{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.0,
  "global_step": 6525,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "eval_loss": 3.629563570022583,
      "eval_runtime": 72.0137,
      "eval_samples_per_second": 22.871,
      "eval_steps_per_second": 2.861,
      "eval_wer": 1.0,
      "step": 100
    },
    {
      "epoch": 0.46,
      "eval_loss": 3.158773899078369,
      "eval_runtime": 70.3627,
      "eval_samples_per_second": 23.407,
      "eval_steps_per_second": 2.928,
      "eval_wer": 0.9998978653865795,
      "step": 200
    },
    {
      "epoch": 0.69,
      "eval_loss": 2.3110687732696533,
      "eval_runtime": 70.6406,
      "eval_samples_per_second": 23.315,
      "eval_steps_per_second": 2.916,
      "eval_wer": 1.0082729036870595,
      "step": 300
    },
    {
      "epoch": 0.92,
      "eval_loss": 0.9851930141448975,
      "eval_runtime": 70.8686,
      "eval_samples_per_second": 23.24,
      "eval_steps_per_second": 2.907,
      "eval_wer": 0.7980798692676948,
      "step": 400
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002988,
      "loss": 3.6643,
      "step": 500
    },
    {
      "epoch": 1.15,
      "eval_loss": 0.7055537104606628,
      "eval_runtime": 71.6734,
      "eval_samples_per_second": 22.979,
      "eval_steps_per_second": 2.874,
      "eval_wer": 0.7362884281482994,
      "step": 500
    },
    {
      "epoch": 1.38,
      "eval_loss": 0.6146486401557922,
      "eval_runtime": 71.5696,
      "eval_samples_per_second": 23.013,
      "eval_steps_per_second": 2.878,
      "eval_wer": 0.6287406802165254,
      "step": 600
    },
    {
      "epoch": 1.61,
      "eval_loss": 0.5582674741744995,
      "eval_runtime": 71.6454,
      "eval_samples_per_second": 22.988,
      "eval_steps_per_second": 2.875,
      "eval_wer": 0.6195485650086815,
      "step": 700
    },
    {
      "epoch": 1.84,
      "eval_loss": 0.5528766512870789,
      "eval_runtime": 71.319,
      "eval_samples_per_second": 23.093,
      "eval_steps_per_second": 2.888,
      "eval_wer": 0.5677663160044939,
      "step": 800
    },
    {
      "epoch": 2.07,
      "eval_loss": 0.5279656052589417,
      "eval_runtime": 71.9775,
      "eval_samples_per_second": 22.882,
      "eval_steps_per_second": 2.862,
      "eval_wer": 0.5373302012051885,
      "step": 900
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.00027520331950207466,
      "loss": 0.5896,
      "step": 1000
    },
    {
      "epoch": 2.3,
      "eval_loss": 0.5253266096115112,
      "eval_runtime": 72.6484,
      "eval_samples_per_second": 22.671,
      "eval_steps_per_second": 2.836,
      "eval_wer": 0.5348789704830967,
      "step": 1000
    },
    {
      "epoch": 2.53,
      "eval_loss": 0.4802704155445099,
      "eval_runtime": 71.3938,
      "eval_samples_per_second": 23.069,
      "eval_steps_per_second": 2.885,
      "eval_wer": 0.5056684710448371,
      "step": 1100
    },
    {
      "epoch": 2.76,
      "eval_loss": 0.4561997056007385,
      "eval_runtime": 71.654,
      "eval_samples_per_second": 22.985,
      "eval_steps_per_second": 2.875,
      "eval_wer": 0.5132264324379532,
      "step": 1200
    },
    {
      "epoch": 2.99,
      "eval_loss": 0.42518743872642517,
      "eval_runtime": 72.1751,
      "eval_samples_per_second": 22.82,
      "eval_steps_per_second": 2.854,
      "eval_wer": 0.4872842406291492,
      "step": 1300
    },
    {
      "epoch": 3.22,
      "eval_loss": 0.4428342878818512,
      "eval_runtime": 71.8698,
      "eval_samples_per_second": 22.916,
