Aidan Thompson
Aidan Thompson
@Luthaf and @akohlmey Thanks for the detailed dialogue. I learned quite a bit from reading this. ML-METATENSOR is a really nice addition to LAMMPS that enables a lot of new...
@ceriottm You are free to maintain your fork separately. However, as @akohlmey indicated, it will probably be less work for you and users if you go the plugin route, because...
I took a close look at this, and it looks promising, most of the basic functionality seems to be there. I made a test script for a Lennard-Jones system with...
This is now running free of obvious code errors on 2xN partitions. I will continue to test it, clean it, and document it.
Doc page added.
I think this is ready to be merged now
Hi @bnebgen-LANL, sorry for not looking at this sooner. Based on this discussion, in order for this to be accepted in to LAMMPS, I see two alternatives: 1. Leverage existing...
@ssande7 This sounds like a significant improvement in the consistency of SLLOD in LAMMPS, particularly recovering the correct work production rate and also time reversibility. For people who want to...
@ssande7 We prefer not to keep legacy versions hanging around. If there is a clean way to provide old functionality, then good, if not, let's not do it.