[Feature Request] Additional Voices: Wheatley, etc.

[dannyvfilms]

Your voice models are the perfect format for use in Home Assistant through the Piper add-on. I haven’t researched the process of training voice models yet, but since yours worked so well, I had to at least ask if you were open to creating more voice models for use in this or other applications?

[YoshiFan501]

I feel like Cave Johnson would also be a good addition as well as Wheatley, but Wheatley would make a perfect assistant.

[dnhkng]

New voices are being planned. I'm currently looking into new TTS models, as VITS is pretty old now.

[youcefs21]

@dnhkng
can you add training code to the repo? The GlaDOS voice is very on point, I'm curious how you trained/fine tuned the model and what params you used, and any other details would be great!

[dnhkng]

I will add the training code at some stage! I would like to refactor the code a bit first though.

[s5cubed]

Can you give her the voice she makes for "[SUBJECT_NAME_HERE]" and "[SUBJECT_HOMETOWN_HERE]" those were cool.

[je-miller]

New voices are being planned. I'm currently looking into new TTS models, as VITS is pretty old now.

Have you taken a look at F5-TTS yet? If so, what was the verdict/performance?