Comment on Larian CEO Responds to Divinity Gen AI Backlash: 'We Are Neither Releasing a Game With Any AI Components, Nor Are We Looking at Trimming Down Teams to Replace Them With AI'

<- View Parent
Hackworth@piefed.ca ⁨11⁩ ⁨hours⁩ ago

As I understand it, CLIP (and other text encoders in diffusion models) aren’t trained like LLMs, exactly. They’re trained on image/text pairing, which ya get from the metadata creators upload with their photos in Adobe Stock. That said, Adobe hasn’t published their entire architecture.

source
Sort:hotnewtop