      "eval_steps_per_second": 2.866,
      "eval_wer": 0.4830967214789092,
      "step": 1400
    },
    {
      "epoch": 3.45,
      "learning_rate": 0.0002503070539419087,
      "loss": 0.368,
      "step": 1500
    },
    {
      "epoch": 3.45,
      "eval_loss": 0.4509741961956024,
      "eval_runtime": 72.2303,
      "eval_samples_per_second": 22.802,
      "eval_steps_per_second": 2.852,
      "eval_wer": 0.4778878561944643,
      "step": 1500
    },
    {
      "epoch": 3.68,
      "eval_loss": 0.44042012095451355,
      "eval_runtime": 72.6999,
      "eval_samples_per_second": 22.655,
      "eval_steps_per_second": 2.834,
      "eval_wer": 0.4946379327954244,
      "step": 1600
    },
    {
      "epoch": 3.91,
      "eval_loss": 0.43302613496780396,
      "eval_runtime": 72.0106,
      "eval_samples_per_second": 22.872,
      "eval_steps_per_second": 2.861,
      "eval_wer": 0.47850066387498724,
      "step": 1700
    },
    {
      "epoch": 4.14,
      "eval_loss": 0.4357718229293823,
      "eval_runtime": 71.8959,
      "eval_samples_per_second": 22.908,
      "eval_steps_per_second": 2.865,
      "eval_wer": 0.45582677969563884,
      "step": 1800
    },
    {
      "epoch": 4.37,
      "eval_loss": 0.41258642077445984,
      "eval_runtime": 72.242,
      "eval_samples_per_second": 22.798,
      "eval_steps_per_second": 2.852,
      "eval_wer": 0.4643039526095394,
      "step": 1900
    },
    {
      "epoch": 4.6,
      "learning_rate": 0.00022541078838174273,
      "loss": 0.2629,
      "step": 2000
    },
    {
      "epoch": 4.6,
      "eval_loss": 0.4196760654449463,
      "eval_runtime": 71.8725,
      "eval_samples_per_second": 22.916,
      "eval_steps_per_second": 2.866,
      "eval_wer": 0.4528648759064447,
      "step": 2000
    },
    {
      "epoch": 4.83,
      "eval_loss": 0.4063510298728943,
      "eval_runtime": 72.3443,
      "eval_samples_per_second": 22.766,
      "eval_steps_per_second": 2.847,
      "eval_wer": 0.44091512613624756,
      "step": 2100
    },
    {
      "epoch": 5.06,
      "eval_loss": 0.4285483658313751,
      "eval_runtime": 72.2808,
      "eval_samples_per_second": 22.786,
      "eval_steps_per_second": 2.85,
      "eval_wer": 0.45143499131855785,
      "step": 2200
    },
    {
      "epoch": 5.29,
      "eval_loss": 0.41930803656578064,
      "eval_runtime": 72.0969,
      "eval_samples_per_second": 22.844,
      "eval_steps_per_second": 2.857,
      "eval_wer": 0.4203860688387294,
      "step": 2300
    },
    {
      "epoch": 5.52,
      "eval_loss": 0.4301486909389496,
      "eval_runtime": 72.1535,
      "eval_samples_per_second": 22.826,
      "eval_steps_per_second": 2.855,
      "eval_wer": 0.4219180880400368,
      "step": 2400
    },
    {
      "epoch": 5.75,
      "learning_rate": 0.00020051452282157675,
      "loss": 0.2072,
      "step": 2500
    },
    {
      "epoch": 5.75,
      "eval_loss": 0.4222012758255005,
      "eval_runtime": 71.9455,
      "eval_samples_per_second": 22.892,
      "eval_steps_per_second": 2.863,
      "eval_wer": 0.4334592993565519,
      "step": 2500
    },
    {
      "epoch": 5.98,
      "eval_loss": 0.4076746702194214,
      "eval_runtime": 71.6876,
      "eval_samples_per_second": 22.975,
      "eval_steps_per_second": 2.874,
      "eval_wer": 0.42314370340108265,
      "step": 2600
    },
    {
      "epoch": 6.21,
      "eval_loss": 0.4132108688354492,
      "eval_runtime": 71.6712,
      "eval_samples_per_second": 22.98,
      "eval_steps_per_second": 2.874,
