{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.0,
  "global_step": 10050,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 9.832425638877252e-05,
      "loss": 16.2547,
      "step": 670
    },
    {
      "epoch": 1.0,
      "eval_cer": 1.0,
      "eval_loss": 3.4289255142211914,
      "eval_new_wer": 1.0,
      "eval_old_wer": 1.0,
      "eval_runtime": 6.8022,
      "eval_samples_per_second": 26.462,
      "eval_steps_per_second": 3.381,
      "step": 670
    },
    {
      "epoch": 2.0,
      "learning_rate": 9.130708001675745e-05,
      "loss": 2.7033,
      "step": 1340
    },
    {
      "epoch": 2.0,
      "eval_cer": 0.9161616161616162,
      "eval_loss": 2.50858211517334,
      "eval_new_wer": 0.8666666666666667,
      "eval_old_wer": 1.0,
      "eval_runtime": 6.7934,
      "eval_samples_per_second": 26.496,
      "eval_steps_per_second": 3.386,
      "step": 1340
    },
    {
      "epoch": 3.0,
      "learning_rate": 8.428990364474236e-05,
      "loss": 1.5437,
      "step": 2010
    },
    {
      "epoch": 3.0,
      "eval_cer": 0.5686868686868687,
      "eval_loss": 2.278925657272339,
      "eval_new_wer": 0.2611111111111111,
      "eval_old_wer": 0.9888888888888889,
      "eval_runtime": 6.812,
      "eval_samples_per_second": 26.424,
      "eval_steps_per_second": 3.376,
      "step": 2010
    },
    {
      "epoch": 4.0,
      "learning_rate": 7.727272727272727e-05,
      "loss": 0.8812,
      "step": 2680
    },
    {
      "epoch": 4.0,
      "eval_cer": 0.5575757575757576,
      "eval_loss": 2.634910821914673,
      "eval_new_wer": 0.2722222222222222,
      "eval_old_wer": 0.95,
      "eval_runtime": 6.7955,
      "eval_samples_per_second": 26.488,
      "eval_steps_per_second": 3.385,
      "step": 2680
    },
    {
      "epoch": 5.0,
      "learning_rate": 7.02555509007122e-05,
      "loss": 0.5765,
      "step": 3350
    },
    {
      "epoch": 5.0,
      "eval_cer": 0.5323232323232323,
      "eval_loss": 2.771409511566162,
      "eval_new_wer": 0.24444444444444444,
      "eval_old_wer": 0.9555555555555556,
      "eval_runtime": 6.8175,
      "eval_samples_per_second": 26.403,
      "eval_steps_per_second": 3.374,
      "step": 3350
    },
    {
      "epoch": 6.0,
      "learning_rate": 6.323837452869711e-05,
      "loss": 0.4065,
      "step": 4020
    },
    {
      "epoch": 6.0,
      "eval_cer": 0.509090909090909,
      "eval_loss": 2.738285779953003,
      "eval_new_wer": 0.2388888888888889,
      "eval_old_wer": 0.9222222222222223,
      "eval_runtime": 6.8211,
      "eval_samples_per_second": 26.389,
      "eval_steps_per_second": 3.372,
      "step": 4020
    },
    {
      "epoch": 7.0,
      "learning_rate": 5.622119815668203e-05,
      "loss": 0.2945,
      "step": 4690
    },
    {
      "epoch": 7.0,
      "eval_cer": 0.5454545454545454,
      "eval_loss": 3.8365838527679443,
      "eval_new_wer": 0.2388888888888889,
      "eval_old_wer": 0.9444444444444444,
      "eval_runtime": 6.7769,
      "eval_samples_per_second": 26.561,
      "eval_steps_per_second": 3.394,
      "step": 4690
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.920402178466695e-05,
      "loss": 0.2316,
      "step": 5360
    },
    {
      "epoch": 8.0,
      "eval_cer": 0.51010101010101,
      "eval_loss": 3.244825601577759,
      "eval_new_wer": 0.16111111111111112,
      "eval_old_wer": 0.9722222222222222,
