fairseq audio text-to-speech

tts_transformer-mn-mbspeech

Transformer text-to-speech model from fairseq S^2 (paper/code):