{
  "best_metric": 0.7733333333333333,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-20/checkpoint-960",
  "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": 2.297672986984253,
      "learning_rate": 0.0004503757705353633,
      "loss": 0.5065,
      "step": 160
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6909803921568628,
      "eval_f1": 0.286231884057971,
      "eval_loss": 0.46740272641181946,
      "eval_mcc": 0.20370649805068908,
      "eval_precision": 0.6220472440944882,
      "eval_recall": 0.18588235294117647,
      "eval_runtime": 1.8709,
      "eval_samples_per_second": 681.504,
      "eval_steps_per_second": 21.381,
      "step": 160
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.163440227508545,
      "learning_rate": 0.0003753131421128028,
      "loss": 0.472,
      "step": 320
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7090196078431372,
      "eval_f1": 0.29601518026565465,
      "eval_loss": 0.447980672121048,
      "eval_mcc": 0.26984218289739087,
      "eval_precision": 0.7647058823529411,
      "eval_recall": 0.18352941176470589,
      "eval_runtime": 2.1203,
      "eval_samples_per_second": 601.317,
      "eval_steps_per_second": 18.865,
      "step": 320
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.110410690307617,
      "learning_rate": 0.0003002505136902422,
      "loss": 0.4616,
      "step": 480
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7262745098039216,
      "eval_f1": 0.5146036161335187,
      "eval_loss": 0.445524126291275,
      "eval_mcc": 0.34365078524294146,
      "eval_precision": 0.6292517006802721,
      "eval_recall": 0.43529411764705883,
      "eval_runtime": 1.8755,
      "eval_samples_per_second": 679.828,
      "eval_steps_per_second": 21.328,
      "step": 480
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1434326171875,
      "learning_rate": 0.00022518788526768165,
      "loss": 0.452,
      "step": 640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7341176470588235,
      "eval_f1": 0.4593301435406698,
      "eval_loss": 0.44121384620666504,
      "eval_mcc": 0.3493311330281174,
      "eval_precision": 0.7128712871287128,
      "eval_recall": 0.3388235294117647,
      "eval_runtime": 1.8788,
      "eval_samples_per_second": 678.621,
      "eval_steps_per_second": 21.29,
      "step": 640
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.2687256336212158,
      "learning_rate": 0.0001501252568451211,
      "loss": 0.4436,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7607843137254902,
      "eval_f1": 0.5970937912813739,
      "eval_loss": 0.43906503915786743,
      "eval_mcc": 0.4372564381540906,
      "eval_precision": 0.6807228915662651,
      "eval_recall": 0.5317647058823529,
      "eval_runtime": 1.8778,
      "eval_samples_per_second": 678.984,
      "eval_steps_per_second": 21.301,
      "step": 800
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.7441729307174683,
      "learning_rate": 7.506262842256055e-05,
      "loss": 0.4334,
      "step": 960
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7733333333333333,
      "eval_f1": 0.611036339165545,
      "eval_loss": 0.4335384964942932,
      "eval_mcc": 0.46528826388603933,
      "eval_precision": 0.7138364779874213,
      "eval_recall": 0.5341176470588235,
      "eval_runtime": 1.8888,
      "eval_samples_per_second": 675.037,
      "eval_steps_per_second": 21.178,
      "step": 960
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.9164094924926758,
      "learning_rate": 0.0,
      "loss": 0.4319,
      "step": 1120
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7694117647058824,
      "eval_f1": 0.5870786516853933,
      "eval_loss": 0.429920494556427,
      "eval_mcc": 0.45148621894108926,
      "eval_precision": 0.7282229965156795,
      "eval_recall": 0.49176470588235294,
      "eval_runtime": 1.8741,
      "eval_samples_per_second": 680.328,
      "eval_steps_per_second": 21.344,
      "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": 0.796338716906447,
    "learning_rate": 0.0005254383989579239,
    "num_train_epochs": 7,
    "temperature": 7
  }
}