Here's the repo:
https://github.com/SamReynoso/pfspeak
It's pretty sick. The usage is just:
1) Change the voice in the conifg
2) Restart the service (for today - restart is required)
3) echo "Hello World!" > "$SPEAK"
I spent a fair amount of time into building the runtime, and the application setup logic. I'm looking for testers interested in building a with the package, so if anyone has had a TTS project sitting on the shelf...
I have a handle on the direction I want to take this now. That's keep working on the runtime, and application framework part.
But checkout the daemon. With how relatively small the model is you'll probably keep it running all the time, and call to it from all over the place.
[link] [comments]