{
  "best_metric": 0.6702544031311155,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-2/checkpoint-576",
  "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": 0.663398027420044,
      "learning_rate": 0.00027867701883546326,
      "loss": 0.2787,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5499021526418787,
      "eval_f1": 0.24092409240924093,
      "eval_loss": 0.25340619683265686,
      "eval_precision": 0.7684210526315789,
      "eval_recall": 0.14285714285714285,
      "eval_runtime": 28.5207,
      "eval_samples_per_second": 35.834,
      "eval_steps_per_second": 1.122,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7603411078453064,
      "learning_rate": 0.00024384239148103037,
      "loss": 0.2582,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5675146771037182,
      "eval_f1": 0.300632911392405,
      "eval_loss": 0.2482573688030243,
      "eval_precision": 0.7851239669421488,
      "eval_recall": 0.18590998043052837,
      "eval_runtime": 28.5782,
      "eval_samples_per_second": 35.762,
      "eval_steps_per_second": 1.12,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6440157294273376,
      "learning_rate": 0.00020900776412659743,
      "loss": 0.2508,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6027397260273972,
      "eval_f1": 0.40117994100294985,
      "eval_loss": 0.25017455220222473,
      "eval_precision": 0.8143712574850299,
      "eval_recall": 0.26614481409001955,
      "eval_runtime": 28.8169,
      "eval_samples_per_second": 35.465,
      "eval_steps_per_second": 1.11,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.8489454388618469,
      "learning_rate": 0.00017417313677216454,
      "loss": 0.2472,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.576320939334638,
      "eval_f1": 0.304975922953451,
      "eval_loss": 0.2455543428659439,
      "eval_precision": 0.8482142857142857,
      "eval_recall": 0.18590998043052837,
      "eval_runtime": 28.3606,
      "eval_samples_per_second": 36.036,
      "eval_steps_per_second": 1.128,
      "step": 384
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6578030586242676,
      "learning_rate": 0.00013933850941773163,
      "loss": 0.2444,
      "step": 480
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5949119373776908,
      "eval_f1": 0.37082066869300917,
      "eval_loss": 0.24357673525810242,
      "eval_precision": 0.8299319727891157,
      "eval_recall": 0.23874755381604695,
      "eval_runtime": 28.5893,
      "eval_samples_per_second": 35.748,
      "eval_steps_per_second": 1.119,
      "step": 480
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.576878547668457,
      "learning_rate": 0.00010450388206329872,
      "loss": 0.244,
      "step": 576
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6702544031311155,
      "eval_f1": 0.5536423841059602,
      "eval_loss": 0.24780145287513733,
      "eval_precision": 0.8565573770491803,
      "eval_recall": 0.4090019569471624,
      "eval_runtime": 28.8228,
      "eval_samples_per_second": 35.458,
      "eval_steps_per_second": 1.11,
      "step": 576
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.42660632729530334,
      "learning_rate": 6.966925470886582e-05,
      "loss": 0.2418,
      "step": 672
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6350293542074364,
      "eval_f1": 0.4709219858156028,
      "eval_loss": 0.2434568703174591,
      "eval_precision": 0.8556701030927835,
      "eval_recall": 0.324853228962818,
      "eval_runtime": 28.6835,
      "eval_samples_per_second": 35.63,
      "eval_steps_per_second": 1.116,
      "step": 672
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.7299667000770569,
      "learning_rate": 3.483462735443291e-05,
      "loss": 0.2396,
      "step": 768
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6203522504892368,
      "eval_f1": 0.43440233236151604,
      "eval_loss": 0.24256332218647003,
      "eval_precision": 0.8514285714285714,
      "eval_recall": 0.29158512720156554,
      "eval_runtime": 29.1745,
      "eval_samples_per_second": 35.031,
      "eval_steps_per_second": 1.097,
      "step": 768
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.48326733708381653,
      "learning_rate": 0.0,
      "loss": 0.2396,
      "step": 864
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6291585127201565,
      "eval_f1": 0.4562410329985653,
      "eval_loss": 0.2423153966665268,
      "eval_precision": 0.8548387096774194,
      "eval_recall": 0.3111545988258317,
      "eval_runtime": 28.281,
      "eval_samples_per_second": 36.137,
      "eval_steps_per_second": 1.131,
      "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": 0.3483609408663828,
    "learning_rate": 0.0003135116461898962,
    "num_train_epochs": 9,
    "temperature": 7
  }
}