TF-GNN: Graph Neural Networks in TensorFlow

2022/07/07 by Oleksandr Ferludin, Arno Eigenwillig, Ferludin, Oleksandr +48
Computer Science · #Advanced Graph Neural Networks #Big Data and Digital Economy #FOS: Computer and information sciences #FOS: Physical sciences #Graph Theory and Algorithms #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Neural and Evolutionary Computing (cs.NE) #Physics and Society (physics.soc-ph) #Social and Information Networks (cs.SI)

paper · pdf · doi:10.48550/arxiv.2207.03522

openalex publication_date 2022/07/07 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

TensorFlow-GNN (TF-GNN) is a scalable library for Graph Neural Networks in TensorFlow. It is designed from the bottom up to support the kinds of rich heterogeneous graph data that occurs in today's information ecosystems. In addition to enabling machine learning researchers and advanced developers, TF-GNN offers low-code solutions to empower the broader developer community in graph learning. Many production models at Google use TF-GNN, and it has been recently released as an open source project. In this paper we describe the TF-GNN data model, its Keras message passing API, and relevant capabilities such as graph sampling and distributed training.

Related

• Efficient Graph Deep Learning in TensorFlow with tf_geometric
• Graph Tensor Networks: An Intuitive Framework for Designing Large-Scale Neural Learning Systems on Multiple Domains
• TC-GNN: Bridging Sparse GNN Computation and Dense Tensor Cores on GPUs
• TensorFlow: A system for large-scale machine learning
• Deep Graph Library: A Graph-Centric, Highly-Performant Package for Graph Neural Networks