{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.8312570781426953,
  "eval_steps": 100,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.056625141562853906,
      "eval_loss": 3.596607208251953,
      "eval_runtime": 152.7562,
      "eval_samples_per_second": 37.026,
      "eval_steps_per_second": 4.628,
      "eval_wer": 1.0,
      "step": 100
    },
    {
      "epoch": 0.11325028312570781,
      "eval_loss": 3.5807790756225586,
      "eval_runtime": 150.6166,
      "eval_samples_per_second": 37.552,
      "eval_steps_per_second": 4.694,
      "eval_wer": 1.0,
      "step": 200
    },
    {
      "epoch": 0.16987542468856173,
      "eval_loss": 2.264209270477295,
      "eval_runtime": 151.9721,
      "eval_samples_per_second": 37.217,
      "eval_steps_per_second": 4.652,
      "eval_wer": 0.9883969122626823,
      "step": 300
    },
    {
      "epoch": 0.22650056625141562,
      "eval_loss": 1.0231246948242188,
      "eval_runtime": 152.2425,
      "eval_samples_per_second": 37.151,
      "eval_steps_per_second": 4.644,
      "eval_wer": 0.7425976151883296,
      "step": 400
    },
    {
      "epoch": 0.28312570781426954,
      "grad_norm": 1.5508322715759277,
      "learning_rate": 0.0002982,
      "loss": 3.6826,
      "step": 500
    },
    {
      "epoch": 0.28312570781426954,
      "eval_loss": 0.9161245822906494,
      "eval_runtime": 152.0295,
      "eval_samples_per_second": 37.203,
      "eval_steps_per_second": 4.65,
      "eval_wer": 0.645680537946751,
      "step": 500
    },
    {
      "epoch": 0.33975084937712347,
      "eval_loss": 0.7942017316818237,
      "eval_runtime": 151.7127,
      "eval_samples_per_second": 37.281,
      "eval_steps_per_second": 4.66,
      "eval_wer": 0.5827382003177609,
      "step": 600
    },
    {
      "epoch": 0.39637599093997733,
      "eval_loss": 0.7551485300064087,
      "eval_runtime": 152.436,
      "eval_samples_per_second": 37.104,
      "eval_steps_per_second": 4.638,
      "eval_wer": 0.5417021071720884,
      "step": 700
    },
    {
      "epoch": 0.45300113250283125,
      "eval_loss": 0.7242529392242432,
      "eval_runtime": 151.7599,
      "eval_samples_per_second": 37.269,
      "eval_steps_per_second": 4.659,
      "eval_wer": 0.5352345492770136,
      "step": 800
    },
    {
      "epoch": 0.5096262740656852,
      "eval_loss": 0.666481077671051,
      "eval_runtime": 152.1768,
      "eval_samples_per_second": 37.167,
      "eval_steps_per_second": 4.646,
      "eval_wer": 0.49851551090497664,
      "step": 900
    },
    {
      "epoch": 0.5662514156285391,
      "grad_norm": 1.319555401802063,
      "learning_rate": 0.00026693333333333333,
      "loss": 0.4714,
      "step": 1000
    },
    {
      "epoch": 0.5662514156285391,
      "eval_loss": 0.6402752995491028,
      "eval_runtime": 152.9026,
      "eval_samples_per_second": 36.991,
      "eval_steps_per_second": 4.624,
      "eval_wer": 0.48628653046813564,
      "step": 1000
    },
    {
      "epoch": 0.622876557191393,
      "eval_loss": 0.6219765543937683,
      "eval_runtime": 153.2975,
      "eval_samples_per_second": 36.896,
      "eval_steps_per_second": 4.612,
      "eval_wer": 0.4727415705092199,
      "step": 1100
    },
    {
      "epoch": 0.6795016987542469,
      "eval_loss": 0.6100151538848877,
      "eval_runtime": 152.7807,
      "eval_samples_per_second": 37.02,
      "eval_steps_per_second": 4.628,
      "eval_wer": 0.46418770361573397,
      "step": 1200
    },
    {
      "epoch": 0.7361268403171007,
      "eval_loss": 0.586254358291626,
      "eval_runtime": 152.4833,
      "eval_samples_per_second": 37.093,
      "eval_steps_per_second": 4.637,
      "eval_wer": 0.4441912342925005,
      "step": 1300
    },
    {
