multimodalart/diffusion

Update app.py

Attempt to fix the clip guidance

Update README.md

Update app.py

Remove model small

swap gallery for carousel for gr.interface.load compatibility

Update app.py

Adjust parameters to work with CLIP Guidance

Load the small model correctly

Update app.py

Add spherical dist loss

Update app.py

CLIP on GPU

devices fix

initialize target embeds and weights

Add more logic to clip embeds

Update app.py

Add nn

Cutouts function

Add normalisation

Update app.py

Update app.py

Update README.md

Update README.md

Update app.py

Update app.py

Update README.md

Update app.py

Update app.py

Update README.md

Set steps

Add references

Remove dumb requirements

Initial diffusion

try vqgan on this space

Initial

initial commit