README.md

mdpu-kit Manual

Table of contents

• mdpu-kit Manual
• Table of contents
• About mdpu-kit
  • Highlighted features
  • License and credits
• Download and install
• Use mdpu-kit
• Advanced
• Code structure

About mdpu-kit

mdpu-kit is a package written in Python/C++, designed to minimize the effort required to build deep learning-based model of interatomic potential energy and force field and to perform molecular dynamics (MD). This brings new hopes to addressing the accuracy-versus-efficiency dilemma in molecular simulations. Applications of mdpu-kit span from finite molecules to extended systems and from metallic systems to chemically bonded systems.

Highlighted features

• interfaced with TensorFlow, one of the most popular deep learning frameworks, making the training process highly automatic and efficient, in addition, Tensorboard can be used to visualize training procedures.
• interfaced with high-performance classical MD packages, i.e., LAMMPS.
• implements the Deep Potential series models, which have been successfully applied to finite and extended systems including organic molecules, metals, semiconductors, insulators, etc.
• implements MPI supports, making it highly efficient for high-performance parallel and distributed computing.
• highly modularized, easy to adapt to different descriptors for deep learning-based potential energy models.

License and credits

The project mdpu-kit is licensed under GNU LGPLv3.0. If you use this code in any future publications, please cite the following publications for general purpose: Breaking the Speed, Power, Cost and Size Limits of Molecular Dynamics with Ab Initio Accuracy # NATCOMPUTSCI-23-0729

In addition, please follow the bib file to cite the methods you used.

Download and install

One may manually install mdpu-kit by following the instructions on installing the Python interface and installing the C++ interface. The C++ interface is necessary when using mdpu-kit with LAMMPS.

Use mdpu-kit

A quick start on using mdpu-kit can be found here.

Advanced

• Installation
  • Install from source code
• Data
  • System
  • Formats of a system
  • Prepare data with dpdata
• Training
  • Training a model
• Test
  • Test a model
• Inference
  • Python interface
  • C++ interface
• Integrate with third-party packages
  • Run MD with LAMMPS
  • LAMMPS commands
• Use MDPU

Code structure

The code is organized as follows:

• data/raw: tools manipulating the raw data files.
• examples: examples.
• mdpukit: mdpu-kit python modules.
• source/api_cc: source code of mdpu-kit C++ API.
• source/lib: source code of mdpu-kit library.
• source/lmp: source code of Lammps module.
• source/op: TensorFlow op implementation. working with the library.