{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12499782989878648,
  "eval_steps": 1000,
  "global_step": 7200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.736080970816479e-05,
      "grad_norm": 10.5625,
      "learning_rate": 2e-06,
      "loss": 1.0,
      "step": 1
    },
    {
      "epoch": 0.001736080970816479,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0002,
      "loss": 0.3311,
      "step": 100
    },
    {
      "epoch": 0.003472161941632958,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0004,
      "loss": 0.2169,
      "step": 200
    },
    {
      "epoch": 0.005208242912449436,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0006,
      "loss": 0.2032,
      "step": 300
    },
    {
      "epoch": 0.006944323883265916,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.0008,
      "loss": 0.188,
      "step": 400
    },
    {
      "epoch": 0.008680404854082394,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.001,
      "loss": 0.1758,
      "step": 500
    },
    {
      "epoch": 0.010416485824898873,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0012,
      "loss": 0.1637,
      "step": 600
    },
    {
      "epoch": 0.012152566795715351,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0014,
      "loss": 0.1518,
      "step": 700
    },
    {
      "epoch": 0.013888647766531832,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0016,
      "loss": 0.1485,
      "step": 800
    },
    {
      "epoch": 0.01562472873734831,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0018000000000000002,
      "loss": 0.1433,
      "step": 900
    },
    {
      "epoch": 0.01736080970816479,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.002,
      "loss": 0.139,
      "step": 1000
    },
    {
      "epoch": 0.01736080970816479,
      "eval_covost2-en-de_loss": 1.896493673324585,
      "eval_covost2-en-de_model_preparation_time": 0.0057,
      "eval_covost2-en-de_runtime": 9.8697,
      "eval_covost2-en-de_samples_per_second": 6.485,
      "eval_covost2-en-de_steps_per_second": 0.811,
      "step": 1000
    },
    {
      "epoch": 0.01736080970816479,
      "eval_covost2-zh-en_loss": 3.1452860832214355,
      "eval_covost2-zh-en_model_preparation_time": 0.0057,
      "eval_covost2-zh-en_runtime": 8.3732,
      "eval_covost2-zh-en_samples_per_second": 7.643,
      "eval_covost2-zh-en_steps_per_second": 0.955,
      "step": 1000
    },
    {
      "epoch": 0.01736080970816479,
      "eval_peoplespeech-clean-transcription_loss": 3.2206106185913086,
      "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0057,
      "eval_peoplespeech-clean-transcription_runtime": 9.6941,
      "eval_peoplespeech-clean-transcription_samples_per_second": 6.602,
      "eval_peoplespeech-clean-transcription_steps_per_second": 0.825,
      "step": 1000
    },
    {
      "epoch": 0.01909689067898127,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.001999725185109816,
      "loss": 0.1334,
      "step": 1100
    },
    {
      "epoch": 0.020832971649797746,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0019989008914857113,
      "loss": 0.1288,
      "step": 1200
    },
    {
      "epoch": 0.022569052620614226,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00199752757218401,
      "loss": 0.1262,
      "step": 1300
    },
    {
      "epoch": 0.024305133591430703,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.001995605982021898,
      "loss": 0.1222,
      "step": 1400
    },
    {
      "epoch": 0.026041214562247183,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0019931371771625545,
      "loss": 0.1193,
      "step": 1500
    },
    {
      "epoch": 0.027777295533063663,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.001990122514534651,
      "loss": 0.1196,
      "step": 1600
    },
    {
      "epoch": 0.02951337650388014,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0019865636510865464,
      "loss": 0.115,
      "step": 1700
    },
    {
      "epoch": 0.03124945747469662,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.001982462542875576,
      "loss": 0.115,
      "step": 1800
    },
    {
      "epoch": 0.0329855384455131,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.001977821443992945,
      "loss": 0.1125,
      "step": 1900
    },
    {
      "epoch": 0.03472161941632958,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.001972642905324813,
      "loss": 0.1094,
      "step": 2000
    },
    {
      "epoch": 0.03472161941632958,
      "eval_covost2-en-de_loss": 1.6700351238250732,
      "eval_covost2-en-de_model_preparation_time": 0.0057,
      "eval_covost2-en-de_runtime": 8.1279,
      "eval_covost2-en-de_samples_per_second": 7.874,
      "eval_covost2-en-de_steps_per_second": 0.984,
      "step": 2000
    },
    {
      "epoch": 0.03472161941632958,
      "eval_covost2-zh-en_loss": 3.093877077102661,
