{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 29.997830802603037,
  "global_step": 6900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.43,
      "eval_loss": 4.412876605987549,
      "eval_runtime": 167.4094,
      "eval_samples_per_second": 29.264,
      "eval_steps_per_second": 0.92,
      "eval_wer": 1.0,
      "step": 100
    },
    {
      "epoch": 0.87,
      "eval_loss": 3.592686653137207,
      "eval_runtime": 162.2391,
      "eval_samples_per_second": 30.196,
      "eval_steps_per_second": 0.949,
      "eval_wer": 1.0,
      "step": 200
    },
    {
      "epoch": 1.3,
      "eval_loss": 3.3779923915863037,
      "eval_runtime": 162.9178,
      "eval_samples_per_second": 30.07,
      "eval_steps_per_second": 0.945,
      "eval_wer": 1.0,
      "step": 300
    },
    {
      "epoch": 1.74,
      "eval_loss": 3.0830302238464355,
      "eval_runtime": 165.7412,
      "eval_samples_per_second": 29.558,
      "eval_steps_per_second": 0.929,
      "eval_wer": 1.0,
      "step": 400
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.00029759999999999997,
      "loss": 5.3551,
      "step": 500
    },
    {
      "epoch": 2.17,
      "eval_loss": 2.6277503967285156,
      "eval_runtime": 162.7347,
      "eval_samples_per_second": 30.104,
      "eval_steps_per_second": 0.946,
      "eval_wer": 0.999864842947505,
      "step": 500
    },
    {
      "epoch": 2.61,
      "eval_loss": 1.8358852863311768,
      "eval_runtime": 165.3543,
      "eval_samples_per_second": 29.627,
      "eval_steps_per_second": 0.931,
      "eval_wer": 0.999972968589501,
      "step": 600
    },
    {
      "epoch": 3.04,
      "eval_loss": 1.7877882719039917,
      "eval_runtime": 161.7494,
      "eval_samples_per_second": 30.288,
      "eval_steps_per_second": 0.952,
      "eval_wer": 0.9914040114613181,
      "step": 700
    },
    {
      "epoch": 3.48,
      "eval_loss": 1.5219440460205078,
      "eval_runtime": 163.3701,
      "eval_samples_per_second": 29.987,
      "eval_steps_per_second": 0.943,
      "eval_wer": 0.9875114883494621,
      "step": 800
    },
    {
      "epoch": 3.91,
      "eval_loss": 1.4348260164260864,
      "eval_runtime": 163.9354,
      "eval_samples_per_second": 29.884,
      "eval_steps_per_second": 0.939,
      "eval_wer": 0.9879439909174461,
      "step": 900
    },
    {
      "epoch": 4.35,
      "learning_rate": 0.00027675,
      "loss": 1.7199,
      "step": 1000
    },
    {
      "epoch": 4.35,
      "eval_loss": 1.4353642463684082,
      "eval_runtime": 167.0923,
      "eval_samples_per_second": 29.319,
      "eval_steps_per_second": 0.922,
      "eval_wer": 0.9644266637833162,
      "step": 1000
    },
    {
      "epoch": 4.78,
      "eval_loss": 1.5210459232330322,
      "eval_runtime": 174.6297,
      "eval_samples_per_second": 28.054,
      "eval_steps_per_second": 0.882,
      "eval_wer": 0.9518840893117803,
      "step": 1100
    },
    {
      "epoch": 5.22,
      "eval_loss": 1.3606946468353271,
      "eval_runtime": 164.1085,
      "eval_samples_per_second": 29.852,
      "eval_steps_per_second": 0.938,
      "eval_wer": 0.9474779694004433,
      "step": 1200
    },
    {
      "epoch": 5.65,
      "eval_loss": 1.383901834487915,
      "eval_runtime": 163.0064,
      "eval_samples_per_second": 30.054,
      "eval_steps_per_second": 0.945,
      "eval_wer": 0.9343136724874304,
      "step": 1300
    },
    {
      "epoch": 6.09,
      "eval_loss": 1.2805912494659424,
      "eval_runtime": 161.8354,
      "eval_samples_per_second": 30.271,
      "eval_steps_per_second": 0.952,
      "eval_wer": 0.8944423420014056,
      "step": 1400
    },
    {
      "epoch": 6.52,
      "learning_rate": 0.00025331249999999996,
