trainer_state.json

{
  "best_metric": 1.240515112876892,
  "best_model_checkpoint": "sean_test_out_large/checkpoint-1000",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "grad_norm": 971037.375,
      "learning_rate": 3e-05,
      "loss": 1.7354,
      "step": 500
    },
    {
      "epoch": 1.0,
      "grad_norm": 247651.375,
      "learning_rate": 0.0,
      "loss": 1.3372,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4668,
      "eval_loss": 1.240515112876892,
      "eval_runtime": 573.2037,
      "eval_samples_per_second": 17.446,
      "eval_steps_per_second": 8.723,
      "step": 1000
    }
  ],
  "logging_steps": 500,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 4315155738240000.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}