search

SimonBoothroyd / descent

Optimize classical force field parameters against reference data
https://simonboothroyd.github.io/descent/
MIT License
8 stars 3 forks source link

readme

DESCENT

Optimize force field parameters against reference data

ci coverage license

The descent framework aims to offer a modern API for training classical force field parameters (either from a traditional format such as SMIRNOFF or from some ML model) against reference data using pytorch.

This framework benefited hugely from ForceBalance, and a significant number of learning from that project, and from Lee-Ping, have influenced the design of this one.

Warning: This code is currently experimental and under active development. If you are using this it, please be aware that it is not guaranteed to provide correct results, the documentation and testing maybe be incomplete, and the API can change without notice.

Installation

This package can be installed using conda (or mamba, a faster version of conda):

mamba install -c conda-forge descent

The example notebooks further require you install jupyter:

mamba install -c conda-forge jupyter

Getting Started

To get started, see the examples.

Copyright

Copyright (c) 2023, Simon Boothroyd