      "loss": 1.2342,
      "step": 1500
    },
    {
      "epoch": 6.52,
      "eval_loss": 1.3035786151885986,
      "eval_runtime": 161.6796,
      "eval_samples_per_second": 30.301,
      "eval_steps_per_second": 0.953,
      "eval_wer": 0.9011191003946586,
      "step": 1500
    },
    {
      "epoch": 6.95,
      "eval_loss": 1.3704484701156616,
      "eval_runtime": 162.7459,
      "eval_samples_per_second": 30.102,
      "eval_steps_per_second": 0.946,
      "eval_wer": 0.9072011677569336,
      "step": 1600
    },
    {
      "epoch": 7.39,
      "eval_loss": 1.298081636428833,
      "eval_runtime": 160.2526,
      "eval_samples_per_second": 30.57,
      "eval_steps_per_second": 0.961,
      "eval_wer": 0.8890630913121047,
      "step": 1700
    },
    {
      "epoch": 7.82,
      "eval_loss": 1.278607964515686,
      "eval_runtime": 160.6135,
      "eval_samples_per_second": 30.502,
      "eval_steps_per_second": 0.959,
      "eval_wer": 0.8733308104016868,
      "step": 1800
    },
    {
      "epoch": 8.26,
      "eval_loss": 1.2897309064865112,
      "eval_runtime": 160.1151,
      "eval_samples_per_second": 30.597,
      "eval_steps_per_second": 0.962,
      "eval_wer": 0.8866843271881927,
      "step": 1900
    },
    {
      "epoch": 8.69,
      "learning_rate": 0.00022987499999999996,
      "loss": 0.9831,
      "step": 2000
    },
    {
      "epoch": 8.69,
      "eval_loss": 1.4436370134353638,
      "eval_runtime": 161.3545,
      "eval_samples_per_second": 30.362,
      "eval_steps_per_second": 0.954,
      "eval_wer": 0.8779531815970157,
      "step": 2000
    },
    {
      "epoch": 9.13,
      "eval_loss": 1.3680068254470825,
      "eval_runtime": 162.428,
      "eval_samples_per_second": 30.161,
      "eval_steps_per_second": 0.948,
      "eval_wer": 0.8872519868086717,
      "step": 2100
    },
    {
      "epoch": 9.56,
      "eval_loss": 1.347055435180664,
      "eval_runtime": 168.1105,
      "eval_samples_per_second": 29.142,
      "eval_steps_per_second": 0.916,
      "eval_wer": 0.8692490674163378,
      "step": 2200
    },
    {
      "epoch": 10.0,
      "eval_loss": 1.3724839687347412,
      "eval_runtime": 161.7991,
      "eval_samples_per_second": 30.278,
      "eval_steps_per_second": 0.952,
      "eval_wer": 0.8729253392442018,
      "step": 2300
    },
    {
      "epoch": 10.43,
      "eval_loss": 1.4439136981964111,
      "eval_runtime": 164.3202,
      "eval_samples_per_second": 29.814,
      "eval_steps_per_second": 0.937,
      "eval_wer": 0.8770881764610478,
      "step": 2400
    },
    {
      "epoch": 10.87,
      "learning_rate": 0.00020643749999999997,
      "loss": 0.8071,
      "step": 2500
    },
    {
      "epoch": 10.87,
      "eval_loss": 1.5114314556121826,
      "eval_runtime": 160.915,
      "eval_samples_per_second": 30.445,
      "eval_steps_per_second": 0.957,
      "eval_wer": 0.8928474887819646,
      "step": 2500
    },
    {
      "epoch": 11.3,
      "eval_loss": 1.6155508756637573,
      "eval_runtime": 162.3974,
      "eval_samples_per_second": 30.167,
      "eval_steps_per_second": 0.948,
      "eval_wer": 0.8958479753473536,
      "step": 2600
    },
    {
      "epoch": 11.74,
      "eval_loss": 1.4381186962127686,
      "eval_runtime": 160.766,
      "eval_samples_per_second": 30.473,
      "eval_steps_per_second": 0.958,
      "eval_wer": 0.8748716008001297,
      "step": 2700
    },
    {
      "epoch": 12.17,
      "eval_loss": 1.508760929107666,
      "eval_runtime": 161.5608,
      "eval_samples_per_second": 30.323,
      "eval_steps_per_second": 0.953,
      "eval_wer": 0.8716818943612478,
      "step": 2800
    },
    {
      "epoch": 12.61,
      "eval_loss": 1.5485790967941284,
