> If you’re sceptical, go look at one of the forums where people are building derivatives of Stable Diffusion (possibly NSFW, I’m not providing any links). Doe

> If you’re sceptical, go look at one of the forums where people are building derivatives of Stable Diffusion (possibly NSFW, I’m not providing any links). Does anybody have a link to a relevant discussion here? I would like to read about the creative process that goes into defining model weights, and how it differs from the mechanical output of running the training algorithm.