This repository contains some tutorials and experiments to learn about Graph Neural Networks, using PyTorch Geometric as the main framework. Currently, only one learning task is covered:
- graph classification
Each folder under experiments
contains a main.py
file which runs the corresponding experiment. The
main
file imports classes from:
src/<experiment>/utils
for generic data loading & preprocessingsrc/<experiment>/datasets
for functions specific to a certain datasetsrc/<experiment>/models
for model initialization, training & inference