FLARE: Active Learning Bayesian Force Fields

We have a few tutorial notebooks that you can check out and play with.

FLARE (ACE descriptors + sparse GP) with LAMMPS. This tutorial demonstrates how to construct Bayesian force field based on sparse Gaussian Process (SGP) model for aluminum. We cover both “online” (on-the-fly) active learning and “offline” training methodologies using ASE MD engine. Additionally, the later section guides users through compiling LAMMPS with FLARE pair styles to enable scalable Bayesian active learning and uncertainty-aware molecular dynamics.

FLARE (LAMMPS active learning). This tutorial demonstrates new functionality for running active learning all within LAMMPS, with LAMMPS running the dynamics to allow arbitrarily complex molecular dynamics workflows while maintaining a simple interface. This also demonstrates how to use the C++ API directly from Python through pybind11. Finally, there’s a simple demonstration of phonon calculations with FLARE using phonopy.

Compute thermal conductivity from FLARE and Boltzmann transport equations. The tutorial shows how to use FLARE (LAMMPS) potential to compute lattice thermal conductivity from Boltzmann transport equation method, with Phono3py for force constants calculations and Phoebe for thermal conductivities.

Using your own customized descriptors with FLARE. The tutorial shows how to attach your own descriptors with FLARE sparse GP model and do training and testing.

All the tutorials take a few minutes to run on a normal desktop computer or laptop (excluding installation time).