      "eval_wer": 0.4121131651516699,
      "step": 2700
    },
    {
      "epoch": 6.44,
      "eval_loss": 0.4112888276576996,
      "eval_runtime": 71.9406,
      "eval_samples_per_second": 22.894,
      "eval_steps_per_second": 2.863,
      "eval_wer": 0.42202022265345723,
      "step": 2800
    },
    {
      "epoch": 6.67,
      "eval_loss": 0.41012728214263916,
      "eval_runtime": 71.6608,
      "eval_samples_per_second": 22.983,
      "eval_steps_per_second": 2.875,
      "eval_wer": 0.4175262996629558,
      "step": 2900
    },
    {
      "epoch": 6.9,
      "learning_rate": 0.0001756182572614108,
      "loss": 0.1731,
      "step": 3000
    },
    {
      "epoch": 6.9,
      "eval_loss": 0.42397525906562805,
      "eval_runtime": 71.8112,
      "eval_samples_per_second": 22.935,
      "eval_steps_per_second": 2.869,
      "eval_wer": 0.4122152997650904,
      "step": 3000
    },
    {
      "epoch": 7.13,
      "eval_loss": 0.43093591928482056,
      "eval_runtime": 71.844,
      "eval_samples_per_second": 22.925,
      "eval_steps_per_second": 2.867,
      "eval_wer": 0.40230824226330303,
      "step": 3100
    },
    {
      "epoch": 7.36,
      "eval_loss": 0.4274670481681824,
      "eval_runtime": 71.3955,
      "eval_samples_per_second": 23.069,
      "eval_steps_per_second": 2.885,
      "eval_wer": 0.3987335307935859,
      "step": 3200
    },
    {
      "epoch": 7.59,
      "eval_loss": 0.42890244722366333,
      "eval_runtime": 71.437,
      "eval_samples_per_second": 23.055,
      "eval_steps_per_second": 2.884,
      "eval_wer": 0.4062914921867021,
      "step": 3300
    },
    {
      "epoch": 7.82,
      "eval_loss": 0.4181130528450012,
      "eval_runtime": 71.7843,
      "eval_samples_per_second": 22.944,
      "eval_steps_per_second": 2.87,
      "eval_wer": 0.402512511490144,
      "step": 3400
    },
    {
      "epoch": 8.05,
      "learning_rate": 0.00015077178423236512,
      "loss": 0.1397,
      "step": 3500
    },
    {
      "epoch": 8.05,
      "eval_loss": 0.4489935040473938,
      "eval_runtime": 71.3608,
      "eval_samples_per_second": 23.08,
      "eval_steps_per_second": 2.887,
      "eval_wer": 0.3885200694515371,
      "step": 3500
    },
    {
      "epoch": 8.28,
      "eval_loss": 0.419808954000473,
      "eval_runtime": 70.9991,
      "eval_samples_per_second": 23.197,
      "eval_steps_per_second": 2.901,
      "eval_wer": 0.3871923194770708,
      "step": 3600
    },
    {
      "epoch": 8.51,
      "eval_loss": 0.3980425298213959,
      "eval_runtime": 70.9302,
      "eval_samples_per_second": 23.22,
      "eval_steps_per_second": 2.904,
      "eval_wer": 0.38423041568787664,
      "step": 3700
    },
    {
      "epoch": 8.74,
      "eval_loss": 0.4051460921764374,
      "eval_runtime": 71.3919,
      "eval_samples_per_second": 23.07,
      "eval_steps_per_second": 2.885,
      "eval_wer": 0.3876008579307527,
      "step": 3800
    },
    {
      "epoch": 8.97,
      "eval_loss": 0.4080096185207367,
      "eval_runtime": 71.5572,
      "eval_samples_per_second": 23.017,
      "eval_steps_per_second": 2.879,
      "eval_wer": 0.3911755694004698,
      "step": 3900
    },
    {
      "epoch": 9.2,
      "learning_rate": 0.00012587551867219917,
      "loss": 0.1224,
      "step": 4000
    },
    {
      "epoch": 9.2,
      "eval_loss": 0.41799476742744446,
      "eval_runtime": 71.2885,
      "eval_samples_per_second": 23.103,
      "eval_steps_per_second": 2.89,
      "eval_wer": 0.3773873965887039,
