run-1/checkpoint-594/trainer_state.json

{
  "best_metric": 0.5801980198019802,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/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": 0.8816747665405273,
      "learning_rate": 0.0001319007893143593,
      "loss": 0.4812,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5188118811881188,
      "eval_loss": 0.4719693958759308,
      "eval_runtime": 14.5929,
      "eval_samples_per_second": 34.606,
      "eval_steps_per_second": 1.096,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.699450671672821,
      "learning_rate": 0.00011724514605720826,
      "loss": 0.4696,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5801980198019802,
      "eval_loss": 0.46854168176651,
      "eval_runtime": 15.1826,
      "eval_samples_per_second": 33.262,
      "eval_steps_per_second": 1.054,
      "step": 594
    }
  ],
  "logging_steps": 500,
  "max_steps": 2970,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1555686566280.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6312656697409034,
    "learning_rate": 0.00014655643257151032,
    "num_train_epochs": 10,
    "temperature": 21
  }
}