neuraloperator/neuraloperator

Health v2 · maintenance only

Adoption not part of the score

3836 stars · 919 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

A PyTorch library for learning neural operators, which map between function spaces rather than finite-dimensional vectors. It provides the official implementation of the Fourier Neural Operator (FNO) and related architectures, with resolution-invariant models for PDE learning.

Use cases

• learn mappings between function spaces with neural networks
• solve partial differential equations with machine learning
• train a Fourier Neural Operator on my own data
• build PDE surrogates that work at any grid resolution
• compress FNO models with tensor factorization
• run scientific ML experiments on Darcy flow datasets

When to choose

• you need resolution-invariant neural operators for PDE or physics problems
• you want the official, PyTorch-ecosystem FNO implementation
• you want tensorized (Tucker) variants to reduce model parameters
• you work in scientific machine learning with PyTorch

When to avoid

• you need standard vision/NLP models rather than operator learning
• you prefer TensorFlow or JAX instead of PyTorch
• you need a no-code or GUI tool rather than a Python library

Facets

library · maturity active

machine-learning deep-learning simulation math machine-learning deep-learning simulation data-science python neural-operator fourier-neural-operator fno pde pytorch-ecosystem scientific-machine-learning tensor-factorization gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="neuraloperator/neuraloperator")