{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 50.0,
  "eval_steps": 500,
  "global_step": 4650,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 4.410810810810811,
      "grad_norm": 2.1526288986206055,
      "learning_rate": 4.987804878048781e-06,
      "loss": 1.5,
      "step": 410
    },
    {
      "epoch": 5.378378378378378,
      "eval_loss": 1.1479133367538452,
      "eval_runtime": 16.0814,
      "eval_samples_per_second": 20.396,
      "eval_steps_per_second": 2.55,
      "eval_wer": 0.9441779328390755,
      "step": 500
    },
    {
      "epoch": 8.821621621621622,
      "grad_norm": 1.133339524269104,
      "learning_rate": 4.517688679245283e-06,
      "loss": 1.2719,
      "step": 820
    },
    {
      "epoch": 10.756756756756756,
      "eval_loss": 1.126757025718689,
      "eval_runtime": 16.315,
      "eval_samples_per_second": 20.104,
      "eval_steps_per_second": 2.513,
      "eval_wer": 0.8634976013955517,
      "step": 1000
    },
    {
      "epoch": 13.227027027027027,
      "grad_norm": 2.221667766571045,
      "learning_rate": 4.034198113207547e-06,
      "loss": 1.1462,
      "step": 1230
    },
    {
      "epoch": 16.12972972972973,
      "eval_loss": 1.1268436908721924,
      "eval_runtime": 16.4501,
      "eval_samples_per_second": 19.939,
      "eval_steps_per_second": 2.492,
      "eval_wer": 0.8495420846053205,
      "step": 1500
    },
    {
      "epoch": 17.63783783783784,
      "grad_norm": 1.2538217306137085,
      "learning_rate": 3.5507075471698116e-06,
      "loss": 1.0711,
      "step": 1640
    },
    {
      "epoch": 21.508108108108107,
      "eval_loss": 1.1400920152664185,
      "eval_runtime": 16.4996,
      "eval_samples_per_second": 19.879,
      "eval_steps_per_second": 2.485,
      "eval_wer": 0.85041430440471,
      "step": 2000
    },
    {
      "epoch": 22.043243243243243,
      "grad_norm": 1.2463653087615967,
      "learning_rate": 3.067216981132076e-06,
      "loss": 1.0059,
      "step": 2050
    },
    {
      "epoch": 26.454054054054055,
      "grad_norm": 2.5332655906677246,
      "learning_rate": 2.5837264150943397e-06,
      "loss": 0.9435,
      "step": 2460
    },
    {
      "epoch": 26.886486486486486,
      "eval_loss": 1.1470223665237427,
      "eval_runtime": 16.2457,
      "eval_samples_per_second": 20.19,
      "eval_steps_per_second": 2.524,
      "eval_wer": 0.8438726559092892,
      "step": 2500
    },
    {
      "epoch": 30.864864864864863,
      "grad_norm": 1.5317901372909546,
      "learning_rate": 2.100235849056604e-06,
      "loss": 0.9056,
      "step": 2870
    },
    {
      "epoch": 32.25945945945946,
      "eval_loss": 1.1596176624298096,
      "eval_runtime": 15.9807,
      "eval_samples_per_second": 20.525,
      "eval_steps_per_second": 2.566,
      "eval_wer": 0.8360226777147841,
      "step": 3000
    },
    {
      "epoch": 35.270270270270274,
      "grad_norm": 1.1806330680847168,
      "learning_rate": 1.6167452830188683e-06,
      "loss": 0.8754,
      "step": 3280
    },
    {
      "epoch": 37.637837837837836,
      "eval_loss": 1.1706100702285767,
      "eval_runtime": 16.3502,
      "eval_samples_per_second": 20.061,
      "eval_steps_per_second": 2.508,
      "eval_wer": 0.840819886611426,
      "step": 3500
    },
    {
      "epoch": 39.68108108108108,
      "grad_norm": 1.8476966619491577,
      "learning_rate": 1.133254716981132e-06,
      "loss": 0.8491,
      "step": 3690
    },
    {
      "epoch": 43.01081081081081,
      "eval_loss": 1.1760590076446533,
      "eval_runtime": 16.1267,
      "eval_samples_per_second": 20.339,
      "eval_steps_per_second": 2.542,
      "eval_wer": 0.8403837767117314,
      "step": 4000
    },
    {
      "epoch": 44.086486486486486,
      "grad_norm": 1.2304221391677856,
      "learning_rate": 6.497641509433964e-07,
      "loss": 0.8404,
      "step": 4100
    },
    {
      "epoch": 48.38918918918919,
      "eval_loss": 1.1790771484375,
      "eval_runtime": 16.6689,
      "eval_samples_per_second": 19.677,
      "eval_steps_per_second": 2.46,
      "eval_wer": 0.8412559965111208,
      "step": 4500
    },
    {
      "epoch": 48.497297297297294,
      "grad_norm": 2.8257575035095215,
      "learning_rate": 1.6745283018867927e-07,
      "loss": 0.827,
      "step": 4510
    }
  ],
  "logging_steps": 410,
  "max_steps": 4650,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.9474157216000532e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}