# google-deepmind/materials_discovery

Repository: https://github.com/google-deepmind/materials_discovery
Canonical: https://ross.abutalabs.com/products/materials_discovery
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Last push: 2026-06-23T21:05:43+00:00

## Health v2 (maintenance only)
Score: 67/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 89, release rhythm 35, longevity 72
- inputs: {"age_days": 1009, "days_push": 71, "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 1230, forks 195 (observed 2026-08-28T04:04:03.827031+00:00)

## What it is
Google DeepMind's GNoME repository sharing a dataset of over 520,000 predicted stable inorganic crystal structures plus model definitions (GNoME and Nequip) and Colab notebooks. It supports machine learning-driven discovery of novel materials in materials science research.

## Use cases
- find novel stable inorganic crystal structures for battery research
- train graph neural networks on materials project data
- explore convex hull data for a chemical family
- reproduce GNoME materials discovery paper results
- load DFT formation energy data as CSV
- visualize crystal structures in a colab notebook

## When to choose
- you need a large dataset of predicted stable inorganic crystals
- you want to reproduce or extend GNoME machine learning models for materials discovery
- you research materials science with graph networks and DFT data

## When to avoid
- you need production-supported software rather than a research codebase
- you need general-purpose chemistry simulation tools rather than crystal discovery datasets
- you expect polished APIs and official product support

## Facets
- artifact type: dataset
- maturity: active
- function: machine-learning, data-science
- domain: chemistry, machine-learning
- platform: python
- tags: gnome, materials-discovery, crystal-structures, dft, convex-hull, graph-networks, research-dataset, materials-science

## Member repositories
- google-deepmind/materials_discovery (main) score 67

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:03.827031+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 2026-08-30T06:15:01.629473+00:00, confidence not recorded.
  - readme: https://github.com/google-deepmind/materials_discovery (fetched 2026-08-28T04:04:03.827031+00:00, sha fcec4611274c)
- Data as of 2026-08-30T08:39:29.467469+00:00.