1. datasets module: capable of downloading/caching famous NLP datasets.
2. embeddings module: capable of downloadin/caching famous pretrained embeddings or creating own embeddings
3. text_encoders module: various text encoding utils 
4. metrics module: various metrics utility functions
5. utils module: various helper functions (also include download utils)
7. models.sklearn.base_sklearn: sklearn based classifiers
8. models.torch.under torch build various nn modules
