trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 12630,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.2956,
      "step": 4210
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8818807339449541,
      "eval_loss": 0.3128007650375366,
      "eval_runtime": 6.3233,
      "eval_samples_per_second": 137.902,
      "eval_steps_per_second": 17.238,
      "step": 4210
    },
    {
      "epoch": 2.0,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.1746,
      "step": 8420
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8979357798165137,
      "eval_loss": 0.38503018021583557,
      "eval_runtime": 6.3387,
      "eval_samples_per_second": 137.568,
      "eval_steps_per_second": 17.196,
      "step": 8420
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0,
      "loss": 0.1204,
      "step": 12630
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8944954128440367,
      "eval_loss": 0.4674423038959503,
      "eval_runtime": 6.3626,
      "eval_samples_per_second": 137.05,
      "eval_steps_per_second": 17.131,
      "step": 12630
    },
    {
      "epoch": 3.0,
      "step": 12630,
      "total_flos": 3.5931840480405504e+16,
      "train_loss": 0.0,
      "train_runtime": 0.0496,
      "train_samples_per_second": 4071932.598,
      "train_steps_per_second": 254537.354
    }
  ],
  "max_steps": 12630,
  "num_train_epochs": 3,
  "total_flos": 3.5931840480405504e+16,
  "trial_name": null,
  "trial_params": null
}