{
  "best_metric": 0.49605429292929293,
  "best_model_checkpoint": "/scratch/elec/puhe/p/palp3/sami_ASR/base_model_output/wav2vec2-base-sami-22k/allps_cap20h/checkpoint-2457",
  "epoch": 60.0,
  "eval_steps": 500,
  "global_step": 49140,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 991.008056640625,
      "learning_rate": 3.264143264143264e-05,
      "loss": 4052.2546,
      "step": 819
    },
    {
      "epoch": 1.0,
      "eval_cer": 0.7619917996344415,
      "eval_loss": 868.5341186523438,
      "eval_runtime": 24.3338,
      "eval_samples_per_second": 36.575,
      "eval_steps_per_second": 4.603,
      "eval_wer": 1.0,
      "step": 819
    },
    {
      "epoch": 2.0,
      "grad_norm": 551.8701171875,
      "learning_rate": 6.597476597476598e-05,
      "loss": 1110.0161,
      "step": 1638
    },
    {
      "epoch": 2.0,
      "eval_cer": 0.16704539840932667,
      "eval_loss": 284.3064880371094,
      "eval_runtime": 24.1337,
      "eval_samples_per_second": 36.878,
      "eval_steps_per_second": 4.641,
      "eval_wer": 0.5340909090909091,
      "step": 1638
    },
    {
      "epoch": 3.0,
      "grad_norm": 952.0059204101562,
      "learning_rate": 9.93080993080993e-05,
      "loss": 792.8359,
      "step": 2457
    },
    {
      "epoch": 3.0,
      "eval_cer": 0.1561280442622141,
      "eval_loss": 281.24151611328125,
      "eval_runtime": 23.8418,
      "eval_samples_per_second": 37.329,
      "eval_steps_per_second": 4.698,
      "eval_wer": 0.49605429292929293,
      "step": 2457
    },
    {
      "epoch": 4.0,
      "grad_norm": 2196.81884765625,
      "learning_rate": 0.00013264143264143263,
      "loss": 752.4135,
      "step": 3276
    },
    {
      "epoch": 4.0,
      "eval_cer": 0.17707355629106358,
      "eval_loss": 294.49365234375,
      "eval_runtime": 23.5949,
      "eval_samples_per_second": 37.72,
      "eval_steps_per_second": 4.747,
      "eval_wer": 0.5666035353535354,
      "step": 3276
    },
    {
      "epoch": 5.0,
      "grad_norm": 1654.1439208984375,
      "learning_rate": 0.000165974765974766,
      "loss": 705.1918,
      "step": 4095
    },
    {
      "epoch": 5.0,
      "eval_cer": 0.17862964975547102,
      "eval_loss": 332.3864440917969,
      "eval_runtime": 23.9822,
      "eval_samples_per_second": 37.111,
      "eval_steps_per_second": 4.67,
      "eval_wer": 0.5135732323232324,
      "step": 4095
    },
    {
      "epoch": 6.0,
      "grad_norm": 786.24609375,
      "learning_rate": 0.00019930809930809932,
      "loss": 726.7231,
      "step": 4914
    },
    {
      "epoch": 6.0,
      "eval_cer": 0.2265474485007163,
      "eval_loss": 395.57098388671875,
      "eval_runtime": 24.1262,
      "eval_samples_per_second": 36.889,
      "eval_steps_per_second": 4.642,
      "eval_wer": 0.6213699494949495,
      "step": 4914
    },
    {
      "epoch": 7.0,
      "grad_norm": 412.29510498046875,
      "learning_rate": 0.00023264143264143265,
      "loss": 742.4047,
      "step": 5733
    },
    {
      "epoch": 7.0,
      "eval_cer": 0.21772958553574076,
      "eval_loss": 339.9690246582031,
      "eval_runtime": 23.5379,
      "eval_samples_per_second": 37.811,
      "eval_steps_per_second": 4.758,
      "eval_wer": 0.602114898989899,
      "step": 5733
    },
    {
      "epoch": 8.0,
      "grad_norm": 403.2104797363281,
      "learning_rate": 0.000265974765974766,
