{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 29.998738965952082,
  "global_step": 11880,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "learning_rate": 5.9999999999999995e-05,
      "loss": NaN,
      "step": 100
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00011999999999999999,
      "loss": NaN,
      "step": 200
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00017999999999999998,
      "loss": NaN,
      "step": 300
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00023999999999999998,
      "loss": NaN,
      "step": 400
    },
    {
      "epoch": 1.01,
      "eval_loss": 2.7731735706329346,
      "eval_runtime": 603.0564,
      "eval_samples_per_second": 8.576,
      "eval_wer": 1.0,
      "step": 400
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0003,
      "loss": NaN,
      "step": 500
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00029736379613356763,
      "loss": NaN,
      "step": 600
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002947275922671353,
      "loss": NaN,
      "step": 700
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.00029209138840070295,
      "loss": NaN,
      "step": 800
    },
    {
      "epoch": 2.02,
      "eval_loss": 0.2804177403450012,
      "eval_runtime": 605.6208,
      "eval_samples_per_second": 8.54,
      "eval_wer": 0.516512825876631,
      "step": 800
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002894551845342706,
      "loss": NaN,
      "step": 900
    },
    {
      "epoch": 2.52,
      "learning_rate": 0.00028681898066783827,
      "loss": NaN,
      "step": 1000
    },
    {
      "epoch": 2.78,
      "learning_rate": 0.000284182776801406,
      "loss": NaN,
      "step": 1100
    },
    {
      "epoch": 3.03,
      "learning_rate": 0.00028154657293497364,
      "loss": NaN,
      "step": 1200
    },
    {
      "epoch": 3.03,
      "eval_loss": 0.19836615025997162,
      "eval_runtime": 613.5068,
      "eval_samples_per_second": 8.43,
      "eval_wer": 0.41406270428575165,
      "step": 1200
    },
    {
      "epoch": 3.28,
      "learning_rate": 0.00027891036906854124,
      "loss": NaN,
      "step": 1300
    },
    {
      "epoch": 3.53,
      "learning_rate": 0.00027627416520210896,
      "loss": NaN,
      "step": 1400
    },
    {
      "epoch": 3.79,
      "learning_rate": 0.0002736379613356766,
      "loss": NaN,
      "step": 1500
    },
    {
      "epoch": 4.04,
      "learning_rate": 0.00027100175746924427,
      "loss": NaN,
      "step": 1600
    },
    {
      "epoch": 4.04,
      "eval_loss": 0.18502908945083618,
      "eval_runtime": 614.3947,
      "eval_samples_per_second": 8.418,
      "eval_wer": 0.3799911094840886,
      "step": 1600
    },
    {
      "epoch": 4.29,
      "learning_rate": 0.00026836555360281193,
      "loss": NaN,
      "step": 1700
    },
    {
      "epoch": 4.54,
      "learning_rate": 0.0002657293497363796,
      "loss": NaN,
      "step": 1800
    },
    {
      "epoch": 4.8,
      "learning_rate": 0.00026309314586994725,
      "loss": NaN,
      "step": 1900
    },
    {
      "epoch": 5.05,
      "learning_rate": 0.0002604569420035149,
      "loss": NaN,
      "step": 2000
    },
    {
      "epoch": 5.05,
      "eval_loss": 0.16938243806362152,
      "eval_runtime": 609.4739,
      "eval_samples_per_second": 8.486,
      "eval_wer": 0.36521716392542425,
