ntrouve-onera

I have a lot of memory issue as of late and I can't explain them either. This is using img2img on large scale image (2560x1440) using 24gb VRAM (TITAN RTX)....

When I reduce drastically the image size (to 1024 / 576) I still get out of memory issues. Tried to delete venv/reinstall fully. No success again.

False, 'Just Resize', False, True, 64, 64, 64, 1, 1, False, 0.9, 5, '0.0001', False, 'None', '', 0.1, False, '\nCFG Scale should be 2 or lower.\n\n', True, True, '', '',...

Pressing generate again after a out of memory error and eventually it works without changing input...puzzling. It seems to always error after a webui restart or a change of image...

I did test with a premade controlnet image (generated outside the webui with blender) and it works just fine. As a workaround until it gets fixed you could find a...

Bump, no update on this issue? From my understanding Inpainting models use a mask as an extra input, the one fed to the inpainting model is wrong for some reason...

@JoeNavark and with 50% blend you get roughly what you could have expected from 0.5 denoise? (more or less?).

 This does somewhat work, you need to add an extra latent mask or it just inpaint everything. Still some feathering/edge issues to solve but it appears to be doing...

I may have switched samples1 and 2 before doing the screenshot to clean up so it might be because of that.

Thanks a lot, supporting the idea of the PR. Indeed the workaround is still not as good as a true partial denoise because the latent blend appears to mess up...