      "eval_runtime": 163.3707,
      "eval_samples_per_second": 29.987,
      "eval_steps_per_second": 0.943,
      "eval_wer": 0.8812510136778937,
      "step": 2900
    },
    {
      "epoch": 13.04,
      "learning_rate": 0.00018299999999999998,
      "loss": 0.6321,
      "step": 3000
    },
    {
      "epoch": 13.04,
      "eval_loss": 1.4535671472549438,
      "eval_runtime": 162.8299,
      "eval_samples_per_second": 30.087,
      "eval_steps_per_second": 0.946,
      "eval_wer": 0.8884143374601287,
      "step": 3000
    },
    {
      "epoch": 13.48,
      "eval_loss": 1.4679176807403564,
      "eval_runtime": 163.238,
      "eval_samples_per_second": 30.011,
      "eval_steps_per_second": 0.943,
      "eval_wer": 0.8947126561063956,
      "step": 3100
    },
    {
      "epoch": 13.91,
      "eval_loss": 1.5627696514129639,
      "eval_runtime": 163.36,
      "eval_samples_per_second": 29.989,
      "eval_steps_per_second": 0.943,
      "eval_wer": 0.9117424447207655,
      "step": 3200
    },
    {
      "epoch": 14.35,
      "eval_loss": 1.58307683467865,
      "eval_runtime": 166.5717,
      "eval_samples_per_second": 29.411,
      "eval_steps_per_second": 0.925,
      "eval_wer": 0.8716008001297507,
      "step": 3300
    },
    {
      "epoch": 14.78,
      "eval_loss": 1.6733046770095825,
      "eval_runtime": 162.8481,
      "eval_samples_per_second": 30.083,
      "eval_steps_per_second": 0.946,
      "eval_wer": 0.8701951667838028,
      "step": 3400
    },
    {
      "epoch": 15.22,
      "learning_rate": 0.00015956249999999998,
      "loss": 0.4998,
      "step": 3500
    },
    {
      "epoch": 15.22,
      "eval_loss": 1.82253897190094,
      "eval_runtime": 166.0482,
      "eval_samples_per_second": 29.503,
      "eval_steps_per_second": 0.927,
      "eval_wer": 0.8664648321349409,
      "step": 3500
    },
    {
      "epoch": 15.65,
      "eval_loss": 1.8557839393615723,
      "eval_runtime": 164.7566,
      "eval_samples_per_second": 29.735,
      "eval_steps_per_second": 0.935,
      "eval_wer": 0.8732226847596908,
      "step": 3600
    },
    {
      "epoch": 16.09,
      "eval_loss": 1.7512831687927246,
      "eval_runtime": 164.2229,
      "eval_samples_per_second": 29.831,
      "eval_steps_per_second": 0.938,
      "eval_wer": 0.8765745796615667,
      "step": 3700
    },
    {
      "epoch": 16.52,
      "eval_loss": 1.8561654090881348,
      "eval_runtime": 163.759,
      "eval_samples_per_second": 29.916,
      "eval_steps_per_second": 0.94,
      "eval_wer": 0.8753041033681137,
      "step": 3800
    },
    {
      "epoch": 16.95,
      "eval_loss": 1.9017548561096191,
      "eval_runtime": 163.9631,
      "eval_samples_per_second": 29.879,
      "eval_steps_per_second": 0.939,
      "eval_wer": 0.8703573552467968,
      "step": 3900
    },
    {
      "epoch": 17.39,
      "learning_rate": 0.000136125,
      "loss": 0.4421,
      "step": 4000
    },
    {
      "epoch": 17.39,
      "eval_loss": 1.9341310262680054,
      "eval_runtime": 162.0015,
      "eval_samples_per_second": 30.24,
      "eval_steps_per_second": 0.951,
      "eval_wer": 0.8788992809644808,
      "step": 4000
    },
    {
      "epoch": 17.82,
      "eval_loss": 1.9581764936447144,
      "eval_runtime": 163.9276,
      "eval_samples_per_second": 29.885,
      "eval_steps_per_second": 0.939,
      "eval_wer": 0.8781424014705087,
      "step": 4100
    },
    {
      "epoch": 18.26,
      "eval_loss": 1.8863332271575928,
      "eval_runtime": 162.6969,
      "eval_samples_per_second": 30.111,
      "eval_steps_per_second": 0.947,
      "eval_wer": 0.8820889874033627,
      "step": 4200
    },
    {
      "epoch": 18.69,
