Inference code with Torch C++. (under construction)

Installation

LibTorch

• Download LibTorch from here.

• Add path to LibTorch to environment variable Torch_DIR.

echo %Torch_DIR%
> C:\libs\libtorch-win-shared-with-deps-1.4.0\share\cmake\Torch

• Add path to *lib and *dll to environment variable Path.

echo %Path%
> foo;bar;C:\libs\libtorch-win-shared-with-deps-1.4.0\lib;

PyTorch BCNNs

• See the https://github.com/yuta-hi/pytorch_bayesian_unet

Dump model

python dump_model.py

Build

• CMake

mkdir build
cd build
cmake -G "Visual Studio 15 2017 Win64" ..

• Open ./infer-app.sln