{
  "best_global_step": 750,
  "best_metric": 0.48618392545595784,
  "best_model_checkpoint": "results/finetuned/ML-ENG-LUG-FULL-A40/checkpoint-750",
  "epoch": 3.613048978854148,
  "eval_steps": 250,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.036146755828664376,
      "grad_norm": 12.125,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.2254,
      "step": 25
    },
    {
      "epoch": 0.07229351165732875,
      "grad_norm": 9.5625,
      "learning_rate": 3.266666666666667e-06,
      "loss": 1.1396,
      "step": 50
    },
    {
      "epoch": 0.10844026748599313,
      "grad_norm": 9.0,
      "learning_rate": 4.933333333333334e-06,
      "loss": 1.0373,
      "step": 75
    },
    {
      "epoch": 0.1445870233146575,
      "grad_norm": 8.125,
      "learning_rate": 6.600000000000001e-06,
      "loss": 0.8837,
      "step": 100
    },
    {
      "epoch": 0.1807337791433219,
      "grad_norm": 8.25,
      "learning_rate": 8.266666666666667e-06,
      "loss": 0.8438,
      "step": 125
    },
    {
      "epoch": 0.21688053497198626,
      "grad_norm": 6.65625,
      "learning_rate": 9.933333333333334e-06,
      "loss": 0.8197,
      "step": 150
    },
    {
      "epoch": 0.2530272908006506,
      "grad_norm": 6.71875,
      "learning_rate": 9.897872340425532e-06,
      "loss": 0.6942,
      "step": 175
    },
    {
      "epoch": 0.289174046629315,
      "grad_norm": 6.75,
      "learning_rate": 9.79148936170213e-06,
      "loss": 0.8162,
      "step": 200
    },
    {
      "epoch": 0.3253208024579794,
      "grad_norm": 7.25,
      "learning_rate": 9.685106382978723e-06,
      "loss": 0.7682,
      "step": 225
    },
    {
      "epoch": 0.3614675582866438,
      "grad_norm": 4.75,
      "learning_rate": 9.57872340425532e-06,
      "loss": 0.7765,
      "step": 250
    },
    {
      "epoch": 0.3614675582866438,
      "eval_cer": 0.3216240827923205,
      "eval_loss": 0.8622527718544006,
      "eval_runtime": 1448.6289,
      "eval_samples_per_second": 0.957,
      "eval_steps_per_second": 0.24,
      "eval_wer": 0.5111065130432537,
      "step": 250
    },
    {
      "epoch": 0.3976143141153082,
      "grad_norm": 7.65625,
      "learning_rate": 9.472340425531916e-06,
      "loss": 0.646,
      "step": 275
    },
    {
      "epoch": 0.4337610699439725,
      "grad_norm": 5.5625,
      "learning_rate": 9.365957446808511e-06,
      "loss": 0.6401,
      "step": 300
    },
    {
      "epoch": 0.4699078257726369,
      "grad_norm": 6.78125,
      "learning_rate": 9.259574468085107e-06,
      "loss": 0.7117,
      "step": 325
    },
    {
      "epoch": 0.5060545816013012,
      "grad_norm": 7.25,
      "learning_rate": 9.153191489361702e-06,
      "loss": 0.7722,
      "step": 350
    },
    {
      "epoch": 0.5422013374299657,
      "grad_norm": 5.875,
      "learning_rate": 9.046808510638298e-06,
      "loss": 0.6399,
      "step": 375
    },
    {
      "epoch": 0.57834809325863,
      "grad_norm": 5.125,
      "learning_rate": 8.940425531914895e-06,
      "loss": 0.695,
      "step": 400
    },
    {
      "epoch": 0.6144948490872945,
      "grad_norm": 4.8125,
      "learning_rate": 8.83404255319149e-06,
      "loss": 0.6709,
      "step": 425
    },
