trainer_state.json

{
  "best_metric": 2.2883481979370117,
  "best_model_checkpoint": "models/basic_NADI2023/checkpoint-1689",
  "epoch": 3.0,
  "global_step": 1689,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.89,
      "learning_rate": 3.5198342214328006e-05,
      "loss": 2.425,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_NADI2023_dev_loss": 2.349003314971924,
      "eval_NADI2023_dev_runtime": 3.2106,
      "eval_NADI2023_dev_samples_per_second": 560.648,
      "eval_NADI2023_dev_steps_per_second": 70.081,
      "step": 563
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.039668442865601e-05,
      "loss": 2.2599,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_NADI2023_dev_loss": 2.310096502304077,
      "eval_NADI2023_dev_runtime": 3.2198,
      "eval_NADI2023_dev_samples_per_second": 559.041,
      "eval_NADI2023_dev_steps_per_second": 69.88,
      "step": 1126
    },
    {
      "epoch": 2.66,
      "learning_rate": 5.595026642984015e-06,
      "loss": 2.1854,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_NADI2023_dev_loss": 2.2883481979370117,
      "eval_NADI2023_dev_runtime": 3.2189,
      "eval_NADI2023_dev_samples_per_second": 559.191,
      "eval_NADI2023_dev_steps_per_second": 69.899,
      "step": 1689
    }
  ],
  "max_steps": 1689,
  "num_train_epochs": 3,
  "total_flos": 2275826410512000.0,
  "trial_name": null,
  "trial_params": null
}