{
  "best_metric": 0.7301960784313726,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-10/checkpoint-800",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 960,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.012742280960083,
      "learning_rate": 0.0007591954920690624,
      "loss": 0.5869,
      "step": 160
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.5785399675369263,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 1.8697,
      "eval_samples_per_second": 681.924,
      "eval_steps_per_second": 21.394,
      "step": 160
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.057482957839966,
      "learning_rate": 0.0006642960555604296,
      "loss": 0.5715,
      "step": 320
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.676078431372549,
      "eval_f1": 0.4411366711772666,
      "eval_loss": 0.5550761222839355,
      "eval_mcc": 0.22526645932553852,
      "eval_precision": 0.5191082802547771,
      "eval_recall": 0.3835294117647059,
      "eval_runtime": 1.8672,
      "eval_samples_per_second": 682.827,
      "eval_steps_per_second": 21.422,
      "step": 320
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.645397663116455,
      "learning_rate": 0.0005693966190517968,
      "loss": 0.5736,
      "step": 480
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7043137254901961,
      "eval_f1": 0.28733459357277885,
      "eval_loss": 0.5566152334213257,
      "eval_mcc": 0.25125318174069416,
      "eval_precision": 0.7307692307692307,
      "eval_recall": 0.17882352941176471,
      "eval_runtime": 1.8663,
      "eval_samples_per_second": 683.154,
      "eval_steps_per_second": 21.432,
      "step": 480
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3315376043319702,
      "learning_rate": 0.000474497182543164,
      "loss": 0.5569,
      "step": 640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7113725490196079,
      "eval_f1": 0.4121405750798722,
      "eval_loss": 0.5367588400840759,
      "eval_mcc": 0.2830725629191131,
      "eval_precision": 0.6417910447761194,
      "eval_recall": 0.3035294117647059,
      "eval_runtime": 1.8722,
      "eval_samples_per_second": 681.014,
      "eval_steps_per_second": 21.365,
      "step": 640
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.5584187507629395,
      "learning_rate": 0.0003795977460345312,
      "loss": 0.5469,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7301960784313726,
      "eval_f1": 0.50997150997151,
      "eval_loss": 0.5246202945709229,
      "eval_mcc": 0.3496658305237892,
      "eval_precision": 0.6462093862815884,
      "eval_recall": 0.4211764705882353,
      "eval_runtime": 1.8851,
      "eval_samples_per_second": 676.35,
      "eval_steps_per_second": 21.219,
      "step": 800
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.2840497493743896,
      "learning_rate": 0.0002846983095258984,
      "loss": 0.5382,
      "step": 960
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7168627450980393,
      "eval_f1": 0.5305591677503251,
      "eval_loss": 0.5253874659538269,
      "eval_mcc": 0.3348618979447744,
      "eval_precision": 0.5930232558139535,
      "eval_recall": 0.48,
      "eval_runtime": 1.8686,
      "eval_samples_per_second": 682.333,
      "eval_steps_per_second": 21.407,
      "step": 960
    }
  ],
  "logging_steps": 500,
  "max_steps": 1440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 1750532627520.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9116955099903541,
    "learning_rate": 0.0008540949285776952,
    "num_train_epochs": 9,
    "temperature": 21
  }
}