Hey Nitro,thanks for the model.Here is how people can make it work in Automatic's Web UI #8

by Enigmator - opened
  1. https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Features#stable-diffusion-20 download the config file, from there.
  2. rename it to "future-diffusion-v1.yaml" and copy it next to your model.
  3. edit the yaml file with a text editor and under "params:" you will see ->>> "parameterization: "v" <<<- written. delete "parameterization: "v" and save the file.
  4. relauch the webui completely from beginning.

Edit : this makes the model to work in eps mode,but however the style didn't work. I need to investigate this one.

@Enigmator Thank you for your tip. We figured this out as well but the issue right now is not the way Auto loads the model but the conversion of the diffusers to weights.
The script needs to be updated by the Huggingface Team. Hopefully they can get it to work next week.

Sure I was checking the parameters 1 by 1 if those make some difference and since I wasn't sure of the parameters that was used when training, that was my only hope,but you saved me from trouble. I liked this model as your other models from the results provided.Anyway I am excited for the updates.

Sign up or log in to comment