The accompanying code for Data-Driven Self-Supervised Graph Representation Learning

Requirements!

• Python 3.6+
• PyTorch 1.10+
• PyG 2.0+
• Numpy 1.17.2+
• Networkx 2.3+
• SciPy 1.5.4+
• (OPTINAL) OPTUNA 2.8.0+ If you wish to tune the hyper-parameters of DSGRL

Example usage

For node classification experiment

$ cd node_classification
$ python main.py

For graph property prediction experiment

$ cd graph_property_pred
$ python main.py

More instructions for both experiments are available in the readme file contained in the respective directory.