      "eval_runtime": 6.796,
      "eval_samples_per_second": 26.486,
      "eval_steps_per_second": 3.384,
      "step": 5360
    },
    {
      "epoch": 9.0,
      "learning_rate": 4.2186845412651864e-05,
      "loss": 0.1838,
      "step": 6030
    },
    {
      "epoch": 9.0,
      "eval_cer": 0.494949494949495,
      "eval_loss": 3.7096900939941406,
      "eval_new_wer": 0.20555555555555555,
      "eval_old_wer": 0.9222222222222223,
      "eval_runtime": 6.8118,
      "eval_samples_per_second": 26.425,
      "eval_steps_per_second": 3.376,
      "step": 6030
    },
    {
      "epoch": 10.0,
      "learning_rate": 3.516966904063678e-05,
      "loss": 0.1523,
      "step": 6700
    },
    {
      "epoch": 10.0,
      "eval_cer": 0.5363636363636364,
      "eval_loss": 3.9365477561950684,
      "eval_new_wer": 0.24444444444444444,
      "eval_old_wer": 0.9222222222222223,
      "eval_runtime": 6.8589,
      "eval_samples_per_second": 26.243,
      "eval_steps_per_second": 3.353,
      "step": 6700
    },
    {
      "epoch": 11.0,
      "learning_rate": 2.8152492668621706e-05,
      "loss": 0.1235,
      "step": 7370
    },
    {
      "epoch": 11.0,
      "eval_cer": 0.5131313131313131,
      "eval_loss": 3.7984564304351807,
      "eval_new_wer": 0.2111111111111111,
      "eval_old_wer": 0.9444444444444444,
      "eval_runtime": 6.876,
      "eval_samples_per_second": 26.178,
      "eval_steps_per_second": 3.345,
      "step": 7370
    },
    {
      "epoch": 12.0,
      "learning_rate": 2.113531629660662e-05,
      "loss": 0.1013,
      "step": 8040
    },
    {
      "epoch": 12.0,
      "eval_cer": 0.5151515151515151,
      "eval_loss": 3.8724520206451416,
      "eval_new_wer": 0.25,
      "eval_old_wer": 0.9111111111111111,
      "eval_runtime": 6.8166,
      "eval_samples_per_second": 26.406,
      "eval_steps_per_second": 3.374,
      "step": 8040
    },
    {
      "epoch": 13.0,
      "learning_rate": 1.4118139924591539e-05,
      "loss": 0.0799,
      "step": 8710
    },
    {
      "epoch": 13.0,
      "eval_cer": 0.4919191919191919,
      "eval_loss": 3.5901904106140137,
      "eval_new_wer": 0.2111111111111111,
      "eval_old_wer": 0.9277777777777778,
      "eval_runtime": 6.8474,
      "eval_samples_per_second": 26.288,
      "eval_steps_per_second": 3.359,
      "step": 8710
    },
    {
      "epoch": 14.0,
      "learning_rate": 7.100963552576456e-06,
      "loss": 0.0617,
      "step": 9380
    },
    {
      "epoch": 14.0,
      "eval_cer": 0.5060606060606061,
      "eval_loss": 3.7018513679504395,
      "eval_new_wer": 0.2111111111111111,
      "eval_old_wer": 0.9055555555555556,
      "eval_runtime": 6.824,
      "eval_samples_per_second": 26.378,
      "eval_steps_per_second": 3.37,
      "step": 9380
    },
    {
      "epoch": 15.0,
      "learning_rate": 8.378718056137412e-08,
      "loss": 0.054,
      "step": 10050
    },
    {
      "epoch": 15.0,
      "eval_cer": 0.48787878787878786,
      "eval_loss": 3.622342824935913,
      "eval_new_wer": 0.19444444444444445,
      "eval_old_wer": 0.8833333333333333,
      "eval_runtime": 6.8734,
      "eval_samples_per_second": 26.188,
      "eval_steps_per_second": 3.346,
      "step": 10050
    }
  ],
  "max_steps": 10050,
  "num_train_epochs": 15,
  "total_flos": 7.467056434063227e+18,
  "trial_name": null,
  "trial_params": null
}