Tokens/words not being counted properly

There is a bug where the software is not counting the words/tokens properly, resulting in inacurrate estimations.