TF-GNN: Graph Neural Networks in TensorFlow
This provides a practical tool for machine learning practitioners working with graph data, but it is incremental as it builds on existing GNN frameworks.
The paper introduces TF-GNN, a scalable library for Graph Neural Networks in TensorFlow designed to handle heterogeneous graph data, enabling both researchers and developers with low-code solutions and used in Google production models.
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.