Cuda toolkit version 11.0 and 11.1 have compatibility issues,
it could make error when compiling cuda codes with RTX30 series.
So my solution was installing CUDA version 11.2 with Pytorch stable version 1.8.0 for CUDA 11.1.
pip install mmcv-full -f \
https://download.openmmlab.com/mmcv/dist/{cu_version}/{torch_version}/index.html
I have tested CUDA-toolkit 11.2 + Pytorch 1.7.1 (cuda 11.0 version)
by using command and it works fine.
pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/11.2/1.7.1/index.html
To check installation python -c 'import mmcv; import mmcv.ops'
Referenced to https://mmcv.readthedocs.io/en/latest/#installation
git clone https://github.com/open-mmlab/mmdetection.git
cd mmdetection
pip install -r requirements/build.txt
pip install -v -e .
cd ..
git clone https://github.com/open-mmlab/mmtracking.git
cd mmtracking
pip install -r requirements/build.txt
pip install -v -e .
cd ..
git clone https://github.com/open-mmlab/mmpose.git
cd mmpose
pip install -r requirements.txt
pip install -v -e .
cd ..