Closed aedocw closed 9 months ago
This is in progress and working great with XTTSv2. It breaks backward compatibility as I took this work as a good opportunity to switch to an OOP approach, and as of now this only works with XTTS and epubs.
I'll add back in existing functionality (text files, VITS model, etc) over the next few weeks.
This was completed and merged.
Some post-processing of of the audio would be nice, experiment with https://github.com/spotify/pedalboard.
Also try inference_stream https://discord.com/channels/1037326658807533628/1062887209352581151/1177002620339114046