We develop software to enable modularity and collaboration in atomistic machine learning; through data, models and architecture exchange.

We have three main projects:

• metatensor is a generic self-describing sparse data format, that can be used by multiple pieces of code to exchange data without knowing about each other;
• metatomic is a interface to execute arbitrary atomistic machine learning models in arbitrary simulation engines;
• metatrain is a command line tool to train state of the art atomistic machine learning models on custom datasets, and expand the capabilities of existing model architectures.

We also have a couple of projects built on top of the above three:

• featomic is a tool to compute machine learning features for atomistic systems, which can then be used in larger models
• lammps-metatomic is our fork of LAMMPS with added support for metatomic models