Published June 11, 2025 | Version v1
Software Open

JAX-LaB: A Python-based, Accelerated, Differentiable Massively Parallel Lattice Boltzmann Library for Modeling Multiphase and Multiphysics Flows & Physics-Based Machine Learning

Creators

  • 1. ROR icon Columbia University

Description

This repository contains all the code, example scripts, and data required to reproduce the results shown in the manuscript: JAX-LaB: A High-Performance, Differentiable, Lattice Boltzmann Library for Modeling Multiphase Fluid Dynamics in Geosciences and Engineering, soon to be published in the Journal of Advances in Modeling Earth Systems (JAMES). Please refer to README.md and src/examples for detailed usage instructions.

Files

JAX-LaB.zip

Files (16.5 MB)

Name Size Download all
md5:cd2199fdc44252b60a995176ec67e2f0
16.5 MB Preview Download

Additional details

Related works

Is supplement to
Publication: arXiv:2506.17713 (arXiv)

Dates

Available
2025-06-24

Software

Repository URL
https://github.com/piyush-ppradhan/JAX-LaB
Programming language
Python
Development Status
Active

References