Zills Fabian, Schäfer Moritz René, Segreto Nico, Kästner Johannes, Holm Christian, Tovey Samuel
Institute for Computational Physics, University of Stuttgart, 70569 Stuttgart, Germany.
Institute for Theoretical Chemistry, University of Stuttgart, 70569 Stuttgart, Germany.
J Phys Chem B. 2024 Apr 18;128(15):3662-3676. doi: 10.1021/acs.jpcb.3c07187. Epub 2024 Apr 3.
The field of machine learning potentials has experienced a rapid surge in progress, thanks to advances in machine learning theory, algorithms, and hardware capabilities. While the underlying methods are continuously evolving, the infrastructure for their deployment has lagged. The community, due to these rapid developments, frequently finds itself split into groups built around different implementations of machine-learned potentials. In this work, we introduce , a Python-driven software package designed to connect different methods and algorithms from the comprehensive field of machine-learned potentials into a single platform while also providing a collaborative infrastructure, helping ensure reproducibility. Furthermore, the data management infrastructure of the code enables simple model sharing and deployment in simulations. Currently, supports six state-of-the-art machine learning approaches for the fitting of interatomic potentials as well as a variety of methods for the selection of training data, running of calculations, learning-on-the-fly strategies, model evaluation, and simulation deployment.
得益于机器学习理论、算法和硬件能力的进步，机器学习势能领域取得了迅速进展。虽然其基础方法在不断发展，但其部署基础设施却滞后了。由于这些快速发展，该领域的研究人员常分成围绕机器学习势能不同实现方式的不同群体。在这项工作中，我们介绍了[具体软件名称]，这是一个由Python驱动的软件包，旨在将机器学习势能综合领域的不同方法和算法连接到一个单一平台，同时还提供协作基础设施，有助于确保可重复性。此外，[具体软件名称]代码的数据管理基础设施使模型在模拟中能够简单地共享和部署。目前，[具体软件名称]支持六种用于拟合原子间势能的先进机器学习方法，以及多种用于选择训练数据、运行[具体软件名称]计算、实时学习策略、模型评估和模拟部署的方法。