      "loss": 749.9364,
      "step": 6552
    },
    {
      "epoch": 8.0,
      "eval_cer": 0.2160252926937707,
      "eval_loss": 407.9001770019531,
      "eval_runtime": 24.6629,
      "eval_samples_per_second": 36.087,
      "eval_steps_per_second": 4.541,
      "eval_wer": 0.6063762626262627,
      "step": 6552
    },
    {
      "epoch": 9.0,
      "grad_norm": 777.958251953125,
      "learning_rate": 0.0002993080993080993,
      "loss": 766.2681,
      "step": 7371
    },
    {
      "epoch": 9.0,
      "eval_cer": 0.25771871758138615,
      "eval_loss": 397.77618408203125,
      "eval_runtime": 24.7874,
      "eval_samples_per_second": 35.905,
      "eval_steps_per_second": 4.518,
      "eval_wer": 0.6685606060606061,
      "step": 7371
    },
    {
      "epoch": 10.0,
      "grad_norm": 726.7489013671875,
      "learning_rate": 0.00033264143264143266,
      "loss": 798.0908,
      "step": 8190
    },
    {
      "epoch": 10.0,
      "eval_cer": 0.3210492515931433,
      "eval_loss": 429.9941711425781,
      "eval_runtime": 23.8326,
      "eval_samples_per_second": 37.344,
      "eval_steps_per_second": 4.699,
      "eval_wer": 0.7534722222222222,
      "step": 8190
    },
    {
      "epoch": 11.0,
      "grad_norm": 423.5915832519531,
      "learning_rate": 0.00036597476597476596,
      "loss": 824.5388,
      "step": 9009
    },
    {
      "epoch": 11.0,
      "eval_cer": 0.3149483772168157,
      "eval_loss": 471.0012512207031,
      "eval_runtime": 24.2588,
      "eval_samples_per_second": 36.688,
      "eval_steps_per_second": 4.617,
      "eval_wer": 0.7575757575757576,
      "step": 9009
    },
    {
      "epoch": 12.0,
      "grad_norm": 665.407958984375,
      "learning_rate": 0.0003993080993080993,
      "loss": 858.5173,
      "step": 9828
    },
    {
      "epoch": 12.0,
      "eval_cer": 0.27846663044015213,
      "eval_loss": 433.6759033203125,
      "eval_runtime": 23.8147,
      "eval_samples_per_second": 37.372,
      "eval_steps_per_second": 4.703,
      "eval_wer": 0.7313762626262627,
      "step": 9828
    },
    {
      "epoch": 13.0,
      "grad_norm": 1094.12109375,
      "learning_rate": 0.0004326007326007326,
      "loss": 886.2443,
      "step": 10647
    },
    {
      "epoch": 13.0,
      "eval_cer": 0.31334288395988735,
      "eval_loss": 451.7483825683594,
      "eval_runtime": 23.8973,
      "eval_samples_per_second": 37.243,
      "eval_steps_per_second": 4.687,
      "eval_wer": 0.7547348484848485,
      "step": 10647
    },
    {
      "epoch": 14.0,
      "grad_norm": 748.5809326171875,
      "learning_rate": 0.000465934065934066,
      "loss": 914.3147,
      "step": 11466
    },
    {
      "epoch": 14.0,
      "eval_cer": 0.3175418663241614,
      "eval_loss": 480.8207092285156,
      "eval_runtime": 23.8907,
      "eval_samples_per_second": 37.253,
      "eval_steps_per_second": 4.688,
      "eval_wer": 0.8120265151515151,
      "step": 11466
    },
    {
      "epoch": 15.0,
      "grad_norm": 304.9002380371094,
      "learning_rate": 0.0004992673992673993,
      "loss": 962.0846,
      "step": 12285
    },
    {
      "epoch": 15.0,
      "eval_cer": 0.32016005532776765,
      "eval_loss": 529.720703125,
      "eval_runtime": 23.8442,
      "eval_samples_per_second": 37.326,
      "eval_steps_per_second": 4.697,
      "eval_wer": 0.7960858585858586,
      "step": 12285
