trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0012890750886239124,
  "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": 2.0686,
      "step": 5
    },
    {
      "epoch": 0.0,
      "eval_accuracy": 0.6686630088781412,
      "eval_loss": 1.7577065229415894,
      "eval_runtime": 76.2786,
      "eval_samples_per_second": 203.465,
      "eval_steps_per_second": 25.433,
      "step": 5
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0,
      "loss": 1.8893,
      "step": 10
    },
    {
      "epoch": 0.0,
      "eval_accuracy": 0.6742729158223068,
      "eval_loss": 1.6961551904678345,
      "eval_runtime": 76.8494,
      "eval_samples_per_second": 201.953,
      "eval_steps_per_second": 25.244,
      "step": 10
    },
    {
      "epoch": 0.0,
      "step": 10,
      "total_flos": 42112770048000.0,
      "train_loss": 1.9789411544799804,
      "train_runtime": 156.255,
      "train_samples_per_second": 1.024,
      "train_steps_per_second": 0.064
    }
  ],
  "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
}