OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Graphormer

Graphormer is an open-source Python deep-learning package that researchers and developers can use to train custom models for molecule-modeling tasks, including research in drug and materials discovery.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Graphormer — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Graphormer when you need a code-based toolkit for training or adapting molecular-modeling models, and can work with its documented Fairseq-based environment.

Who it is for

  • Researchers developing molecular-modeling methods or applications.
  • Developers training custom models for molecule-science tasks.

Documented uses

  • Train custom deep-learning models for molecular modeling research.
  • Use available PCQM4M or PCQM4Mv2 pre-trained models as research starting points.
  • Explore molecule-modeling workflows for drug discovery, materials discovery or Open Catalyst Project tasks.
Before you choose

Requirements & limits

The repository is public and identifies its license as MIT. Advanced pre-trained versions are offered exclusively through Azure Quantum Elements.

  • The standard repository supports custom model training; advanced pre-trained versions are offered through Azure Quantum Elements.
  • The documented container pins a legacy Python and PyTorch stack, which the project does not claim is compatible with current releases.
  • GPU execution requires the NVIDIA Container Toolkit; the documented distributed setup is single-host only.
  • The README lists support for PyG, DGL, OGB and OCP interfaces and datasets, along with a Fairseq backbone.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Build the documented legacy container

Prepare Graphormer's pinned Fairseq-based environment and verify CPU imports and graph preprocessing.

Before you start

  • Docker
  • The Graphormer repository and its Fairseq submodule

Steps

  1. Initialize the Fairseq submodule recursively with git submodule update --init --recursive.

  2. Build the image with docker build --tag graphormer:legacy .

  3. Run python docker/verify_environment.py inside the image to verify CPU imports and graph preprocessing.

Expected outcome

A Graphormer legacy container is built and the documented CPU environment verification is run.

Guide limitations

  • The image pins Python 3.9, PyTorch 1.9.1, CUDA 11.1 user-space libraries and pip 23.3.2; it does not claim compatibility with current Python or PyTorch releases.
  • GPU execution requires the NVIDIA Container Toolkit.
  • The documented distributed-training example covers one host, not multi-node training.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
14 October 2026
Original website title
microsoft/Graphormer · Official homepage title · retained in its source language
How we review sourcesSuggest a correction