    },
    {
      "epoch": 16.0,
      "grad_norm": 587.8988037109375,
      "learning_rate": 0.0004891330891330891,
      "loss": 973.3332,
      "step": 13104
    },
    {
      "epoch": 16.0,
      "eval_cer": 0.3679543545917107,
      "eval_loss": 518.808837890625,
      "eval_runtime": 24.7703,
      "eval_samples_per_second": 35.93,
      "eval_steps_per_second": 4.522,
      "eval_wer": 0.8304924242424242,
      "step": 13104
    },
    {
      "epoch": 17.0,
      "grad_norm": 536.6315307617188,
      "learning_rate": 0.00047803554470221137,
      "loss": 954.7767,
      "step": 13923
    },
    {
      "epoch": 17.0,
      "eval_cer": 0.30805710616015414,
      "eval_loss": 507.0257873535156,
      "eval_runtime": 24.2842,
      "eval_samples_per_second": 36.649,
      "eval_steps_per_second": 4.612,
      "eval_wer": 0.7891414141414141,
      "step": 13923
    },
    {
      "epoch": 18.0,
      "grad_norm": 754.9872436523438,
      "learning_rate": 0.00046692443359110027,
      "loss": 935.9705,
      "step": 14742
    },
    {
      "epoch": 18.0,
      "eval_cer": 0.32087635231932027,
      "eval_loss": 477.90582275390625,
      "eval_runtime": 24.8782,
      "eval_samples_per_second": 35.774,
      "eval_steps_per_second": 4.502,
      "eval_wer": 0.7940340909090909,
      "step": 14742
    },
    {
      "epoch": 19.0,
      "grad_norm": 514.4397583007812,
      "learning_rate": 0.0004558133224799891,
      "loss": 901.4906,
      "step": 15561
    },
    {
      "epoch": 19.0,
      "eval_cer": 0.3500963295954157,
      "eval_loss": 513.8383178710938,
      "eval_runtime": 26.7781,
      "eval_samples_per_second": 33.236,
      "eval_steps_per_second": 4.183,
      "eval_wer": 0.8093434343434344,
      "step": 15561
    },
    {
      "epoch": 20.0,
      "grad_norm": 1911.3572998046875,
      "learning_rate": 0.0004447022113688781,
      "loss": 886.2501,
      "step": 16380
    },
    {
      "epoch": 20.0,
      "eval_cer": 0.3455515486834955,
      "eval_loss": 471.8292236328125,
      "eval_runtime": 24.3743,
      "eval_samples_per_second": 36.514,
      "eval_steps_per_second": 4.595,
      "eval_wer": 0.7859848484848485,
      "step": 16380
    },
    {
      "epoch": 21.0,
      "grad_norm": 1233.0408935546875,
      "learning_rate": 0.0004335911002577669,
      "loss": 853.8701,
      "step": 17199
    },
    {
      "epoch": 21.0,
      "eval_cer": 0.3084523045003211,
      "eval_loss": 505.87554931640625,
      "eval_runtime": 24.8737,
      "eval_samples_per_second": 35.781,
      "eval_steps_per_second": 4.503,
      "eval_wer": 0.8118686868686869,
      "step": 17199
    },
    {
      "epoch": 22.0,
      "grad_norm": 911.884033203125,
      "learning_rate": 0.00042249355582688915,
      "loss": 835.6011,
      "step": 18018
    },
    {
      "epoch": 22.0,
      "eval_cer": 0.31450377908412785,
      "eval_loss": 488.744873046875,
      "eval_runtime": 24.8755,
      "eval_samples_per_second": 35.778,
      "eval_steps_per_second": 4.502,
      "eval_wer": 0.7542613636363636,
      "step": 18018
    },
    {
      "epoch": 23.0,
      "grad_norm": 426.70599365234375,
      "learning_rate": 0.00041140957807624475,
      "loss": 834.0464,
      "step": 18837
    },
    {
      "epoch": 23.0,
      "eval_cer": 0.2940769648767475,
      "eval_loss": 465.870849609375,
      "eval_runtime": 25.0917,
