{
  "best_metric": 0.6829745596868885,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-4/checkpoint-1536",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 1536,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.896482527256012,
      "learning_rate": 2.4429947101690063e-05,
      "loss": 0.3642,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.525440313111546,
      "eval_f1": 0.14762741652021089,
      "eval_loss": 0.33992624282836914,
      "eval_mcc": 0.1099564814603399,
      "eval_precision": 0.7241379310344828,
      "eval_recall": 0.0821917808219178,
      "eval_runtime": 66.5832,
      "eval_samples_per_second": 15.349,
      "eval_steps_per_second": 0.481,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.451338529586792,
      "learning_rate": 2.1376203713978805e-05,
      "loss": 0.3285,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5753424657534246,
      "eval_f1": 0.31761006289308175,
      "eval_loss": 0.31206491589546204,
      "eval_mcc": 0.2299532209711479,
      "eval_precision": 0.808,
      "eval_recall": 0.19765166340508805,
      "eval_runtime": 66.7282,
      "eval_samples_per_second": 15.316,
      "eval_steps_per_second": 0.48,
      "step": 384
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4082952737808228,
      "learning_rate": 1.8322460326267547e-05,
      "loss": 0.314,
      "step": 576
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6526418786692759,
      "eval_f1": 0.5310435931307793,
      "eval_loss": 0.30446648597717285,
      "eval_mcc": 0.3570476542530334,
      "eval_precision": 0.8170731707317073,
      "eval_recall": 0.3933463796477495,
      "eval_runtime": 66.749,
      "eval_samples_per_second": 15.311,
      "eval_steps_per_second": 0.479,
      "step": 576
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.84216570854187,
      "learning_rate": 1.526871693855629e-05,
      "loss": 0.3104,
      "step": 768
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6193737769080235,
      "eval_f1": 0.44028776978417267,
      "eval_loss": 0.303169846534729,
      "eval_mcc": 0.3106911694942452,
      "eval_precision": 0.8315217391304348,
      "eval_recall": 0.299412915851272,
      "eval_runtime": 66.7305,
      "eval_samples_per_second": 15.315,
      "eval_steps_per_second": 0.48,
      "step": 768
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.591246485710144,
      "learning_rate": 1.2214973550845031e-05,
      "loss": 0.3085,
      "step": 960
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.62720156555773,
      "eval_f1": 0.4715672676837725,
      "eval_loss": 0.301248162984848,
      "eval_mcc": 0.31481519810966196,
      "eval_precision": 0.8095238095238095,
      "eval_recall": 0.33268101761252444,
      "eval_runtime": 66.7765,
      "eval_samples_per_second": 15.305,
      "eval_steps_per_second": 0.479,
      "step": 960
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.543092966079712,
      "learning_rate": 9.161230163133773e-06,
      "loss": 0.3072,
      "step": 1152
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6663405088062623,
      "eval_f1": 0.5721455457967377,
      "eval_loss": 0.3020024299621582,
      "eval_mcc": 0.3705330895438953,
      "eval_precision": 0.7972027972027972,
      "eval_recall": 0.4461839530332681,
      "eval_runtime": 66.3804,
      "eval_samples_per_second": 15.396,
      "eval_steps_per_second": 0.482,
      "step": 1152
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.1360045671463013,
      "learning_rate": 6.107486775422516e-06,
      "loss": 0.3054,
      "step": 1344
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6761252446183953,
      "eval_f1": 0.5867665418227216,
      "eval_loss": 0.3013523519039154,
      "eval_mcc": 0.39067710929989413,
      "eval_precision": 0.8103448275862069,
      "eval_recall": 0.4598825831702544,
      "eval_runtime": 66.3545,
      "eval_samples_per_second": 15.402,
      "eval_steps_per_second": 0.482,
      "step": 1344
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.4480931758880615,
      "learning_rate": 3.053743387711258e-06,
      "loss": 0.3049,
      "step": 1536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6829745596868885,
      "eval_f1": 0.6029411764705882,
      "eval_loss": 0.3014501929283142,
      "eval_mcc": 0.39988223338971185,
      "eval_precision": 0.8065573770491803,
      "eval_recall": 0.48140900195694714,
      "eval_runtime": 66.4924,
      "eval_samples_per_second": 15.37,
      "eval_steps_per_second": 0.481,
      "step": 1536
    }
  ],
  "logging_steps": 500,
  "max_steps": 1728,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 1885561578240.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.4423735218743273,
    "learning_rate": 2.748369048940132e-05,
    "num_train_epochs": 9,
    "per_device_train_batch_size": 16,
    "temperature": 18
  }
}