Text-to-Speech

[snxraven]

I think it would be cool to add a switch to pipe the responses from the models out to esay or another kind of terminal text to speech method.

This could be very useful to folks with eye sight issues.

[Sovenok-Hacker]

I think we can use Silero TTS to do it, but it is written in Python.

[xor2003]

RHVoice

[Sovenok-Hacker]

RHVoice
It is also good, but Silero have better quality.

[Sovenok-Hacker]

As I know, Torch works in C++

[zenixls2]

mimic3 https://github.com/MycroftAI/mimic3

[zenixls2]

piper from Mycroft former employee

[gogo2464]

Let me just try with

Llamacpp -m model.gguf > out.mp3.

We just have to document tve command. No need to add extra option imho.

[gianpaj]

there's some progress tracked here #13759