      "step": 2000
    },
    {
      "epoch": 5.3,
      "learning_rate": 0.00025782073813708257,
      "loss": NaN,
      "step": 2100
    },
    {
      "epoch": 5.55,
      "learning_rate": 0.0002551845342706502,
      "loss": NaN,
      "step": 2200
    },
    {
      "epoch": 5.81,
      "learning_rate": 0.00025254833040421794,
      "loss": NaN,
      "step": 2300
    },
    {
      "epoch": 6.06,
      "learning_rate": 0.00024991212653778554,
      "loss": NaN,
      "step": 2400
    },
    {
      "epoch": 6.06,
      "eval_loss": 0.17296069860458374,
      "eval_runtime": 608.666,
      "eval_samples_per_second": 8.497,
      "eval_wer": 0.35112308134822057,
      "step": 2400
    },
    {
      "epoch": 6.31,
      "learning_rate": 0.00024727592267135325,
      "loss": NaN,
      "step": 2500
    },
    {
      "epoch": 6.56,
      "learning_rate": 0.0002446397188049209,
      "loss": NaN,
      "step": 2600
    },
    {
      "epoch": 6.82,
      "learning_rate": 0.00024200351493848854,
      "loss": NaN,
      "step": 2700
    },
    {
      "epoch": 7.07,
      "learning_rate": 0.00023936731107205623,
      "loss": NaN,
      "step": 2800
    },
    {
      "epoch": 7.07,
      "eval_loss": 0.17802052199840546,
      "eval_runtime": 612.551,
      "eval_samples_per_second": 8.443,
      "eval_wer": 0.3466778233925163,
      "step": 2800
    },
    {
      "epoch": 7.32,
      "learning_rate": 0.00023673110720562386,
      "loss": NaN,
      "step": 2900
    },
    {
      "epoch": 7.58,
      "learning_rate": 0.00023409490333919155,
      "loss": NaN,
      "step": 3000
    },
    {
      "epoch": 7.83,
      "learning_rate": 0.0002314586994727592,
      "loss": NaN,
      "step": 3100
    },
    {
      "epoch": 8.08,
      "learning_rate": 0.0002288224956063269,
      "loss": NaN,
      "step": 3200
    },
    {
      "epoch": 8.08,
      "eval_loss": 0.17137496173381805,
      "eval_runtime": 621.5301,
      "eval_samples_per_second": 8.321,
      "eval_wer": 0.343304657061423,
      "step": 3200
    },
    {
      "epoch": 8.33,
      "learning_rate": 0.00022618629173989452,
      "loss": NaN,
      "step": 3300
    },
    {
      "epoch": 8.59,
      "learning_rate": 0.00022355008787346218,
      "loss": NaN,
      "step": 3400
    },
    {
      "epoch": 8.84,
      "learning_rate": 0.00022091388400702987,
      "loss": NaN,
      "step": 3500
    },
    {
      "epoch": 9.09,
      "learning_rate": 0.0002182776801405975,
      "loss": NaN,
      "step": 3600
    },
    {
      "epoch": 9.09,
      "eval_loss": 0.16793519258499146,
      "eval_runtime": 613.7107,
      "eval_samples_per_second": 8.427,
      "eval_wer": 0.3368721073137568,
      "step": 3600
    },
    {
      "epoch": 9.34,
      "learning_rate": 0.00021564147627416518,
      "loss": NaN,
      "step": 3700
    },
    {
      "epoch": 9.6,
      "learning_rate": 0.00021300527240773284,
      "loss": NaN,
      "step": 3800
    },
    {
      "epoch": 9.85,
      "learning_rate": 0.00021036906854130053,
      "loss": NaN,
      "step": 3900
    },
    {
      "epoch": 10.1,
      "learning_rate": 0.00020773286467486816,
      "loss": NaN,
      "step": 4000
    },
    {
      "epoch": 10.1,
      "eval_loss": 0.165544793009758,
      "eval_runtime": 616.3457,
      "eval_samples_per_second": 8.391,
      "eval_wer": 0.3317731349528018,
