This is a set of models used for experiments on the differences in bpe in tokenizers trained on domain specific genomes.