When using DiffusionPipeline to generate images, I run into the error ValueError: You have to specify pixel_values in the transfomers/models/clip/modeling_clip.py file, does anyone have a fix for this?
I'm currently using diffusers 0.31.0 and transformers 4.46.2