    {
      "epoch": 0.6506416049159588,
      "grad_norm": 6.28125,
      "learning_rate": 8.727659574468086e-06,
      "loss": 0.7144,
      "step": 450
    },
    {
      "epoch": 0.6867883607446231,
      "grad_norm": 5.875,
      "learning_rate": 8.621276595744682e-06,
      "loss": 0.7185,
      "step": 475
    },
    {
      "epoch": 0.7229351165732876,
      "grad_norm": 6.84375,
      "learning_rate": 8.514893617021277e-06,
      "loss": 0.6632,
      "step": 500
    },
    {
      "epoch": 0.7229351165732876,
      "eval_cer": 0.32359849188245976,
      "eval_loss": 0.8226236701011658,
      "eval_runtime": 1430.6804,
      "eval_samples_per_second": 0.969,
      "eval_steps_per_second": 0.243,
      "eval_wer": 0.5051111231477887,
      "step": 500
    },
    {
      "epoch": 0.7590818724019519,
      "grad_norm": 5.5,
      "learning_rate": 8.408510638297873e-06,
      "loss": 0.6509,
      "step": 525
    },
    {
      "epoch": 0.7952286282306164,
      "grad_norm": 5.3125,
      "learning_rate": 8.30212765957447e-06,
      "loss": 0.6506,
      "step": 550
    },
    {
      "epoch": 0.8313753840592807,
      "grad_norm": 6.9375,
      "learning_rate": 8.195744680851064e-06,
      "loss": 0.6635,
      "step": 575
    },
    {
      "epoch": 0.867522139887945,
      "grad_norm": 7.90625,
      "learning_rate": 8.08936170212766e-06,
      "loss": 0.6494,
      "step": 600
    },
    {
      "epoch": 0.9036688957166095,
      "grad_norm": 5.1875,
      "learning_rate": 7.982978723404256e-06,
      "loss": 0.6319,
      "step": 625
    },
    {
      "epoch": 0.9398156515452738,
      "grad_norm": 4.125,
      "learning_rate": 7.876595744680852e-06,
      "loss": 0.6252,
      "step": 650
    },
    {
      "epoch": 0.9759624073739381,
      "grad_norm": 3.90625,
      "learning_rate": 7.770212765957447e-06,
      "loss": 0.6612,
      "step": 675
    },
    {
      "epoch": 1.0115669618651726,
      "grad_norm": 5.21875,
      "learning_rate": 7.663829787234043e-06,
      "loss": 0.5836,
      "step": 700
    },
    {
      "epoch": 1.047713717693837,
      "grad_norm": 5.1875,
      "learning_rate": 7.557446808510639e-06,
      "loss": 0.6302,
      "step": 725
    },
    {
      "epoch": 1.0838604735225013,
      "grad_norm": 5.3125,
      "learning_rate": 7.4510638297872355e-06,
      "loss": 0.5672,
      "step": 750
    },
    {
      "epoch": 1.0838604735225013,
      "eval_cer": 0.3044346468112434,
      "eval_loss": 0.8067195415496826,
      "eval_runtime": 1358.1527,
      "eval_samples_per_second": 1.021,
      "eval_steps_per_second": 0.255,
      "eval_wer": 0.48618392545595784,
      "step": 750
    },
    {
      "epoch": 1.1200072293511658,
      "grad_norm": 5.65625,
      "learning_rate": 7.34468085106383e-06,
      "loss": 0.5476,
      "step": 775
    },
    {
      "epoch": 1.1561539851798301,
      "grad_norm": 6.625,
      "learning_rate": 7.2382978723404265e-06,
      "loss": 0.6132,
      "step": 800
    },
    {
      "epoch": 1.1923007410084945,
      "grad_norm": 4.03125,
      "learning_rate": 7.131914893617022e-06,
      "loss": 0.6052,
      "step": 825
    },
    {
      "epoch": 1.2284474968371588,
      "grad_norm": 3.84375,
