{
  "best_metric": 0.6831683168316832,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-18/checkpoint-2079",
  "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": 1.6970789432525635,
      "learning_rate": 0.0001678982407816263,
      "loss": 0.6448,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5287128712871287,
      "eval_f1": 0.22222222222222224,
      "eval_loss": 0.6524360179901123,
      "eval_mcc": 0.09039579598072835,
      "eval_precision": 0.6296296296296297,
      "eval_recall": 0.1349206349206349,
      "eval_runtime": 0.9373,
      "eval_samples_per_second": 538.78,
      "eval_steps_per_second": 17.07,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.302747964859009,
      "learning_rate": 0.0001399152006513553,
      "loss": 0.6287,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6237623762376238,
      "eval_f1": 0.6415094339622641,
      "eval_loss": 0.6215968132019043,
      "eval_mcc": 0.24899823929472167,
      "eval_precision": 0.6115107913669064,
      "eval_recall": 0.6746031746031746,
      "eval_runtime": 0.9284,
      "eval_samples_per_second": 543.954,
      "eval_steps_per_second": 17.234,
      "step": 594
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.1591579914093018,
      "learning_rate": 0.00011193216052108422,
      "loss": 0.5976,
      "step": 891
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6574257425742575,
      "eval_f1": 0.7205169628432957,
      "eval_loss": 0.610277533531189,
      "eval_mcc": 0.3542682585809925,
      "eval_precision": 0.6076294277929155,
      "eval_recall": 0.8849206349206349,
      "eval_runtime": 0.9295,
      "eval_samples_per_second": 543.279,
      "eval_steps_per_second": 17.213,
      "step": 891
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.0118372440338135,
      "learning_rate": 8.394912039081316e-05,
      "loss": 0.5713,
      "step": 1188
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6534653465346535,
      "eval_f1": 0.7107438016528924,
      "eval_loss": 0.6075993776321411,
      "eval_mcc": 0.33543408373963995,
      "eval_precision": 0.6090651558073654,
      "eval_recall": 0.8531746031746031,
      "eval_runtime": 0.9317,
      "eval_samples_per_second": 542.017,
      "eval_steps_per_second": 17.173,
      "step": 1188
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.029730796813965,
      "learning_rate": 5.596608026054211e-05,
      "loss": 0.555,
      "step": 1485
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6653465346534654,
      "eval_f1": 0.70298769771529,
      "eval_loss": 0.585565984249115,
      "eval_mcc": 0.34256466693603754,
      "eval_precision": 0.6309148264984227,
      "eval_recall": 0.7936507936507936,
      "eval_runtime": 0.9318,
      "eval_samples_per_second": 541.94,
      "eval_steps_per_second": 17.17,
      "step": 1485
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.707862138748169,
      "learning_rate": 2.7983040130271054e-05,
      "loss": 0.5461,
      "step": 1782
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6772277227722773,
      "eval_f1": 0.7104795737122557,
      "eval_loss": 0.58236163854599,
      "eval_mcc": 0.3648417915119883,
      "eval_precision": 0.6430868167202572,
      "eval_recall": 0.7936507936507936,
      "eval_runtime": 0.9341,
      "eval_samples_per_second": 540.611,
      "eval_steps_per_second": 17.128,
      "step": 1782
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.284318447113037,
      "learning_rate": 0.0,
      "loss": 0.5394,
      "step": 2079
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6831683168316832,
      "eval_f1": 0.7132616487455198,
      "eval_loss": 0.5821448564529419,
      "eval_mcc": 0.37527743125375096,
      "eval_precision": 0.6503267973856209,
      "eval_recall": 0.7896825396825397,
      "eval_runtime": 0.9426,
      "eval_samples_per_second": 535.728,
      "eval_steps_per_second": 16.974,
      "step": 2079
    }
  ],
  "logging_steps": 500,
  "max_steps": 2079,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 5114908861860.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9135980547715112,
    "learning_rate": 0.00019588128091189738,
    "num_train_epochs": 7,
    "temperature": 19
  }
}