mmdeploy/linux-x86_64.md at master · open-mmlab/mmdeploy · GitHub
cmake和GCC在ubuntu20.04下是符合要求的
教程针对cuda113 torch1.12.0
装cmake 3.20.0
wget https://github.com/Kitware/CMake/releases/download/v3.20.0/cmake-3.20.0-linux-x86_64.tar.gz tar -xzvf cmake-3.20.0-linux-x86_64.tar.gz sudo ln -sf $(pwd)/cmake-3.20.0-linux-x86_64/bin/* /usr/bin/
装mmcv-full
pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu113/torch1.12.0/index.html
装OpenCV
sudo apt-get install libopencv-dev
装pplcv
git clone https://github.com/openppl-public/ppl.cv.git cd ppl.cv export PPLCV_DIR=$(pwd) git checkout tags/v0.7.0 -b v0.7.0 ./build.sh cuda
sudo vim ~/.bashrc
export PPLCV_DIR=$/home/sekiro/ppl.cv
source ~/.bashrc
然后就按部就班装pytorch，mmcv-full，opencv，pplcv，tensorrt（已安装不用再装了），pycuda
然后(如果tensorrt用dpkg安装的 还是默认路径的话)
export TENSORRT_DIR=$/usr/src/tensorrt
export LD_LIBRARY_PATH=$TENSORRT_DIR/lib:$LD_LIBRARY_PATH