run-3/checkpoint-1054/trainer_state.json

{
  "best_metric": 0.9013761467889908,
  "best_model_checkpoint": "electra-distilled-sst\\run-3\\checkpoint-1054",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1054,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 20.69542694091797,
      "learning_rate": 1.4084477147380312e-05,
      "loss": 1.4419,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8704128440366973,
      "eval_loss": 0.8999464511871338,
      "eval_runtime": 1.4417,
      "eval_samples_per_second": 604.854,
      "eval_steps_per_second": 4.855,
      "step": 527
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5.335029222492542e-08,
      "loss": 0.8235,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9013761467889908,
      "eval_loss": 0.7633256912231445,
      "eval_runtime": 1.4453,
      "eval_samples_per_second": 603.342,
      "eval_steps_per_second": 4.843,
      "step": 1054
    }
  ],
  "logging_steps": 500,
  "max_steps": 1054,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 373776249469380.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7273510251657795,
    "learning_rate": 2.8115604002535698e-05,
    "num_train_epochs": 2,
    "temperature": 7
  }
}