      "step": 4000
    },
    {
      "epoch": 9.43,
      "eval_loss": 0.4101611077785492,
      "eval_runtime": 71.7369,
      "eval_samples_per_second": 22.959,
      "eval_steps_per_second": 2.872,
      "eval_wer": 0.38198345419262586,
      "step": 4100
    },
    {
      "epoch": 9.66,
      "eval_loss": 0.3977714776992798,
      "eval_runtime": 71.6699,
      "eval_samples_per_second": 22.98,
      "eval_steps_per_second": 2.874,
      "eval_wer": 0.38800939638443466,
      "step": 4200
    },
    {
      "epoch": 9.89,
      "eval_loss": 0.415744423866272,
      "eval_runtime": 71.3403,
      "eval_samples_per_second": 23.087,
      "eval_steps_per_second": 2.888,
      "eval_wer": 0.37309774282504343,
      "step": 4300
    },
    {
      "epoch": 10.11,
      "eval_loss": 0.4174509644508362,
      "eval_runtime": 70.9527,
      "eval_samples_per_second": 23.213,
      "eval_steps_per_second": 2.903,
      "eval_wer": 0.3741190889592483,
      "step": 4400
    },
    {
      "epoch": 10.34,
      "learning_rate": 0.00010097925311203319,
      "loss": 0.1012,
      "step": 4500
    },
    {
      "epoch": 10.34,
      "eval_loss": 0.38867875933647156,
      "eval_runtime": 71.7567,
      "eval_samples_per_second": 22.953,
      "eval_steps_per_second": 2.871,
      "eval_wer": 0.37054437748953123,
      "step": 4500
    },
    {
      "epoch": 10.57,
      "eval_loss": 0.4063725471496582,
      "eval_runtime": 71.9349,
      "eval_samples_per_second": 22.896,
      "eval_steps_per_second": 2.864,
      "eval_wer": 0.3773873965887039,
      "step": 4600
    },
    {
      "epoch": 10.8,
      "eval_loss": 0.3960791826248169,
      "eval_runtime": 71.333,
      "eval_samples_per_second": 23.089,
      "eval_steps_per_second": 2.888,
      "eval_wer": 0.36216933918905114,
      "step": 4700
    },
    {
      "epoch": 11.03,
      "eval_loss": 0.3912270963191986,
      "eval_runtime": 71.311,
      "eval_samples_per_second": 23.096,
      "eval_steps_per_second": 2.889,
      "eval_wer": 0.3573690123582882,
      "step": 4800
    },
    {
      "epoch": 11.26,
      "eval_loss": 0.40203624963760376,
      "eval_runtime": 71.1318,
      "eval_samples_per_second": 23.154,
      "eval_steps_per_second": 2.896,
      "eval_wer": 0.36380349300377895,
      "step": 4900
    },
    {
      "epoch": 11.49,
      "learning_rate": 7.608298755186722e-05,
      "loss": 0.088,
      "step": 5000
    },
    {
      "epoch": 11.49,
      "eval_loss": 0.4116981625556946,
      "eval_runtime": 71.2238,
      "eval_samples_per_second": 23.124,
      "eval_steps_per_second": 2.892,
      "eval_wer": 0.3560412623838219,
      "step": 5000
    },
    {
      "epoch": 11.72,
      "eval_loss": 0.39163461327552795,
      "eval_runtime": 74.0308,
      "eval_samples_per_second": 22.247,
      "eval_steps_per_second": 2.783,
      "eval_wer": 0.3523644163006843,
      "step": 5100
    },
    {
      "epoch": 11.95,
      "eval_loss": 0.40122881531715393,
      "eval_runtime": 71.4186,
      "eval_samples_per_second": 23.061,
      "eval_steps_per_second": 2.884,
      "eval_wer": 0.3532836278214687,
      "step": 5200
    },
    {
      "epoch": 12.18,
      "eval_loss": 0.40852609276771545,
      "eval_runtime": 71.9077,
      "eval_samples_per_second": 22.904,
      "eval_steps_per_second": 2.865,
      "eval_wer": 0.3583903584924931,
      "step": 5300
    },
    {
      "epoch": 12.41,
      "eval_loss": 0.40001052618026733,
      "eval_runtime": 72.0262,
      "eval_samples_per_second": 22.867,
