{
  "best_metric": 0.7955477445811365,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-16/checkpoint-3210",
  "epoch": 15.0,
  "eval_steps": 500,
  "global_step": 3210,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 7.164041519165039,
      "learning_rate": 0.0004107110800529193,
      "loss": 0.5976,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7182190978324546,
      "eval_f1": 0.49208025343189016,
      "eval_loss": 0.5573095679283142,
      "eval_mcc": 0.32024447377810733,
      "eval_precision": 0.6164021164021164,
      "eval_recall": 0.4094903339191564,
      "eval_runtime": 3.1533,
      "eval_samples_per_second": 541.343,
      "eval_steps_per_second": 17.125,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.810741424560547,
      "learning_rate": 0.00038909470741855514,
      "loss": 0.535,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7299355594610427,
      "eval_f1": 0.5719591457753018,
      "eval_loss": 0.5309674143791199,
      "eval_mcc": 0.3769093274057024,
      "eval_precision": 0.6062992125984252,
      "eval_recall": 0.5413005272407733,
      "eval_runtime": 3.2455,
      "eval_samples_per_second": 525.953,
      "eval_steps_per_second": 16.638,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.5110363960266113,
      "learning_rate": 0.00036747833478419095,
      "loss": 0.51,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7527826596367897,
      "eval_f1": 0.5290178571428571,
      "eval_loss": 0.5084466934204102,
      "eval_mcc": 0.4042060476032326,
      "eval_precision": 0.7247706422018348,
      "eval_recall": 0.4165202108963093,
      "eval_runtime": 3.1369,
      "eval_samples_per_second": 544.167,
      "eval_steps_per_second": 17.214,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.1826329231262207,
      "learning_rate": 0.0003458619621498268,
      "loss": 0.4995,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7451669595782073,
      "eval_f1": 0.49241540256709454,
      "eval_loss": 0.5046694278717041,
      "eval_mcc": 0.38160728386386483,
      "eval_precision": 0.7326388888888888,
      "eval_recall": 0.37082601054481545,
      "eval_runtime": 3.3202,
      "eval_samples_per_second": 514.123,
      "eval_steps_per_second": 16.264,
      "step": 856
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.0804123878479,
      "learning_rate": 0.0003242455895154626,
      "loss": 0.4853,
      "step": 1070
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7574692442882249,
      "eval_f1": 0.6393728222996515,
      "eval_loss": 0.4947313964366913,
      "eval_mcc": 0.45673224487908876,
      "eval_precision": 0.6338514680483592,
      "eval_recall": 0.6449912126537786,
      "eval_runtime": 3.1788,
      "eval_samples_per_second": 536.989,
      "eval_steps_per_second": 16.987,
      "step": 1070
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.1514229774475098,
      "learning_rate": 0.0003026292168810984,
      "loss": 0.4724,
      "step": 1284
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7650849443468073,
      "eval_f1": 0.6598812553011025,
      "eval_loss": 0.477894127368927,
      "eval_mcc": 0.4814730312998762,
      "eval_precision": 0.6377049180327868,
      "eval_recall": 0.6836555360281195,
      "eval_runtime": 3.2099,
      "eval_samples_per_second": 531.785,
      "eval_steps_per_second": 16.823,
      "step": 1284
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.599643230438232,
      "learning_rate": 0.0002810128442467343,
      "loss": 0.46,
      "step": 1498
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.773286467486819,
      "eval_f1": 0.5798045602605862,
      "eval_loss": 0.47905173897743225,
      "eval_mcc": 0.45971678817188244,
      "eval_precision": 0.7585227272727273,
      "eval_recall": 0.46924428822495606,
      "eval_runtime": 3.1625,
      "eval_samples_per_second": 539.761,
      "eval_steps_per_second": 17.075,
      "step": 1498
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.2664639949798584,
      "learning_rate": 0.0002593964716123701,
      "loss": 0.4428,
      "step": 1712
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7885178676039836,
      "eval_f1": 0.6386386386386386,
      "eval_loss": 0.4652605950832367,
      "eval_mcc": 0.5028818263800968,
      "eval_precision": 0.7418604651162791,
