DP Compress: A Model Compression Scheme for Generating Efficient Deep Potential Models

Autor:	Denghui Lu, Wanrun Jiang, Yixiao Chen, Linfeng Zhang, Weile Jia, Han Wang, Mohan Chen
Rok vydání:	2022
Předmět:	Machine Learning Chemical Physics (physics.chem-ph) Physics - Chemical Physics FOS: Physical sciences Neural Networks Computer Molecular Dynamics Simulation Computational Physics (physics.comp-ph) Physical and Theoretical Chemistry Physics - Computational Physics Computer Science Applications
Zdroj:	Journal of Chemical Theory and Computation. 18:5559-5567
ISSN:	1549-9626 1549-9618
Popis:	Machine-learning-based interatomic potential energy surface (PES) models are revolutionizing the field of molecular modeling. However, although much faster than electronic structure schemes, these models suffer from costly computations via deep neural networks to predict the energy and atomic forces, resulting in lower running efficiency as compared to the typical empirical force fields. Herein, we report a model compression scheme for boosting the performance of the Deep Potential (DP) model, a deep learning based PES model. This scheme, we call DP Compress, is an efficient post-processing step after the training of DP models (DP Train). DP Compress combines several DP-specific compression techniques, which typically speed up DP-based molecular dynamics simulations by an order of magnitude faster, and consume an order of magnitude less memory. We demonstrate that DP Compress is sufficiently accurate by testing a variety of physical properties of Cu, H2O, and Al-Cu-Mg systems. DP Compress applies to both CPU and GPU machines and is publicly available online.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_dedup___::38eff435cefb1c766e4c0d621b477c6a https://doi.org/10.1021/acs.jctc.2c00102 Zobrazit plný text záznamu