OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Developer Tools & SDKs

Pytorch Geometric

PyTorch Geometric is an open-source Python library built on PyTorch for writing and training graph neural networks and other geometric deep-learning models on structured and irregular data.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Pytorch Geometric — OpenViST project profile in AI Developer Tools & SDKs.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use PyG when a machine-learning project needs graph or other irregular-structure data to be modeled in a PyTorch workflow, including experiments with graph convolution or custom GNN layers.

Who it is for

  • Machine-learning researchers developing or adapting graph neural network architectures.
  • Python developers building models for graph-structured data.
  • Practitioners working with graphs, 3D meshes or point clouds.

Documented uses

  • Build and train graph neural network models using PyTorch.
  • Experiment with graph convolutional networks and other published GNN methods.
  • Create custom graph-learning layers with a message-passing API.
  • Load and batch many small graphs or work with a single large graph.
  • Apply data transformations to graphs, 3D meshes and point clouds.
Before you choose

Requirements & limits

The project repository is public and lists an MIT license. Its README documents Python 3.10–3.14 and a basic pip installation requiring PyTorch; optional extensions provide additional features.

  • PyTorch is required for basic use; the library follows a tensor-centric workflow close to standard PyTorch.
  • Basic installation does not require the optional extension libraries.
  • Optional packages enable specific capabilities, including heterogeneous GNN operators, sampling, clustering, sparse-tensor support and accelerated sparse reductions.
  • Documented capabilities include multi-GPU support and support for torch.compile and DataPipe.
  • The README describes use with graphs ranging from many small graphs to large graphs, and examples involving dynamic and heterogeneous graphs.

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

Install PyTorch Geometric for basic use

Add PyG to a Python environment for graph-learning development.

Before you start

  • Python 3.10–3.14
  • PyTorch
  • pip

Steps

  1. Install PyTorch in the Python environment.

  2. Install PyG with the command below.

    Documented command
    pip install torch_geometric

  3. Consult the installation documentation if a project needs optional extension features.

Expected outcome

PyG is installed for basic use with PyTorch; optional packages can be added for specific capabilities.

Guide limitations

  • The installation instructions specify PyTorch as a prerequisite but do not provide its installation steps here.
  • Additional libraries are optional for basic use; extension and wheel choices depend on the selected features and PyTorch/accelerator configuration.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
pyg-team/pytorch_geometric · Official homepage title · retained in its source language
How we review sourcesSuggest a correction