>>29095684
She's an LLM trained on Twitch data. She hears with speech to text and speaks with text to speech. Her model is rigged to respond to her outputs. Also she has integration with Twitch to read chat and some games as well, and a vision model so she can see. Vedal987 did pretty much all of the programming and training, Annytf did the model, QueenPB helps with the songs.