Graph Convolutions Enrich the Self-Attention in Transformers!

Open in new window