https://github.com/NVIDIA/cuEquivariance
cuEquivariance is a math library that is a collective of low-level primitives and tensor ops to accelerate widely-used models, like DiffDock, MACE, Allegro and NEQUIP, based on equivariant neural networks.
https://github.com/NVIDIA/cuEquivariance
Last synced: 5 days ago
JSON representation
cuEquivariance is a math library that is a collective of low-level primitives and tensor ops to accelerate widely-used models, like DiffDock, MACE, Allegro and NEQUIP, based on equivariant neural networks.
- Host: GitHub
- URL: https://github.com/NVIDIA/cuEquivariance
- Owner: NVIDIA
- Created: 2024-10-22T15:36:00.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2025-04-08T07:13:22.000Z (8 days ago)
- Last Synced: 2025-04-08T07:37:59.748Z (8 days ago)
- Language: Python
- Size: 808 KB
- Stars: 208
- Watchers: 14
- Forks: 11
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- Contributing: CONTRIBUTING.md
- License: LICENSES/LICENSE
Awesome Lists containing this project
- best-of-atomistic-machine-learning - GitHub - 28% open · ⏱️ 09.04.2025): (Mathematical tools)
README
# cuEquivariance
cuEquivariance is an NVIDIA Python library designed to facilitate the construction of high-performance equivariant neural networks using segmented tensor products. cuEquivariance provides a comprehensive API for describing segmented tensor products and optimized CUDA kernels for their execution. Additionally, cuEquivariance offers bindings for both PyTorch and JAX, ensuring broad compatibility and ease of integration.
Equivariance is the mathematical formalization of the concept of “respecting symmetries.” Robust physical models exhibit equivariance with respect to rotations and translations in three-dimensional space. Artificial intelligence models that incorporate equivariance are often more data-efficient.
## Documentation
Please refer to the project documentation for more information [https://docs.nvidia.com/cuda/cuequivariance/](https://docs.nvidia.com/cuda/cuequivariance/).
## Installation
```bash
# Choose the frontend you want to use
pip install cuequivariance-jax
pip install cuequivariance-torch
pip install cuequivariance # Installs only the core non-ML components
# CUDA kernels
pip install cuequivariance-ops-jax-cu12
pip install cuequivariance-ops-torch-cu12 # or cu11
```
## License
All files hosted in this repository are subject to the Apache 2.0 license.
## Disclaimer
cuEquivariance is in a Beta state. Beta products may not be fully functional, may contain errors or design flaws, and may be changed at any time without notice. We appreciate your feedback to improve and iterate on our Beta products.