      "eval_samples_per_second": 35.47,
      "eval_steps_per_second": 4.464,
      "eval_wer": 0.7578914141414141,
      "step": 18837
    },
    {
      "epoch": 24.0,
      "grad_norm": 613.6903686523438,
      "learning_rate": 0.00040029846696513366,
      "loss": 783.4934,
      "step": 19656
    },
    {
      "epoch": 24.0,
      "eval_cer": 0.3116138912216569,
      "eval_loss": 448.39312744140625,
      "eval_runtime": 24.9209,
      "eval_samples_per_second": 35.713,
      "eval_steps_per_second": 4.494,
      "eval_wer": 0.7634154040404041,
      "step": 19656
    },
    {
      "epoch": 25.0,
      "grad_norm": 262.5715637207031,
      "learning_rate": 0.00038918735585402256,
      "loss": 766.9646,
      "step": 20475
    },
    {
      "epoch": 25.0,
      "eval_cer": 0.294669762386998,
      "eval_loss": 465.7940979003906,
      "eval_runtime": 25.194,
      "eval_samples_per_second": 35.326,
      "eval_steps_per_second": 4.446,
      "eval_wer": 0.7376893939393939,
      "step": 20475
    },
    {
      "epoch": 26.0,
      "grad_norm": 865.4219970703125,
      "learning_rate": 0.0003780762447429114,
      "loss": 740.5047,
      "step": 21294
    },
    {
      "epoch": 26.0,
      "eval_cer": 0.28654349651731464,
      "eval_loss": 481.7481689453125,
      "eval_runtime": 24.4084,
      "eval_samples_per_second": 36.463,
      "eval_steps_per_second": 4.589,
      "eval_wer": 0.7323232323232324,
      "step": 21294
    },
    {
      "epoch": 27.0,
      "grad_norm": 644.6862182617188,
      "learning_rate": 0.0003669651336318003,
      "loss": 724.7147,
      "step": 22113
    },
    {
      "epoch": 27.0,
      "eval_cer": 0.2947191621795188,
      "eval_loss": 447.7026062011719,
      "eval_runtime": 24.2348,
      "eval_samples_per_second": 36.724,
      "eval_steps_per_second": 4.621,
      "eval_wer": 0.7174873737373737,
      "step": 22113
    },
    {
      "epoch": 28.0,
      "grad_norm": 462.7465515136719,
      "learning_rate": 0.0003558540225206892,
      "loss": 694.9597,
      "step": 22932
    },
    {
      "epoch": 28.0,
      "eval_cer": 0.27029096477794795,
      "eval_loss": 439.9412841796875,
      "eval_runtime": 25.1492,
      "eval_samples_per_second": 35.389,
      "eval_steps_per_second": 4.453,
      "eval_wer": 0.7050189393939394,
      "step": 22932
    },
    {
      "epoch": 29.0,
      "grad_norm": 696.3271484375,
      "learning_rate": 0.0003447429114095781,
      "loss": 678.131,
      "step": 23751
    },
    {
      "epoch": 29.0,
      "eval_cer": 0.2723657560638245,
      "eval_loss": 439.17047119140625,
      "eval_runtime": 24.9686,
      "eval_samples_per_second": 35.645,
      "eval_steps_per_second": 4.486,
      "eval_wer": 0.694760101010101,
      "step": 23751
    },
    {
      "epoch": 30.0,
      "grad_norm": 348.0857849121094,
      "learning_rate": 0.0003336453669787003,
      "loss": 669.4257,
      "step": 24570
    },
    {
      "epoch": 30.0,
      "eval_cer": 0.2678456750481648,
      "eval_loss": 463.5548400878906,
      "eval_runtime": 24.2758,
      "eval_samples_per_second": 36.662,
      "eval_steps_per_second": 4.614,
      "eval_wer": 0.6933396464646465,
      "step": 24570
    },
    {
      "epoch": 31.0,
      "grad_norm": 679.9598999023438,
      "learning_rate": 0.00032253425586758924,
      "loss": 640.7192,
      "step": 25389
    },
