I tested it but the jaggies are noticeable, it's best to update the code in the python package just to be on the safe side
(although the code doesn't seem to have changed much when I checked it)
@huchenlei

OK，they change dino model,too

OK，they change dino model,too

I don't think that dino model change is the cause. The commit happens on Jan 15th before I fork the repo on Jan 22nd.

The output from this is really pixilated when used as a drop in replacement

maybe this is what changed LiheYoung/Depth-Anything#81

I have update the sd-webui-controlnet repo to support depth_anything_v2 in this forked branch if you approve the pull request I'm about to submit. Forked Repo - https://github.com/MackinationsAi/sd-webui-controlnet Pull Request - #2969

I have update the sd-webui-controlnet repo to support depth_anything_v2 in this forked branch if you approve the pull request I'm about to submit. Forked Repo - https://github.com/MackinationsAi/sd-webui-controlnet

Thanks for your contribution, I will review as well as test it as soon as possible.

I have update the sd-webui-controlnet repo to support depth_anything_v2 in this forked branch if you approve the pull request I'm about to submit. Forked Repo - https://github.com/MackinationsAi/sd-webui-controlnet

Thanks for your contribution, I will review as well as test it as soon as possible.

I still need to upload the new hf depth_anything_v2 transformers (which I will be doing next). I have tested it locally on my machine & it works really well!