{
  "best_metric": 0.5009784735812133,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-384",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 864,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.101290702819824,
      "learning_rate": 1.1962059637335337e-06,
      "loss": 0.5173,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.6666666666666666,
      "eval_loss": 0.5046124458312988,
      "eval_precision": 0.5,
      "eval_recall": 1.0,
      "eval_runtime": 28.251,
      "eval_samples_per_second": 36.176,
      "eval_steps_per_second": 1.133,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6350829005241394,
      "learning_rate": 1.0466802182668422e-06,
      "loss": 0.4989,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4843444227005871,
      "eval_f1": 0.6216798277099785,
      "eval_loss": 0.48798972368240356,
      "eval_precision": 0.4909297052154195,
      "eval_recall": 0.8473581213307241,
      "eval_runtime": 28.492,
      "eval_samples_per_second": 35.87,
      "eval_steps_per_second": 1.123,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8015046119689941,
      "learning_rate": 8.971544728001503e-07,
      "loss": 0.4869,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.48238747553816047,
      "eval_f1": 0.09572649572649572,
      "eval_loss": 0.4779175817966461,
      "eval_precision": 0.3783783783783784,
      "eval_recall": 0.0547945205479452,
      "eval_runtime": 28.2537,
      "eval_samples_per_second": 36.172,
      "eval_steps_per_second": 1.133,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.0801610946655273,
      "learning_rate": 7.476287273334586e-07,
      "loss": 0.4795,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5009784735812133,
      "eval_f1": 0.00390625,
      "eval_loss": 0.47173693776130676,
      "eval_precision": 1.0,
      "eval_recall": 0.0019569471624266144,
      "eval_runtime": 28.2359,
      "eval_samples_per_second": 36.195,
      "eval_steps_per_second": 1.133,
      "step": 384
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.289751410484314,
      "learning_rate": 5.981029818667669e-07,
      "loss": 0.4737,
      "step": 480
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5009784735812133,
      "eval_f1": 0.00390625,
      "eval_loss": 0.468070924282074,
      "eval_precision": 1.0,
      "eval_recall": 0.0019569471624266144,
      "eval_runtime": 28.7218,
      "eval_samples_per_second": 35.583,
      "eval_steps_per_second": 1.114,
      "step": 480
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.6734181642532349,
      "learning_rate": 4.4857723640007514e-07,
      "loss": 0.4709,
      "step": 576
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0,
      "eval_loss": 0.4659123420715332,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 28.369,
      "eval_samples_per_second": 36.025,
      "eval_steps_per_second": 1.128,
      "step": 576
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.6338440179824829,
      "learning_rate": 2.9905149093338343e-07,
      "loss": 0.4686,
      "step": 672
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0,
      "eval_loss": 0.46463751792907715,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 28.015,
      "eval_samples_per_second": 36.48,
      "eval_steps_per_second": 1.142,
      "step": 672
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.5559163093566895,
      "learning_rate": 1.4952574546669171e-07,
      "loss": 0.4678,
      "step": 768
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0,
      "eval_loss": 0.4639304578304291,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 28.4868,
      "eval_samples_per_second": 35.876,
      "eval_steps_per_second": 1.123,
      "step": 768
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.9788464307785034,
      "learning_rate": 0.0,
      "loss": 0.467,
      "step": 864
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.5,
      "eval_f1": 0.0,
      "eval_loss": 0.4637199938297272,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 31.2173,
      "eval_samples_per_second": 32.738,
      "eval_steps_per_second": 1.025,
      "step": 864
    }
  ],
  "logging_steps": 500,
  "max_steps": 864,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 2121256775520.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.586704262013931,
    "learning_rate": 1.3457317092002255e-06,
    "num_train_epochs": 9,
    "temperature": 21
  }
}