Skip to content

Latest commit

 

History

History
42 lines (34 loc) · 1.57 KB

BUILD.md

File metadata and controls

42 lines (34 loc) · 1.57 KB

Building Homing-Pigeon

Homing-Pigeon is using cmake for the build. First you need to install Pytorch C++ API in the directory libs/torch:

  • cd libs
  • wget https://download.pytorch.org/libtorch/cpu/libtorch-cxx11-abi-shared-with-deps-1.10.2%2Bcpu.zip
  • unzip libtorch-cxx11-abi-shared-with-deps-1.10.2%2Bcpu.zip
  • mv libtorch torch
  • rm libtorch-cxx11-abi-shared-with-deps-1.10.2%2Bcpu.zip
  • cd ..

Second, install OpenCV:

  • sudo apt-get install libopencv-dev

Next, if boost is not installed on your system, run the following command:

  • sudo apt-get install libboost-all-dev

Similarly, if gnuplot is not installed on your system, run the following command:

  • sudo apt-get install -y gnuplot libgnuplot-iostream-dev

Then, use cmake to build the project as follows:

  • mkdir build
  • cd build
  • cmake ..
  • make
  • cd ..

Next, if you want to run the simulation involving the dSprites dataset run the following commands:

  • cd examples/d_sprites
  • git clone https://github.com/deepmind/dsprites-dataset.git
  • python3 ./d_sprites_from_npz_to_pickle.py
  • cd ../..

The above set of commands will build the Homing-Pigeon library, the unit tests as well as the examples of the project. Note that the example named deep_learning_mnist requires the mnist dataset to be present in the build directory. You can download the dataset using the following command:

  • mkdir build/mnist
  • cd build/mnist
  • git clone https://github.com/HIPS/hypergrad.git
  • mv hypergrad/data/mnist/* .
  • rm -r hypergrad mnist_data.pkl
  • gunzip t*-ubyte.gz
  • cd ../..