summary

multilingual tokenizer trained on multilingual data by using the SentencePiece library and the BPE algorithm.