Training a Machine Learning Interatomic Potential with metatensor
This repository contains the notebook for the tutorial on training fully custom MLIP using metatensor, presented during the CECAM/Psi-k school on MLIP of 2023 (https://www.mlip-workshop.xyz).
Running the tutorials
You should go through the different notebooks in this repository in order. Everytime you see this icon
This means you need to modify the code to make it work and make a small test pass. All the modifications should only require basic Python knowledge, and no specific PyTorch/Metatensor/Rascaline knowledge.
Installation
You don't need to install the code if you are participating to the school on-site, we are already providing you with an environment for this on the Puhti supercalculator.
If you are joining the school remotely, or want to re-run this tutorial after the school ends, here is how you can install everything needed!
You'll need a working conda installation, and to run the following commands:
git clone https://github.com/Luthaf/mlip-2023-tutorial
cd mlip-2023-tutorial
conda env create --file install/environment.yaml
conda activate metatensor-mlip-2023
./install/extra-pip-packages.sh
# OPTIONAL, only run if you want to run the last notebook
# (running metatensor models inside LAMMPS)
./install/build-lammps.sh