run-1/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": 0.6135483870967742,
      "eval_loss": 0.23216405510902405,
      "eval_runtime": 14.5556,
      "eval_samples_per_second": 212.976,
      "eval_steps_per_second": 26.656,
      "step": 318
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.5930498838424683,
      "learning_rate": 1.606918238993711e-05,
      "loss": 0.3696,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8325806451612904,
      "eval_loss": 0.1069631278514862,
      "eval_runtime": 15.7006,
      "eval_samples_per_second": 197.444,
      "eval_steps_per_second": 24.712,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8890322580645161,
      "eval_loss": 0.0675421878695488,
      "eval_runtime": 15.266,
      "eval_samples_per_second": 203.065,
      "eval_steps_per_second": 25.416,
      "step": 954
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.593765914440155,
      "learning_rate": 1.2138364779874214e-05,
      "loss": 0.1241,
      "step": 1000
    }
  ],
  "logging_steps": 500,
  "max_steps": 2544,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 260941334653608.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.3077189989743373,
    "num_train_epochs": 8,
    "temperature": 5
  }
}