{
  "best_metric": 0.5247524752475248,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-1782",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2079,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8027682304382324,
      "learning_rate": 0.0002796956441517031,
      "loss": 0.2656,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5128712871287129,
      "eval_loss": 0.27001720666885376,
      "eval_runtime": 55.7019,
      "eval_samples_per_second": 9.066,
      "eval_steps_per_second": 0.287,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2776716649532318,
      "learning_rate": 0.00023307970345975263,
      "loss": 0.2582,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5108910891089109,
      "eval_loss": 0.25422945618629456,
      "eval_runtime": 56.0184,
      "eval_samples_per_second": 9.015,
      "eval_steps_per_second": 0.286,
      "step": 594
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.49428918957710266,
      "learning_rate": 0.0001864637627678021,
      "loss": 0.255,
      "step": 891
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5168316831683168,
      "eval_loss": 0.2529114782810211,
      "eval_runtime": 56.4466,
      "eval_samples_per_second": 8.947,
      "eval_steps_per_second": 0.283,
      "step": 891
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3203761875629425,
      "learning_rate": 0.00013984782207585156,
      "loss": 0.2523,
      "step": 1188
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5148514851485149,
      "eval_loss": 0.2504001557826996,
      "eval_runtime": 56.4336,
      "eval_samples_per_second": 8.949,
      "eval_steps_per_second": 0.284,
      "step": 1188
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3831195831298828,
      "learning_rate": 9.323188138390105e-05,
      "loss": 0.2499,
      "step": 1485
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5188118811881188,
      "eval_loss": 0.24836017191410065,
      "eval_runtime": 58.52,
      "eval_samples_per_second": 8.63,
      "eval_steps_per_second": 0.273,
      "step": 1485
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5231035947799683,
      "learning_rate": 4.661594069195052e-05,
      "loss": 0.2486,
      "step": 1782
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5247524752475248,
      "eval_loss": 0.24778248369693756,
      "eval_runtime": 56.1683,
      "eval_samples_per_second": 8.991,
      "eval_steps_per_second": 0.285,
      "step": 1782
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.4036513566970825,
      "learning_rate": 0.0,
      "loss": 0.2466,
      "step": 2079
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5247524752475248,
      "eval_loss": 0.24736006557941437,
      "eval_runtime": 54.9354,
      "eval_samples_per_second": 9.193,
      "eval_steps_per_second": 0.291,
      "step": 2079
    }
  ],
  "logging_steps": 500,
  "max_steps": 2079,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 5444902981980.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.3139802761523254,
    "learning_rate": 0.0003263115848436537,
    "num_train_epochs": 7,
    "temperature": 8
  }
}