Splitting into Categorical features

https://github.com/harshraj22/DeepLearning/blob/60b90a6cd9a79025cbb658691dcdc329dbe26b0a/implementations/TabTransformer/data_loader/datasets.py#L21

Since the dataset is specific, it seems ok to use variables to store categorical and continuous variables. For generic case, you may want to use [fastai’s cont_cat_split](https://khuyentran1401.github.io/Efficient_Python_tricks_and_tools_for_data_scientists/Chapter5/feature_extraction.html#fastais-cont-cat-split-get-a-dataframes-continuous-and-categorical-variables-based-on-their-cardinality)