SpConv: PyTorch Spatially Sparse Convolution Library
直接下载官方镜像(没试过)
docker pull scrin/dev-spconv
自个儿折腾
git clone git://github.com/traveller59/spconv.git
cd spconv/third_party
删除其中的 cutlass, mp11, pybind11
git clone git://github.com/NVIDIA/cutlass.git
git clone git://github.com/boostorg/mp11.git
git clone git://github.com/pybind/pybind11.git
1.命令行下载安装
apt-get install libboost-all-dev
2.软件包安装
官网(https://www.boost.org/)下载压缩包
wget https://boostorg.jfrog.io/artifactory/main/release/1.76.0/source/boost_1_76_0.tar.gz
tar -zxvf boost_1_76_0.tar.gz
cd boost_1_76_0
sh ./bootstrap.sh --prefix=/usr
./b2 install
cmake >= 3.13.2,若已满足则跳过。
官网(https://cmake.org/download/)下载压缩包
wget https://github.com/Kitware/CMake/releases/download/v3.19.8/cmake-3.19.8-Linux-x86_64.tar.gz
tar -zxvf cmake-3.19.8-Linux-x86_64.tar.gz
vi /etc/profile
添加export PATH="cmake的bin文件夹路径:$PATH"
source /etc/profile
返回spconv文件夹
python setup.py bdist_wheel
cd ./dist
安装生成的whl文件
pip install ×.whl