{
  "best_metric": 0.9197247706422018,
  "best_model_checkpoint": "electra-distilled-sst\\run-30\\checkpoint-1581",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1581,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 26.225385665893555,
      "learning_rate": 6.561713559027068e-05,
      "loss": 0.6075,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9139908256880734,
      "eval_loss": 0.47316521406173706,
      "eval_runtime": 1.4499,
      "eval_samples_per_second": 601.43,
      "eval_steps_per_second": 4.828,
      "step": 527
    },
    {
      "epoch": 2.0,
      "grad_norm": 13.330994606018066,
      "learning_rate": 5.741499364148685e-05,
      "loss": 0.3214,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8956422018348624,
      "eval_loss": 0.6051578521728516,
      "eval_runtime": 1.4493,
      "eval_samples_per_second": 601.653,
      "eval_steps_per_second": 4.83,
      "step": 1054
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.15234661102295,
      "learning_rate": 4.9212851692703006e-05,
      "loss": 0.2348,
      "step": 1581
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9197247706422018,
      "eval_loss": 0.4945127069950104,
      "eval_runtime": 1.4486,
      "eval_samples_per_second": 601.961,
      "eval_steps_per_second": 4.832,
      "step": 1581
    }
  ],
  "logging_steps": 500,
  "max_steps": 4743,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 561038813831964.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9134230134636561,
    "learning_rate": 7.381927753905452e-05,
    "num_train_epochs": 9,
    "temperature": 18
  }
}