I am trying to run the small tiny-stable-diffusion-pipe model from HuggingFace but it looks like it’s broken and won’t sample properly.

import os import torch from diffusers import StableDiffusionPipeline from diffusers import DiffusionPipeline

pipe = StableDiffusionPipeline.from_pretrained("hf-internal-testing/tiny-stable-diffusion-pipe", torch_dtype=torch.float16) pipe = pipe.to("cuda") prompt = "a picture of a rabbit" images = pipe(prompt, num_images_per_prompt=1).images

gives

RuntimeError: The size of tensor a (12545) must match the size of tensor b (226) at non-singleton dimension 1

Try it yourself in this google colab:

https://colab.research.google.com/drive/1AE5uA68myGvyzEi-vm4w806EN9rdjzq4?usp=sharing

I was expecting the model to sample when using

images = pipe(prompt, num_images_per_prompt=1).images