      "eval_covost2-zh-en_model_preparation_time": 0.0057,
      "eval_covost2-zh-en_runtime": 8.1488,
      "eval_covost2-zh-en_samples_per_second": 7.854,
      "eval_covost2-zh-en_steps_per_second": 0.982,
      "step": 2000
    },
    {
      "epoch": 0.03472161941632958,
      "eval_peoplespeech-clean-transcription_loss": 2.478968620300293,
      "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0057,
      "eval_peoplespeech-clean-transcription_runtime": 9.5507,
      "eval_peoplespeech-clean-transcription_samples_per_second": 6.701,
      "eval_peoplespeech-clean-transcription_steps_per_second": 0.838,
      "step": 2000
    },
    {
      "epoch": 0.036457700387146054,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0019669297731502505,
      "loss": 0.1077,
      "step": 2100
    },
    {
      "epoch": 0.03819378135796254,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00196068518757684,
      "loss": 0.1069,
      "step": 2200
    },
    {
      "epoch": 0.039929862328779014,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.001953912580814779,
      "loss": 0.1043,
      "step": 2300
    },
    {
      "epoch": 0.04166594329959549,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0019466156752904343,
      "loss": 0.1035,
      "step": 2400
    },
    {
      "epoch": 0.043402024270411975,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0019387984816003866,
      "loss": 0.1033,
      "step": 2500
    },
    {
      "epoch": 0.04513810524122845,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0019304652963070869,
      "loss": 0.102,
      "step": 2600
    },
    {
      "epoch": 0.04687418621204493,
      "grad_norm": 0.046875,
      "learning_rate": 0.0019216206995773372,
      "loss": 0.0998,
      "step": 2700
    },
    {
      "epoch": 0.048610267182861405,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0019122695526648968,
      "loss": 0.1002,
      "step": 2800
    },
    {
      "epoch": 0.05034634815367789,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0019024169952385887,
      "loss": 0.0978,
      "step": 2900
    },
    {
      "epoch": 0.052082429124494366,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0018920684425573864,
      "loss": 0.097,
      "step": 3000
    },
    {
      "epoch": 0.052082429124494366,
      "eval_covost2-en-de_loss": 1.749150276184082,
      "eval_covost2-en-de_model_preparation_time": 0.0057,
      "eval_covost2-en-de_runtime": 8.1948,
      "eval_covost2-en-de_samples_per_second": 7.81,
      "eval_covost2-en-de_steps_per_second": 0.976,
      "step": 3000
    },
    {
      "epoch": 0.052082429124494366,
      "eval_covost2-zh-en_loss": 3.198117971420288,
      "eval_covost2-zh-en_model_preparation_time": 0.0057,
      "eval_covost2-zh-en_runtime": 8.1979,
      "eval_covost2-zh-en_samples_per_second": 7.807,
      "eval_covost2-zh-en_steps_per_second": 0.976,
      "step": 3000
    },
    {
      "epoch": 0.052082429124494366,
      "eval_peoplespeech-clean-transcription_loss": 2.345036506652832,
      "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0057,
      "eval_peoplespeech-clean-transcription_runtime": 11.4402,
      "eval_peoplespeech-clean-transcription_samples_per_second": 5.594,
      "eval_peoplespeech-clean-transcription_steps_per_second": 0.699,
      "step": 3000
    },
    {
      "epoch": 0.05381851009531084,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0018812295824940284,
      "loss": 0.0955,
      "step": 3100
    },
    {
      "epoch": 0.055554591066127326,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0018699063724087904,
      "loss": 0.0951,
      "step": 3200
    },
    {
      "epoch": 0.0572906720369438,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0018581050358751443,
      "loss": 0.0947,
      "step": 3300
    },
    {
      "epoch": 0.05902675300776028,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0018458320592590974,
      "loss": 0.0939,
      "step": 3400
    },
    {
      "epoch": 0.060762833978576763,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0018330941881540914,
      "loss": 0.0941,
      "step": 3500
    },
    {
      "epoch": 0.06249891494939324,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0018198984236734246,
      "loss": 0.0927,
      "step": 3600
    },
    {
      "epoch": 0.06423499592020972,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0018062520186022297,
      "loss": 0.0948,
      "step": 3700
    },
    {
      "epoch": 0.0659710768910262,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0017921624734111292,
      "loss": 0.09,
      "step": 3800
    },
    {
      "epoch": 0.06770715786184267,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.001777637532133752,
      "loss": 0.0926,
      "step": 3900
    },
    {
      "epoch": 0.06944323883265915,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0017626851781103819,
