{
  "best_metric": 0.7715289982425307,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-8/checkpoint-2140",
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 2568,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.357285261154175,
      "learning_rate": 0.00022367746295285276,
      "loss": 0.4396,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6842413591095489,
      "eval_f1": 0.10909090909090909,
      "eval_loss": 0.41522327065467834,
      "eval_mcc": 0.1816293698914945,
      "eval_precision": 0.9166666666666666,
      "eval_recall": 0.05799648506151142,
      "eval_runtime": 3.1634,
      "eval_samples_per_second": 539.601,
      "eval_steps_per_second": 17.07,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.425903558731079,
      "learning_rate": 0.00020334314813895706,
      "loss": 0.4062,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7416520210896309,
      "eval_f1": 0.449438202247191,
      "eval_loss": 0.4018629789352417,
      "eval_mcc": 0.37230229731288733,
      "eval_precision": 0.7758620689655172,
      "eval_recall": 0.3163444639718805,
      "eval_runtime": 3.8997,
      "eval_samples_per_second": 437.73,
      "eval_steps_per_second": 13.847,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.631404161453247,
      "learning_rate": 0.00018300883332506135,
      "loss": 0.3929,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7322788517867604,
      "eval_f1": 0.3898531375166889,
      "eval_loss": 0.3961062431335449,
      "eval_mcc": 0.347975715943952,
      "eval_precision": 0.8111111111111111,
      "eval_recall": 0.2565905096660808,
      "eval_runtime": 3.1833,
      "eval_samples_per_second": 536.238,
      "eval_steps_per_second": 16.964,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.494969367980957,
      "learning_rate": 0.00016267451851116564,
      "loss": 0.3877,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7492677211482133,
      "eval_f1": 0.45959595959595967,
      "eval_loss": 0.38814693689346313,
      "eval_mcc": 0.3970252633604229,
      "eval_precision": 0.8161434977578476,
      "eval_recall": 0.31985940246045697,
      "eval_runtime": 3.8771,
      "eval_samples_per_second": 440.28,
      "eval_steps_per_second": 13.928,
      "step": 856
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.2848458290100098,
      "learning_rate": 0.00014234020369726996,
      "loss": 0.3808,
      "step": 1070
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7621558289396603,
      "eval_f1": 0.5458612975391499,
      "eval_loss": 0.3853852450847626,
      "eval_mcc": 0.4294214256069951,
      "eval_precision": 0.7507692307692307,
      "eval_recall": 0.4288224956063269,
      "eval_runtime": 3.119,
      "eval_samples_per_second": 547.282,
      "eval_steps_per_second": 17.313,
      "step": 1070
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.14024019241333,
      "learning_rate": 0.00012200588888337424,
      "loss": 0.379,
      "step": 1284
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7656707674282367,
      "eval_f1": 0.5771670190274841,
      "eval_loss": 0.38638636469841003,
      "eval_mcc": 0.4413780950203647,
      "eval_precision": 0.7241379310344828,
      "eval_recall": 0.4797891036906854,
      "eval_runtime": 3.2685,
      "eval_samples_per_second": 522.261,
      "eval_steps_per_second": 16.521,
      "step": 1284
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.743562698364258,
      "learning_rate": 0.00010167157406947853,
      "loss": 0.3744,
      "step": 1498
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7521968365553603,
      "eval_f1": 0.4705882352941177,
      "eval_loss": 0.38443464040756226,
      "eval_mcc": 0.405207339687894,
      "eval_precision": 0.8173913043478261,
      "eval_recall": 0.3304042179261863,
      "eval_runtime": 3.1595,
      "eval_samples_per_second": 540.284,
      "eval_steps_per_second": 17.092,
      "step": 1498
    },
    {
      "epoch": 8.0,
      "grad_norm": 4.100239276885986,
      "learning_rate": 8.133725925558282e-05,
      "loss": 0.3683,
      "step": 1712
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7639132981839485,
      "eval_f1": 0.5351787773933103,
      "eval_loss": 0.38164493441581726,
      "eval_mcc": 0.43431461671833843,
      "eval_precision": 0.7785234899328859,
      "eval_recall": 0.4077328646748682,
      "eval_runtime": 3.2702,
      "eval_samples_per_second": 521.994,
      "eval_steps_per_second": 16.513,
      "step": 1712
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.055349111557007,
      "learning_rate": 6.100294444168712e-05,
      "loss": 0.3696,
      "step": 1926
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7697715289982425,
      "eval_f1": 0.5769644779332616,
      "eval_loss": 0.3795158863067627,
      "eval_mcc": 0.4508508824307292,
      "eval_precision": 0.7444444444444445,
      "eval_recall": 0.4710017574692443,
      "eval_runtime": 3.1704,
      "eval_samples_per_second": 538.426,
      "eval_steps_per_second": 17.033,
      "step": 1926
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.974280595779419,
      "learning_rate": 4.066862962779141e-05,
      "loss": 0.3649,
      "step": 2140
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7715289982425307,
      "eval_f1": 0.5666666666666667,
      "eval_loss": 0.37987595796585083,
      "eval_mcc": 0.4547278125926345,
      "eval_precision": 0.770392749244713,
      "eval_recall": 0.44815465729349735,
      "eval_runtime": 3.3339,
      "eval_samples_per_second": 512.008,
      "eval_steps_per_second": 16.197,
      "step": 2140
    },
    {
      "epoch": 11.0,
      "grad_norm": 3.0384652614593506,
      "learning_rate": 2.0334314813895705e-05,
      "loss": 0.3626,
      "step": 2354
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.7709431751611013,
      "eval_f1": 0.5679558011049725,
      "eval_loss": 0.37754446268081665,
      "eval_mcc": 0.45319581737950565,
      "eval_precision": 0.7648809523809523,
      "eval_recall": 0.45166959578207383,
      "eval_runtime": 3.1573,
      "eval_samples_per_second": 540.652,
      "eval_steps_per_second": 17.103,
      "step": 2354
    },
    {
      "epoch": 12.0,
      "grad_norm": 1.8090615272521973,
      "learning_rate": 0.0,
      "loss": 0.3599,
      "step": 2568
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7715289982425307,
      "eval_f1": 0.5859872611464968,
      "eval_loss": 0.3779904842376709,
      "eval_mcc": 0.45610419653874357,
      "eval_precision": 0.739946380697051,
      "eval_recall": 0.4850615114235501,
      "eval_runtime": 3.2799,
      "eval_samples_per_second": 520.449,
      "eval_steps_per_second": 16.464,
      "step": 2568
    }
  ],
  "logging_steps": 500,
  "max_steps": 2568,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 500,
  "total_flos": 6297307973280.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6983600329595915,
    "learning_rate": 0.00024401177776674847,
    "num_train_epochs": 12,
    "temperature": 36
  }
}