"TypeError: OnnxStableDiffusionPipeline.__init__() missing 1 required positional argument: 'vae_encoder'"

[ChaseFlorom]

Hey, I'm getting this error:

"TypeError: OnnxStableDiffusionPipeline.init() missing 1 required positional argument: 'vae_encoder'"

Earlier in the error it also says:

"The config attributes {'vae_encoder': ['diffusers', 'OnnxRuntimeModel']} were passed to StableDiffusionOnnxPipeline, but are not expected and will be ignored. Please verify your model_index.json configuration file."

I have followed the instructions as listed, and the test script of the txt2img works just fine. Any help appreciated, thanks!

[azuritecoin]

it looks like they updated the conversion script in between me writing the guide and you trying to setup your system. -__-

Try this version of the script instead: https://raw.githubusercontent.com/huggingface/diffusers/75bb6d2d466d742feb1e2be15f74d2605d11f0e9/scripts/convert_stable_diffusion_checkpoint_to_onnx.py

[ChaseFlorom]

Hey thanks for responding! So i deleted my folder that it created and tried the conversion again with the new script you provided...and now I'm getting that same error on the conversion instead of on opening the GUI.

[azuritecoin]

You might have a too new version of diffusers. try installing 0.5.1:
pip uninstall diffusers
pip install diffusers==0.5.1

[ChaseFlorom]

That did it! Thanks for your help.

[azuritecoin]

Yay! Glad to help