trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 8,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 3.4252,
      "step": 1
    },
    {
      "epoch": 0.62,
      "learning_rate": 7.774790660436857e-06,
      "loss": 5.1898,
      "step": 5
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.842633008956909,
      "eval_runtime": 6.6261,
      "eval_samples_per_second": 31.089,
      "eval_steps_per_second": 0.302,
      "step": 8
    },
    {
      "epoch": 1.0,
      "step": 8,
      "total_flos": 6700148981760.0,
      "train_loss": 4.295631617307663,
      "train_runtime": 136.5276,
      "train_samples_per_second": 13.66,
      "train_steps_per_second": 0.059
    }
  ],
  "logging_steps": 5,
  "max_steps": 8,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 6700148981760.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}