{
  "best_metric": 0.8395303326810176,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-19/checkpoint-960",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 960,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.265742301940918,
      "learning_rate": 0.00016552182978108178,
      "loss": 0.5708,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7592954990215264,
      "eval_f1": 0.7980295566502462,
      "eval_loss": 0.4895075857639313,
      "eval_precision": 0.6874115983026874,
      "eval_recall": 0.9510763209393346,
      "eval_runtime": 24.8356,
      "eval_samples_per_second": 41.151,
      "eval_steps_per_second": 1.288,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.506702899932861,
      "learning_rate": 0.00014713051536096157,
      "loss": 0.4657,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.812133072407045,
      "eval_f1": 0.8318739054290717,
      "eval_loss": 0.42986905574798584,
      "eval_precision": 0.7527733755942948,
      "eval_recall": 0.9295499021526419,
      "eval_runtime": 25.1401,
      "eval_samples_per_second": 40.652,
      "eval_steps_per_second": 1.273,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.952416181564331,
      "learning_rate": 0.00012873920094084136,
      "loss": 0.4303,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8209393346379648,
      "eval_f1": 0.8300835654596102,
      "eval_loss": 0.42889997363090515,
      "eval_precision": 0.7897526501766784,
      "eval_recall": 0.8747553816046967,
      "eval_runtime": 25.3264,
      "eval_samples_per_second": 40.353,
      "eval_steps_per_second": 1.264,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.971947193145752,
      "learning_rate": 0.00011034788652072118,
      "loss": 0.4148,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.815068493150685,
      "eval_f1": 0.8156097560975609,
      "eval_loss": 0.4303808808326721,
      "eval_precision": 0.8132295719844358,
      "eval_recall": 0.8180039138943248,
      "eval_runtime": 25.4322,
      "eval_samples_per_second": 40.185,
      "eval_steps_per_second": 1.258,
      "step": 384
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.9225187301635742,
      "learning_rate": 9.195657210060098e-05,
      "loss": 0.4076,
      "step": 480
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8277886497064579,
      "eval_f1": 0.8414414414414414,
      "eval_loss": 0.40675535798072815,
      "eval_precision": 0.7796327212020033,
      "eval_recall": 0.913894324853229,
      "eval_runtime": 25.6271,
      "eval_samples_per_second": 39.88,
      "eval_steps_per_second": 1.249,
      "step": 480
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.180634498596191,
      "learning_rate": 7.356525768048079e-05,
      "loss": 0.4009,
      "step": 576
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8238747553816047,
      "eval_f1": 0.8432055749128919,
      "eval_loss": 0.4179611802101135,
      "eval_precision": 0.7598116169544741,
      "eval_recall": 0.9471624266144814,
      "eval_runtime": 25.4907,
      "eval_samples_per_second": 40.093,
      "eval_steps_per_second": 1.255,
      "step": 576
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.1518468856811523,
      "learning_rate": 5.517394326036059e-05,
      "loss": 0.3956,
      "step": 672
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8268101761252447,
      "eval_f1": 0.8456843940714908,
      "eval_loss": 0.412363737821579,
      "eval_precision": 0.7625786163522013,
      "eval_recall": 0.949119373776908,
      "eval_runtime": 25.9899,
      "eval_samples_per_second": 39.323,
      "eval_steps_per_second": 1.231,
      "step": 672
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.885427236557007,
      "learning_rate": 3.678262884024039e-05,
      "loss": 0.3914,
      "step": 768
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8287671232876712,
      "eval_f1": 0.8466257668711656,
      "eval_loss": 0.4102531671524048,
      "eval_precision": 0.7666666666666667,
      "eval_recall": 0.9452054794520548,
      "eval_runtime": 25.478,
      "eval_samples_per_second": 40.113,
      "eval_steps_per_second": 1.256,
      "step": 768
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.2045633792877197,
      "learning_rate": 1.8391314420120196e-05,
      "loss": 0.3868,
      "step": 864
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8365949119373777,
      "eval_f1": 0.8460829493087557,
      "eval_loss": 0.4009145498275757,
      "eval_precision": 0.7996515679442509,
      "eval_recall": 0.898238747553816,
      "eval_runtime": 24.6734,
      "eval_samples_per_second": 41.421,
      "eval_steps_per_second": 1.297,
      "step": 864
    },
    {
      "epoch": 10.0,
      "grad_norm": 4.398197650909424,
      "learning_rate": 0.0,
      "loss": 0.3889,
      "step": 960
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8395303326810176,
      "eval_f1": 0.8509090909090908,
      "eval_loss": 0.3997047245502472,
      "eval_precision": 0.7945670628183361,
      "eval_recall": 0.9158512720156555,
      "eval_runtime": 25.0131,
      "eval_samples_per_second": 40.859,
      "eval_steps_per_second": 1.279,
      "step": 960
    }
  ],
  "logging_steps": 500,
  "max_steps": 960,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 2356951972800.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8822091648395545,
    "learning_rate": 0.00018391314420120196,
    "num_train_epochs": 10,
    "temperature": 8
  }
}