      "eval_recall": 0.5606326889279437,
      "eval_runtime": 3.2551,
      "eval_samples_per_second": 524.404,
      "eval_steps_per_second": 16.589,
      "step": 1712
    },
    {
      "epoch": 9.0,
      "grad_norm": 5.06269645690918,
      "learning_rate": 0.00023778009897800593,
      "loss": 0.4442,
      "step": 1926
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7773872290568249,
      "eval_f1": 0.6008403361344538,
      "eval_loss": 0.47534701228141785,
      "eval_mcc": 0.4716667178372095,
      "eval_precision": 0.7467362924281984,
      "eval_recall": 0.5026362038664324,
      "eval_runtime": 3.1553,
      "eval_samples_per_second": 541.002,
      "eval_steps_per_second": 17.114,
      "step": 1926
    },
    {
      "epoch": 10.0,
      "grad_norm": 5.0700459480285645,
      "learning_rate": 0.00021616372634364174,
      "loss": 0.4307,
      "step": 2140
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7838312829525483,
      "eval_f1": 0.6666666666666666,
      "eval_loss": 0.47036686539649963,
      "eval_mcc": 0.5073397319926002,
      "eval_precision": 0.6858736059479554,
      "eval_recall": 0.648506151142355,
      "eval_runtime": 3.3854,
      "eval_samples_per_second": 504.231,
      "eval_steps_per_second": 15.951,
      "step": 2140
    },
    {
      "epoch": 11.0,
      "grad_norm": 6.991399765014648,
      "learning_rate": 0.00019454735370927757,
      "loss": 0.4243,
      "step": 2354
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.7867603983596954,
      "eval_f1": 0.636,
      "eval_loss": 0.4695989191532135,
      "eval_mcc": 0.49868088180967785,
      "eval_precision": 0.7378190255220418,
      "eval_recall": 0.5588752196836555,
      "eval_runtime": 3.1405,
      "eval_samples_per_second": 543.539,
      "eval_steps_per_second": 17.195,
      "step": 2354
    },
    {
      "epoch": 12.0,
      "grad_norm": 5.208732604980469,
      "learning_rate": 0.0001729309810749134,
      "loss": 0.4198,
      "step": 2568
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7896895137668424,
      "eval_f1": 0.666046511627907,
      "eval_loss": 0.45460793375968933,
      "eval_mcc": 0.5152127188733497,
      "eval_precision": 0.7075098814229249,
      "eval_recall": 0.6291739894551845,
      "eval_runtime": 3.8567,
      "eval_samples_per_second": 442.612,
      "eval_steps_per_second": 14.002,
      "step": 2568
    },
    {
      "epoch": 13.0,
      "grad_norm": 16.56888771057129,
      "learning_rate": 0.0001513146084405492,
      "loss": 0.4199,
      "step": 2782
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.7908611599297012,
      "eval_f1": 0.6641580432737535,
      "eval_loss": 0.46026891469955444,
      "eval_mcc": 0.5161067927608803,
      "eval_precision": 0.7145748987854251,
      "eval_recall": 0.6203866432337434,
      "eval_runtime": 3.1632,
      "eval_samples_per_second": 539.639,
      "eval_steps_per_second": 17.071,
      "step": 2782
    },
    {
      "epoch": 14.0,
      "grad_norm": 6.45969820022583,
      "learning_rate": 0.00012969823580618505,
      "loss": 0.4135,
      "step": 2996
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.789103690685413,
      "eval_f1": 0.6622889305816135,
      "eval_loss": 0.4559510350227356,
      "eval_mcc": 0.5124487501229456,
      "eval_precision": 0.710261569416499,
      "eval_recall": 0.6203866432337434,
      "eval_runtime": 3.2281,
      "eval_samples_per_second": 528.794,
      "eval_steps_per_second": 16.728,
      "step": 2996
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.7556620836257935,
      "learning_rate": 0.00010808186317182087,
      "loss": 0.4066,
      "step": 3210
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.7955477445811365,
      "eval_f1": 0.6666666666666667,
      "eval_loss": 0.4586973190307617,
      "eval_mcc": 0.5249372343364972,
      "eval_precision": 0.7301255230125523,
      "eval_recall": 0.6133567662565905,
      "eval_runtime": 3.1701,
      "eval_samples_per_second": 538.473,
      "eval_steps_per_second": 17.034,
      "step": 3210
    }
  ],
  "logging_steps": 500,
  "max_steps": 4280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 7871634966600.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9950627760257441,
    "learning_rate": 0.0004323274526872835,
    "num_train_epochs": 20,
    "temperature": 49
  }
}