      "loss": 0.0906,
      "step": 4000
    },
    {
      "epoch": 0.06944323883265915,
      "eval_covost2-en-de_loss": 1.7936017513275146,
      "eval_covost2-en-de_model_preparation_time": 0.0057,
      "eval_covost2-en-de_runtime": 8.0356,
      "eval_covost2-en-de_samples_per_second": 7.965,
      "eval_covost2-en-de_steps_per_second": 0.996,
      "step": 4000
    },
    {
      "epoch": 0.06944323883265915,
      "eval_covost2-zh-en_loss": 3.2699265480041504,
      "eval_covost2-zh-en_model_preparation_time": 0.0057,
      "eval_covost2-zh-en_runtime": 9.5779,
      "eval_covost2-zh-en_samples_per_second": 6.682,
      "eval_covost2-zh-en_steps_per_second": 0.835,
      "step": 4000
    },
    {
      "epoch": 0.06944323883265915,
      "eval_peoplespeech-clean-transcription_loss": 2.3380110263824463,
      "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0057,
      "eval_peoplespeech-clean-transcription_runtime": 9.5943,
      "eval_peoplespeech-clean-transcription_samples_per_second": 6.671,
      "eval_peoplespeech-clean-transcription_steps_per_second": 0.834,
      "step": 4000
    },
    {
      "epoch": 0.07117931980347564,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.001747313629600077,
      "loss": 0.0926,
      "step": 4100
    },
    {
      "epoch": 0.07291540077429211,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.001731531335263669,
      "loss": 0.0907,
      "step": 4200
    },
    {
      "epoch": 0.07465148174510859,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0017153469695201276,
      "loss": 0.0898,
      "step": 4300
    },
    {
      "epoch": 0.07638756271592508,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0016987694277788418,
      "loss": 0.0876,
      "step": 4400
    },
    {
      "epoch": 0.07812364368674155,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.001681807821550438,
      "loss": 0.0874,
      "step": 4500
    },
    {
      "epoch": 0.07985972465755803,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0016644714734388218,
      "loss": 0.0865,
      "step": 4600
    },
    {
      "epoch": 0.08159580562837451,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0016467699120171987,
      "loss": 0.0866,
      "step": 4700
    },
    {
      "epoch": 0.08333188659919098,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.001628712866590885,
      "loss": 0.0864,
      "step": 4800
    },
    {
      "epoch": 0.08506796757000747,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0016103102618497923,
      "loss": 0.0862,
      "step": 4900
    },
    {
      "epoch": 0.08680404854082395,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0015915722124135226,
      "loss": 0.0855,
      "step": 5000
    },
    {
      "epoch": 0.08680404854082395,
      "eval_covost2-en-de_loss": 1.7862941026687622,
      "eval_covost2-en-de_model_preparation_time": 0.0057,
      "eval_covost2-en-de_runtime": 8.2861,
      "eval_covost2-en-de_samples_per_second": 7.724,
      "eval_covost2-en-de_steps_per_second": 0.965,
      "step": 5000
    },
    {
      "epoch": 0.08680404854082395,
      "eval_covost2-zh-en_loss": 3.33290433883667,
      "eval_covost2-zh-en_model_preparation_time": 0.0057,
      "eval_covost2-zh-en_runtime": 8.4063,
      "eval_covost2-zh-en_samples_per_second": 7.613,
      "eval_covost2-zh-en_steps_per_second": 0.952,
      "step": 5000
    },
    {
      "epoch": 0.08680404854082395,
      "eval_peoplespeech-clean-transcription_loss": 2.2601113319396973,
      "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0057,
      "eval_peoplespeech-clean-transcription_runtime": 9.4946,
      "eval_peoplespeech-clean-transcription_samples_per_second": 6.741,
      "eval_peoplespeech-clean-transcription_steps_per_second": 0.843,
      "step": 5000
    },
    {
      "epoch": 0.08854012951164042,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.001572509017272072,
      "loss": 0.0872,
      "step": 5100
    },
    {
      "epoch": 0.0902762104824569,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0015531311541251993,
      "loss": 0.0859,
      "step": 5200
    },
    {
      "epoch": 0.09201229145327339,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0015334492736235703,
      "loss": 0.085,
      "step": 5300
    },
    {
      "epoch": 0.09374837242408986,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0015134741935148419,
      "loss": 0.0844,
      "step": 5400
    },
    {
      "epoch": 0.09548445339490634,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0014932168926979072,
      "loss": 0.0844,
      "step": 5500
    },
    {
      "epoch": 0.09722053436572281,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0014726885051885652,
      "loss": 0.0856,
      "step": 5600
    },
    {
      "epoch": 0.0989566153365393,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0014519003139999338,
