I\'m confused by the sklearn documentation of Naive Bayes Classification. Do I need to specify whether I\'m throwing in counts, TF, or TF-IDF