{
  "best_metric": 0.7089108910891089,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-28/checkpoint-1782",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2079,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.3590941429138184,
      "learning_rate": 0.00033268893750406807,
      "loss": 0.6415,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5861386138613861,
      "eval_f1": 0.473551637279597,
      "eval_loss": 0.6401379108428955,
      "eval_mcc": 0.1894629867942258,
      "eval_precision": 0.6482758620689655,
      "eval_recall": 0.373015873015873,
      "eval_runtime": 0.9371,
      "eval_samples_per_second": 538.913,
      "eval_steps_per_second": 17.074,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.476292848587036,
      "learning_rate": 0.00028516194643205835,
      "loss": 0.613,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6514851485148515,
      "eval_f1": 0.68,
      "eval_loss": 0.5918833017349243,
      "eval_mcc": 0.30837572082433856,
      "eval_precision": 0.62751677852349,
      "eval_recall": 0.7420634920634921,
      "eval_runtime": 0.9445,
      "eval_samples_per_second": 534.7,
      "eval_steps_per_second": 16.941,
      "step": 594
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.6527726650238037,
      "learning_rate": 0.0002376349553600486,
      "loss": 0.5836,
      "step": 891
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6594059405940594,
      "eval_f1": 0.693950177935943,
      "eval_loss": 0.5902799367904663,
      "eval_mcc": 0.3278781101029775,
      "eval_precision": 0.6290322580645161,
      "eval_recall": 0.7738095238095238,
      "eval_runtime": 0.9292,
      "eval_samples_per_second": 543.474,
      "eval_steps_per_second": 17.219,
      "step": 891
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.4662635326385498,
      "learning_rate": 0.0001901079642880389,
      "loss": 0.5692,
      "step": 1188
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6613861386138614,
      "eval_f1": 0.6994727592267135,
      "eval_loss": 0.5874853730201721,
      "eval_mcc": 0.33437205674458564,
      "eval_precision": 0.6277602523659306,
      "eval_recall": 0.7896825396825397,
      "eval_runtime": 0.9276,
      "eval_samples_per_second": 544.403,
      "eval_steps_per_second": 17.248,
      "step": 1188
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.2937889099121094,
      "learning_rate": 0.00014258097321602918,
      "loss": 0.5565,
      "step": 1485
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6712871287128713,
      "eval_f1": 0.6844106463878327,
      "eval_loss": 0.5748322010040283,
      "eval_mcc": 0.34399282923991964,
      "eval_precision": 0.656934306569343,
      "eval_recall": 0.7142857142857143,
      "eval_runtime": 0.9341,
      "eval_samples_per_second": 540.619,
      "eval_steps_per_second": 17.129,
      "step": 1485
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.6975440979003906,
      "learning_rate": 9.505398214401945e-05,
      "loss": 0.5445,
      "step": 1782
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7089108910891089,
      "eval_f1": 0.7379679144385027,
      "eval_loss": 0.5586543083190918,
      "eval_mcc": 0.4291472704190268,
      "eval_precision": 0.6699029126213593,
      "eval_recall": 0.8214285714285714,
      "eval_runtime": 0.932,
      "eval_samples_per_second": 541.831,
      "eval_steps_per_second": 17.167,
      "step": 1782
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.01635479927063,
      "learning_rate": 4.752699107200972e-05,
      "loss": 0.5314,
      "step": 2079
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7069306930693069,
      "eval_f1": 0.7279411764705882,
      "eval_loss": 0.5484515428543091,
      "eval_mcc": 0.419334756687555,
      "eval_precision": 0.678082191780822,
      "eval_recall": 0.7857142857142857,
      "eval_runtime": 0.937,
      "eval_samples_per_second": 538.933,
      "eval_steps_per_second": 17.075,
      "step": 2079
    }
  ],
  "logging_steps": 500,
  "max_steps": 2376,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 5114908861860.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9092941270097592,
    "learning_rate": 0.0003802159285760778,
    "num_train_epochs": 8,
    "temperature": 11
  }
}