grave cedar
So, I'm currnetly watching the livestream (although obviously not live) and have to say it's awesome. I'm curious if it's possible to have multiple models trained for it? So, during a "normal" voice control, obviously the default is used. However, i could see it might be useful to have a model that was just trained with "yes" and "no". That way, automations could be a little more conversational. If I leave the house, the VA could say "Do you want me to vacuum", then start listening for just YES or NO.
Side note, don't think there is a tag for speech to phrase yet