Title question mostly. I’ve played with XTTS-v2 and it worked pretty well, but I’m wondering if folks are using anything else special. I’d like to train my own voice finetune which is what I did with XTTS-v2, and then use it with home assistant’s voice feature. Welcome all opinions on it!
You must log in or register to comment.
if you need English - right now it’s kokoro-fastapi https://github.com/remsky/Kokoro-FastAPI set this container up and use it as an openai TTS endpoint using this hacs integration https://github.com/sfortis/openai_tts
Very nice! I’ll check this out!
Piper works pretty well. I’m only using it because it was easier to find a custom glados voice.
Kokoro has good default voices. I also started trying out Speaches recently. It provides an open ai api wrapper around several options
Any tips on getting speaches to work with Home assistant? Got speaches working but haven’t gone the next step yet.
Don’t know much about the training side of things, but I have Piper set up with home assistant using the Wyoming protocol and it just goes. Some of the out-of-the-box voices are pretty decent too.
Pretty much just personal preference at this point. XTTS is certainly not the most efficient though.
any personal preferences you recommend?
Pico, Piper, Mary, and Google all run locally and off of CPU only.
I think all the rest require cloud accounts or acceleration hardware to work quickly.
I’m personally fine with Mary or Piper, but I know some people like the fancier ones.
Google? Have you verified that?
Yes. Have a look at the docs: https://www.home-assistant.io/integrations/google_translate/
The docs don’t say it’s completely offline. Can you turn off your LAN connection and it still works? Have you tried this? Or just firewall off out bound access to Google services?
This comment:
Contrary to what the name suggests, the integration only does text-to-speech and does not translate messages sent to it.
doesn’t say it doesn’t call out to Google services; it says only that it doesn’t use translation services. I didn’t see anything else that implies it doesn’t send data to Google.
If it doesn’t require an API key in the config, it’s offline. My HA works totally offline unless I need to do updates, and it’s always worked for me.
You can also view the code, cuz open source.
👍 Thanks. I’m surprised, and still skeptical, but thanks.
Yeah, so I dug into it, and it’s definitely not offline. It uses gtts, which ultimately makes calls to google.com for the tts. You can track it down yourself, but you’ll eventually end up here, which talks about how to change the google host name in case it’s blocked.
I’m not sure why you believe not needing an API key means it isn’t calling a Google API, especially in this case where it clearly states it’s using an unofficial channel - which is the same trick third party YouTube clients use to access YouTube videos without using API keys.
You might consider asking this in the !homeassistant@lemmy.world community, too.
