论文标题
TF-GNN：TensorFlow中的图形神经网络
TF-GNN: Graph Neural Networks in TensorFlow
论文作者
论文摘要
TensorFlow-GNN（TF-GNN）是用于张量的图形神经网络的可扩展库。它是从自下而上设计的，以支持当今信息生态系统中发生的丰富的异质图数据。除了使机器学习研究人员和高级开发人员提供能力之外，TF-GNN还提供低代码解决方案，以增强图形学习中更广泛的开发人员社区的能力。 Google的许多生产模型都使用TF-GNN，并且最近已作为开源项目发布。在本文中，我们描述了TF-GNN数据模型，其KERAS消息传递API以及相关功能，例如图形采样和分布式培训。
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.
