run-2/checkpoint-594/trainer_state.json

{
  "best_metric": 0.5207920792079208,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-2/checkpoint-594",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 594,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.46941760182380676,
      "learning_rate": 1.0066413527942762e-05,
      "loss": 0.2155,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.497029702970297,
      "eval_loss": 0.18417517840862274,
      "eval_runtime": 14.6034,
      "eval_samples_per_second": 34.581,
      "eval_steps_per_second": 1.096,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3071568012237549,
      "learning_rate": 8.808111836949918e-06,
      "loss": 0.1831,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5207920792079208,
      "eval_loss": 0.17807520925998688,
      "eval_runtime": 14.6044,
      "eval_samples_per_second": 34.579,
      "eval_steps_per_second": 1.096,
      "step": 594
    }
  ],
  "logging_steps": 500,
  "max_steps": 2673,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 1555686566280.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.20650329892275032,
    "learning_rate": 1.1324715218935609e-05,
    "num_train_epochs": 9,
    "temperature": 10
  }
}