{
  "best_metric": 0.9038386971694455,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-20/checkpoint-3870",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 4515,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.6259269714355469,
      "learning_rate": 0.0003431069228198171,
      "loss": 0.3055,
      "step": 645
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8889104303993796,
      "eval_f1": 0.19409282700421943,
      "eval_loss": 0.2940736413002014,
      "eval_mcc": 0.20172457395808635,
      "eval_precision": 0.4859154929577465,
      "eval_recall": 0.12126537785588752,
      "eval_runtime": 9.9362,
      "eval_samples_per_second": 519.111,
      "eval_steps_per_second": 16.304,
      "step": 645
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.638690710067749,
      "learning_rate": 0.00028592243568318097,
      "loss": 0.2894,
      "step": 1290
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8964715005816207,
      "eval_f1": 0.18844984802431614,
      "eval_loss": 0.2831971049308777,
      "eval_mcc": 0.24798827118495087,
      "eval_precision": 0.6966292134831461,
      "eval_recall": 0.10896309314586995,
      "eval_runtime": 9.9132,
      "eval_samples_per_second": 520.316,
      "eval_steps_per_second": 16.342,
      "step": 1290
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.7561137080192566,
      "learning_rate": 0.00022873794854654474,
      "loss": 0.2795,
      "step": 1935
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8964715005816207,
      "eval_f1": 0.165625,
      "eval_loss": 0.2780698239803314,
      "eval_mcc": 0.2399021362225046,
      "eval_precision": 0.7464788732394366,
      "eval_recall": 0.09314586994727592,
      "eval_runtime": 10.0176,
      "eval_samples_per_second": 514.893,
      "eval_steps_per_second": 16.172,
      "step": 1935
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1933393478393555,
      "learning_rate": 0.00017155346140990856,
      "loss": 0.2738,
      "step": 2580
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8986041101202016,
      "eval_f1": 0.309114927344782,
      "eval_loss": 0.28069812059402466,
      "eval_mcc": 0.3178770819197392,
      "eval_precision": 0.6223404255319149,
      "eval_recall": 0.2056239015817223,
      "eval_runtime": 10.0583,
      "eval_samples_per_second": 512.809,
      "eval_steps_per_second": 16.106,
      "step": 2580
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.9380802512168884,
      "learning_rate": 0.00011436897427327237,
      "loss": 0.27,
      "step": 3225
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9011244668476154,
      "eval_f1": 0.32180851063829785,
      "eval_loss": 0.2702881097793579,
      "eval_mcc": 0.3372551401109897,
      "eval_precision": 0.6612021857923497,
      "eval_recall": 0.2126537785588752,
      "eval_runtime": 10.0143,
      "eval_samples_per_second": 515.061,
      "eval_steps_per_second": 16.177,
      "step": 3225
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.2645069360733032,
      "learning_rate": 5.7184487136636185e-05,
      "loss": 0.2641,
      "step": 3870
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9038386971694455,
      "eval_f1": 0.35078534031413616,
      "eval_loss": 0.26536738872528076,
      "eval_mcc": 0.36499409965660234,
      "eval_precision": 0.6871794871794872,
      "eval_recall": 0.23550087873462214,
      "eval_runtime": 10.0214,
      "eval_samples_per_second": 514.7,
      "eval_steps_per_second": 16.165,
      "step": 3870
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.633353054523468,
      "learning_rate": 0.0,
      "loss": 0.2619,
      "step": 4515
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9026754556029469,
      "eval_f1": 0.335978835978836,
      "eval_loss": 0.2652672529220581,
      "eval_mcc": 0.35216632630481853,
      "eval_precision": 0.679144385026738,
      "eval_recall": 0.22319859402460457,
      "eval_runtime": 9.9714,
      "eval_samples_per_second": 517.279,
      "eval_steps_per_second": 16.246,
      "step": 4515
    }
  ],
  "logging_steps": 500,
  "max_steps": 4515,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 11108597728200.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8743805278290786,
    "learning_rate": 0.0004002914099564533,
    "num_train_epochs": 7,
    "temperature": 27
  }
}