    {
      "epoch": 31.0,
      "eval_cer": 0.26994516623030185,
      "eval_loss": 449.90155029296875,
      "eval_runtime": 24.2781,
      "eval_samples_per_second": 36.659,
      "eval_steps_per_second": 4.613,
      "eval_wer": 0.6968118686868687,
      "step": 25389
    },
    {
      "epoch": 32.0,
      "grad_norm": 276.22564697265625,
      "learning_rate": 0.0003114231447564781,
      "loss": 616.687,
      "step": 26208
    },
    {
      "epoch": 32.0,
      "eval_cer": 0.27528034382255595,
      "eval_loss": 445.01812744140625,
      "eval_runtime": 24.4977,
      "eval_samples_per_second": 36.33,
      "eval_steps_per_second": 4.572,
      "eval_wer": 0.6966540404040404,
      "step": 26208
    },
    {
      "epoch": 33.0,
      "grad_norm": 249.17897033691406,
      "learning_rate": 0.000300312033645367,
      "loss": 606.6529,
      "step": 27027
    },
    {
      "epoch": 33.0,
      "eval_cer": 0.2614484019167119,
      "eval_loss": 424.00262451171875,
      "eval_runtime": 24.2299,
      "eval_samples_per_second": 36.732,
      "eval_steps_per_second": 4.622,
      "eval_wer": 0.6897095959595959,
      "step": 27027
    },
    {
      "epoch": 34.0,
      "grad_norm": 728.5042724609375,
      "learning_rate": 0.0002892009225342559,
      "loss": 588.8339,
      "step": 27846
    },
    {
      "epoch": 34.0,
      "eval_cer": 0.2601640073111693,
      "eval_loss": 425.9888916015625,
      "eval_runtime": 23.7857,
      "eval_samples_per_second": 37.417,
      "eval_steps_per_second": 4.709,
      "eval_wer": 0.7024936868686869,
      "step": 27846
    },
    {
      "epoch": 35.0,
      "grad_norm": 235.7451934814453,
      "learning_rate": 0.0002781033781033781,
      "loss": 575.3545,
      "step": 28665
    },
    {
      "epoch": 35.0,
      "eval_cer": 0.28026972286716395,
      "eval_loss": 458.68701171875,
      "eval_runtime": 25.4135,
      "eval_samples_per_second": 35.021,
      "eval_steps_per_second": 4.407,
      "eval_wer": 0.6829229797979798,
      "step": 28665
    },
    {
      "epoch": 36.0,
      "grad_norm": 363.6673278808594,
      "learning_rate": 0.00026700583367250035,
      "loss": 558.9386,
      "step": 29484
    },
    {
      "epoch": 36.0,
      "eval_cer": 0.27476164600108677,
      "eval_loss": 456.4928283691406,
      "eval_runtime": 24.0819,
      "eval_samples_per_second": 36.957,
      "eval_steps_per_second": 4.651,
      "eval_wer": 0.6876578282828283,
      "step": 29484
    },
    {
      "epoch": 37.0,
      "grad_norm": 305.44427490234375,
      "learning_rate": 0.00025589472256138925,
      "loss": 530.0468,
      "step": 30303
    },
    {
      "epoch": 37.0,
      "eval_cer": 0.2480610581435558,
      "eval_loss": 434.17950439453125,
      "eval_runtime": 23.6026,
      "eval_samples_per_second": 37.708,
      "eval_steps_per_second": 4.745,
      "eval_wer": 0.6513573232323232,
      "step": 30303
    },
    {
      "epoch": 38.0,
      "grad_norm": 457.92230224609375,
      "learning_rate": 0.0002447836114502781,
      "loss": 520.7852,
      "step": 31122
    },
    {
      "epoch": 38.0,
      "eval_cer": 0.25776811737390704,
      "eval_loss": 458.8715515136719,
      "eval_runtime": 24.0205,
      "eval_samples_per_second": 37.052,
      "eval_steps_per_second": 4.663,
      "eval_wer": 0.6710858585858586,
      "step": 31122
    },
    {
