This repository provides ready-to-use software for high-dimensional neural network potentials in computational physics and chemistry.
This package uses automatic documentation generation via Sphinx, Breathe and doxygen. An online version of the documentation which is automatically updated with the master branch of the repository can be found here.
The pair_style hdnnp interface now forwards per-atom charges predicted by
Q-HDNNP models directly into LAMMPS. Ensure atom_style charge (or a derived
style) is active so the q attribute is allocated, then inspect the values via
standard tooling, for example:
compute myCharge all property/atom qdump 1 all custom 1 charges.dump id type qthermo_style custom step pe etotal(for checking energy consistency)
The charges are updated every time pair_hdnnp invokes the network evaluation,
which allows downstream polarization or post-processing workflows to consume
self-consistent NN charge predictions without any additional coupling code.
See AUTHORS.rst for a list of contributions.
This software is licensed under the GNU General Public License version 3 or any later version (GPL-3.0-or-later).