[butz]

Generating audio takes a bit, but wow, 92MB model for really decent sounding speech. Is there a way to plug this thing into speech dispatcher on Linux and use for accessibility?