{
  "best_metric": 0.5089108910891089,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-2/checkpoint-297",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 891,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.6225994825363159,
      "learning_rate": 0.0001586738890732749,
      "loss": 0.0237,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5089108910891089,
      "eval_loss": 0.015105457045137882,
      "eval_runtime": 57.753,
      "eval_samples_per_second": 8.744,
      "eval_steps_per_second": 0.277,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.24433593451976776,
      "learning_rate": 7.933694453663745e-05,
      "loss": 0.0161,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.504950495049505,
      "eval_loss": 0.014346625655889511,
      "eval_runtime": 56.8115,
      "eval_samples_per_second": 8.889,
      "eval_steps_per_second": 0.282,
      "step": 594
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.19921059906482697,
      "learning_rate": 0.0,
      "loss": 0.0151,
      "step": 891
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.504950495049505,
      "eval_loss": 0.013619398698210716,
      "eval_runtime": 56.0103,
      "eval_samples_per_second": 9.016,
      "eval_steps_per_second": 0.286,
      "step": 891
    }
  ],
  "logging_steps": 500,
  "max_steps": 891,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2333529849420.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.007640219223822298,
    "learning_rate": 0.00023801083360991236,
    "num_train_epochs": 3,
    "temperature": 6
  }
}