{
  "best_metric": 0.7662745098039215,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-21/checkpoint-800",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 1120,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.839683175086975,
      "learning_rate": 0.0004864196945990802,
      "loss": 0.5129,
      "step": 160
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.5049448609352112,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 1.8683,
      "eval_samples_per_second": 682.449,
      "eval_steps_per_second": 21.41,
      "step": 160
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1274871826171875,
      "learning_rate": 0.0004053497454992336,
      "loss": 0.4794,
      "step": 320
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7050980392156863,
      "eval_f1": 0.26848249027237353,
      "eval_loss": 0.45544198155403137,
      "eval_mcc": 0.2568204849073838,
      "eval_precision": 0.7752808988764045,
      "eval_recall": 0.1623529411764706,
      "eval_runtime": 2.1369,
      "eval_samples_per_second": 596.661,
      "eval_steps_per_second": 18.719,
      "step": 320
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5591638088226318,
      "learning_rate": 0.00032427979639938683,
      "loss": 0.4627,
      "step": 480
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7333333333333333,
      "eval_f1": 0.470404984423676,
      "eval_loss": 0.44455307722091675,
      "eval_mcc": 0.34827660665065835,
      "eval_precision": 0.695852534562212,
      "eval_recall": 0.3552941176470588,
      "eval_runtime": 1.8876,
      "eval_samples_per_second": 675.477,
      "eval_steps_per_second": 21.191,
      "step": 480
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6579329371452332,
      "learning_rate": 0.0002432098472995401,
      "loss": 0.4529,
      "step": 640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7239215686274509,
      "eval_f1": 0.38461538461538464,
      "eval_loss": 0.4378909468650818,
      "eval_mcc": 0.31777730290612743,
      "eval_precision": 0.7482993197278912,
      "eval_recall": 0.25882352941176473,
      "eval_runtime": 1.8818,
      "eval_samples_per_second": 677.548,
      "eval_steps_per_second": 21.256,
      "step": 640
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.6731197834014893,
      "learning_rate": 0.00016213989819969341,
      "loss": 0.4444,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7662745098039215,
      "eval_f1": 0.5906593406593407,
      "eval_loss": 0.4348558187484741,
      "eval_mcc": 0.4456117783291848,
      "eval_precision": 0.7095709570957096,
      "eval_recall": 0.5058823529411764,
      "eval_runtime": 1.8669,
      "eval_samples_per_second": 682.95,
      "eval_steps_per_second": 21.426,
      "step": 800
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.8448941707611084,
      "learning_rate": 8.106994909984671e-05,
      "loss": 0.4415,
      "step": 960
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7623529411764706,
      "eval_f1": 0.5738396624472574,
      "eval_loss": 0.4331771433353424,
      "eval_mcc": 0.4334324670996623,
      "eval_precision": 0.7132867132867133,
      "eval_recall": 0.48,
      "eval_runtime": 1.8638,
      "eval_samples_per_second": 684.078,
      "eval_steps_per_second": 21.461,
      "step": 960
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.8716546297073364,
      "learning_rate": 0.0,
      "loss": 0.436,
      "step": 1120
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7654901960784314,
      "eval_f1": 0.5672937771345876,
      "eval_loss": 0.4309416115283966,
      "eval_mcc": 0.4394955402199731,
      "eval_precision": 0.7368421052631579,
      "eval_recall": 0.4611764705882353,
      "eval_runtime": 1.8706,
      "eval_samples_per_second": 681.61,
      "eval_steps_per_second": 21.384,
      "step": 1120
    }
  ],
  "logging_steps": 500,
  "max_steps": 1120,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 2042288065440.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7976867614913634,
    "learning_rate": 0.000567489643698927,
    "num_train_epochs": 7,
    "temperature": 7
  }
}