zero-resource-speech-tokenizer(zrst) Library that transcribes wav files into sequences of subword symbols. Requires HTK, SRILM, matlab and python, numpy, scipy.