Skip to content

Instantly share code, notes, and snippets.

@SalahAdDin
Last active May 16, 2019 14:32
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save SalahAdDin/de8ab87803b0714a128154f043c64d43 to your computer and use it in GitHub Desktop.
Save SalahAdDin/de8ab87803b0714a128154f043c64d43 to your computer and use it in GitHub Desktop.
Setup to Darknet and Deeplearning

First, we need to install allr required libraries:
sudo pacman -S python-numpy vtk hdf5 openmp opencv cuda cudnn
It installs all requirements to work with the GPU and the OpenCv's support to Darknet and other libraries; hdf5 is required to work with opencv in python; vtk is required to compile Darknet.
By default Darknet handles libraries' references on Debians based linux systems, then, it requires opencv's as well as cuda references to work; in ArchLinux, opencv is in its 4.1.0 and it used opencv4 as environment variable, we need change the reference then:
sudo cp /usr/lib/pkgconfig/opencv4.pc /usr/lib/pkgconfig/opencv.pc.
While the cuda library is installed in a different folder than Debian's based systems, we need to created a symbolic link to solve this problem:
sudo ln -s /opt/cuda/ /usr/local/cuda.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment