Introduction

The Orchestrator is intended to be used as an integrated software package for building, training, testing, augmenting, running, and analyzing interatomic potentials (IAPs) and their simulations. It provides a uniform API via abstract classes, enabling the use of common codes relevant to the IAP development and deployment cycle alongside novel Orchestrator functionality. The shared API also enables drop-in replacement of different concrete instances of a single abstract class, adding to the flexibility of the framework.

Beyond the module design, the code base supports multiple usage modes, either as a library for custom built scripts or through the Executor module which facilitates an application-like style.

Note

The Executor module is still under active development and is not included in the 0.5 version release.

Examples for running the code are covered in the examples section and the test suite.

In the subsequent sections of the introduction, we cover the key steps needed to start using the Orchestrator, including instructions and tips for installation, how to generate a Jupyer kernel, and details on how to best leverage the Orchestrator’s integration with KIM suite tools from the KIM initiative.

Jump to a specific topic:

• Installation
• Integrating Orchestrator with Jupyter Notebooks
• Using KIM Software
• KIMkit Setup
• AiiDA