add zonos3-ivy voice: embedding + sample render (desktop CPU expressive TTS)

#1

192-d MECT-B2 speaker embedding + 9.6s sample cloned on-device from a 15s clip. For use with Zyphra ZONOS3-Mini ggml server (Apache 2.0).

Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment