Word Tokenization: move Cree off non-whitespace languages list
It seems on further inspection that though Cree can mix whitespace and non-whitespace delineation, the language edition has largely used whitespace thusfar and a mixture of Cree syllabics and latin characters. More overview: https://en.wikipedia.org/wiki/Cree_syllabics
Thus, we can move it off of the non-whitespace languages and retire its sentencepiece model.