      "eval_steps_per_second": 2.86,
      "eval_wer": 0.35471351240935556,
      "step": 5400
    },
    {
      "epoch": 12.64,
      "learning_rate": 5.118672199170124e-05,
      "loss": 0.0775,
      "step": 5500
    },
    {
      "epoch": 12.64,
      "eval_loss": 0.41365721821784973,
      "eval_runtime": 72.4646,
      "eval_samples_per_second": 22.728,
      "eval_steps_per_second": 2.843,
      "eval_wer": 0.3524665509141048,
      "step": 5500
    },
    {
      "epoch": 12.87,
      "eval_loss": 0.4004632234573364,
      "eval_runtime": 72.0964,
      "eval_samples_per_second": 22.844,
      "eval_steps_per_second": 2.857,
      "eval_wer": 0.34664487794913695,
      "step": 5600
    },
    {
      "epoch": 13.1,
      "eval_loss": 0.39855289459228516,
      "eval_runtime": 72.3063,
      "eval_samples_per_second": 22.778,
      "eval_steps_per_second": 2.849,
      "eval_wer": 0.3478704933101828,
      "step": 5700
    },
    {
      "epoch": 13.33,
      "eval_loss": 0.3983455300331116,
      "eval_runtime": 72.5729,
      "eval_samples_per_second": 22.694,
      "eval_steps_per_second": 2.839,
      "eval_wer": 0.34695128178939844,
      "step": 5800
    },
    {
      "epoch": 13.56,
      "eval_loss": 0.39398157596588135,
      "eval_runtime": 72.6508,
      "eval_samples_per_second": 22.67,
      "eval_steps_per_second": 2.835,
      "eval_wer": 0.3428658972525789,
      "step": 5900
    },
    {
      "epoch": 13.79,
      "learning_rate": 2.6290456431535268e-05,
      "loss": 0.0716,
      "step": 6000
    },
    {
      "epoch": 13.79,
      "eval_loss": 0.3872474133968353,
      "eval_runtime": 72.1424,
      "eval_samples_per_second": 22.83,
      "eval_steps_per_second": 2.855,
      "eval_wer": 0.3382698396486569,
      "step": 6000
    },
    {
      "epoch": 14.02,
      "eval_loss": 0.40051060914993286,
      "eval_runtime": 72.2601,
      "eval_samples_per_second": 22.793,
      "eval_steps_per_second": 2.851,
      "eval_wer": 0.33837197426207744,
      "step": 6100
    },
    {
      "epoch": 14.25,
      "eval_loss": 0.4004524350166321,
      "eval_runtime": 72.3811,
      "eval_samples_per_second": 22.755,
      "eval_steps_per_second": 2.846,
      "eval_wer": 0.33632928199366763,
      "step": 6200
    },
    {
      "epoch": 14.48,
      "eval_loss": 0.3973444402217865,
      "eval_runtime": 72.4916,
      "eval_samples_per_second": 22.72,
      "eval_steps_per_second": 2.842,
      "eval_wer": 0.3357164743131447,
      "step": 6300
    },
    {
      "epoch": 14.71,
      "eval_loss": 0.39567238092422485,
      "eval_runtime": 71.9912,
      "eval_samples_per_second": 22.878,
      "eval_steps_per_second": 2.861,
      "eval_wer": 0.3346951281789398,
      "step": 6400
    },
    {
      "epoch": 14.94,
      "learning_rate": 1.3941908713692947e-06,
      "loss": 0.0639,
      "step": 6500
    },
    {
      "epoch": 14.94,
      "eval_loss": 0.3942253589630127,
      "eval_runtime": 72.1115,
      "eval_samples_per_second": 22.84,
      "eval_steps_per_second": 2.857,
      "eval_wer": 0.33398018588499645,
      "step": 6500
    },
    {
      "epoch": 15.0,
      "step": 6525,
      "total_flos": 6.375652211627562e+18,
      "train_loss": 0.4546195959770816,
      "train_runtime": 9820.6394,
      "train_samples_per_second": 5.312,
      "train_steps_per_second": 0.664
    }
  ],
  "max_steps": 6525,
  "num_train_epochs": 15,
  "total_flos": 6.375652211627562e+18,
  "trial_name": null,
  "trial_params": null
}