Evaluation metrics, e.g., accuracy and auc, can NOT be used at the same time

For a classification task, cannot use all metrics at the same time.

Examples:
Accuracy and auc_roc requires different prediction return values. Therefore they cannot be used in one training/evaluation epoch.

Reasons:
GS uses `return_proba` to control the types of prediction values, which means the predict function can NOT return both types of values.

Suggested solution:
- Option 1: modify predict function to return both;
- Option 2: based on the metrics given, run one or twice of the predict function.