fairseq S2T: Fast Speech-to-Text Modeling with fairseq Paper • 2010.05171 • Published Oct 11, 2020 • 2