{
  "best_metric": 0.8258317025440313,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-8/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": 3.2297720909118652,
      "learning_rate": 0.0003356185870926135,
      "loss": 0.5529,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7602739726027398,
      "eval_f1": 0.7973531844499586,
      "eval_loss": 0.4773445129394531,
      "eval_precision": 0.6905444126074498,
      "eval_recall": 0.9432485322896281,
      "eval_runtime": 28.1534,
      "eval_samples_per_second": 36.301,
      "eval_steps_per_second": 1.137,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.981153130531311,
      "learning_rate": 0.00026849486967409075,
      "loss": 0.4542,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8199608610567515,
      "eval_f1": 0.8365896980461812,
      "eval_loss": 0.42821842432022095,
      "eval_precision": 0.7658536585365854,
      "eval_recall": 0.9217221135029354,
      "eval_runtime": 28.2265,
      "eval_samples_per_second": 36.207,
      "eval_steps_per_second": 1.134,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.941507339477539,
      "learning_rate": 0.0002013711522555681,
      "loss": 0.4191,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.821917808219178,
      "eval_f1": 0.8311688311688312,
      "eval_loss": 0.44406670331954956,
      "eval_precision": 0.7901234567901234,
      "eval_recall": 0.8767123287671232,
      "eval_runtime": 28.0469,
      "eval_samples_per_second": 36.439,
      "eval_steps_per_second": 1.141,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.470360279083252,
      "learning_rate": 0.00013424743483704538,
      "loss": 0.4059,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8170254403131115,
      "eval_f1": 0.8186226964112512,
      "eval_loss": 0.4284485876560211,
      "eval_precision": 0.8115384615384615,
      "eval_recall": 0.8258317025440313,
      "eval_runtime": 28.1941,
      "eval_samples_per_second": 36.249,
      "eval_steps_per_second": 1.135,
      "step": 384
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.5780177116394043,
      "learning_rate": 6.712371741852269e-05,
      "loss": 0.398,
      "step": 480
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8258317025440313,
      "eval_f1": 0.8378870673952642,
      "eval_loss": 0.4101215600967407,
      "eval_precision": 0.7836456558773425,
      "eval_recall": 0.9001956947162426,
      "eval_runtime": 28.0686,
      "eval_samples_per_second": 36.411,
      "eval_steps_per_second": 1.14,
      "step": 480
    }
  ],
  "logging_steps": 500,
  "max_steps": 576,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 1178475986400.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8957226795607325,
    "learning_rate": 0.0004027423045111362,
    "num_train_epochs": 6,
    "temperature": 3
  }
}