Source Code

We release the source code of the entire deep learning infrastructure, comprising:

  • AutoenCODE - Learning Infrastructure

  • Graph Embedding - Hope

AutoenCODE - Learning Infrastructure

We publicly release our learning infrastructure called AutoenCODE. The GitHub repository contains source code, clear instructions and steps, as well as example data.

Configuration Settings

  • Embedding size: 300

  • Max training iteration: 30

  • Max length size: 3000 tokens

  • Training time: 12-70 hours (depending on the project)

Graph Embedding - HOPE

We rely on GEM implementation for the graph embedding technique HOPE. Below is the link to the GEM GitHub code, as well as the code and scripts we modified in our work.