      "epoch": 0.7927519818799547,
      "eval_loss": 0.5667628645896912,
      "eval_runtime": 154.0765,
      "eval_samples_per_second": 36.709,
      "eval_steps_per_second": 4.589,
      "eval_wer": 0.43844586028149124,
      "step": 1400
    },
    {
      "epoch": 0.8493771234428086,
      "grad_norm": 1.6093908548355103,
      "learning_rate": 0.00023366666666666664,
      "loss": 0.3661,
      "step": 1500
    },
    {
      "epoch": 0.8493771234428086,
      "eval_loss": 0.5753483176231384,
      "eval_runtime": 154.2844,
      "eval_samples_per_second": 36.66,
      "eval_steps_per_second": 4.582,
      "eval_wer": 0.44213702235560337,
      "step": 1500
    },
    {
      "epoch": 0.9060022650056625,
      "eval_loss": 0.5641067624092102,
      "eval_runtime": 153.3007,
      "eval_samples_per_second": 36.895,
      "eval_steps_per_second": 4.612,
      "eval_wer": 0.43111168172553804,
      "step": 1600
    },
    {
      "epoch": 0.9626274065685164,
      "eval_loss": 0.5466340780258179,
      "eval_runtime": 154.2149,
      "eval_samples_per_second": 36.676,
      "eval_steps_per_second": 4.585,
      "eval_wer": 0.41525573333761295,
      "step": 1700
    },
    {
      "epoch": 1.0192525481313703,
      "eval_loss": 0.5303572416305542,
      "eval_runtime": 152.6129,
      "eval_samples_per_second": 37.061,
      "eval_steps_per_second": 4.633,
      "eval_wer": 0.40233666607822055,
      "step": 1800
    },
    {
      "epoch": 1.0758776896942241,
      "eval_loss": 0.5279697179794312,
      "eval_runtime": 153.0429,
      "eval_samples_per_second": 36.957,
      "eval_steps_per_second": 4.62,
      "eval_wer": 0.3938469933077627,
      "step": 1900
    },
    {
      "epoch": 1.1325028312570782,
      "grad_norm": 0.6171467900276184,
      "learning_rate": 0.00020046666666666667,
      "loss": 0.3069,
      "step": 2000
    },
    {
      "epoch": 1.1325028312570782,
      "eval_loss": 0.5219519138336182,
      "eval_runtime": 153.2427,
      "eval_samples_per_second": 36.909,
      "eval_steps_per_second": 4.614,
      "eval_wer": 0.3902200253566786,
      "step": 2000
    },
    {
      "epoch": 1.189127972819932,
      "eval_loss": 0.50642991065979,
      "eval_runtime": 153.5923,
      "eval_samples_per_second": 36.825,
      "eval_steps_per_second": 4.603,
      "eval_wer": 0.3928519844008281,
      "step": 2100
    },
    {
      "epoch": 1.245753114382786,
      "eval_loss": 0.5032854676246643,
      "eval_runtime": 152.9375,
      "eval_samples_per_second": 36.982,
      "eval_steps_per_second": 4.623,
      "eval_wer": 0.38232414822423005,
      "step": 2200
    },
    {
      "epoch": 1.3023782559456398,
      "eval_loss": 0.4920494556427002,
      "eval_runtime": 153.4126,
      "eval_samples_per_second": 36.868,
      "eval_steps_per_second": 4.608,
      "eval_wer": 0.37728490956652916,
      "step": 2300
    },
    {
      "epoch": 1.3590033975084936,
      "eval_loss": 0.49545198678970337,
      "eval_runtime": 153.0386,
      "eval_samples_per_second": 36.958,
      "eval_steps_per_second": 4.62,
      "eval_wer": 0.3736739901462021,
      "step": 2400
    },
    {
      "epoch": 1.4156285390713477,
      "grad_norm": 0.5370103716850281,
      "learning_rate": 0.0001672,
      "loss": 0.264,
      "step": 2500
    },
    {
      "epoch": 1.4156285390713477,
      "eval_loss": 0.4954512119293213,
      "eval_runtime": 152.6004,
      "eval_samples_per_second": 37.064,
      "eval_steps_per_second": 4.633,
      "eval_wer": 0.3739147181075573,
      "step": 2500
    },
    {
      "epoch": 1.4722536806342017,
      "eval_loss": 0.4845733046531677,
      "eval_runtime": 150.9099,
      "eval_samples_per_second": 37.479,
      "eval_steps_per_second": 4.685,
