This is a library for training Graph Neural Networks (GNNs). It supports more than 30 datasets, 5 models, and various ways to augment node features based on graph topology.

Example usage:

python train.py --dataset ogbn-arxiv --model GT