libtorch1.9_opencv_win10_cuda11.1_vs2019配置

libtorch1.9_opencv_win10_cuda11.1_vs2019配置

Libtorch是c++接口的Pytorch，提供了与python版本基本一致的接口，使用很方便，而且更适合实际部署。这里介绍Libtorch1.9的在VS2019下的部署和简单测试使用。

从官网下载

https://pytorch.org/get-started/locally/

目前可以下载到Libtorch1.9版本，如果电脑支持Gpu，且安装了Cuda，可以下载Gpu版本，否则下载cpu版本。

vs2019工程配置(release)

项目->属性页面

调试->环境

PATH=D:\libtorch\lib;%CUDA_PATH%;%PATH%

vc++ ->包含目录

D:\opencv\build\include;D:\libtorch\include;
D:\libtorch\include\torch\csrc\api\include;
C:\Program Files\NVIDIA Corporation\NvToolsExt\include;
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.1\include;$(IncludePath)

vc++ ->库目录

D:\opencv\build\x64\vc15\lib;
D:\libtorch\lib;
C:\Program Files\NVIDIA Corporation\NvToolsExt\lib\x64;
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.1\lib\x64;$(LibraryPath)

C/C++ ->附加包含目录

D:\opencv\build\include;
D:\libtorch\include;
D:\libtorch\include\torch\csrc\api\include;$(IncludePath)

链接器->常规->附加库目录

D:\opencv\build\x64\vc15\lib;D:\libtorch\lib;$(LibraryPath)

链接器->输入->附加依赖项

opencv_world342.lib;c10.lib;nvToolsExt64_1.lib;cudart_static.lib;caffe2_nvrtc.lib;c10_cuda.lib;torch.lib;torch_cuda.lib;torch_cuda_cu.lib;-INCLUDE:?searchsorted_cuda@native@at@@YA?AVTensor@2@AEBV32@0_N1@Z;torch_cuda_cpp.lib;torch_cpu.lib;cufft.lib;curand.lib;cublas.lib;cudnn.lib;-INCLUDE:?warp_size@cuda@at@@YAHXZ;%(AdditionalDependencies)