stabilityai

/

stable-diffusion-2-1-unclip-small

like 32

We're using unCLIP for img2img and the StableUnCLIPImg2Img2Pipeline works well enough but we think it would perform better if fine tuned on our dataset. It's not clear to me how to do this. From poking around the docs, it would seem like the necessary component models (TextEncoder, ImageEncoder, Tokenizer..) could be loaded somehow but I haven't been able to determine how.