate": 7.1688888888888895e-06,
      "loss": 0.0007,
      "step": 1775
    },
    {
      "epoch": 17.6,
      "grad_norm": 0.1542888581752777,
      "learning_rate": 7.113333333333334e-06,
      "loss": 0.001,
      "step": 1800
    },
    {
      "epoch": 17.85,
      "grad_norm": 0.624476432800293,
      "learning_rate": 7.057777777777778e-06,
      "loss": 0.0016,
      "step": 1825
    },
    {
      "epoch": 18.09,
      "grad_norm": 0.3069753050804138,
      "learning_rate": 7.0022222222222225e-06,
      "loss": 0.0013,
      "step": 1850
    },
    {
      "epoch": 18.34,
      "grad_norm": 0.3797595500946045,
      "learning_rate": 6.946666666666667e-06,
      "loss": 0.0018,
      "step": 1875
    },
    {
      "epoch": 18.58,
      "grad_norm": 0.3482624590396881,
      "learning_rate": 6.891111111111111e-06,
      "loss": 0.0014,
      "step": 1900
    },
    {
      "epoch": 18.83,
      "grad_norm": 0.05667097494006157,
      "learning_rate": 6.835555555555556e-06,
      "loss": 0.0008,
      "step": 1925
    },
    {
      "epoch": 19.07,
      "grad_norm": 0.09759561717510223,
      "learning_rate": 6.780000000000001e-06,
      "loss": 0.0011,
      "step": 1950
    },
    {
      "epoch": 19.32,
      "grad_norm": 0.5752401947975159,
      "learning_rate": 6.724444444444444e-06,
      "loss": 0.0014,
      "step": 1975
    },
    {
      "epoch": 19.56,
      "grad_norm": 0.1028558760881424,
      "learning_rate": 6.668888888888889e-06,
      "loss": 0.0009,
      "step": 2000
    },
    {
      "epoch": 19.56,
      "eval_loss": 0.4927152395248413,
      "eval_runtime": 1461.0654,
      "eval_samples_per_second": 1.981,
      "eval_steps_per_second": 0.496,
      "eval_wer": 18.301895430821354,
      "step": 2000
    }
  ],
  "logging_steps": 25,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 1000,
  "total_flos": 3.691699875053568e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}