      "eval_loss": 1.9366161823272705,
      "eval_runtime": 163.4934,
      "eval_samples_per_second": 29.965,
      "eval_steps_per_second": 0.942,
      "eval_wer": 0.8846569714007677,
      "step": 4300
    },
    {
      "epoch": 19.13,
      "eval_loss": 2.1901650428771973,
      "eval_runtime": 166.8301,
      "eval_samples_per_second": 29.365,
      "eval_steps_per_second": 0.923,
      "eval_wer": 0.8721414283397307,
      "step": 4400
    },
    {
      "epoch": 19.56,
      "learning_rate": 0.00011268749999999998,
      "loss": 0.3712,
      "step": 4500
    },
    {
      "epoch": 19.56,
      "eval_loss": 2.164060592651367,
      "eval_runtime": 162.9234,
      "eval_samples_per_second": 30.069,
      "eval_steps_per_second": 0.945,
      "eval_wer": 0.8670054603449208,
      "step": 4500
    },
    {
      "epoch": 20.0,
      "eval_loss": 2.163888931274414,
      "eval_runtime": 164.0497,
      "eval_samples_per_second": 29.863,
      "eval_steps_per_second": 0.939,
      "eval_wer": 0.8776017732605287,
      "step": 4600
    },
    {
      "epoch": 20.43,
      "eval_loss": 2.269531726837158,
      "eval_runtime": 161.5615,
      "eval_samples_per_second": 30.323,
      "eval_steps_per_second": 0.953,
      "eval_wer": 0.9029842677190896,
      "step": 4700
    },
    {
      "epoch": 20.87,
      "eval_loss": 2.1909375190734863,
      "eval_runtime": 159.4363,
      "eval_samples_per_second": 30.727,
      "eval_steps_per_second": 0.966,
      "eval_wer": 0.8936584310969347,
      "step": 4800
    },
    {
      "epoch": 21.3,
      "eval_loss": 2.160627603530884,
      "eval_runtime": 160.8614,
      "eval_samples_per_second": 30.455,
      "eval_steps_per_second": 0.957,
      "eval_wer": 0.8959290695788507,
      "step": 4900
    },
    {
      "epoch": 21.74,
      "learning_rate": 8.924999999999999e-05,
      "loss": 0.3067,
      "step": 5000
    },
    {
      "epoch": 21.74,
      "eval_loss": 2.1755802631378174,
      "eval_runtime": 160.8247,
      "eval_samples_per_second": 30.462,
      "eval_steps_per_second": 0.958,
      "eval_wer": 0.8943342163594097,
      "step": 5000
    },
    {
      "epoch": 22.17,
      "eval_loss": 2.409205913543701,
      "eval_runtime": 160.8564,
      "eval_samples_per_second": 30.456,
      "eval_steps_per_second": 0.957,
      "eval_wer": 0.8772503649240417,
      "step": 5100
    },
    {
      "epoch": 22.61,
      "eval_loss": 2.499131202697754,
      "eval_runtime": 163.2327,
      "eval_samples_per_second": 30.012,
      "eval_steps_per_second": 0.943,
      "eval_wer": 0.8721143969292318,
      "step": 5200
    },
    {
      "epoch": 23.04,
      "eval_loss": 2.3339521884918213,
      "eval_runtime": 160.7999,
      "eval_samples_per_second": 30.466,
      "eval_steps_per_second": 0.958,
      "eval_wer": 0.8910363842785316,
      "step": 5300
    },
    {
      "epoch": 23.48,
      "eval_loss": 2.3566715717315674,
      "eval_runtime": 161.7409,
      "eval_samples_per_second": 30.289,
      "eval_steps_per_second": 0.952,
      "eval_wer": 0.8946045304643996,
      "step": 5400
    },
    {
      "epoch": 23.91,
      "learning_rate": 6.58125e-05,
      "loss": 0.2764,
      "step": 5500
    },
    {
      "epoch": 23.91,
      "eval_loss": 2.3214945793151855,
      "eval_runtime": 165.7293,
      "eval_samples_per_second": 29.56,
      "eval_steps_per_second": 0.929,
      "eval_wer": 0.8897118451640806,
      "step": 5500
    },
    {
      "epoch": 24.35,
      "eval_loss": 2.482433795928955,
      "eval_runtime": 158.7566,
      "eval_samples_per_second": 30.859,
      "eval_steps_per_second": 0.97,
      "eval_wer": 0.9002270638481916,
      "step": 5600
    },
    {
      "epoch": 24.78,
