TGraphX: Tensor-Aware Graph Neural Network for Multi-Dimensional Feature Learning

Sajjadi, Arash; Eramian, Mark

TGraphX — Tensor-Aware Graph Neural Networks for PyTorch

TGraphX is an open-source PyTorch research framework where node and edge features can be multi-dimensional tensors — image patches, volumetric blocks, sequences — rather than flat vectors. It ships shape-aware validation, scalable mini-batch samplers, knowledge graph embedding, graph reinforcement learning, graph generation, and reproducibility tooling. MIT-licensed. arXiv:2504.03953.

Install: pip install tgraphx — Python 3.10+, PyTorch required.

Learn TGraphX Articles & Tutorials Compare Frameworks VisionServeX Package