[BUG] ELLA has some bug and quality issues

ModelsLab / diffusers_plus_plus

Diffusers++: State-of-the-art diffusion models for image and audio generation in PyTorch

https://huggingface.co/docs/diffusers

Apache License 2.0

12 stars 2 forks source link

Open shauray8 opened 4 months ago

shauray8 commented 4 months ago

[X] pipeline loading utils needs to be fixed
[X] gives off No module named 'diffusers.plus_pipelines.stable_diffusion' on some models and not others
[ ] ~~CLIP adds to HBM on every shape change~~
[ ] quality has to be checked (seems a little off)
[ ] sometimes the device does not match - UNET does not match
[ ] needs flex token support added
[ ] move CLIP, look at SD3 implimentation