{
  "best_metric": 0.8287671232876712,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-15/checkpoint-480",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 480,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.341264247894287,
      "learning_rate": 0.0003257774454972516,
      "loss": 0.5225,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7906066536203522,
      "eval_f1": 0.8078994614003591,
      "eval_loss": 0.4404396712779999,
      "eval_precision": 0.746268656716418,
      "eval_recall": 0.8806262230919765,
      "eval_runtime": 24.9228,
      "eval_samples_per_second": 41.007,
      "eval_steps_per_second": 1.284,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.7467563152313232,
      "learning_rate": 0.00028957995155311255,
      "loss": 0.4443,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8189823874755382,
      "eval_f1": 0.8322756119673618,
      "eval_loss": 0.40899789333343506,
      "eval_precision": 0.7753378378378378,
      "eval_recall": 0.898238747553816,
      "eval_runtime": 24.9291,
      "eval_samples_per_second": 40.996,
      "eval_steps_per_second": 1.284,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.382880210876465,
      "learning_rate": 0.00025338245760897347,
      "loss": 0.4106,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8238747553816047,
      "eval_f1": 0.8324022346368715,
      "eval_loss": 0.4271247386932373,
      "eval_precision": 0.7939609236234458,
      "eval_recall": 0.8747553816046967,
      "eval_runtime": 26.7353,
      "eval_samples_per_second": 38.227,
      "eval_steps_per_second": 1.197,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.9181997776031494,
      "learning_rate": 0.00021718496366483439,
      "loss": 0.4035,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8170254403131115,
      "eval_f1": 0.8172043010752689,
      "eval_loss": 0.41200369596481323,
      "eval_precision": 0.81640625,
      "eval_recall": 0.8180039138943248,
      "eval_runtime": 26.179,
      "eval_samples_per_second": 39.039,
      "eval_steps_per_second": 1.222,
      "step": 384
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.339964509010315,
      "learning_rate": 0.00018098746972069533,
      "loss": 0.3903,
      "step": 480
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8287671232876712,
      "eval_f1": 0.8362956033676333,
      "eval_loss": 0.40260037779808044,
      "eval_precision": 0.8010752688172043,
      "eval_recall": 0.8747553816046967,
      "eval_runtime": 25.7166,
      "eval_samples_per_second": 39.741,
      "eval_steps_per_second": 1.244,
      "step": 480
    }
  ],
  "logging_steps": 500,
  "max_steps": 960,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1178475986400.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7955319016044536,
    "learning_rate": 0.00036197493944139066,
    "num_train_epochs": 10,
    "temperature": 8
  }
}