{
  "best_metric": 0.1275225579738617,
  "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-nyagen-male-model/checkpoint-1800",
  "epoch": 5.172413793103448,
  "eval_steps": 100,
  "global_step": 2100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.24630541871921183,
      "grad_norm": 1.8344100713729858,
      "learning_rate": 0.00028799999999999995,
      "loss": 6.345,
      "step": 100
    },
    {
      "epoch": 0.24630541871921183,
      "eval_loss": 0.45057007670402527,
      "eval_runtime": 16.3753,
      "eval_samples_per_second": 11.236,
      "eval_steps_per_second": 2.809,
      "eval_wer": 0.4025974025974026,
      "step": 100
    },
    {
      "epoch": 0.49261083743842365,
      "grad_norm": 1.3397903442382812,
      "learning_rate": 0.0002976158940397351,
      "loss": 0.353,
      "step": 200
    },
    {
      "epoch": 0.49261083743842365,
      "eval_loss": 0.2066199630498886,
      "eval_runtime": 16.5098,
      "eval_samples_per_second": 11.145,
      "eval_steps_per_second": 2.786,
      "eval_wer": 0.2924045651318379,
      "step": 200
    },
    {
      "epoch": 0.7389162561576355,
      "grad_norm": 68.3929672241211,
      "learning_rate": 0.00029513245033112583,
      "loss": 0.2989,
      "step": 300
    },
    {
      "epoch": 0.7389162561576355,
      "eval_loss": 0.18291255831718445,
      "eval_runtime": 16.4355,
      "eval_samples_per_second": 11.195,
      "eval_steps_per_second": 2.799,
      "eval_wer": 0.2715466351829988,
      "step": 300
    },
    {
      "epoch": 0.9852216748768473,
      "grad_norm": 1.0996179580688477,
      "learning_rate": 0.0002926490066225165,
      "loss": 0.2402,
      "step": 400
    },
    {
      "epoch": 0.9852216748768473,
      "eval_loss": 0.16742299497127533,
      "eval_runtime": 16.3654,
      "eval_samples_per_second": 11.243,
      "eval_steps_per_second": 2.811,
      "eval_wer": 0.2546241637150728,
      "step": 400
    },
    {
      "epoch": 1.2315270935960592,
      "grad_norm": 0.8030900359153748,
      "learning_rate": 0.00029016556291390727,
      "loss": 0.2302,
      "step": 500
    },
    {
      "epoch": 1.2315270935960592,
      "eval_loss": 0.15821850299835205,
      "eval_runtime": 16.4101,
      "eval_samples_per_second": 11.213,
      "eval_steps_per_second": 2.803,
      "eval_wer": 0.24360487996851632,
      "step": 500
    },
    {
      "epoch": 1.477832512315271,
      "grad_norm": 0.7671955227851868,
      "learning_rate": 0.00028768211920529796,
      "loss": 0.2279,
      "step": 600
    },
    {
      "epoch": 1.477832512315271,
      "eval_loss": 0.16019868850708008,
      "eval_runtime": 16.6478,
      "eval_samples_per_second": 11.053,
      "eval_steps_per_second": 2.763,
      "eval_wer": 0.24281778827233372,
      "step": 600
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 0.7692348957061768,
      "learning_rate": 0.0002851986754966887,
      "loss": 0.2264,
      "step": 700
    },
    {
      "epoch": 1.7241379310344827,
      "eval_loss": 0.15070533752441406,
      "eval_runtime": 16.3912,
      "eval_samples_per_second": 11.226,
      "eval_steps_per_second": 2.806,
      "eval_wer": 0.22865013774104684,
      "step": 700
    },
    {
      "epoch": 1.9704433497536946,
      "grad_norm": 0.5601118206977844,
      "learning_rate": 0.00028271523178807946,
      "loss": 0.2276,
      "step": 800
    },
    {
      "epoch": 1.9704433497536946,
      "eval_loss": 0.14960433542728424,
      "eval_runtime": 16.3656,
      "eval_samples_per_second": 11.243,
      "eval_steps_per_second": 2.811,
      "eval_wer": 0.23534041715859896,
      "step": 800
    },
    {
      "epoch": 2.2167487684729066,
      "grad_norm": 0.7260679006576538,
      "learning_rate": 0.00028023178807947015,
      "loss": 0.2088,
      "step": 900
    },
    {
      "epoch": 2.2167487684729066,
      "eval_loss": 0.1460462063550949,
      "eval_runtime": 16.4242,
      "eval_samples_per_second": 11.203,
      "eval_steps_per_second": 2.801,
      "eval_wer": 0.22077922077922077,
      "step": 900
    },
    {
      "epoch": 2.4630541871921183,
      "grad_norm": 0.4174855649471283,
      "learning_rate": 0.0002777483443708609,
      "loss": 0.1881,
      "step": 1000
    },
    {
      "epoch": 2.4630541871921183,
      "eval_loss": 0.14552360773086548,
      "eval_runtime": 16.4841,
      "eval_samples_per_second": 11.162,
      "eval_steps_per_second": 2.791,
      "eval_wer": 0.21645021645021645,
      "step": 1000
    },
    {
      "epoch": 2.70935960591133,
      "grad_norm": 2.906888484954834,
      "learning_rate": 0.00027526490066225165,
      "loss": 0.2079,
      "step": 1100
    },
    {
      "epoch": 2.70935960591133,
      "eval_loss": 0.14178700745105743,
      "eval_runtime": 16.3484,
      "eval_samples_per_second": 11.255,
      "eval_steps_per_second": 2.814,
