keras-nlp icon indicating copy to clipboard operation
keras-nlp copied to clipboard

Polish StableDiffusion3

Open james77777778 opened this issue 1 year ago • 0 comments

I'm aware that the original private implementation is missing the encoder in VAE model. However, the encoder is necessary for image2image, inpainting and finetuning.

I'm working on bringing it back and we can expect the following features for SD3 in the future:

  • image2image
  • inpainting
  • finetuning

Are these features planned for the KerasHub release?

TODO:

  • Bring back VAE: #1892
  • Add Image2Image: #1909
  • Add Inpaint: #1909
  • Add T5 text encoder preset
  • Add finetuning recipe for SD3

james77777778 avatar Sep 27 '24 09:09 james77777778