run-4/checkpoint-594/trainer_state.json

{
  "best_metric": 0.500990099009901,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-4/checkpoint-297",
  "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": 1.3486188650131226,
      "learning_rate": 0.0008534021470265506,
      "loss": 0.5548,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.500990099009901,
      "eval_loss": 0.5633774399757385,
      "eval_runtime": 55.4108,
      "eval_samples_per_second": 9.114,
      "eval_steps_per_second": 0.289,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8194136023521423,
      "learning_rate": 0.0007467268786482318,
      "loss": 0.5553,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.500990099009901,
      "eval_loss": 0.5545336604118347,
      "eval_runtime": 54.9924,
      "eval_samples_per_second": 9.183,
      "eval_steps_per_second": 0.291,
      "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": 0.7480049651920854,
    "learning_rate": 0.0009600774154048695,
    "num_train_epochs": 9,
    "temperature": 2
  }
}