      "epoch": 39.0,
      "grad_norm": 327.72235107421875,
      "learning_rate": 0.000233672500339167,
      "loss": 503.4493,
      "step": 31941
    },
    {
      "epoch": 39.0,
      "eval_cer": 0.2560144247394161,
      "eval_loss": 450.276123046875,
      "eval_runtime": 23.7188,
      "eval_samples_per_second": 37.523,
      "eval_steps_per_second": 4.722,
      "eval_wer": 0.6609848484848485,
      "step": 31941
    },
    {
      "epoch": 40.0,
      "grad_norm": 708.286376953125,
      "learning_rate": 0.00022257495590828925,
      "loss": 484.1686,
      "step": 32760
    },
    {
      "epoch": 40.0,
      "eval_cer": 0.25863261374302227,
      "eval_loss": 438.8575134277344,
      "eval_runtime": 24.0264,
      "eval_samples_per_second": 37.043,
      "eval_steps_per_second": 4.662,
      "eval_wer": 0.6541982323232324,
      "step": 32760
    },
    {
      "epoch": 41.0,
      "grad_norm": 273.99664306640625,
      "learning_rate": 0.00021146384479717815,
      "loss": 467.5615,
      "step": 33579
    },
    {
      "epoch": 41.0,
      "eval_cer": 0.2578175171664279,
      "eval_loss": 469.5018615722656,
      "eval_runtime": 25.2591,
      "eval_samples_per_second": 35.235,
      "eval_steps_per_second": 4.434,
      "eval_wer": 0.6687184343434344,
      "step": 33579
    },
    {
      "epoch": 42.0,
      "grad_norm": 649.657958984375,
      "learning_rate": 0.00020035273368606703,
      "loss": 454.8417,
      "step": 34398
    },
    {
      "epoch": 42.0,
      "eval_cer": 0.25369263449093515,
      "eval_loss": 439.1047058105469,
      "eval_runtime": 24.7496,
      "eval_samples_per_second": 35.96,
      "eval_steps_per_second": 4.525,
      "eval_wer": 0.6535669191919192,
      "step": 34398
    },
    {
      "epoch": 43.0,
      "grad_norm": 460.6722717285156,
      "learning_rate": 0.00018924162257495593,
      "loss": 441.654,
      "step": 35217
    },
    {
      "epoch": 43.0,
      "eval_cer": 0.25929951094205406,
      "eval_loss": 486.397216796875,
      "eval_runtime": 23.9495,
      "eval_samples_per_second": 37.162,
      "eval_steps_per_second": 4.677,
      "eval_wer": 0.6527777777777778,
      "step": 35217
    },
    {
      "epoch": 44.0,
      "grad_norm": 714.3244018554688,
      "learning_rate": 0.00017813051146384478,
      "loss": 437.2555,
      "step": 36036
    },
    {
      "epoch": 44.0,
      "eval_cer": 0.2559897248431557,
      "eval_loss": 454.9423828125,
      "eval_runtime": 24.3558,
      "eval_samples_per_second": 36.542,
      "eval_steps_per_second": 4.598,
      "eval_wer": 0.6589330808080808,
      "step": 36036
    },
    {
      "epoch": 45.0,
      "grad_norm": 315.14154052734375,
      "learning_rate": 0.00016701940035273368,
      "loss": 413.9407,
      "step": 36855
    },
    {
      "epoch": 45.0,
      "eval_cer": 0.24739416094452404,
      "eval_loss": 478.7567443847656,
      "eval_runtime": 24.1415,
      "eval_samples_per_second": 36.866,
      "eval_steps_per_second": 4.639,
      "eval_wer": 0.6437815656565656,
      "step": 36855
    },
    {
      "epoch": 46.0,
      "grad_norm": 559.75048828125,
      "learning_rate": 0.00015592185592185593,
      "loss": 398.7603,
      "step": 37674
    },
    {
      "epoch": 46.0,
      "eval_cer": 0.2493701526453589,
      "eval_loss": 461.5147399902344,
      "eval_runtime": 24.4344,
      "eval_samples_per_second": 36.424,
