Acly

What original repo are you referring to? The reason it's blury is that it runs at low resolution (256 for LaMA, 512 for MAT). You can technically run it at...

I don't use VAEEncodeForInpaint and I downscale/crop images which makes the low resolution less of an issue. Lama still helps at 1.0 denoise as a base for conservative inpainting (remove...

It doesn't, unless you change the inputs like image/mask.

Feeding the image/mask to the new model also changes the model. Attaching the image/mask to the model is the source of the problem, it would make more sense if they...

For diffusion fp16 is normal. RTX 30xx and 40xx use bf16. Nobody uses fp32 if they can help it. I don't think there is a rule for step count, it...

I don't see anything suspicious. It's a 2.4 GB download, if your connection is slow it would take very long. Usually not hours, but you can try the link directly...

It might work to download https://download.pytorch.org/whl/cu121/torch-2.3.1%2Bcu121-cp310-cp310-win_amd64.whl manually and install via ``` python\python.exe -m pip install torch-2.3.1+cu121-cp310-cp310-win_amd64.whl ``` - assuming you're in the server install directory, eg C:\Users\YOURNAME\AppData\Roaming\krita\ai_diffusion

It's checkpoints only

See https://github.com/Acly/krita-ai-diffusion/discussions/1176 for wider support of various Flux models.

Internet connection is required to retrieve available languages and download language packs. Please check [log files](https://github.com/Acly/krita-ai-diffusion/wiki/Common-Issues#log-files) for errors. Try http://127.0.0.1:8188/api/etn/languages (while Krita/ComfyUI is running).