#Does Voice PE need HAOS?
1 messages · Page 1 of 1 (latest)
It works with any installation type
Amazing, thank you
Do you also happen to know if an i7-8700 would be enough for local voice recognition?
See this if you want to use containers for whisper, piper, and/or openwakeword: #1319677976392110143 message
I use NC cloud. I'll leave that question to others
No problem, thanks for the help!
I use Whisper/Piper in docker on i5-9500U. It's ok with base models, but lags on anything better...
Thanks! Guess I'll do some trial and error
Dive into that rabbit hole, it's funny here 🙂
I have whisper running with cuda acceleration in docker and piper just using cpu in docker 🙂
Running large-v3-turbo model with around 250ms response times
Wait, how do you use CUDA acceleration on pure CPU?
No my whisper container is cuda, piper is just cpu