run-4/checkpoint-1000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.1446540880503147,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6445161290322581,
      "eval_loss": 0.29249975085258484,
      "eval_runtime": 5.1547,
      "eval_samples_per_second": 601.395,
      "eval_steps_per_second": 12.61,
      "step": 318
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.6633772850036621,
      "learning_rate": 1.550763701707098e-05,
      "loss": 0.4626,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8380645161290322,
      "eval_loss": 0.12248263508081436,
      "eval_runtime": 5.377,
      "eval_samples_per_second": 576.528,
      "eval_steps_per_second": 12.088,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8887096774193548,
      "eval_loss": 0.07343784719705582,
      "eval_runtime": 5.2352,
      "eval_samples_per_second": 592.149,
      "eval_steps_per_second": 12.416,
      "step": 954
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.59869384765625,
      "learning_rate": 1.101527403414196e-05,
      "loss": 0.1438,
      "step": 1000
    }
  ],
  "logging_steps": 500,
  "max_steps": 2226,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 260941334653608.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.3308965359344769,
    "num_train_epochs": 7,
    "temperature": 3
  }
}