torch_sim.models.orb

ORB: PyTorch implementation of ORB models for atomistic simulations.

This module provides a TorchSim wrapper of the ORB models for computing energies, forces, and stresses of atomistic systems. It serves as a wrapper around the ORB models library, integrating it with the torch_sim framework to enable seamless simulation of atomistic systems with machine learning potentials.

The OrbModel class adapts ORB models to the ModelInterface protocol, allowing them to be used within the broader torch_sim simulation framework.

Notes

This implementation requires orb_models to be installed and accessible. It supports various model configurations through model instances or model paths.

Functions

state_to_atom_graphs

Convert a SimState object into AtomGraphs format, ready for use in an ORB model.

Classes

OrbModel

Computes atomistic energies, forces and stresses using an ORB model.