trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1001,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2e-05,
      "loss": 1.8159,
      "step": 1
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.0548953398643276e-05,
      "loss": 1.2753,
      "step": 500
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.5952586291849703e-08,
      "loss": 1.2094,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.2099609375,
      "eval_runtime": 157.1367,
      "eval_samples_per_second": 90.221,
      "eval_steps_per_second": 0.356,
      "step": 1001
    },
    {
      "epoch": 1.0,
      "step": 1001,
      "total_flos": 2649857382678528.0,
      "train_loss": 1.2428761424122752,
      "train_runtime": 6169.2421,
      "train_samples_per_second": 20.769,
      "train_steps_per_second": 0.162
    }
  ],
  "logging_steps": 500,
  "max_steps": 1001,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 2649857382678528.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}