trainer_state.json

{
  "best_metric": 56.786645065858984,
  "best_model_checkpoint": "./checkpoint-200",
  "epoch": 18.01,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.02,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 5.7448,
      "step": 25
    },
    {
      "epoch": 4.03,
      "learning_rate": 4.5e-06,
      "loss": 2.3533,
      "step": 50
    },
    {
      "epoch": 6.04,
      "learning_rate": 7e-06,
      "loss": 1.1794,
      "step": 75
    },
    {
      "epoch": 9.01,
      "learning_rate": 9.5e-06,
      "loss": 0.6219,
      "step": 100
    },
    {
      "epoch": 9.01,
      "eval_loss": 1.075757622718811,
      "eval_runtime": 695.7008,
      "eval_samples_per_second": 1.228,
      "eval_steps_per_second": 0.039,
      "eval_wer": 62.020145101077695,
      "step": 100
    },
    {
      "epoch": 11.02,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.2866,
      "step": 125
    },
    {
      "epoch": 13.04,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.1211,
      "step": 150
    },
    {
      "epoch": 15.05,
      "learning_rate": 3e-06,
      "loss": 0.0538,
      "step": 175
    },
    {
      "epoch": 18.01,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.0318,
      "step": 200
    },
    {
      "epoch": 18.01,
      "eval_loss": 1.1142734289169312,
      "eval_runtime": 624.7196,
      "eval_samples_per_second": 1.367,
      "eval_steps_per_second": 0.043,
      "eval_wer": 56.786645065858984,
      "step": 200
    },
    {
      "epoch": 18.01,
      "step": 200,
      "total_flos": 3.5380570079232e+18,
      "train_loss": 1.2990843003988266,
      "train_runtime": 5166.6053,
      "train_samples_per_second": 2.477,
      "train_steps_per_second": 0.039
    }
  ],
  "max_steps": 200,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 3.5380570079232e+18,
  "trial_name": null,
  "trial_params": null
}