trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "global_step": 725,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.69,
      "learning_rate": 3.75e-07,
      "loss": 1.2459,
      "step": 100
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.5e-07,
      "loss": 1.2298,
      "step": 200
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.125e-06,
      "loss": 1.19,
      "step": 300
    },
    {
      "epoch": 2.76,
      "learning_rate": 1.5e-06,
      "loss": 1.1671,
      "step": 400
    },
    {
      "epoch": 3.45,
      "learning_rate": 1.875e-06,
      "loss": 1.1224,
      "step": 500
    },
    {
      "epoch": 3.45,
      "eval_loss": 0.1640530526638031,
      "eval_runtime": 91.0313,
      "eval_samples_per_second": 18.093,
      "eval_steps_per_second": 9.052,
      "eval_wer": 0.13961801654580738,
      "step": 500
    },
    {
      "epoch": 4.14,
      "learning_rate": 2.25e-06,
      "loss": 1.1193,
      "step": 600
    },
    {
      "epoch": 4.83,
      "learning_rate": 2.625e-06,
      "loss": 1.1034,
      "step": 700
    },
    {
      "epoch": 5.0,
      "step": 725,
      "total_flos": 6.315878094644183e+17,
      "train_loss": 1.1678272247314454,
      "train_runtime": 1371.4783,
      "train_samples_per_second": 12.68,
      "train_steps_per_second": 0.529
    }
  ],
  "max_steps": 725,
  "num_train_epochs": 5,
  "total_flos": 6.315878094644183e+17,
  "trial_name": null,
  "trial_params": null
}