cuEquivariance
GPU-accelerated equivariant tensor operations for geometric neural networks
All Skills in This Repository (3)
Pure Emerald Level Indicators
NVIDIAcuequivariance-torch
Execute equivariant tensor-product computations on GPUs from PyTorch using cuEquivariance.
NVIDIAcuequivariance
Define custom groups and build segmented tensor products with Clebsch-Gordan coefficients.
NVIDIAcuequivariance-jax
Execute equivariant polynomials in JAX using cuequivariance primitives.
Frequently Asked Questions
FAQPage SchemaHow to install cuEquivariance?▼
Run `npx skills add NVIDIA/cuEquivariance --all -g -y` in your terminal to install all skills in this suite globally.
What is cuEquivariance used for?▼
It is an NVIDIA library for building equivariant neural networks that respect 3D rotation and translation symmetries, commonly used in molecular modeling and physics-informed machine learning.
Does cuEquivariance work with PyTorch and JAX?▼
Yes. It provides dedicated bindings for both PyTorch and JAX, including ready-made layers like tensor products, spherical harmonics, and batch normalization.
Do I need a GPU to use cuEquivariance?▼
The naive fallback methods run on any platform, but the high-performance CUDA kernels require an NVIDIA GPU for full speed.
Can I use cuEquivariance without group theory knowledge?▼
Yes. Built-in descriptors handle the Clebsch-Gordan math for you, so you can define tensor products and layers with simple irreps strings.
Related Repositories in Software Engineering
View All in Software Engineering→
openclawopenclaw
Run a personal AI assistant across your devices and chat apps
obrasuperpowers
Gives coding agents a disciplined workflow from idea to merged code
facebookreact
AI agent skills for building, testing, and porting React core