Conan package recipe for darknet
Darknet is an open source neural network framework written in C and CUDA. It is fast, easy to install, and supports CPU and GPU computation.
The packages generated with this conanfile can be found on Bintray
If you wish to report an issue or make a request for a package, please do so here:
Issues Tracker → aks for support and access
$ conan install darknet/xxx@conan/stable
If you handle multiple dependencies in your project is better to add a conanfile.txt
[requires]
darknet/xxx@conan/stable
[generators]
cmake
Complete the installation of requirements for your project running:
$ mkdir build && cd build && conan install ..
Note: It is recommended that you run conan install from a build directory and not the root of the project directory. This is because conan generates conanbuildinfo files specific to a single build configuration which by default comes from an autodetected default profile located in ~/.conan/profiles/default . If you pass different build configuration options to conan install, it will generate different conanbuildinfo files. Thus, they should not be added to the root of the project, nor committed to git.
The following command both runs all the steps of the conan file, and publishes the package to the local system cache. This includes downloading dependencies from "build_requires" and "requires" , and then running the build() method.
$ conan create . conan/stable
Option | Default | Possible Values | Description |
---|---|---|---|
shared | False | [True, False] | Build shared libraries only |
gpu | False | [True, False] | Compile with NVidai Cuda GPU support (Linux only) |
cudnn | False | [True, False] | Include NVidia Cuda Runtime support |
opencv | False | [True, False] | Build with OpenCV support |
openmp | False | [True, False] | Build with OpenMP |
Conan Community has its own Bintray repository, however, we are working to distribute all package in the Conan Center:
$ conan remote add conan-center "https://conan.bintray.com"
NOTE: The conan recipe license applies only to the files of this recipe, which can be used to build and package darknet. It does not in any way apply or is related to the actual software being packaged.
- Build and package stand-alone
- Pick and place the right libraries
- Create CMakeLists.txt for darknet project
- Build with OpenCV conan package