      "eval_steps_per_second": 4.584,
      "eval_wer": 0.6463068181818182,
      "step": 37674
    },
    {
      "epoch": 47.0,
      "grad_norm": 181.94717407226562,
      "learning_rate": 0.00014481074481074484,
      "loss": 386.0169,
      "step": 38493
    },
    {
      "epoch": 47.0,
      "eval_cer": 0.24331867806155213,
      "eval_loss": 462.0274658203125,
      "eval_runtime": 24.4708,
      "eval_samples_per_second": 36.37,
      "eval_steps_per_second": 4.577,
      "eval_wer": 0.6365214646464646,
      "step": 38493
    },
    {
      "epoch": 48.0,
      "grad_norm": 429.9681701660156,
      "learning_rate": 0.0001336996336996337,
      "loss": 374.3441,
      "step": 39312
    },
    {
      "epoch": 48.0,
      "eval_cer": 0.24242948179617646,
      "eval_loss": 470.66925048828125,
      "eval_runtime": 24.2843,
      "eval_samples_per_second": 36.649,
      "eval_steps_per_second": 4.612,
      "eval_wer": 0.6289457070707071,
      "step": 39312
    },
    {
      "epoch": 49.0,
      "grad_norm": 8201.0947265625,
      "learning_rate": 0.00012258852258852259,
      "loss": 363.7627,
      "step": 40131
    },
    {
      "epoch": 49.0,
      "eval_cer": 0.25260583905547596,
      "eval_loss": 465.7492980957031,
      "eval_runtime": 24.917,
      "eval_samples_per_second": 35.719,
      "eval_steps_per_second": 4.495,
      "eval_wer": 0.6412563131313131,
      "step": 40131
    },
    {
      "epoch": 50.0,
      "grad_norm": 343.2039489746094,
      "learning_rate": 0.00011147741147741149,
      "loss": 349.4398,
      "step": 40950
    },
    {
      "epoch": 50.0,
      "eval_cer": 0.25156844341253765,
      "eval_loss": 478.899169921875,
      "eval_runtime": 25.7381,
      "eval_samples_per_second": 34.579,
      "eval_steps_per_second": 4.352,
      "eval_wer": 0.6346275252525253,
      "step": 40950
    },
    {
      "epoch": 51.0,
      "grad_norm": 1325.235107421875,
      "learning_rate": 0.00010039343372676705,
      "loss": 337.0751,
      "step": 41769
    },
    {
      "epoch": 51.0,
      "eval_cer": 0.23669910586375537,
      "eval_loss": 467.36248779296875,
      "eval_runtime": 23.3669,
      "eval_samples_per_second": 38.088,
      "eval_steps_per_second": 4.793,
      "eval_wer": 0.6174242424242424,
      "step": 41769
    },
    {
      "epoch": 52.0,
      "grad_norm": 234.5354766845703,
      "learning_rate": 8.928232261565595e-05,
      "loss": 330.7304,
      "step": 42588
    },
    {
      "epoch": 52.0,
      "eval_cer": 0.2420342834560095,
      "eval_loss": 497.57745361328125,
      "eval_runtime": 23.9343,
      "eval_samples_per_second": 37.185,
      "eval_steps_per_second": 4.679,
      "eval_wer": 0.6212121212121212,
      "step": 42588
    },
    {
      "epoch": 53.0,
      "grad_norm": 472.48504638671875,
      "learning_rate": 7.817121150454484e-05,
      "loss": 312.5302,
      "step": 43407
    },
    {
      "epoch": 53.0,
      "eval_cer": 0.24297287951390603,
      "eval_loss": 496.43292236328125,
      "eval_runtime": 26.1733,
      "eval_samples_per_second": 34.004,
      "eval_steps_per_second": 4.279,
      "eval_wer": 0.6242108585858586,
      "step": 43407
    },
    {
      "epoch": 54.0,
      "grad_norm": 168.4958953857422,
      "learning_rate": 6.706010039343373e-05,
      "loss": 309.1301,
      "step": 44226
    },
    {
