- Ready-to-go GUI via colab
- Stable diffusion via webui
- Huggingface's Diffusers library
- Disable from the original CompVis's repository
- Disable from inside a Jupyter notebook code cell, paste this below the pipe generation command (source):
from diffusers.pipelines.stable_diffusion import safety_checker
def sc(self, clip_input, images) :
return images, [False for i in images]
# edit StableDiffusionSafetyChecker class so that, when called, it just returns the images and an array of True values
safety_checker.StableDiffusionSafetyChecker.forward = sc
The diffusers-interpret library can generate token attribution and gif of the denoising process.