trainer_state.json

{
  "best_metric": 10.110413057544196,
  "best_model_checkpoint": "./checkpoint-100",
  "epoch": 1.2425,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.5086,
      "step": 25
    },
    {
      "epoch": 0.5,
      "learning_rate": 5.555555555555555e-07,
      "loss": 0.1889,
      "step": 50
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.1992899626493454,
      "eval_runtime": 1843.7009,
      "eval_samples_per_second": 1.685,
      "eval_steps_per_second": 0.106,
      "eval_wer": 10.373400680428293,
      "step": 50
    },
    {
      "epoch": 0.75,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 0.1663,
      "step": 75
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0,
      "loss": 0.1597,
      "step": 100
    },
    {
      "epoch": 1.24,
      "eval_loss": 0.1836123764514923,
      "eval_runtime": 1847.4417,
      "eval_samples_per_second": 1.682,
      "eval_steps_per_second": 0.106,
      "eval_wer": 10.110413057544196,
      "step": 100
    },
    {
      "epoch": 1.24,
      "step": 100,
      "total_flos": 1.84636939124736e+18,
      "train_loss": 0.25586883068084715,
      "train_runtime": 4627.7909,
      "train_samples_per_second": 1.383,
      "train_steps_per_second": 0.022
    }
  ],
  "logging_steps": 25,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 50,
  "total_flos": 1.84636939124736e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}