      "step": 4000
    },
    {
      "epoch": 10.35,
      "learning_rate": 0.00020509666080843582,
      "loss": NaN,
      "step": 4100
    },
    {
      "epoch": 10.61,
      "learning_rate": 0.0002024604569420035,
      "loss": NaN,
      "step": 4200
    },
    {
      "epoch": 10.86,
      "learning_rate": 0.00019982425307557116,
      "loss": NaN,
      "step": 4300
    },
    {
      "epoch": 11.11,
      "learning_rate": 0.00019718804920913882,
      "loss": NaN,
      "step": 4400
    },
    {
      "epoch": 11.11,
      "eval_loss": 0.17516674101352692,
      "eval_runtime": 616.3789,
      "eval_samples_per_second": 8.391,
      "eval_wer": 0.32222890463614257,
      "step": 4400
    },
    {
      "epoch": 11.36,
      "learning_rate": 0.00019455184534270648,
      "loss": NaN,
      "step": 4500
    },
    {
      "epoch": 11.62,
      "learning_rate": 0.00019191564147627417,
      "loss": NaN,
      "step": 4600
    },
    {
      "epoch": 11.87,
      "learning_rate": 0.0001892794376098418,
      "loss": NaN,
      "step": 4700
    },
    {
      "epoch": 12.12,
      "learning_rate": 0.00018664323374340946,
      "loss": NaN,
      "step": 4800
    },
    {
      "epoch": 12.12,
      "eval_loss": 0.18037526309490204,
      "eval_runtime": 612.0676,
      "eval_samples_per_second": 8.45,
      "eval_wer": 0.3176790523755981,
      "step": 4800
    },
    {
      "epoch": 12.37,
      "learning_rate": 0.00018400702987697714,
      "loss": NaN,
      "step": 4900
    },
    {
      "epoch": 12.63,
      "learning_rate": 0.0001813708260105448,
      "loss": NaN,
      "step": 5000
    },
    {
      "epoch": 12.88,
      "learning_rate": 0.00017873462214411246,
      "loss": NaN,
      "step": 5100
    },
    {
      "epoch": 13.13,
      "learning_rate": 0.00017609841827768012,
      "loss": NaN,
      "step": 5200
    },
    {
      "epoch": 13.13,
      "eval_loss": 0.1690378040075302,
      "eval_runtime": 616.3617,
      "eval_samples_per_second": 8.391,
      "eval_wer": 0.31728682373244776,
      "step": 5200
    },
    {
      "epoch": 13.38,
      "learning_rate": 0.0001734622144112478,
      "loss": NaN,
      "step": 5300
    },
    {
      "epoch": 13.64,
      "learning_rate": 0.00017082601054481546,
      "loss": NaN,
      "step": 5400
    },
    {
      "epoch": 13.89,
      "learning_rate": 0.0001681898066783831,
      "loss": NaN,
      "step": 5500
    },
    {
      "epoch": 14.14,
      "learning_rate": 0.00016555360281195078,
      "loss": NaN,
      "step": 5600
    },
    {
      "epoch": 14.14,
      "eval_loss": 0.17649142444133759,
      "eval_runtime": 619.1476,
      "eval_samples_per_second": 8.353,
      "eval_wer": 0.31932641267682976,
      "step": 5600
    },
    {
      "epoch": 14.39,
      "learning_rate": 0.00016291739894551844,
      "loss": NaN,
      "step": 5700
    },
    {
      "epoch": 14.65,
      "learning_rate": 0.00016028119507908612,
      "loss": NaN,
      "step": 5800
    },
    {
      "epoch": 14.9,
      "learning_rate": 0.00015764499121265376,
      "loss": NaN,
      "step": 5900
    },
    {
      "epoch": 15.15,
      "learning_rate": 0.00015500878734622144,
      "loss": NaN,
      "step": 6000
    },
    {
      "epoch": 15.15,
      "eval_loss": 0.17734038829803467,
      "eval_runtime": 622.5524,
      "eval_samples_per_second": 8.308,