      "eval_wer": 0.3651040747219592,
      "step": 2600
    },
    {
      "epoch": 1.5288788221970555,
      "eval_loss": 0.47606581449508667,
      "eval_runtime": 150.159,
      "eval_samples_per_second": 37.667,
      "eval_steps_per_second": 4.708,
      "eval_wer": 0.36447818202243587,
      "step": 2700
    },
    {
      "epoch": 1.5855039637599093,
      "eval_loss": 0.4672282934188843,
      "eval_runtime": 151.2978,
      "eval_samples_per_second": 37.383,
      "eval_steps_per_second": 4.673,
      "eval_wer": 0.36266469804689383,
      "step": 2800
    },
    {
      "epoch": 1.6421291053227633,
      "eval_loss": 0.4644935131072998,
      "eval_runtime": 151.1778,
      "eval_samples_per_second": 37.413,
      "eval_steps_per_second": 4.677,
      "eval_wer": 0.36399672609972555,
      "step": 2900
    },
    {
      "epoch": 1.6987542468856174,
      "grad_norm": 1.252485990524292,
      "learning_rate": 0.00013386666666666665,
      "loss": 0.2498,
      "step": 3000
    },
    {
      "epoch": 1.6987542468856174,
      "eval_loss": 0.46515557169914246,
      "eval_runtime": 152.0904,
      "eval_samples_per_second": 37.188,
      "eval_steps_per_second": 4.649,
      "eval_wer": 0.3588772447882396,
      "step": 3000
    },
    {
      "epoch": 1.7553793884484712,
      "eval_loss": 0.45957261323928833,
      "eval_runtime": 153.0761,
      "eval_samples_per_second": 36.949,
      "eval_steps_per_second": 4.619,
      "eval_wer": 0.35335655020782847,
      "step": 3100
    },
    {
      "epoch": 1.812004530011325,
      "eval_loss": 0.450503408908844,
      "eval_runtime": 153.7448,
      "eval_samples_per_second": 36.788,
      "eval_steps_per_second": 4.599,
      "eval_wer": 0.35098136765579113,
      "step": 3200
    },
    {
      "epoch": 1.8686296715741788,
      "eval_loss": 0.4490334093570709,
      "eval_runtime": 152.7992,
      "eval_samples_per_second": 37.016,
      "eval_steps_per_second": 4.627,
      "eval_wer": 0.3522010559933238,
      "step": 3300
    },
    {
      "epoch": 1.9252548131370328,
      "eval_loss": 0.44600149989128113,
      "eval_runtime": 153.6597,
      "eval_samples_per_second": 36.809,
      "eval_steps_per_second": 4.601,
      "eval_wer": 0.34557301279067903,
      "step": 3400
    },
    {
      "epoch": 1.9818799546998869,
      "grad_norm": 0.38641658425331116,
      "learning_rate": 0.00010059999999999999,
      "loss": 0.2304,
      "step": 3500
    },
    {
      "epoch": 1.9818799546998869,
      "eval_loss": 0.44401755928993225,
      "eval_runtime": 152.5541,
      "eval_samples_per_second": 37.075,
      "eval_steps_per_second": 4.634,
      "eval_wer": 0.3404535314791931,
      "step": 3500
    },
    {
      "epoch": 2.0385050962627407,
      "eval_loss": 0.44931888580322266,
      "eval_runtime": 151.9569,
      "eval_samples_per_second": 37.221,
      "eval_steps_per_second": 4.653,
      "eval_wer": 0.3424756463545762,
      "step": 3600
    },
    {
      "epoch": 2.0951302378255945,
      "eval_loss": 0.4386073052883148,
      "eval_runtime": 152.2426,
      "eval_samples_per_second": 37.151,
      "eval_steps_per_second": 4.644,
      "eval_wer": 0.33743640769687533,
      "step": 3700
    },
    {
      "epoch": 2.1517553793884483,
      "eval_loss": 0.4424333870410919,
      "eval_runtime": 152.4948,
      "eval_samples_per_second": 37.09,
      "eval_steps_per_second": 4.636,
      "eval_wer": 0.33806230039639873,
      "step": 3800
    },
    {
      "epoch": 2.2083805209513026,
      "eval_loss": 0.43851807713508606,
      "eval_runtime": 151.5663,
      "eval_samples_per_second": 37.317,
      "eval_steps_per_second": 4.665,
      "eval_wer": 0.3372598738585483,
      "step": 3900
    },
    {
