pizzadequeso its awesome to see honestly, way to much of AI has:
No real function (ai art, most of GPT)
Resource consumption to the point that its training and deployment have destroyed other resources and serious environmental impact.
No source or transparency on training data or its weights
And that its resource requirements make it impossible to actually functionally run on any kind of personally owned hardware.
Seeing GOS make an AI that assumedly doesn't cost a lot of resources & has a very good accessibility based purpose is something I like to see.
I'm really happy to hear it will also be text to speech, vision impaired people will have a lot to get out of that. I'm assuming the TTS model can eventually be used for accessibility features?