      "eval_wer": 0.3148288575687054,
      "step": 6000
    },
    {
      "epoch": 15.4,
      "learning_rate": 0.0001523725834797891,
      "loss": NaN,
      "step": 6100
    },
    {
      "epoch": 15.66,
      "learning_rate": 0.00014973637961335676,
      "loss": NaN,
      "step": 6200
    },
    {
      "epoch": 15.91,
      "learning_rate": 0.00014710017574692442,
      "loss": NaN,
      "step": 6300
    },
    {
      "epoch": 16.16,
      "learning_rate": 0.00014446397188049208,
      "loss": NaN,
      "step": 6400
    },
    {
      "epoch": 16.16,
      "eval_loss": 0.187224343419075,
      "eval_runtime": 615.2372,
      "eval_samples_per_second": 8.407,
      "eval_wer": 0.3136260230630442,
      "step": 6400
    },
    {
      "epoch": 16.41,
      "learning_rate": 0.00014182776801405973,
      "loss": NaN,
      "step": 6500
    },
    {
      "epoch": 16.67,
      "learning_rate": 0.0001391915641476274,
      "loss": NaN,
      "step": 6600
    },
    {
      "epoch": 16.92,
      "learning_rate": 0.00013655536028119505,
      "loss": NaN,
      "step": 6700
    },
    {
      "epoch": 17.17,
      "learning_rate": 0.00013391915641476274,
      "loss": NaN,
      "step": 6800
    },
    {
      "epoch": 17.17,
      "eval_loss": 0.1843162626028061,
      "eval_runtime": 620.0086,
      "eval_samples_per_second": 8.342,
      "eval_wer": 0.30912846795491983,
      "step": 6800
    },
    {
      "epoch": 17.42,
      "learning_rate": 0.0001312829525483304,
      "loss": NaN,
      "step": 6900
    },
    {
      "epoch": 17.68,
      "learning_rate": 0.00012864674868189805,
      "loss": NaN,
      "step": 7000
    },
    {
      "epoch": 17.93,
      "learning_rate": 0.0001260105448154657,
      "loss": NaN,
      "step": 7100
    },
    {
      "epoch": 18.18,
      "learning_rate": 0.00012337434094903337,
      "loss": NaN,
      "step": 7200
    },
    {
      "epoch": 18.18,
      "eval_loss": 0.1754978895187378,
      "eval_runtime": 616.3908,
      "eval_samples_per_second": 8.391,
      "eval_wer": 0.3102005595795309,
      "step": 7200
    },
    {
      "epoch": 18.43,
      "learning_rate": 0.00012073813708260104,
      "loss": NaN,
      "step": 7300
    },
    {
      "epoch": 18.69,
      "learning_rate": 0.0001181019332161687,
      "loss": NaN,
      "step": 7400
    },
    {
      "epoch": 18.94,
      "learning_rate": 0.00011546572934973637,
      "loss": NaN,
      "step": 7500
    },
    {
      "epoch": 19.19,
      "learning_rate": 0.00011282952548330403,
      "loss": NaN,
      "step": 7600
    },
    {
      "epoch": 19.19,
      "eval_loss": 0.18895429372787476,
      "eval_runtime": 623.3336,
      "eval_samples_per_second": 8.297,
      "eval_wer": 0.30520618152341605,
      "step": 7600
    },
    {
      "epoch": 19.44,
      "learning_rate": 0.0001101933216168717,
      "loss": NaN,
      "step": 7700
    },
    {
      "epoch": 19.7,
      "learning_rate": 0.00010755711775043936,
      "loss": NaN,
      "step": 7800
    },
    {
      "epoch": 19.95,
      "learning_rate": 0.00010492091388400701,
      "loss": NaN,
      "step": 7900
    },
    {
      "epoch": 20.2,
      "learning_rate": 0.00010228471001757468,
      "loss": NaN,
      "step": 8000
    },
    {
      "epoch": 20.2,
      "eval_loss": 0.19008345901966095,
      "eval_runtime": 614.7127,
      "eval_samples_per_second": 8.414,
