run-0/checkpoint-1032/trainer_state.json

{
  "best_metric": 0.802,
  "best_model_checkpoint": "Bert_tinybert-distilled/run-0/checkpoint-1032",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1032,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.40991276502609253,
      "learning_rate": 0.00012592472624312588,
      "loss": 0.3541,
      "step": 172
    },
    {
      "epoch": 1.0,
      "eval_AUC": 0.8165582222222223,
      "eval_BAC": 0.7313333333333334,
      "eval_MCC": 0.4637400760214038,
      "eval_PR_AUC": 0.7669251828483298,
      "eval_accuracy": 0.7313333333333333,
      "eval_loss": 0.33297714591026306,
      "eval_runtime": 5.88,
      "eval_samples_per_second": 255.102,
      "eval_steps_per_second": 7.993,
      "step": 172
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7852521538734436,
      "learning_rate": 0.0001007397809945007,
      "loss": 0.3395,
      "step": 344
    },
    {
      "epoch": 2.0,
      "eval_AUC": 0.8291591111111111,
      "eval_BAC": 0.7453333333333334,
      "eval_MCC": 0.5307731566207068,
      "eval_PR_AUC": 0.7773307731479302,
      "eval_accuracy": 0.7453333333333333,
      "eval_loss": 0.32892802357673645,
      "eval_runtime": 5.8962,
      "eval_samples_per_second": 254.401,
      "eval_steps_per_second": 7.971,
      "step": 344
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8841129541397095,
      "learning_rate": 7.555483574587553e-05,
      "loss": 0.3354,
      "step": 516
    },
    {
      "epoch": 3.0,
      "eval_AUC": 0.8464124444444445,
      "eval_BAC": 0.7846666666666666,
      "eval_MCC": 0.585494859343166,
      "eval_PR_AUC": 0.8098393230420249,
      "eval_accuracy": 0.7846666666666666,
      "eval_loss": 0.3203962743282318,
      "eval_runtime": 6.0014,
      "eval_samples_per_second": 249.942,
      "eval_steps_per_second": 7.832,
      "step": 516
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.4307535886764526,
      "learning_rate": 5.036989049725035e-05,
      "loss": 0.3283,
      "step": 688
    },
    {
      "epoch": 4.0,
      "eval_AUC": 0.853079111111111,
      "eval_BAC": 0.7986666666666666,
      "eval_MCC": 0.6048666115791026,
      "eval_PR_AUC": 0.8161612377554972,
      "eval_accuracy": 0.7986666666666666,
      "eval_loss": 0.3148361146450043,
      "eval_runtime": 6.1071,
      "eval_samples_per_second": 245.616,
      "eval_steps_per_second": 7.696,
      "step": 688
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.2096748352050781,
      "learning_rate": 2.5184945248625176e-05,
      "loss": 0.3238,
      "step": 860
    },
    {
      "epoch": 5.0,
      "eval_AUC": 0.8653653333333333,
      "eval_BAC": 0.8013333333333333,
      "eval_MCC": 0.6179940224891067,
      "eval_PR_AUC": 0.8210118227054324,
      "eval_accuracy": 0.8013333333333333,
      "eval_loss": 0.31179723143577576,
      "eval_runtime": 6.1725,
      "eval_samples_per_second": 243.014,
      "eval_steps_per_second": 7.614,
      "step": 860
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.405503273010254,
      "learning_rate": 0.0,
      "loss": 0.3214,
      "step": 1032
    },
    {
      "epoch": 6.0,
      "eval_AUC": 0.8693102222222221,
      "eval_BAC": 0.802,
      "eval_MCC": 0.6195541306150731,
      "eval_PR_AUC": 0.8205258347742059,
      "eval_accuracy": 0.802,
      "eval_loss": 0.31147220730781555,
      "eval_runtime": 6.2204,
      "eval_samples_per_second": 241.144,
      "eval_steps_per_second": 7.556,
      "step": 1032
    }
  ],
  "logging_steps": 500,
  "max_steps": 1032,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2702266380000.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.5274993372148938,
    "learning_rate": 0.00015110967149175106,
    "num_train_epochs": 6,
    "temperature": 12
  }
}