{
  "best_metric": 0.8238747553816047,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-14/checkpoint-1152",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 1344,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.0995702743530273,
      "learning_rate": 0.00024019499178328863,
      "loss": 0.4158,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7318982387475538,
      "eval_f1": 0.7078891257995735,
      "eval_loss": 0.38118404150009155,
      "eval_mcc": 0.4701927470442256,
      "eval_precision": 0.7775175644028103,
      "eval_recall": 0.649706457925636,
      "eval_runtime": 66.3838,
      "eval_samples_per_second": 15.395,
      "eval_steps_per_second": 0.482,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.5503435134887695,
      "learning_rate": 0.00020016249315274053,
      "loss": 0.3871,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7798434442270059,
      "eval_f1": 0.7619047619047619,
      "eval_loss": 0.3572150468826294,
      "eval_mcc": 0.566151295376701,
      "eval_precision": 0.8294930875576036,
      "eval_recall": 0.7045009784735812,
      "eval_runtime": 66.9799,
      "eval_samples_per_second": 15.258,
      "eval_steps_per_second": 0.478,
      "step": 384
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.7390501499176025,
      "learning_rate": 0.00016012999452219242,
      "loss": 0.3651,
      "step": 576
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8072407045009785,
      "eval_f1": 0.8047571853320119,
      "eval_loss": 0.36672884225845337,
      "eval_mcc": 0.6146803546891556,
      "eval_precision": 0.8152610441767069,
      "eval_recall": 0.7945205479452054,
      "eval_runtime": 66.7321,
      "eval_samples_per_second": 15.315,
      "eval_steps_per_second": 0.48,
      "step": 576
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.3112893104553223,
      "learning_rate": 0.00012009749589164432,
      "loss": 0.3601,
      "step": 768
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8131115459882583,
      "eval_f1": 0.811451135241856,
      "eval_loss": 0.35847175121307373,
      "eval_mcc": 0.626320242099439,
      "eval_precision": 0.8187250996015937,
      "eval_recall": 0.8043052837573386,
      "eval_runtime": 66.7668,
      "eval_samples_per_second": 15.307,
      "eval_steps_per_second": 0.479,
      "step": 768
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.440983295440674,
      "learning_rate": 8.006499726109621e-05,
      "loss": 0.3558,
      "step": 960
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.812133072407045,
      "eval_f1": 0.8095238095238095,
      "eval_loss": 0.36063674092292786,
      "eval_mcc": 0.6245005669570931,
      "eval_precision": 0.8209255533199196,
      "eval_recall": 0.7984344422700587,
      "eval_runtime": 66.686,
      "eval_samples_per_second": 15.326,
      "eval_steps_per_second": 0.48,
      "step": 960
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.304095506668091,
      "learning_rate": 4.0032498630548105e-05,
      "loss": 0.349,
      "step": 1152
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8238747553816047,
      "eval_f1": 0.8311444652908067,
      "eval_loss": 0.3584487736225128,
      "eval_mcc": 0.6501642144850927,
      "eval_precision": 0.7981981981981981,
      "eval_recall": 0.8669275929549902,
      "eval_runtime": 66.5615,
      "eval_samples_per_second": 15.354,
      "eval_steps_per_second": 0.481,
      "step": 1152
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.643546223640442,
      "learning_rate": 0.0,
      "loss": 0.3446,
      "step": 1344
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8189823874755382,
      "eval_f1": 0.8198636806231744,
      "eval_loss": 0.3578638434410095,
      "eval_mcc": 0.637995316854455,
      "eval_precision": 0.8158914728682171,
      "eval_recall": 0.8238747553816047,
      "eval_runtime": 68.6894,
      "eval_samples_per_second": 14.879,
      "eval_steps_per_second": 0.466,
      "step": 1344
    }
  ],
  "logging_steps": 500,
  "max_steps": 1344,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 1649866380960.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6115137465963567,
    "learning_rate": 0.00028022749041383674,
    "num_train_epochs": 7,
    "per_device_train_batch_size": 16,
    "temperature": 33
  }
}