Supports hyperparameter tuning for nn methods

[qychen2001]

Very good work!
Hyperparameter tuning is a common strategy for nn-based methods. This will also give better results for nn-based methods. But I didn't find a simple implementation. May I ask if support will be considered in the future? Or is there any simple way to implement this part?

[weihua916]

We have a script that does optuna-based hyper-param tuning for a variety of deep tabular models; https://github.com/pyg-team/pytorch-frame/blob/master/benchmark/data_frame_benchmark.py

[qychen2001]

Thanks for the answer! My problem has been solved.