      "learning_rate": 7.0255319148936175e-06,
      "loss": 0.5723,
      "step": 850
    },
    {
      "epoch": 1.2645942526658231,
      "grad_norm": 5.375,
      "learning_rate": 6.919148936170214e-06,
      "loss": 0.5791,
      "step": 875
    },
    {
      "epoch": 1.3007410084944877,
      "grad_norm": 5.25,
      "learning_rate": 6.8127659574468085e-06,
      "loss": 0.5981,
      "step": 900
    },
    {
      "epoch": 1.336887764323152,
      "grad_norm": 5.28125,
      "learning_rate": 6.706382978723405e-06,
      "loss": 0.5917,
      "step": 925
    },
    {
      "epoch": 1.3730345201518164,
      "grad_norm": 5.65625,
      "learning_rate": 6.600000000000001e-06,
      "loss": 0.633,
      "step": 950
    },
    {
      "epoch": 1.4091812759804807,
      "grad_norm": 4.78125,
      "learning_rate": 6.493617021276596e-06,
      "loss": 0.617,
      "step": 975
    },
    {
      "epoch": 1.445328031809145,
      "grad_norm": 4.875,
      "learning_rate": 6.387234042553192e-06,
      "loss": 0.6058,
      "step": 1000
    },
    {
      "epoch": 1.445328031809145,
      "eval_cer": 0.3142144715579215,
      "eval_loss": 0.7990756034851074,
      "eval_runtime": 1373.1379,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.253,
      "eval_wer": 0.49487802577000256,
      "step": 1000
    },
    {
      "epoch": 1.4814747876378096,
      "grad_norm": 5.09375,
      "learning_rate": 6.2808510638297885e-06,
      "loss": 0.6011,
      "step": 1025
    },
    {
      "epoch": 1.517621543466474,
      "grad_norm": 4.625,
      "learning_rate": 6.174468085106383e-06,
      "loss": 0.5811,
      "step": 1050
    },
    {
      "epoch": 1.5537682992951383,
      "grad_norm": 5.625,
      "learning_rate": 6.0680851063829795e-06,
      "loss": 0.6012,
      "step": 1075
    },
    {
      "epoch": 1.5899150551238026,
      "grad_norm": 4.5,
      "learning_rate": 5.961702127659575e-06,
      "loss": 0.5822,
      "step": 1100
    },
    {
      "epoch": 1.626061810952467,
      "grad_norm": 6.65625,
      "learning_rate": 5.8553191489361705e-06,
      "loss": 0.6002,
      "step": 1125
    },
    {
      "epoch": 1.6622085667811315,
      "grad_norm": 4.78125,
      "learning_rate": 5.748936170212767e-06,
      "loss": 0.5965,
      "step": 1150
    },
    {
      "epoch": 1.6983553226097958,
      "grad_norm": 5.34375,
      "learning_rate": 5.6425531914893615e-06,
      "loss": 0.6277,
      "step": 1175
    },
    {
      "epoch": 1.7345020784384602,
      "grad_norm": 4.40625,
      "learning_rate": 5.536170212765958e-06,
      "loss": 0.5448,
      "step": 1200
    },
    {
      "epoch": 1.7706488342671245,
      "grad_norm": 5.0625,
      "learning_rate": 5.429787234042554e-06,
      "loss": 0.5645,
      "step": 1225
    },
    {
      "epoch": 1.8067955900957888,
      "grad_norm": 4.71875,
      "learning_rate": 5.323404255319149e-06,
      "loss": 0.6589,
      "step": 1250
    },
    {
      "epoch": 1.8067955900957888,
      "eval_cer": 0.31060312404039775,
      "eval_loss": 0.7972104549407959,
      "eval_runtime": 1367.9273,
      "eval_samples_per_second": 1.014,
      "eval_steps_per_second": 0.254,
      "eval_wer": 0.49013358950339214,
      "step": 1250
