LLaMA 3v

Software / App

Meta's multimodal model, for which training data is implied to be Instagram Reels, and which is being worked on for speech capabilities.

Mentioned in 1 video