Command line interface

SevenNet provides five commands for preprocessing, training, and deployment: sevenn preset, sevenn graph_build, sevenn, sevenn inference, and sevenn get_model.

sevenn preset

With the sevenn preset command, the input file setting the training parameters is generated automatically.

sevenn preset {preset keyword} > input.yaml

Available preset keywords are: base, fine_tune, multi_modal, sevennet-0, and sevennet-l3i5. Check comments in the preset YAML files for explanations. For fine-tuning, be aware that most model hyperparameters cannot be modified unless explicitly indicated. To reuse a preprocessed training set, you can specify sevenn_data/${dataset_name}.pt for the load_trainset_path: in the input.yaml.

sevenn graph_build

To obtain the preprocessed data, sevenn_data/graph.pt, sevenn graph_build command can be used. The output files can be used for training (sevenn) or inference (sevenn inference) to skip the graph build stage.

sevenn graph_build {dataset path} {cutoff radius}

The output sevenn_data/graph.yaml contains statistics and meta information about the dataset. These files must be located in the sevenn_data directory. If you move the dataset, move the entire sevenn_data directory without changing the contents.

See sevenn graph_build --help for more information.

sevenn train

Given that input.yaml and sevenn_data/graph.pt are prepared, SevenNet can be trained by the following command:

sevenn train input.yaml -s

We support multi-GPU training using PyTorch DDP (distributed data parallel) with a single process (or a CPU core) per GPU.

torchrun --standalone --nnodes {number of nodes} --nproc_per_node {number of GPUs} --no_python sevenn input.yaml -d

Enable cuEquivariance or flashTP while training by adding the –enable_cueq or –enable_flashTP.

sevenn train input.yaml -s --enable_cueq # or --enable_flashTP

Please note that batch_size in input.yaml refers to the per-GPU batch size.

sevenn inference

Using the checkpoint after training, the properties such as energy, force, and stress can be inferred directly.

sevenn inference checkpoint_best.pth path_to_my_structures/*

This will create the inference_results directory, where CSV files contain predicted energy, force, stress, and their references (if available). See sevenn inference --help for more information.

sevenn get_model

The command is for LAMMPS integration of SevenNet. It deploys a model into a LAMMPS readable file(s).

See LAMMPS: PyTorch or LAMMPS: ML-IAP for installation.

LAMMPS/Torch

The checkpoint can be deployed as LAMMPS potentials. The argument is either the path to the checkpoint or the name of a pretrained potential.

sevenn get_model 7net-0  # For pre-trained models
sevenn get_model {checkpoint path}  # For user-trained models

This will create deployed_serial.pt, which can be used as a LAMMPS potential with the e3gnn pair_style in LAMMPS.

The potential for parallel MD simulation can be obtained similarly.

sevenn get_model 7net-0 -p
sevenn get_model {checkpoint path} -p

This will create a directory with several deployed_parallel_*.pt files. The directory path itself is an argument for the LAMMPS script. Please do not modify or remove files in the directory. These models can be used as LAMMPS potentials to run parallel MD simulations with a GNN potential across multiple GPUs.

LAMMPS/ML-IAP

sevenn get_model 7net-0 --use_mliap  # For pre-trained models
sevenn get_model {checkpoint path} --use_mliap  # For user-trained models

This will create deployed_serial_mliap.pt, which can be used as a LAMMPS potential with the mliap pair_style in LAMMPS.

sevenn cp

This is an utility command. You can check model’s complexity, metadata, and its modalities (or tasks) at glance.

sevenn cp 7net-mf-ompa

Sevennet version                                0.11.0
When                                  2025-03-10 23:32
Hash                  ce5b3f90adf546499e528174fbc72fce
Cutoff                                             6.0
Channel                                            128
Lmax                                                 3
Group (parity)                                      O3
Interaction layers                                   5
Self connection type                            nequip
Last epoch                                        None
Elements                                           119
cuEquivariance used                              False
FlashTP used                                     False
Modality                                   omat24, mpa