    },
    {
      "epoch": 1.8429423459244534,
      "grad_norm": 4.0625,
      "learning_rate": 5.217021276595745e-06,
      "loss": 0.6027,
      "step": 1275
    },
    {
      "epoch": 1.8790891017531175,
      "grad_norm": 5.46875,
      "learning_rate": 5.110638297872342e-06,
      "loss": 0.6039,
      "step": 1300
    },
    {
      "epoch": 1.915235857581782,
      "grad_norm": 4.59375,
      "learning_rate": 5.004255319148936e-06,
      "loss": 0.6226,
      "step": 1325
    },
    {
      "epoch": 1.9513826134104464,
      "grad_norm": 5.59375,
      "learning_rate": 4.897872340425533e-06,
      "loss": 0.5832,
      "step": 1350
    },
    {
      "epoch": 1.9875293692391107,
      "grad_norm": 5.65625,
      "learning_rate": 4.791489361702128e-06,
      "loss": 0.5714,
      "step": 1375
    },
    {
      "epoch": 2.023133923730345,
      "grad_norm": 5.25,
      "learning_rate": 4.685106382978724e-06,
      "loss": 0.553,
      "step": 1400
    },
    {
      "epoch": 2.0592806795590097,
      "grad_norm": 4.3125,
      "learning_rate": 4.57872340425532e-06,
      "loss": 0.58,
      "step": 1425
    },
    {
      "epoch": 2.095427435387674,
      "grad_norm": 5.34375,
      "learning_rate": 4.4723404255319155e-06,
      "loss": 0.5211,
      "step": 1450
    },
    {
      "epoch": 2.1315741912163384,
      "grad_norm": 5.3125,
      "learning_rate": 4.365957446808511e-06,
      "loss": 0.5708,
      "step": 1475
    },
    {
      "epoch": 2.1677209470450025,
      "grad_norm": 3.984375,
      "learning_rate": 4.259574468085107e-06,
      "loss": 0.5959,
      "step": 1500
    },
    {
      "epoch": 2.1677209470450025,
      "eval_cer": 0.3117614241560084,
      "eval_loss": 0.7976789474487305,
      "eval_runtime": 1424.2468,
      "eval_samples_per_second": 0.974,
      "eval_steps_per_second": 0.244,
      "eval_wer": 0.4926164455085106,
      "step": 1500
    },
    {
      "epoch": 2.203867702873667,
      "grad_norm": 5.125,
      "learning_rate": 4.153191489361703e-06,
      "loss": 0.5694,
      "step": 1525
    },
    {
      "epoch": 2.2400144587023316,
      "grad_norm": 4.53125,
      "learning_rate": 4.046808510638298e-06,
      "loss": 0.5767,
      "step": 1550
    },
    {
      "epoch": 2.2761612145309957,
      "grad_norm": 5.3125,
      "learning_rate": 3.940425531914894e-06,
      "loss": 0.5827,
      "step": 1575
    },
    {
      "epoch": 2.3123079703596603,
      "grad_norm": 4.75,
      "learning_rate": 3.83404255319149e-06,
      "loss": 0.5597,
      "step": 1600
    },
    {
      "epoch": 2.3484547261883244,
      "grad_norm": 4.96875,
      "learning_rate": 3.7276595744680857e-06,
      "loss": 0.5605,
      "step": 1625
    },
    {
      "epoch": 2.384601482016989,
      "grad_norm": 4.75,
      "learning_rate": 3.621276595744681e-06,
      "loss": 0.5523,
      "step": 1650
    },
    {
      "epoch": 2.4207482378456535,
      "grad_norm": 4.84375,
      "learning_rate": 3.5148936170212767e-06,
      "loss": 0.5674,
      "step": 1675
    },
    {
      "epoch": 2.4568949936743176,
      "grad_norm": 4.21875,
      "learning_rate": 3.4085106382978726e-06,
      "loss": 0.5801,
      "step": 1700
    },
    {
