{
  "best_metric": 0.8337155963302753,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-2108",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 3689,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 14.515488624572754,
      "learning_rate": 0.00014545600250112875,
      "loss": 1.8614,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8176605504587156,
      "eval_loss": 1.578995704650879,
      "eval_runtime": 2.8296,
      "eval_samples_per_second": 308.168,
      "eval_steps_per_second": 2.474,
      "step": 527
    },
    {
      "epoch": 2.0,
      "grad_norm": 19.460262298583984,
      "learning_rate": 0.00012121333541760731,
      "loss": 0.862,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8302752293577982,
      "eval_loss": 1.6238842010498047,
      "eval_runtime": 2.8061,
      "eval_samples_per_second": 310.75,
      "eval_steps_per_second": 2.495,
      "step": 1054
    },
    {
      "epoch": 3.0,
      "grad_norm": NaN,
      "learning_rate": 9.701666959989899e-05,
      "loss": 0.6401,
      "step": 1581
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8325688073394495,
      "eval_loss": 1.6516802310943604,
      "eval_runtime": 2.8353,
      "eval_samples_per_second": 307.546,
      "eval_steps_per_second": 2.469,
      "step": 1581
    },
    {
      "epoch": 4.0,
      "grad_norm": 11.252717018127441,
      "learning_rate": 7.277400251637752e-05,
      "loss": 0.5157,
      "step": 2108
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8337155963302753,
      "eval_loss": 1.666493535041809,
      "eval_runtime": 2.8206,
      "eval_samples_per_second": 309.151,
      "eval_steps_per_second": 2.482,
      "step": 2108
    },
    {
      "epoch": 5.0,
      "grad_norm": 13.872171401977539,
      "learning_rate": 4.853133543285606e-05,
      "loss": 0.4408,
      "step": 2635
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8256880733944955,
      "eval_loss": 1.715163230895996,
      "eval_runtime": 2.8147,
      "eval_samples_per_second": 309.804,
      "eval_steps_per_second": 2.487,
      "step": 2635
    },
    {
      "epoch": 6.0,
      "grad_norm": 36.86779022216797,
      "learning_rate": 2.4288668349334598e-05,
      "loss": 0.3886,
      "step": 3162
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8279816513761468,
      "eval_loss": 1.7373932600021362,
      "eval_runtime": 2.8309,
      "eval_samples_per_second": 308.029,
      "eval_steps_per_second": 2.473,
      "step": 3162
    },
    {
      "epoch": 7.0,
      "grad_norm": 16.598100662231445,
      "learning_rate": 4.60012658131337e-08,
      "loss": 0.367,
      "step": 3689
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8279816513761468,
      "eval_loss": 1.7568955421447754,
      "eval_runtime": 2.8047,
      "eval_samples_per_second": 310.903,
      "eval_steps_per_second": 2.496,
      "step": 3689
    }
  ],
  "logging_steps": 500,
  "max_steps": 3689,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 56628612372420.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.24080905069528658,
    "learning_rate": 0.00016969866958465023,
    "num_train_epochs": 7,
    "temperature": 25
  }
}