inpainting doesn't work with CPU

[krrishdholakia]

Describe the bug

Using the inpainting example but trying to run it on the cpu instead of gpu (since i'm trying to push to hf space).

Hitting this runtime error: RuntimeError: Input type (torch.FloatTensor) and weight type (torch.HalfTensor) should be the same or input should be a MKLDNN tensor and weight is a dense tensor

Reproduction

Try and run pipeline in cpu

Logs

No response

System Info

using google colab

[krrishdholakia]

cc: @patil-suraj

[krrishdholakia]

https://colab.research.google.com/github/patil-suraj/Notebooks/blob/master/in_painting_with_stable_diffusion_using_diffusers.ipynb

[krrishdholakia]

tried using the updated pipeline - StableDiffusionInpaintPipeline

but it doesn't seem to have been uploaded yet

[patrickvonplaten]

@krrishdholakia does it still not work on CPU? cc'ing our in-painting authors here: @patil-suraj and @anton-l

[patrickvonplaten]

@patil-suraj could you take a look here?

[patil-suraj]

Hey @krrishdholakia , sorry to come to this only now. The pipeline should work on CPU, I suspect that you are setting the torch_dtype=torch.float16 , which won't on CPU as it does as fp16 is not supported on CPU, please pass torch_dtype=torch.float32 to make sure it works on CPU.

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.