{
  "best_metric": 0.7720156555772995,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-10/checkpoint-3830",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 3830,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.387345790863037,
      "learning_rate": 9.274722577647405e-06,
      "loss": 0.6719,
      "step": 766
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6594911937377691,
      "eval_f1": 0.5786924939467312,
      "eval_loss": 0.6500130891799927,
      "eval_precision": 0.7587301587301587,
      "eval_recall": 0.46771037181996084,
      "eval_runtime": 133.0073,
      "eval_samples_per_second": 7.684,
      "eval_steps_per_second": 1.925,
      "step": 766
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.753175735473633,
      "learning_rate": 6.956041933235553e-06,
      "loss": 0.6052,
      "step": 1532
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7250489236790607,
      "eval_f1": 0.7236971484759096,
      "eval_loss": 0.5506593585014343,
      "eval_precision": 0.7272727272727273,
      "eval_recall": 0.7201565557729941,
      "eval_runtime": 131.7286,
      "eval_samples_per_second": 7.758,
      "eval_steps_per_second": 1.943,
      "step": 1532
    },
    {
      "epoch": 3.0,
      "grad_norm": 16.92954444885254,
      "learning_rate": 4.637361288823702e-06,
      "loss": 0.5438,
      "step": 2298
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7583170254403131,
      "eval_f1": 0.7776777677767776,
      "eval_loss": 0.5147304534912109,
      "eval_precision": 0.72,
      "eval_recall": 0.8454011741682974,
      "eval_runtime": 132.3477,
      "eval_samples_per_second": 7.722,
      "eval_steps_per_second": 1.934,
      "step": 2298
    },
    {
      "epoch": 4.0,
      "grad_norm": 12.114532470703125,
      "learning_rate": 2.318680644411851e-06,
      "loss": 0.523,
      "step": 3064
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7710371819960861,
      "eval_f1": 0.7975778546712803,
      "eval_loss": 0.5050157904624939,
      "eval_precision": 0.7147286821705426,
      "eval_recall": 0.9021526418786693,
      "eval_runtime": 130.8317,
      "eval_samples_per_second": 7.812,
      "eval_steps_per_second": 1.957,
      "step": 3064
    },
    {
      "epoch": 5.0,
      "grad_norm": 12.235973358154297,
      "learning_rate": 0.0,
      "loss": 0.517,
      "step": 3830
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7720156555772995,
      "eval_f1": 0.7986171132238549,
      "eval_loss": 0.5030738711357117,
      "eval_precision": 0.7151702786377709,
      "eval_recall": 0.9041095890410958,
      "eval_runtime": 131.0021,
      "eval_samples_per_second": 7.801,
      "eval_steps_per_second": 1.954,
      "step": 3830
    }
  ],
  "logging_steps": 500,
  "max_steps": 3830,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1178475986400.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9677948459483069,
    "learning_rate": 1.1593403222059255e-05,
    "num_train_epochs": 5,
    "temperature": 2
  }
}