Text-to-Image
Diffusers
Safetensors

Unaable to run decoder, expects vqgan

#4
by obviyus - opened

Hi, upon using the code from the readme, I get a:

ValueError: Pipeline <class 'diffusers.pipelines.stable_cascade.pipeline_stable_cascade.StableCascadeDecoderPipeline'> expected {'tokenizer', 'scheduler', 'decoder', 'text_encoder', 'vqgan'}, but only {'tokenizer', 'scheduler', 'text_encoder'} were passed.

I can confirm I'm using the latest commit of the wuerstchen-v3 branch. Is there a parameter I'm missing somewhere?

My apologies, was loading the prior into the decoder.

obviyus changed discussion status to closed

Sign up or log in to comment