trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9968387776606953,
  "eval_steps": 500,
  "global_step": 1422,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9989462592202318,
      "eval_accuracy": 0.7725118483412322,
      "eval_loss": 0.5226067304611206,
      "eval_runtime": 13.0982,
      "eval_samples_per_second": 32.218,
      "eval_steps_per_second": 8.093,
      "step": 474
    },
    {
      "epoch": 1.053740779768177,
      "grad_norm": 7.272853851318359,
      "learning_rate": 0.0006522557186771741,
      "loss": 0.5991,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8317535545023697,
      "eval_loss": 0.4196150302886963,
      "eval_runtime": 13.0934,
      "eval_samples_per_second": 32.23,
      "eval_steps_per_second": 8.096,
      "step": 949
    },
    {
      "epoch": 2.107481559536354,
      "grad_norm": 15.65460205078125,
      "learning_rate": 0.0001818105588338676,
      "loss": 0.4192,
      "step": 1000
    },
    {
      "epoch": 2.9968387776606953,
      "eval_accuracy": 0.8388625592417062,
      "eval_loss": 0.3961751163005829,
      "eval_runtime": 13.1095,
      "eval_samples_per_second": 32.19,
      "eval_steps_per_second": 8.086,
      "step": 1422
    }
  ],
  "logging_steps": 500,
  "max_steps": 1422,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 6754275451469824.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}