run-3/checkpoint-1500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.716981132075472,
  "eval_steps": 500,
  "global_step": 1500,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5706451612903226,
      "eval_loss": 0.2087496519088745,
      "eval_runtime": 1.4434,
      "eval_samples_per_second": 2147.668,
      "eval_steps_per_second": 45.032,
      "step": 318
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.371069182389937e-05,
      "loss": 0.3271,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8041935483870968,
      "eval_loss": 0.1091708391904831,
      "eval_runtime": 1.4332,
      "eval_samples_per_second": 2162.978,
      "eval_steps_per_second": 45.353,
      "step": 636
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.857741935483871,
      "eval_loss": 0.07862202078104019,
      "eval_runtime": 1.4409,
      "eval_samples_per_second": 2151.481,
      "eval_steps_per_second": 45.112,
      "step": 954
    },
    {
      "epoch": 3.14,
      "learning_rate": 7.421383647798742e-06,
      "loss": 0.125,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8787096774193548,
      "eval_loss": 0.0655854269862175,
      "eval_runtime": 1.4459,
      "eval_samples_per_second": 2143.976,
      "eval_steps_per_second": 44.954,
      "step": 1272
    },
    {
      "epoch": 4.72,
      "learning_rate": 1.1320754716981133e-06,
      "loss": 0.0897,
      "step": 1500
    }
  ],
  "logging_steps": 500,
  "max_steps": 1590,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 391111083046572.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.038404641621370916,
    "num_train_epochs": 5,
    "temperature": 11
  }
}