Machine learning (ML)-based framework to train interaction models between atoms
Typical usage scenario:
- create pseudo/ folder and copy your pseudopotential file in it
mkdir pseudo && cp <your-pp-file> pseudo/<your-pp-file>
- create directories storing the simulations i/o data
python init_random_structures.py
-
[optional] specify/implement your interaction model in interaction_models.py file
-
process the simulation data output and train interaction model:
python train_model.py