run-35/checkpoint-856/trainer_state.json

{
  "best_metric": 0.7281780902167545,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-35/checkpoint-856",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 856,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 9.204007148742676,
      "learning_rate": 0.00010073134118996591,
      "loss": 0.5694,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7047451669595782,
      "eval_f1": 0.514450867052023,
      "eval_loss": 0.5304614901542664,
      "eval_mcc": 0.30808910613726637,
      "eval_precision": 0.5692963752665245,
      "eval_recall": 0.46924428822495606,
      "eval_runtime": 3.1372,
      "eval_samples_per_second": 544.118,
      "eval_steps_per_second": 17.213,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.162537097930908,
      "learning_rate": 8.39427843249716e-05,
      "loss": 0.5228,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7041593438781488,
      "eval_f1": 0.45168295331161784,
      "eval_loss": 0.5144214034080505,
      "eval_mcc": 0.27849213002840095,
      "eval_precision": 0.5909090909090909,
      "eval_recall": 0.3655536028119508,
      "eval_runtime": 3.6842,
      "eval_samples_per_second": 463.333,
      "eval_steps_per_second": 14.657,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.3931689262390137,
      "learning_rate": 6.715422745997727e-05,
      "loss": 0.5045,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7170474516695958,
      "eval_f1": 0.455467869222097,
      "eval_loss": 0.5072826743125916,
      "eval_mcc": 0.3064169652874968,
      "eval_precision": 0.6352201257861635,
      "eval_recall": 0.35500878734622143,
      "eval_runtime": 3.1368,
      "eval_samples_per_second": 544.189,
      "eval_steps_per_second": 17.215,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.560317039489746,
      "learning_rate": 5.0365670594982955e-05,
      "loss": 0.4943,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7281780902167545,
      "eval_f1": 0.49892008639308866,
      "eval_loss": 0.49498075246810913,
      "eval_mcc": 0.3422341055571989,
      "eval_precision": 0.6470588235294118,
      "eval_recall": 0.40597539543058,
      "eval_runtime": 3.6672,
      "eval_samples_per_second": 465.48,
      "eval_steps_per_second": 14.725,
      "step": 856
    }
  ],
  "logging_steps": 500,
  "max_steps": 1498,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 2099102657760.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9231832458790068,
    "learning_rate": 0.00011751989805496024,
    "num_train_epochs": 7,
    "temperature": 10
  }
}