There are 2 repositories under fused-gromov-wasserstein topic.
Improving word mover’s distance by leveraging self-attention matrix
A Graph Optimal Transport Python Package