ParAMS Parametrization

ParAMS MLPotential DFTB ReaxFF

Parametrization tutorials using the ParAMS module. The ParAMS tutorials are located inside the ParAMS documentation.

Getting Started: Lennard-Jones Potential for Argon

Keywords: ParAMS GUI, job collection, training set, Nelder-Mead

Import training data (GUI)

Keywords: Charges, forces, bond scan, angle scan, volume scan, reaction energies, trajectory snapshots

ReaxFF: Gas phase H₂O

Keywords: ReaxFF, bond scan, CMA-ES

M3GNet ML Potential

Keywords: M3Gnet, ML Potential

Simple (MD) Active Learning

Keywords: M3Gnet, ML Potential, active learning, molecular dynamics

ReaxFF: Adsorption on ZnS(110)

Keywords: ReaxFF, choose initial parameters, advanced example

ReaxFF: Convert old training sets to ParAMS format

Keywords: ReaxFF, convert geo, trainset.in, ffield, and control

ReaxFF: Training set for cobalt

Keywords: ReaxFF, detailed description of a training set, vacancies, adsorption energies, equation of state

GFN1-xTB: Lithium fluoride

Keywords: Volume scan, repulsive potential, job-dependent engine settings (k-space), ParAMS engine

GFN1-xTB: Silicon Band Structure

Keywords: Bandstructure, GFN1-xTB, Volume scan, repulsive potential, ParAMS engine

Training and validation sets

Keywords: Training set, validation set (test set)

Running parallel optimizers

Keywords: Optimizer stoppers, optimizer spawning control, multiple optimizers

Restarting (continuing) an optimization

Keywords: CMA-ES, checkpoint file

Calculate reference values with ParAMS

Keywords: Reference engine

SCC-DFTB repulsive potential for ZnO

Keywords: SCC-DFTB repulsive potential, tapered double-exponential, Slater-Koster file (.skf)