KLIFF: A framework to develop physics-based and machine learning interatomic potentials

Abstract

Interatomic potentials (IPs) are reduced-order models for calculating the potential energy of a system of atoms given their positions in space and species. IPs treat atoms as classical particles without explicitly modeling electrons and thus are computationally far less expensive than first-principles methods, enabling molecular simulations of significantly larger systems over longer times. Developing an IP is a complex iterative process involving multiple steps: assembling a training set, designing a functional form, optimizing the function parameters, testing model quality, and deployment to molecular simulation packages. This paper introduces the KIM-based learning-integrated fitting framework (KLIFF), a package that facilitates the entire IP development process. KLIFF supports both physics-based and machine learning IPs. It adopts a modular approach whereby various components in the fitting process, such as atomic environment descriptors, functional forms, loss functions, optimizers, quality analyzers, and so on, work seamlessly with each other. This provides a flexible framework for the rapid design of new IP forms. Trained IPs are compatible with the Knowledgebase of Interatomic Models (KIM) application programming interface (API) and can be readily used in major materials simulation packages compatible with KIM, including ASE, DL_POLY, GULP, LAMMPS, and QC. KLIFF is written in Python with computationally intensive components implemented in C++. It is parallelized over data and supports both shared-memory multicore desktop machines and high-performance distributed memory computing clusters. We demonstrate the use of KLIFF by fitting a physics-based Stillinger--Weber potential and a machine learning neural network potential for silicon. The KLIFF package, together with its documentation, is publicly available at: https://github.com/openkim/kliff.

arXiv comment16 pages, 9 figures

References in corpus (32)

• Improving neural networks by preventing co-adaptation of feature detectors
• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• On representing chemical environments
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• Interatomic potentials from first-principles calculations: the force-matching method
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• DScribe: Library of Descriptors for Machine Learning in Materials Science
• Interatomic Potential for Silicon Defects and Disordered Phases
• Environment Dependent Interatomic Potential for Bulk Silicon
• Nucleation mechanism for the direct graphite-to-diamond phase transition
• The geometry of nonlinear least squares with applications to sloppy models and optimization
• Potfit: effective potentials from ab-initio data
• The sloppy model universality class and the Vandermonde matrix
• Improvements to the Levenberg-Marquardt algorithm for nonlinear least-squares minimization
• Representations of molecules and materials for interpolation of quantum-mechanical simulations via machine learning
• Modeling of Covalent Bonding in Solids by Inversion of Cohesive Energy Curves
• Dihedral-angle-corrected registry-dependent interlayer potential for multilayer graphene structures
• Classical interaction potentials for diverse materials from ab initio data: a review of potfit
• Stratified construction of neural network based interatomic models for multicomponent materials
• Hybrid neural network potential for multilayer graphene
• PANNA: Properties from Artificial Neural Network Architectures
• MAISE: Construction of neural network interatomic models and evolutionary structure optimization
• Atomic Permutationally Invariant Polynomials for Fitting Molecular Force Fields
• PyXtal FF: a Python Library for Automated Force Field Generation
• Bayesian, frequentist, and information geometric approaches to parametric uncertainty quantification of classical empirical interatomic potentials
• atomicrex - A general purpose tool for the construction of atomic interaction models
• Uncertainty quantification for classical effective potentials: an extension to potfit
• The OpenKIM Processing Pipeline: A Cloud-Based Automatic Materials Property Computation Engine
• A KIM-compliant potfit for fitting sloppy interatomic potentials: Application to the EDIP model for silicon
• Database optimization for empirical interatomic potential models

Cited by in corpus (3)

• GPUMD: A package for constructing accurate machine-learned potentials and performing highly efficient atomistic simulations
• Bayesian, frequentist, and information geometric approaches to parametric uncertainty quantification of classical empirical interatomic potentials
• Extending OpenKIM with an Uncertainty Quantification Toolkit for Molecular Modeling