A baseline implementation of Graph Conjoint Attention Networks (CATs) for semi-supervised node classification, that has been proposed in our paper:
Tiantian He, Yew-Soon Ong, and Lu Bai, "Learning Conjoint Attentions for Graph Neural Nets," NeurIPS 2021.
Requirements: Python (>=3.8) PyTorch (>=1.8.1) DGL (0.6.1)
As there are no pre/post process or early stopping control and different GPU/CUDA platforms might be used, the performances might be slightly different from those reported in the paper.