CS1o

dont use these options: --opt-sub-quad-attention --lowvram --disable-nan-check --use-zluda --disable-safe-unpickle --update-check --listen --precision full --no-half --device-id 1 A lot of them slow you down massivly. Only use the ones provided by...

> without all then my vram fills up instantly even with lowvram idk why id doesnt get freed up Can you show a cmd log when using only `--use-zluda --update-check...

Why are you now using Linux and the webui.sh ?

Yea you didnt mentioned it. ZLUDA is not that practical for Linux as on Linux you have ROCm that supports AMD GPUs nativly. With a RX570 you would need to...

To setup the Directml webui properly (and without onnx) do the following steps: Open up a cmd and type `pip cache purge` then hit enter and close the cmd. That...

Hey, 12-24GB Vram is enough for Hires Fix, The important parts are the Settings you should use and knowing the Limit. This only works when using my Commandline_args settings mentioned...

I did some tests on my old Nvidia PC (GTX1080) and there Segmentation models work with: :white_check_mark:Auto1111 Ultralytics: 8.3.203 rich: 14.1.0 mediapipe: 0.10.14 :white_check_mark: Forge Ultralytics: 8.3.49 rich: 13.7.1 mediapipe:...

Thanks for the reply. I upgraded my test webui to Zluda 3.9.1 Nightly. reset the venv and .zluda folder. Replaced the .zluda files and torch/lib with the renamed zluda files...

> Fooocus Here you find a install Guide for Fooocus with ZLUDA. I tested it a few minutes ago. https://github.com/CS1o/Stable-Diffusion-Info/wiki/Installation-Guides

> Error loading caffe2_nvrtc.dll or its dependencied after replace the three dll That error is caused by any Python Version that got installed through the Microsoft Store. To fix it,...