      "eval_wer": 0.21684376229830776,
      "step": 1100
    },
    {
      "epoch": 2.955665024630542,
      "grad_norm": 8.858180046081543,
      "learning_rate": 0.0002727814569536424,
      "loss": 0.196,
      "step": 1200
    },
    {
      "epoch": 2.955665024630542,
      "eval_loss": 0.140374094247818,
      "eval_runtime": 16.4414,
      "eval_samples_per_second": 11.191,
      "eval_steps_per_second": 2.798,
      "eval_wer": 0.2085792994883904,
      "step": 1200
    },
    {
      "epoch": 3.2019704433497536,
      "grad_norm": 0.8277586102485657,
      "learning_rate": 0.0002702980132450331,
      "loss": 0.1782,
      "step": 1300
    },
    {
      "epoch": 3.2019704433497536,
      "eval_loss": 0.13734222948551178,
      "eval_runtime": 16.5713,
      "eval_samples_per_second": 11.104,
      "eval_steps_per_second": 2.776,
      "eval_wer": 0.2077922077922078,
      "step": 1300
    },
    {
      "epoch": 3.4482758620689653,
      "grad_norm": 0.6015973091125488,
      "learning_rate": 0.0002678145695364238,
      "loss": 0.1741,
      "step": 1400
    },
    {
      "epoch": 3.4482758620689653,
      "eval_loss": 0.13425125181674957,
      "eval_runtime": 16.5286,
      "eval_samples_per_second": 11.132,
      "eval_steps_per_second": 2.783,
      "eval_wer": 0.1944116489571035,
      "step": 1400
    },
    {
      "epoch": 3.6945812807881775,
      "grad_norm": 0.8872548341751099,
      "learning_rate": 0.00026533112582781453,
      "loss": 0.1948,
      "step": 1500
    },
    {
      "epoch": 3.6945812807881775,
      "eval_loss": 0.1318366825580597,
      "eval_runtime": 16.3284,
      "eval_samples_per_second": 11.269,
      "eval_steps_per_second": 2.817,
      "eval_wer": 0.21369539551357733,
      "step": 1500
    },
    {
      "epoch": 3.9408866995073892,
      "grad_norm": 1.193844199180603,
      "learning_rate": 0.0002628476821192053,
      "loss": 0.1904,
      "step": 1600
    },
    {
      "epoch": 3.9408866995073892,
      "eval_loss": 0.13069555163383484,
      "eval_runtime": 16.3618,
      "eval_samples_per_second": 11.246,
      "eval_steps_per_second": 2.811,
      "eval_wer": 0.20425029515938606,
      "step": 1600
    },
    {
      "epoch": 4.187192118226601,
      "grad_norm": 0.6119738221168518,
      "learning_rate": 0.000260364238410596,
      "loss": 0.1762,
      "step": 1700
    },
    {
      "epoch": 4.187192118226601,
      "eval_loss": 0.13134504854679108,
      "eval_runtime": 16.468,
      "eval_samples_per_second": 11.173,
      "eval_steps_per_second": 2.793,
      "eval_wer": 0.20031483667847305,
      "step": 1700
    },
    {
      "epoch": 4.433497536945813,
      "grad_norm": 0.7769166827201843,
      "learning_rate": 0.0002578807947019867,
      "loss": 0.1718,
      "step": 1800
    },
    {
      "epoch": 4.433497536945813,
      "eval_loss": 0.1275225579738617,
      "eval_runtime": 16.422,
      "eval_samples_per_second": 11.204,
      "eval_steps_per_second": 2.801,
      "eval_wer": 0.19323101141282958,
      "step": 1800
    },
    {
      "epoch": 4.679802955665025,
      "grad_norm": 0.7259223461151123,
      "learning_rate": 0.00025539735099337746,
      "loss": 0.1595,
      "step": 1900
    },
    {
      "epoch": 4.679802955665025,
      "eval_loss": 0.1275661438703537,
      "eval_runtime": 16.3823,
      "eval_samples_per_second": 11.232,
      "eval_steps_per_second": 2.808,
      "eval_wer": 0.1951987406532861,
      "step": 1900
    },
    {
      "epoch": 4.926108374384237,
      "grad_norm": 0.5679446458816528,
      "learning_rate": 0.0002529139072847682,
      "loss": 0.1811,
      "step": 2000
    },
    {
      "epoch": 4.926108374384237,
      "eval_loss": 0.12904822826385498,
      "eval_runtime": 16.3388,
      "eval_samples_per_second": 11.262,
      "eval_steps_per_second": 2.815,
      "eval_wer": 0.19834710743801653,
      "step": 2000
    },
    {
      "epoch": 5.172413793103448,
      "grad_norm": 0.6480135917663574,
      "learning_rate": 0.00025043046357615896,
      "loss": 0.1477,
      "step": 2100
    },
    {
      "epoch": 5.172413793103448,
      "eval_loss": 0.12981154024600983,
      "eval_runtime": 16.4531,
      "eval_samples_per_second": 11.183,
      "eval_steps_per_second": 2.796,
      "eval_wer": 0.1951987406532861,
      "step": 2100
    },
    {
      "epoch": 5.172413793103448,
      "step": 2100,
      "total_flos": 7.341751672651536e+18,
      "train_loss": 0.5011415631430489,
      "train_runtime": 1836.4834,
      "train_samples_per_second": 26.529,
      "train_steps_per_second": 6.632
    }
  ],
  "logging_steps": 100,
  "max_steps": 12180,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 400,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 2
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.341751672651536e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}