      "eval_loss": 2.4584639072418213,
      "eval_runtime": 160.826,
      "eval_samples_per_second": 30.461,
      "eval_steps_per_second": 0.958,
      "eval_wer": 0.8962534465048386,
      "step": 5700
    },
    {
      "epoch": 25.22,
      "eval_loss": 2.580402135848999,
      "eval_runtime": 161.3761,
      "eval_samples_per_second": 30.358,
      "eval_steps_per_second": 0.954,
      "eval_wer": 0.8878737092501486,
      "step": 5800
    },
    {
      "epoch": 25.65,
      "eval_loss": 2.5814249515533447,
      "eval_runtime": 161.7763,
      "eval_samples_per_second": 30.283,
      "eval_steps_per_second": 0.952,
      "eval_wer": 0.8902795047845596,
      "step": 5900
    },
    {
      "epoch": 26.09,
      "learning_rate": 4.237499999999999e-05,
      "loss": 0.2593,
      "step": 6000
    },
    {
      "epoch": 26.09,
      "eval_loss": 2.5374372005462646,
      "eval_runtime": 161.3184,
      "eval_samples_per_second": 30.369,
      "eval_steps_per_second": 0.955,
      "eval_wer": 0.8867654214196897,
      "step": 6000
    },
    {
      "epoch": 26.52,
      "eval_loss": 2.5346157550811768,
      "eval_runtime": 161.3913,
      "eval_samples_per_second": 30.355,
      "eval_steps_per_second": 0.954,
      "eval_wer": 0.8922257663404877,
      "step": 6100
    },
    {
      "epoch": 26.95,
      "eval_loss": 2.546508312225342,
      "eval_runtime": 161.2827,
      "eval_samples_per_second": 30.375,
      "eval_steps_per_second": 0.955,
      "eval_wer": 0.8873060496296696,
      "step": 6200
    },
    {
      "epoch": 27.39,
      "eval_loss": 2.6001508235931396,
      "eval_runtime": 161.5241,
      "eval_samples_per_second": 30.33,
      "eval_steps_per_second": 0.953,
      "eval_wer": 0.8919013894144997,
      "step": 6300
    },
    {
      "epoch": 27.82,
      "eval_loss": 2.610161304473877,
      "eval_runtime": 159.1491,
      "eval_samples_per_second": 30.782,
      "eval_steps_per_second": 0.968,
      "eval_wer": 0.8927663945504677,
      "step": 6400
    },
    {
      "epoch": 28.26,
      "learning_rate": 1.89375e-05,
      "loss": 0.227,
      "step": 6500
    },
    {
      "epoch": 28.26,
      "eval_loss": 2.692467451095581,
      "eval_runtime": 161.7243,
      "eval_samples_per_second": 30.292,
      "eval_steps_per_second": 0.952,
      "eval_wer": 0.8914418554360166,
      "step": 6500
    },
    {
      "epoch": 28.69,
      "eval_loss": 2.6981399059295654,
      "eval_runtime": 164.1482,
      "eval_samples_per_second": 29.845,
      "eval_steps_per_second": 0.938,
      "eval_wer": 0.8913337297940207,
      "step": 6600
    },
    {
      "epoch": 29.13,
      "eval_loss": 2.687201499938965,
      "eval_runtime": 161.4856,
      "eval_samples_per_second": 30.337,
      "eval_steps_per_second": 0.954,
      "eval_wer": 0.8890630913121047,
      "step": 6700
    },
    {
      "epoch": 29.56,
      "eval_loss": 2.7014529705047607,
      "eval_runtime": 161.2779,
      "eval_samples_per_second": 30.376,
      "eval_steps_per_second": 0.955,
      "eval_wer": 0.8896577823430827,
      "step": 6800
    },
    {
      "epoch": 30.0,
      "eval_loss": 2.711408853530884,
      "eval_runtime": 161.2987,
      "eval_samples_per_second": 30.372,
      "eval_steps_per_second": 0.955,
      "eval_wer": 0.8907931015840407,
      "step": 6900
    },
    {
      "epoch": 30.0,
      "step": 6900,
      "total_flos": 4.4600972514253696e+18,
      "train_loss": 0.9629449507118999,
      "train_runtime": 26035.0137,
      "train_samples_per_second": 16.973,
      "train_steps_per_second": 0.265
    }
  ],
  "max_steps": 6900,
  "num_train_epochs": 30,
  "total_flos": 4.4600972514253696e+18,
  "trial_name": null,
  "trial_params": null
}