TGM Documentation

TGM is a research library for temporal graph learning, designed to accelerate training on dynamic graphs while enabling rapid prototyping of new methods. It provides a unified abstraction for both discrete and continuous-time graphs, supporting diverse tasks across link, node, and graph-level prediction.

Installation

From Source (recommended)

pip install git+https://github.com/tgm-team/tgm.git@main

From PyPi

pip install tgm-lib