      "epoch": 2.493041749502982,
      "grad_norm": 5.1875,
      "learning_rate": 3.3021276595744685e-06,
      "loss": 0.5634,
      "step": 1725
    },
    {
      "epoch": 2.5291885053316463,
      "grad_norm": 3.53125,
      "learning_rate": 3.195744680851064e-06,
      "loss": 0.5402,
      "step": 1750
    },
    {
      "epoch": 2.5291885053316463,
      "eval_cer": 0.3114490059221394,
      "eval_loss": 0.7963515520095825,
      "eval_runtime": 1412.53,
      "eval_samples_per_second": 0.982,
      "eval_steps_per_second": 0.246,
      "eval_wer": 0.49261625004199694,
      "step": 1750
    },
    {
      "epoch": 2.565335261160311,
      "grad_norm": 5.84375,
      "learning_rate": 3.08936170212766e-06,
      "loss": 0.5874,
      "step": 1775
    },
    {
      "epoch": 2.6014820169889754,
      "grad_norm": 5.0,
      "learning_rate": 2.9829787234042554e-06,
      "loss": 0.5083,
      "step": 1800
    },
    {
      "epoch": 2.6376287728176395,
      "grad_norm": 4.9375,
      "learning_rate": 2.8765957446808514e-06,
      "loss": 0.5587,
      "step": 1825
    },
    {
      "epoch": 2.673775528646304,
      "grad_norm": 4.4375,
      "learning_rate": 2.770212765957447e-06,
      "loss": 0.5781,
      "step": 1850
    },
    {
      "epoch": 2.7099222844749686,
      "grad_norm": 5.0,
      "learning_rate": 2.663829787234043e-06,
      "loss": 0.5224,
      "step": 1875
    },
    {
      "epoch": 2.7460690403036327,
      "grad_norm": 5.0,
      "learning_rate": 2.5574468085106387e-06,
      "loss": 0.5491,
      "step": 1900
    },
    {
      "epoch": 2.782215796132297,
      "grad_norm": 4.625,
      "learning_rate": 2.4510638297872342e-06,
      "loss": 0.5653,
      "step": 1925
    },
    {
      "epoch": 2.8183625519609614,
      "grad_norm": 6.625,
      "learning_rate": 2.34468085106383e-06,
      "loss": 0.5236,
      "step": 1950
    },
    {
      "epoch": 2.854509307789626,
      "grad_norm": 4.75,
      "learning_rate": 2.2382978723404256e-06,
      "loss": 0.5827,
      "step": 1975
    },
    {
      "epoch": 2.89065606361829,
      "grad_norm": 5.15625,
      "learning_rate": 2.1319148936170216e-06,
      "loss": 0.5934,
      "step": 2000
    },
    {
      "epoch": 2.89065606361829,
      "eval_cer": 0.3117611880607615,
      "eval_loss": 0.7964433431625366,
      "eval_runtime": 1409.3242,
      "eval_samples_per_second": 0.984,
      "eval_steps_per_second": 0.246,
      "eval_wer": 0.4920525130602013,
      "step": 2000
    },
    {
      "epoch": 2.9268028194469546,
      "grad_norm": 5.25,
      "learning_rate": 2.025531914893617e-06,
      "loss": 0.5579,
      "step": 2025
    },
    {
      "epoch": 2.962949575275619,
      "grad_norm": 5.21875,
      "learning_rate": 1.919148936170213e-06,
      "loss": 0.5838,
      "step": 2050
    },
    {
      "epoch": 2.9990963311042833,
      "grad_norm": 4.5625,
      "learning_rate": 1.8127659574468087e-06,
      "loss": 0.5663,
      "step": 2075
    },
    {
      "epoch": 3.0347008855955178,
      "grad_norm": 5.6875,
      "learning_rate": 1.7063829787234042e-06,
      "loss": 0.5703,
      "step": 2100
    },
    {
      "epoch": 3.0708476414241823,
      "grad_norm": 4.875,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.5703,
