{
  "best_metric": 0.8360091743119266,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-527",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2635,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 22.49123764038086,
      "learning_rate": 0.0002780733463240362,
      "loss": 1.7444,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8360091743119266,
      "eval_loss": 1.752740740776062,
      "eval_runtime": 2.7351,
      "eval_samples_per_second": 318.815,
      "eval_steps_per_second": 2.559,
      "step": 527
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.574456214904785,
      "learning_rate": 0.00020855500974302718,
      "loss": 0.7696,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8142201834862385,
      "eval_loss": 1.9697859287261963,
      "eval_runtime": 2.6884,
      "eval_samples_per_second": 324.355,
      "eval_steps_per_second": 2.604,
      "step": 1054
    },
    {
      "epoch": 3.0,
      "grad_norm": NaN,
      "learning_rate": 0.00013916858651416426,
      "loss": 0.5359,
      "step": 1581
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8222477064220184,
      "eval_loss": 1.9042038917541504,
      "eval_runtime": 2.7045,
      "eval_samples_per_second": 322.427,
      "eval_steps_per_second": 2.588,
      "step": 1581
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.347850799560547,
      "learning_rate": 6.965024993315518e-05,
      "loss": 0.4078,
      "step": 2108
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8222477064220184,
      "eval_loss": 2.049978017807007,
      "eval_runtime": 2.713,
      "eval_samples_per_second": 321.42,
      "eval_steps_per_second": 2.58,
      "step": 2108
    },
    {
      "epoch": 5.0,
      "grad_norm": 14.40003490447998,
      "learning_rate": 1.3191335214612724e-07,
      "loss": 0.329,
      "step": 2635
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8279816513761468,
      "eval_loss": 2.0179200172424316,
      "eval_runtime": 2.6918,
      "eval_samples_per_second": 323.947,
      "eval_steps_per_second": 2.6,
      "step": 2635
    }
  ],
  "logging_steps": 500,
  "max_steps": 2635,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 40447180739880.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.08970692378937972,
    "learning_rate": 0.0003475916829050453,
    "num_train_epochs": 5,
    "temperature": 30
  }
}