stable-diffusion text-to-image safetensors

Inspired by unico

I didn't keep good notes on this and I never actually wrote down the mix recipe but iirc it was largely the same as the Castagno recipe but with bpmodel and evt_M mixed in.

I can't really recreate this so I'll leave it up for archival purposes. The results aren't really all that good and I don't recommend you use them.