      "step": 2125
    },
    {
      "epoch": 3.1069943972528464,
      "grad_norm": 4.875,
      "learning_rate": 1.4936170212765956e-06,
      "loss": 0.5774,
      "step": 2150
    },
    {
      "epoch": 3.143141153081511,
      "grad_norm": 4.84375,
      "learning_rate": 1.3872340425531916e-06,
      "loss": 0.5686,
      "step": 2175
    },
    {
      "epoch": 3.179287908910175,
      "grad_norm": 3.96875,
      "learning_rate": 1.2808510638297875e-06,
      "loss": 0.4906,
      "step": 2200
    },
    {
      "epoch": 3.2154346647388397,
      "grad_norm": 4.78125,
      "learning_rate": 1.174468085106383e-06,
      "loss": 0.5839,
      "step": 2225
    },
    {
      "epoch": 3.251581420567504,
      "grad_norm": 5.09375,
      "learning_rate": 1.0680851063829787e-06,
      "loss": 0.5464,
      "step": 2250
    },
    {
      "epoch": 3.251581420567504,
      "eval_cer": 0.3112543561914034,
      "eval_loss": 0.7959883213043213,
      "eval_runtime": 1614.2453,
      "eval_samples_per_second": 0.859,
      "eval_steps_per_second": 0.215,
      "eval_wer": 0.4931348188600785,
      "step": 2250
    },
    {
      "epoch": 3.2877281763961683,
      "grad_norm": 4.40625,
      "learning_rate": 9.617021276595744e-07,
      "loss": 0.5668,
      "step": 2275
    },
    {
      "epoch": 3.323874932224833,
      "grad_norm": 4.3125,
      "learning_rate": 8.553191489361703e-07,
      "loss": 0.5077,
      "step": 2300
    },
    {
      "epoch": 3.3600216880534974,
      "grad_norm": 5.78125,
      "learning_rate": 7.489361702127661e-07,
      "loss": 0.5428,
      "step": 2325
    },
    {
      "epoch": 3.3961684438821615,
      "grad_norm": 5.625,
      "learning_rate": 6.425531914893618e-07,
      "loss": 0.5428,
      "step": 2350
    },
    {
      "epoch": 3.432315199710826,
      "grad_norm": 5.25,
      "learning_rate": 5.361702127659575e-07,
      "loss": 0.5395,
      "step": 2375
    },
    {
      "epoch": 3.46846195553949,
      "grad_norm": 5.1875,
      "learning_rate": 4.2978723404255325e-07,
      "loss": 0.5461,
      "step": 2400
    },
    {
      "epoch": 3.5046087113681548,
      "grad_norm": 5.03125,
      "learning_rate": 3.2340425531914897e-07,
      "loss": 0.5538,
      "step": 2425
    },
    {
      "epoch": 3.540755467196819,
      "grad_norm": 5.4375,
      "learning_rate": 2.170212765957447e-07,
      "loss": 0.5298,
      "step": 2450
    },
    {
      "epoch": 3.5769022230254834,
      "grad_norm": 5.21875,
      "learning_rate": 1.1063829787234043e-07,
      "loss": 0.5959,
      "step": 2475
    },
    {
      "epoch": 3.613048978854148,
      "grad_norm": 5.5,
      "learning_rate": 4.25531914893617e-09,
      "loss": 0.5497,
      "step": 2500
    },
    {
      "epoch": 3.613048978854148,
      "eval_cer": 0.31241486923400996,
      "eval_loss": 0.7960155010223389,
      "eval_runtime": 1533.3812,
      "eval_samples_per_second": 0.905,
      "eval_steps_per_second": 0.226,
      "eval_wer": 0.49320595826893987,
      "step": 2500
    }
  ],
  "logging_steps": 25,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3582858479796224e+20,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}