      "epoch": 54.0,
      "eval_cer": 0.24124388677567554,
      "eval_loss": 521.0546875,
      "eval_runtime": 23.8596,
      "eval_samples_per_second": 37.302,
      "eval_steps_per_second": 4.694,
      "eval_wer": 0.6137941919191919,
      "step": 44226
    },
    {
      "epoch": 55.0,
      "grad_norm": 513.9391479492188,
      "learning_rate": 5.594898928232261e-05,
      "loss": 292.5729,
      "step": 45045
    },
    {
      "epoch": 55.0,
      "eval_cer": 0.23716840389270366,
      "eval_loss": 502.69549560546875,
      "eval_runtime": 23.3761,
      "eval_samples_per_second": 38.073,
      "eval_steps_per_second": 4.791,
      "eval_wer": 0.6122159090909091,
      "step": 45045
    },
    {
      "epoch": 56.0,
      "grad_norm": 518.4885864257812,
      "learning_rate": 4.485144485144485e-05,
      "loss": 286.7145,
      "step": 45864
    },
    {
      "epoch": 56.0,
      "eval_cer": 0.2365015066936719,
      "eval_loss": 504.39935302734375,
      "eval_runtime": 24.3587,
      "eval_samples_per_second": 36.537,
      "eval_steps_per_second": 4.598,
      "eval_wer": 0.6158459595959596,
      "step": 45864
    },
    {
      "epoch": 57.0,
      "grad_norm": 215.2507781982422,
      "learning_rate": 3.374033374033374e-05,
      "loss": 268.4251,
      "step": 46683
    },
    {
      "epoch": 57.0,
      "eval_cer": 0.23455021488909747,
      "eval_loss": 523.27880859375,
      "eval_runtime": 24.3318,
      "eval_samples_per_second": 36.578,
      "eval_steps_per_second": 4.603,
      "eval_wer": 0.6174242424242424,
      "step": 46683
    },
    {
      "epoch": 58.0,
      "grad_norm": 496.5767822265625,
      "learning_rate": 2.262922262922263e-05,
      "loss": 272.3521,
      "step": 47502
    },
    {
      "epoch": 58.0,
      "eval_cer": 0.23504421281430618,
      "eval_loss": 519.0136108398438,
      "eval_runtime": 25.3712,
      "eval_samples_per_second": 35.079,
      "eval_steps_per_second": 4.414,
      "eval_wer": 0.6142676767676768,
      "step": 47502
    },
    {
      "epoch": 59.0,
      "grad_norm": 224.57681274414062,
      "learning_rate": 1.1518111518111518e-05,
      "loss": 263.5849,
      "step": 48321
    },
    {
      "epoch": 59.0,
      "eval_cer": 0.2333399199723361,
      "eval_loss": 521.5694580078125,
      "eval_runtime": 25.3221,
      "eval_samples_per_second": 35.147,
      "eval_steps_per_second": 4.423,
      "eval_wer": 0.6092171717171717,
      "step": 48321
    },
    {
      "epoch": 60.0,
      "grad_norm": 511.5002136230469,
      "learning_rate": 4.205670872337539e-07,
      "loss": 259.7302,
      "step": 49140
    },
    {
      "epoch": 60.0,
      "eval_cer": 0.23321642049103394,
      "eval_loss": 526.7686767578125,
      "eval_runtime": 24.9818,
      "eval_samples_per_second": 35.626,
      "eval_steps_per_second": 4.483,
      "eval_wer": 0.6117424242424242,
      "step": 49140
    },
    {
      "epoch": 60.0,
      "step": 49140,
      "total_flos": 4.128908222553674e+19,
      "train_loss": 675.357091982092,
      "train_runtime": 28715.4368,
      "train_samples_per_second": 27.372,
      "train_steps_per_second": 1.711
    }
  ],
  "logging_steps": 500,
  "max_steps": 49140,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 60,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.128908222553674e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}