Code to paper: EGRAFFBENCH: EVALUATION OF EQUIVARIANT GRAPH NEURAL NETWORK FORCE FIELDS FOR ATOMISTIC SIMULATIONS
see : env
- cd ../
- pip install gdown (install gdown to download data from google drive)
- gdown --folder --id 1PrrKaMBbjMyt3DrXM94XQ-X48nVrf-kY
- cd mdbenchgnn/
- ln -s ../example_mdbenchgnn example
- ln -s path/to/output_dir output_dir_sl
- ln -s path/to/data_dir data_ls
-
Nequip
-
Allegro
-
Mace
-
Botnet
-
Equiformer
-
TorchMDnet
Try on example data: example/lips/data
- to train the models go through: scripts/train
- to run a MD simulation on trained model: scripts/md_simulation
see : preprocessing
- Radial Distribution function scripts/struct_props
Our implementation is based on PyTorch, PyG, nequip, allegro, equiformer, mace, TorchMD-NET and MDsim