trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.001288992008249549,
  "eval_steps": 5,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 5e-05,
      "loss": 2.068,
      "step": 5
    },
    {
      "epoch": 0.0,
      "eval_accuracy": 0.6649938341745655,
      "eval_loss": 1.775834321975708,
      "eval_runtime": 78.6312,
      "eval_samples_per_second": 197.441,
      "eval_steps_per_second": 24.685,
      "step": 5
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0,
      "loss": 1.9159,
      "step": 10
    },
    {
      "epoch": 0.0,
      "eval_accuracy": 0.6735738398990131,
      "eval_loss": 1.719245433807373,
      "eval_runtime": 79.5982,
      "eval_samples_per_second": 195.042,
      "eval_steps_per_second": 24.385,
      "step": 10
    },
    {
      "epoch": 0.0,
      "step": 10,
      "total_flos": 42112770048000.0,
      "train_loss": 1.9919448852539063,
      "train_runtime": 161.1653,
      "train_samples_per_second": 0.993,
      "train_steps_per_second": 0.062
    }
  ],
  "logging_steps": 5,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 42112770048000.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}