search

brian6091 / Dreambooth

Fine-tuning of diffusion models
Apache License 2.0
97 stars 20 forks source link

Workaround for fact SD 2 models don't have feature_extractor #3

Open brian6091 opened 1 year ago

brian6091 commented 1 year ago

https://github.com/huggingface/diffusers/issues/1538

Passing None works for SD 2 models, but I think it wrecks previous models, since they come packaged with a feature_extractor

Related, https://github.com/huggingface/diffusers/issues/1500

brian6091 commented 1 year ago

Setting feature_extractor=None does not seem to affect inference for SD v1-5