MIR@Harvard

NequIP is a code for building E(3)-equivariant interatomic potentials

A high-performance software package for training and evaluating machine-learned XC functionals using the CIDER framework

An open-source Python package for creating fast and accurate interatomic potentials.

Allegro-pol extends the Allegro architecture to model the electric response of materials

Allegro is a code for building highly scalable E(3)-equivariant interatomic potentials

NequIP-NAC extends the NequIP architecture to perform nonadiabatic molecular dynamics.

LAMMPS pair styles for NequIP and Allegro deep learning interatomic potentials

This repo serves as a minimal template for `nequip` extension packages.

Resources for running NequIP foundation models with LAMMPS on Perlmutter