google-deepmind/graph_nets
DeepMind's library for building graph neural networks in TensorFlow and Sonnet.
Not currently ranked — collecting fresh signals.
star history
Graph Nets is a neural network library for building graph network architectures in TensorFlow. It provides tools to create networks that operate on graph-structured data, where nodes, edges, and global attributes are processed to produce updated graph outputs. The library implements the relational inductive biases described in the DeepMind paper on graph networks and integrates with Sonnet, DeepMind’s neural network library.
Frequently asked
- What is google-deepmind/graph_nets?
- DeepMind's library for building graph neural networks in TensorFlow and Sonnet.
- Is graph_nets open source?
- Yes — google-deepmind/graph_nets is open source, released under the Apache-2.0 license.
- What language is graph_nets written in?
- google-deepmind/graph_nets is primarily written in Python.
- How popular is graph_nets?
- google-deepmind/graph_nets has 5.4k stars on GitHub.
- Where can I find graph_nets?
- google-deepmind/graph_nets is on GitHub at https://github.com/google-deepmind/graph_nets.