run-11/checkpoint-1054/trainer_state.json

{
  "best_metric": 0.8337155963302753,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-11/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,
      "learning_rate": 0.00010911592510991879,
      "loss": 1.2248,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8130733944954128,
      "eval_loss": 1.0815833806991577,
      "eval_runtime": 19.1179,
      "eval_samples_per_second": 45.612,
      "eval_steps_per_second": 0.366,
      "step": 527
    },
    {
      "epoch": 2.0,
      "learning_rate": 9.0929937591599e-05,
      "loss": 0.6152,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8337155963302753,
      "eval_loss": 1.116389513015747,
      "eval_runtime": 6.3065,
      "eval_samples_per_second": 138.271,
      "eval_steps_per_second": 1.11,
      "step": 1054
    }
  ],
  "logging_steps": 500,
  "max_steps": 3689,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 16162868129520.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6271968482895725,
    "learning_rate": 0.0001273019126282386,
    "num_train_epochs": 7,
    "temperature": 11
  }
}