      "eval_wer": 0.30334963261250425,
      "step": 8000
    },
    {
      "epoch": 20.45,
      "learning_rate": 9.964850615114234e-05,
      "loss": NaN,
      "step": 8100
    },
    {
      "epoch": 20.71,
      "learning_rate": 9.701230228471001e-05,
      "loss": NaN,
      "step": 8200
    },
    {
      "epoch": 20.96,
      "learning_rate": 9.437609841827767e-05,
      "loss": NaN,
      "step": 8300
    },
    {
      "epoch": 21.21,
      "learning_rate": 9.173989455184534e-05,
      "loss": NaN,
      "step": 8400
    },
    {
      "epoch": 21.21,
      "eval_loss": 0.1935020387172699,
      "eval_runtime": 619.5234,
      "eval_samples_per_second": 8.348,
      "eval_wer": 0.307193473315378,
      "step": 8400
    },
    {
      "epoch": 21.46,
      "learning_rate": 8.9103690685413e-05,
      "loss": NaN,
      "step": 8500
    },
    {
      "epoch": 21.72,
      "learning_rate": 8.646748681898065e-05,
      "loss": NaN,
      "step": 8600
    },
    {
      "epoch": 21.97,
      "learning_rate": 8.383128295254832e-05,
      "loss": NaN,
      "step": 8700
    },
    {
      "epoch": 22.22,
      "learning_rate": 8.119507908611598e-05,
      "loss": NaN,
      "step": 8800
    },
    {
      "epoch": 22.22,
      "eval_loss": 0.19641530513763428,
      "eval_runtime": 611.6485,
      "eval_samples_per_second": 8.456,
      "eval_wer": 0.30502314148994586,
      "step": 8800
    },
    {
      "epoch": 22.47,
      "learning_rate": 7.855887521968365e-05,
      "loss": NaN,
      "step": 8900
    },
    {
      "epoch": 22.73,
      "learning_rate": 7.592267135325131e-05,
      "loss": NaN,
      "step": 9000
    },
    {
      "epoch": 22.98,
      "learning_rate": 7.328646748681898e-05,
      "loss": NaN,
      "step": 9100
    },
    {
      "epoch": 23.23,
      "learning_rate": 7.065026362038664e-05,
      "loss": NaN,
      "step": 9200
    },
    {
      "epoch": 23.23,
      "eval_loss": 0.19613772630691528,
      "eval_runtime": 613.3141,
      "eval_samples_per_second": 8.433,
      "eval_wer": 0.3008132207201318,
      "step": 9200
    },
    {
      "epoch": 23.48,
      "learning_rate": 6.80140597539543e-05,
      "loss": NaN,
      "step": 9300
    },
    {
      "epoch": 23.74,
      "learning_rate": 6.537785588752197e-05,
      "loss": NaN,
      "step": 9400
    },
    {
      "epoch": 23.99,
      "learning_rate": 6.274165202108963e-05,
      "loss": NaN,
      "step": 9500
    },
    {
      "epoch": 24.24,
      "learning_rate": 6.0105448154657286e-05,
      "loss": NaN,
      "step": 9600
    },
    {
      "epoch": 24.24,
      "eval_loss": 0.18839870393276215,
      "eval_runtime": 615.6693,
      "eval_samples_per_second": 8.401,
      "eval_wer": 0.29623721988337737,
      "step": 9600
    },
    {
      "epoch": 24.49,
      "learning_rate": 5.746924428822495e-05,
      "loss": NaN,
      "step": 9700
    },
    {
      "epoch": 24.75,
      "learning_rate": 5.483304042179262e-05,
      "loss": NaN,
      "step": 9800
    },
    {
      "epoch": 25.0,
      "learning_rate": 5.2196836555360276e-05,
      "loss": NaN,
      "step": 9900
    },
    {
      "epoch": 25.25,
      "learning_rate": 4.956063268892794e-05,
      "loss": NaN,
      "step": 10000
    },
    {
      "epoch": 25.25,
      "eval_loss": 0.19206073880195618,
      "eval_runtime": 610.6418,