      "loss": 0.0841,
      "step": 5700
    },
    {
      "epoch": 0.10069269630735578,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0014308637449409706,
      "loss": 0.0841,
      "step": 5800
    },
    {
      "epoch": 0.10242877727817225,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0014095903603365066,
      "loss": 0.0825,
      "step": 5900
    },
    {
      "epoch": 0.10416485824898873,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0013880918526722496,
      "loss": 0.0828,
      "step": 6000
    },
    {
      "epoch": 0.10416485824898873,
      "eval_covost2-en-de_loss": 1.8097732067108154,
      "eval_covost2-en-de_model_preparation_time": 0.0057,
      "eval_covost2-en-de_runtime": 8.2052,
      "eval_covost2-en-de_samples_per_second": 7.8,
      "eval_covost2-en-de_steps_per_second": 0.975,
      "step": 6000
    },
    {
      "epoch": 0.10416485824898873,
      "eval_covost2-zh-en_loss": 3.331326961517334,
      "eval_covost2-zh-en_model_preparation_time": 0.0057,
      "eval_covost2-zh-en_runtime": 8.2653,
      "eval_covost2-zh-en_samples_per_second": 7.743,
      "eval_covost2-zh-en_steps_per_second": 0.968,
      "step": 6000
    },
    {
      "epoch": 0.10416485824898873,
      "eval_peoplespeech-clean-transcription_loss": 2.250232219696045,
      "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0057,
      "eval_peoplespeech-clean-transcription_runtime": 9.4708,
      "eval_peoplespeech-clean-transcription_samples_per_second": 6.758,
      "eval_peoplespeech-clean-transcription_steps_per_second": 0.845,
      "step": 6000
    },
    {
      "epoch": 0.10590093921980522,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0013663800381682463,
      "loss": 0.0819,
      "step": 6100
    },
    {
      "epoch": 0.10763702019062169,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0013444668502843329,
      "loss": 0.08,
      "step": 6200
    },
    {
      "epoch": 0.10937310116143817,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0013223643331611537,
      "loss": 0.0805,
      "step": 6300
    },
    {
      "epoch": 0.11110918213225465,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.001300084635000341,
      "loss": 0.0799,
      "step": 6400
    },
    {
      "epoch": 0.11284526310307112,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0012776400013875004,
      "loss": 0.0807,
      "step": 6500
    },
    {
      "epoch": 0.1145813440738876,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0012550427685616766,
      "loss": 0.0799,
      "step": 6600
    },
    {
      "epoch": 0.11631742504470409,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0012323053566349834,
      "loss": 0.0802,
      "step": 6700
    },
    {
      "epoch": 0.11805350601552056,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0012094402627661448,
      "loss": 0.0796,
      "step": 6800
    },
    {
      "epoch": 0.11978958698633704,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0011864600542916813,
      "loss": 0.0784,
      "step": 6900
    },
    {
      "epoch": 0.12152566795715353,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0011633773618185302,
      "loss": 0.0808,
      "step": 7000
    },
    {
      "epoch": 0.12152566795715353,
      "eval_covost2-en-de_loss": 1.7786378860473633,
      "eval_covost2-en-de_model_preparation_time": 0.0057,
      "eval_covost2-en-de_runtime": 8.0291,
      "eval_covost2-en-de_samples_per_second": 7.971,
      "eval_covost2-en-de_steps_per_second": 0.996,
      "step": 7000
    },
    {
      "epoch": 0.12152566795715353,
      "eval_covost2-zh-en_loss": 3.273571252822876,
      "eval_covost2-zh-en_model_preparation_time": 0.0057,
      "eval_covost2-zh-en_runtime": 8.3234,
      "eval_covost2-zh-en_samples_per_second": 7.689,
      "eval_covost2-zh-en_steps_per_second": 0.961,
      "step": 7000
    },
    {
      "epoch": 0.12152566795715353,
      "eval_peoplespeech-clean-transcription_loss": 2.2290830612182617,
      "eval_peoplespeech-clean-transcription_model_preparation_time": 0.0057,
      "eval_peoplespeech-clean-transcription_runtime": 9.7693,
      "eval_peoplespeech-clean-transcription_samples_per_second": 6.551,
      "eval_peoplespeech-clean-transcription_steps_per_second": 0.819,
      "step": 7000
    },
    {
      "epoch": 0.12326174892797,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0011402048722818862,
      "loss": 0.0786,
      "step": 7100
    },
    {
      "epoch": 0.12499782989878648,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0011169553219720827,
      "loss": 0.0795,
      "step": 7200
    }
  ],
  "logging_steps": 100,
  "max_steps": 14400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3600,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.3101190265194086e+17,
  "train_batch_size": 24,
  "trial_name": null,
  "trial_params": null
}