# google-deepmind/graph_nets

Build Graph Nets in Tensorflow

Repository: https://github.com/google-deepmind/graph_nets
Canonical: https://ross.abutalabs.com/products/graph_nets
Homepage: https://arxiv.org/abs/1806.01261
Language: Python
License: Apache-2.0
License Family: permissive
Topics: graph-networks, graphs, deep-learning, neural-networks, tensorflow, sonnet, artificial-intelligence
Last push: 2022-12-12T11:28:07+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2924, "days_push": 1360, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5406, forks 778 (observed 2026-08-28T04:09:17.667691+00:00)

## What it is
DeepMind's library for building graph networks (graph neural networks) in TensorFlow and Sonnet, based on the 'Relational inductive biases, deep learning, and graph networks' paper. It takes graphs as input and returns graphs with updated edge, node, and global attributes.

## Use cases
- build graph neural networks in tensorflow
- train a model to find shortest paths on graphs
- predict physical dynamics of particle systems
- learn to sort graph-structured data
- implement message-passing neural networks
- experiment with relational inductive biases in deep learning

## When to choose
- you want DeepMind's reference implementation of graph networks
- your project already uses TensorFlow 1/2 and Sonnet
- you need graph-to-graph models with edge, node, and global attributes
- you want demo notebooks for shortest path, sorting, and physics tasks

## When to avoid
- you prefer PyTorch or JAX ecosystems (consider jraph instead)
- you need actively maintained software with recent releases
- you need TF 1.x compatibility beyond v1.0.4 of the library

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, machine-learning
- domain: deep-learning, machine-learning, artificial-intelligence
- platform: python
- tags: graph-neural-networks, tensorflow, sonnet, graph-networks, relational-reasoning, linux, macos, gpu

## Member repositories
- google-deepmind/graph_nets (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:17.667691+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:57:47.574696+00:00, confidence not recorded.
  - readme: https://github.com/google-deepmind/graph_nets (fetched 2026-08-28T04:09:17.667691+00:00, sha 42991775aeab)
  - homepage: https://arxiv.org/abs/1806.01261 (fetched 2026-08-29T08:52:28.975253+00:00, sha 81e01fdd67dd)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T08:52:28.977889+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T08:52:28.983438+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T08:52:28.985910+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T08:52:28.980844+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.