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_loss": 0.2079285830259323,
      "eval_runtime": 3.3536,
      "eval_samples_per_second": 924.366,
      "eval_steps_per_second": 19.382,
      "step": 318
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.5509255528450012,
      "learning_rate": 1.371069182389937e-05,
      "loss": 0.3261,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.10788384079933167,
      "eval_runtime": 3.3691,
      "eval_samples_per_second": 920.129,
      "eval_steps_per_second": 19.293,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.07727167755365372,
      "eval_runtime": 3.3489,
      "eval_samples_per_second": 925.666,
      "eval_steps_per_second": 19.409,
      "step": 954
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.5211280584335327,
      "learning_rate": 7.421383647798742e-06,
      "loss": 0.1246,
      "step": 1000
    }
  ],
  "logging_steps": 500,
  "max_steps": 1590,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 260941334653608.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.3115289936166381,
    "num_train_epochs": 5,
    "temperature": 11
  }
}