Tokenization + Char Ngrams

[Hossam Amer]

(1) What is the subword tokenization done for English FastText input?

I understand that you might refer me to Europarl. However, I cannot see where this preprocessing script occurs.

(2) How can I know the value of n used in the pre-trained model character n-gram?