      "epoch": 2.2650056625141564,
      "grad_norm": 0.3740696609020233,
      "learning_rate": 6.739999999999998e-05,
      "loss": 0.1947,
      "step": 4000
    },
    {
      "epoch": 2.2650056625141564,
      "eval_loss": 0.4393535852432251,
      "eval_runtime": 151.3645,
      "eval_samples_per_second": 37.367,
      "eval_steps_per_second": 4.671,
      "eval_wer": 0.3354624384137632,
      "step": 4000
    },
    {
      "epoch": 2.32163080407701,
      "eval_loss": 0.4353909194469452,
      "eval_runtime": 151.9247,
      "eval_samples_per_second": 37.229,
      "eval_steps_per_second": 4.654,
      "eval_wer": 0.33242926610068846,
      "step": 4100
    },
    {
      "epoch": 2.378255945639864,
      "eval_loss": 0.4344673752784729,
      "eval_runtime": 151.7721,
      "eval_samples_per_second": 37.266,
      "eval_steps_per_second": 4.658,
      "eval_wer": 0.3324774116929595,
      "step": 4200
    },
    {
      "epoch": 2.434881087202718,
      "eval_loss": 0.4342668950557709,
      "eval_runtime": 151.7151,
      "eval_samples_per_second": 37.28,
      "eval_steps_per_second": 4.66,
      "eval_wer": 0.33311935292323985,
      "step": 4300
    },
    {
      "epoch": 2.491506228765572,
      "eval_loss": 0.4516645669937134,
      "eval_runtime": 150.7864,
      "eval_samples_per_second": 37.51,
      "eval_steps_per_second": 4.689,
      "eval_wer": 0.3349328368987819,
      "step": 4400
    },
    {
      "epoch": 2.548131370328426,
      "grad_norm": 1.0771362781524658,
      "learning_rate": 3.4133333333333334e-05,
      "loss": 0.1996,
      "step": 4500
    },
    {
      "epoch": 2.548131370328426,
      "eval_loss": 0.45709919929504395,
      "eval_runtime": 151.5098,
      "eval_samples_per_second": 37.331,
      "eval_steps_per_second": 4.666,
      "eval_wer": 0.3327662852465857,
      "step": 4500
    },
    {
      "epoch": 2.6047565118912797,
      "eval_loss": 0.44231659173965454,
      "eval_runtime": 151.9331,
      "eval_samples_per_second": 37.227,
      "eval_steps_per_second": 4.653,
      "eval_wer": 0.3254963008136605,
      "step": 4600
    },
    {
      "epoch": 2.661381653454134,
      "eval_loss": 0.44050487875938416,
      "eval_runtime": 151.3798,
      "eval_samples_per_second": 37.363,
      "eval_steps_per_second": 4.67,
      "eval_wer": 0.3253518640368474,
      "step": 4700
    },
    {
      "epoch": 2.7180067950169873,
      "eval_loss": 0.4436578154563904,
      "eval_runtime": 152.4074,
      "eval_samples_per_second": 37.111,
      "eval_steps_per_second": 4.639,
      "eval_wer": 0.3237791080226605,
      "step": 4800
    },
    {
      "epoch": 2.7746319365798415,
      "eval_loss": 0.44009852409362793,
      "eval_runtime": 152.2847,
      "eval_samples_per_second": 37.141,
      "eval_steps_per_second": 4.643,
      "eval_wer": 0.3232334579769222,
      "step": 4900
    },
    {
      "epoch": 2.8312570781426953,
      "grad_norm": 0.8818329572677612,
      "learning_rate": 7.999999999999999e-07,
      "loss": 0.2024,
      "step": 5000
    },
    {
      "epoch": 2.8312570781426953,
      "eval_loss": 0.44040822982788086,
      "eval_runtime": 151.6855,
      "eval_samples_per_second": 37.288,
      "eval_steps_per_second": 4.661,
      "eval_wer": 0.32349023446903435,
      "step": 5000
    },
    {
      "epoch": 2.8312570781426953,
      "step": 5000,
      "total_flos": 4.569633767265734e+19,
      "train_loss": 0.6167776947021484,
      "train_runtime": 19530.6251,
      "train_samples_per_second": 16.385,
      "train_steps_per_second": 0.256
    }
  ],
  "logging_steps": 500,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 400,
  "total_flos": 4.569633767265734e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}