# microsoft/Graphormer

Graphormer is a general-purpose deep learning backbone for molecular modeling.

Repository: https://github.com/microsoft/Graphormer
Canonical: https://ross.abutalabs.com/products/graphormer
Language: Python
License: MIT
License Family: permissive
Topics: graph, transformer, deep-learning, ai4science, molecule-simulation
Last push: 2026-06-12T20:05:34+00:00

## Health v2 (maintenance only)
Score: 62/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 87, release rhythm 8, longevity 100
- inputs: {"age_days": 1924, "days_push": 82, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2474, forks 376 (observed 2026-08-28T04:06:54.943827+00:00)

## What it is
Graphormer is a deep learning library from Microsoft providing a graph transformer backbone for molecular modeling tasks. It ships pre-trained models and supports PyG, DGL, OGB, and OCP dataset interfaces on a fairseq backbone.

## Use cases
- train custom models for molecular property prediction
- run drug discovery ML experiments
- predict quantum properties of molecules like PCQM4M
- model catalysts with the Open Catalyst dataset
- fine-tune pretrained graph transformer models on molecule data

## When to choose
- you need a proven graph transformer for molecule-level ML tasks
- you want pretrained molecular models and standard dataset integrations (PyG, DGL, OGB, OCP)
- you are doing AI-for-science research in chemistry or materials

## When to avoid
- you need general-purpose graph learning beyond molecules
- you want a lightweight production inference library rather than a research training framework
- you need the most advanced pretrained versions, which are exclusive to Azure Quantum Elements

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, machine-learning, llm-training
- domain: machine-learning, deep-learning, chemistry, artificial-intelligence
- platform: python
- tags: graph-transformer, molecular-modeling, ai4science, drug-discovery, materials-science, graph-neural-networks, pretrained-models, linux, gpu

## Member repositories
- microsoft/Graphormer (main) score 62

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:54.943827+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 2026-08-30T02:28:25.886155+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/Graphormer (fetched 2026-08-28T04:06:54.943827+00:00, sha 4fd922ba9c68)
- Data as of 2026-08-30T08:39:29.467469+00:00.