Differentiable Physics Python Packages

exponax

Efficient Differentiable n-d PDE Solvers in JAX.

apebench

[Neurips 2024] A benchmark suite for autoregressive neural emulation of PDEs. (≥46 PDEs in 1D, 2D, 3D; Differentiable Physics; Unrolled Training; Rollout Metrics)

tesseract-core

Universal components for differentiable scientific computing. :package:

dmg

Generic framework for building differentiable models.

softtorch

A library for soft differentiable relaxations of common PyTorch functions.

adirondax

A differentiable astrophysics simulator in JAX

hydrodl2

Repository for MHPI differentiable hydrological models.

nefi

Neural-field inversion: recover hidden physical fields from a single measurement with differentiable physics and no training data

softjax

A library for soft differentiable relaxations of common JAX functions.

spicex

A differentiable SPICE circuit solver in JAX

jaxion

A simple JAX-powered simulation library for numerical experiments of fuzzy dark matter, stars, gas + more!

eiko

Eiko: the GPU-accelerated Eikonal equation solver

pycharge

Electromagnetics simulation library for moving point charges built on JAX

mosaic-bench

A benchmark suite and re-usable collection of differentiable physics solvers.

neuralgcm-torch

PyTorch port of NeuralGCM

updes

Universal Partial Differential Equations Simulator

dinosaur-torch

PyTorch port of NeuralGCM

synth-afm

Differentiable High-Speed AFM (HS-AFM) simulation for protein structures. JAX-powered toolkit for generating synthetic movies with realistic tip-dilation and scanning lag.

hydrologeez

Differentiable conceptual hydrological models in PyTorch — end-to-end gradients through the full simulation for parameter calibration.