trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 282,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 6.666666666666667e-05,
      "loss": 3.4338,
      "step": 94
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.4137182235717773,
      "eval_runtime": 36.2713,
      "eval_samples_per_second": 5.514,
      "eval_steps_per_second": 0.689,
      "step": 94
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 2.9565,
      "step": 188
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.173758029937744,
      "eval_runtime": 35.902,
      "eval_samples_per_second": 5.571,
      "eval_steps_per_second": 0.696,
      "step": 188
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0,
      "loss": 2.7101,
      "step": 282
    },
    {
      "epoch": 3.0,
      "eval_loss": 2.012174606323242,
      "eval_runtime": 35.6264,
      "eval_samples_per_second": 5.614,
      "eval_steps_per_second": 0.702,
      "step": 282
    },
    {
      "epoch": 3.0,
      "step": 282,
      "total_flos": 1.07525797056e+18,
      "train_loss": 3.0334743202155363,
      "train_runtime": 718.9206,
      "train_samples_per_second": 3.13,
      "train_steps_per_second": 0.392
    }
  ],
  "max_steps": 282,
  "num_train_epochs": 3,
  "total_flos": 1.07525797056e+18,
  "trial_name": null,
  "trial_params": null
}