It is yet another and a standalone implementation of HiTIN: Hierarchy-aware Tree Isomorphism Network for Hierarchical Text Classification. The code here should be simple and easy to follow, and can be adapted as a standalone module with ease.
Disclaimer: I haven't run any dataset experiment of the original paper.