Loading pipeline components problem

[osamaismail]

I try to use it locally on ubuntu OS, I got this error:

OSError: Error no file named diffusion_pytorch_model.bin found in directory /root/.cache/huggingface/hub/models--SG161222--RealVisXL_V4.0/snapshots/49740684ab2d8f4f5dcf6c644df2b33388a8ba85/unet.

[yliyli]

Got the same issue, it seems like the corresponding repo on hugging face doesn't have all the necessary files. Try sd_model_path = models_dict["Juggernaut"] #"SG161222/RealVisXL_V4.0" in load pipeline instead of the original realvision. I just don't know if this would lead to worse results later as I'm still trying to run it. Also issue of #3 has stuff on this too.

[King-Wood-Shen]

This is because you downloaded the model file in the safetensors format instead of PyTorch's bin format. We need to modify the code to use the safetensors format model file. pipe = StableDiffusionXLPipeline.from_pretrained(sd_model_path, torch_dtype=torch.float16, use_safetensors=True)