      "eval_samples_per_second": 8.47,
      "eval_wer": 0.2963418141882175,
      "step": 10000
    },
    {
      "epoch": 25.5,
      "learning_rate": 4.6924428822495606e-05,
      "loss": NaN,
      "step": 10100
    },
    {
      "epoch": 25.76,
      "learning_rate": 4.4288224956063265e-05,
      "loss": NaN,
      "step": 10200
    },
    {
      "epoch": 26.01,
      "learning_rate": 4.165202108963093e-05,
      "loss": NaN,
      "step": 10300
    },
    {
      "epoch": 26.26,
      "learning_rate": 3.901581722319859e-05,
      "loss": NaN,
      "step": 10400
    },
    {
      "epoch": 26.26,
      "eval_loss": 0.19253745675086975,
      "eval_runtime": 622.5244,
      "eval_samples_per_second": 8.308,
      "eval_wer": 0.2921841905708234,
      "step": 10400
    },
    {
      "epoch": 26.51,
      "learning_rate": 3.6379613356766254e-05,
      "loss": NaN,
      "step": 10500
    },
    {
      "epoch": 26.77,
      "learning_rate": 3.374340949033391e-05,
      "loss": NaN,
      "step": 10600
    },
    {
      "epoch": 27.02,
      "learning_rate": 3.110720562390158e-05,
      "loss": NaN,
      "step": 10700
    },
    {
      "epoch": 27.27,
      "learning_rate": 2.847100175746924e-05,
      "loss": NaN,
      "step": 10800
    },
    {
      "epoch": 27.27,
      "eval_loss": 0.19466331601142883,
      "eval_runtime": 617.6765,
      "eval_samples_per_second": 8.373,
      "eval_wer": 0.292785607823654,
      "step": 10800
    },
    {
      "epoch": 27.52,
      "learning_rate": 2.5834797891036906e-05,
      "loss": NaN,
      "step": 10900
    },
    {
      "epoch": 27.78,
      "learning_rate": 2.3198594024604568e-05,
      "loss": NaN,
      "step": 11000
    },
    {
      "epoch": 28.03,
      "learning_rate": 2.0562390158172233e-05,
      "loss": NaN,
      "step": 11100
    },
    {
      "epoch": 28.28,
      "learning_rate": 1.792618629173989e-05,
      "loss": NaN,
      "step": 11200
    },
    {
      "epoch": 28.28,
      "eval_loss": 0.1931692361831665,
      "eval_runtime": 617.2527,
      "eval_samples_per_second": 8.379,
      "eval_wer": 0.29189655623251315,
      "step": 11200
    },
    {
      "epoch": 28.53,
      "learning_rate": 1.5289982425307557e-05,
      "loss": NaN,
      "step": 11300
    },
    {
      "epoch": 28.79,
      "learning_rate": 1.2653778558875219e-05,
      "loss": NaN,
      "step": 11400
    },
    {
      "epoch": 29.04,
      "learning_rate": 1.0017574692442883e-05,
      "loss": NaN,
      "step": 11500
    },
    {
      "epoch": 29.29,
      "learning_rate": 7.381370826010544e-06,
      "loss": NaN,
      "step": 11600
    },
    {
      "epoch": 29.29,
      "eval_loss": 0.1954003870487213,
      "eval_runtime": 617.8209,
      "eval_samples_per_second": 8.371,
      "eval_wer": 0.2904322359647517,
      "step": 11600
    },
    {
      "epoch": 29.54,
      "learning_rate": 4.745166959578207e-06,
      "loss": NaN,
      "step": 11700
    },
    {
      "epoch": 29.8,
      "learning_rate": 2.1089630931458696e-06,
      "loss": NaN,
      "step": 11800
    },
    {
      "epoch": 30.0,
      "step": 11880,
      "total_flos": 6.1923487227453555e+19,
      "train_runtime": 56662.4324,
      "train_samples_per_second": 0.21
    }
  ],
  "max_steps": 11880,
  "num_train_epochs": 30,
  "total_flos": 6.1923487227453555e+19,
  "trial_name": null,
  "trial_params": null
}