PRODIGE: Probabilistic Differentiable Graph Embeddings

A supplementary code for Beyond Vector Spaces: Compact Data Representation as Differentiable Weighted Graphs. https://arxiv.org/pdf/1910.03524.pdf

What does it do?

It learns weighted graph representation for your data end-to-end by backpropagation.

What do i need to run it?

• Get as many CPUs as you can
  • We do not support GPU pathfinding (yet)
• Use any popular 64-bit Linux operating system
  • Tested on Ubuntu16.04, should work fine on most linux x64 and even MacOS;
  • On other operating systems we recommend using Docker, e.g. pytorch-docker
• Install the libraries required to compile C++ parts of PRODIGE
  • sudo apt-get install gcc g++ libstdc++6 wget curl unzip git
  • sudo apt-get install swig3.0 (or just swig)

How do I run it?

1. Clone or download this repo. cd yourself to it's root directory.
2. Grab or build a working python enviromnent. Anaconda works fine.
3. Install packages from requirements.txt

• It is critical that you use torch >= 1.1, not 1.0 or earlier
• You will also need jupyter or some other way to work with .ipynb files

4. Open jupyter notebook in ./notebooks/ and you're done!