trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.1616842105263158,
  "eval_steps": 500,
  "global_step": 3,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "learning_rate": 1.9848077530122083e-05,
      "loss": 1.9947,
      "step": 1
    },
    {
      "epoch": 0.16,
      "eval_loss": 2.009272813796997,
      "eval_runtime": 16.1807,
      "eval_samples_per_second": 30.901,
      "eval_steps_per_second": 3.894,
      "step": 3
    },
    {
      "epoch": 0.16,
      "step": 3,
      "total_flos": 1.4034501424381952e+17,
      "train_loss": 2.0712431271870932,
      "train_runtime": 1041.0021,
      "train_samples_per_second": 9.126,
      "train_steps_per_second": 0.017
    }
  ],
  "logging_steps": 5,
  "max_steps": 18,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 1.4034501424381952e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}