{
  "best_metric": 0.8307240704500979,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-5/checkpoint-864",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 864,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.850555896759033,
      "learning_rate": 0.00013535007385325417,
      "loss": 0.604,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7485322896281801,
      "eval_f1": 0.7925746569814367,
      "eval_loss": 0.5116264820098877,
      "eval_precision": 0.6744505494505495,
      "eval_recall": 0.9608610567514677,
      "eval_runtime": 25.4927,
      "eval_samples_per_second": 40.09,
      "eval_steps_per_second": 1.255,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.2508227825164795,
      "learning_rate": 0.0001184313146215974,
      "loss": 0.4897,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7984344422700587,
      "eval_f1": 0.8214904679376083,
      "eval_loss": 0.4436704218387604,
      "eval_precision": 0.7371695178849145,
      "eval_recall": 0.9275929549902152,
      "eval_runtime": 26.2196,
      "eval_samples_per_second": 38.979,
      "eval_steps_per_second": 1.22,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.221353530883789,
      "learning_rate": 0.00010151255538994063,
      "loss": 0.4449,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.812133072407045,
      "eval_f1": 0.8181818181818181,
      "eval_loss": 0.43934348225593567,
      "eval_precision": 0.7926605504587156,
      "eval_recall": 0.8454011741682974,
      "eval_runtime": 25.7442,
      "eval_samples_per_second": 39.698,
      "eval_steps_per_second": 1.243,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.505247116088867,
      "learning_rate": 8.459379615828387e-05,
      "loss": 0.4275,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8170254403131115,
      "eval_f1": 0.8257222739981361,
      "eval_loss": 0.41963326930999756,
      "eval_precision": 0.7882562277580071,
      "eval_recall": 0.8669275929549902,
      "eval_runtime": 25.8902,
      "eval_samples_per_second": 39.474,
      "eval_steps_per_second": 1.236,
      "step": 384
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.8876335620880127,
      "learning_rate": 6.767503692662709e-05,
      "loss": 0.4125,
      "step": 480
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.821917808219178,
      "eval_f1": 0.8372093023255814,
      "eval_loss": 0.41102567315101624,
      "eval_precision": 0.771004942339374,
      "eval_recall": 0.9158512720156555,
      "eval_runtime": 26.546,
      "eval_samples_per_second": 38.499,
      "eval_steps_per_second": 1.205,
      "step": 480
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.6481876373291016,
      "learning_rate": 5.075627769497032e-05,
      "loss": 0.4037,
      "step": 576
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8199608610567515,
      "eval_f1": 0.8371681415929204,
      "eval_loss": 0.4179791808128357,
      "eval_precision": 0.7641357027463651,
      "eval_recall": 0.9256360078277887,
      "eval_runtime": 25.6553,
      "eval_samples_per_second": 39.836,
      "eval_steps_per_second": 1.247,
      "step": 576
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.620311260223389,
      "learning_rate": 3.383751846331354e-05,
      "loss": 0.3929,
      "step": 672
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8268101761252447,
      "eval_f1": 0.8426666666666667,
      "eval_loss": 0.4127921462059021,
      "eval_precision": 0.7719869706840391,
      "eval_recall": 0.9275929549902152,
      "eval_runtime": 26.1034,
      "eval_samples_per_second": 39.152,
      "eval_steps_per_second": 1.226,
      "step": 672
    },
    {
      "epoch": 8.0,
      "grad_norm": 4.286402225494385,
      "learning_rate": 1.691875923165677e-05,
      "loss": 0.3895,
      "step": 768
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8277886497064579,
      "eval_f1": 0.842857142857143,
      "eval_loss": 0.4042445123195648,
      "eval_precision": 0.7750410509031199,
      "eval_recall": 0.923679060665362,
      "eval_runtime": 26.1296,
      "eval_samples_per_second": 39.113,
      "eval_steps_per_second": 1.225,
      "step": 768
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.3019790649414062,
      "learning_rate": 0.0,
      "loss": 0.3874,
      "step": 864
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8307240704500979,
      "eval_f1": 0.843155031731641,
      "eval_loss": 0.4016450047492981,
      "eval_precision": 0.785472972972973,
      "eval_recall": 0.9099804305283757,
      "eval_runtime": 25.5652,
      "eval_samples_per_second": 39.976,
      "eval_steps_per_second": 1.252,
      "step": 864
    }
  ],
  "logging_steps": 500,
  "max_steps": 864,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 2121256775520.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9725110110091697,
    "learning_rate": 0.00015226883308491095,
    "num_train_epochs": 9,
    "temperature": 30
  }
}