contains pretrained tokenizers for multilingual language modeling. Designed with a focus on African and other underrepresented languages