DeePMD-kit Training

Use this skill to guide DeePMD-kit model training without loading every model-specific recipe up front. The workflow is intentionally progressive:

1. Understand the user's data, target accuracy, compute budget, and deployment backend.
2. Choose an appropriate model family.
3. Read only the reference file for the selected model under models/.
4. Generate or edit input.json, run training, monitor, freeze, and test.

Progressive disclosure protocol

Do not start by reading every model document. First classify the request:

• If the user already named a model, read only that model reference.
• If the user asks for a recommendation, collect the decision inputs below, choose a model, then read only the selected reference.
• If model-specific parameters are not needed yet, stay in this top-level workflow.

Available model references:

Model selection

Ask only for missing information that changes the choice. Prefer reasonable defaults when the answer is obvious from context.

Key inputs:

• Data format and size: deepmd/npy, deepmd/hdf5, mixed type, number of systems/frames/elements.
• Target: quick baseline, production accuracy, large atomic model, transfer/fine-tuning, or deployment in MD.
• Compute: CPU/GPU, available memory, single-node vs. distributed training.
• Backend/deployment: PyTorch/TensorFlow/JAX/Paddle training; LAMMPS, Python inference, or other downstream use.
• Labels: energy/force only or also virial/stress.
• System diversity: single chemistry/phase vs. diverse multi-domain datasets.

Recommended defaults:

• Choose se_e2_a for a robust baseline, small to medium systems, compatibility-focused workflows, or when compute is limited.
• Choose DPA3 for high accuracy on diverse datasets, LAM-style training, or when the user explicitly asks for DPA3, DPA-3, LiGS, dynamic neighbor selection, or pretrained DPA3 variants.
• Choose DPA4/SeZM when the user explicitly requests it or wants its SO(3)-equivariant message-passing architecture and accepts a GPU-oriented, PyTorch-only workflow. For a new configuration, choose a grade and use a dpa4-<grade>-v20260911 preset.
• Choose DPA4C when the user explicitly requests it or prioritizes compact, high-throughput deployment, including distillation from a DPA4 teacher. For a new configuration, choose a grade and use a dpa4c-<grade>-v20260911 preset.

Common workflow

1. Confirm environment

dp --version

For conventional PyTorch training, use dp --pt ...; DPA4C uses the PyTorch Exportable backend, dp --pt-expt .... For TensorFlow, use dp ...; for other backends, confirm the installed backend first.

2. Confirm training data

Training data should be in DeePMD format, typically deepmd/npy or deepmd/hdf5. If the user has raw electronic-structure outputs, convert them first with dpdata before writing the training input.

Minimum information needed to build input.json:

• type_map
• training system paths
• validation system paths
• whether virial labels are present and should be trained
• target number of steps or accuracy/time budget
• model choice

3. Read the selected model reference

For new DPA4 and DPA4C inputs, follow the preset workflow first and reserve manual architecture blocks for the advanced path.

After selecting a model, read the corresponding file under models/ and apply its model-specific configuration, hyperparameters, and caveats.

4. Train

dp --pt train input.json

Use the backend-specific command if not using PyTorch.

Restart from a checkpoint when needed:

dp --pt train input.json --restart model.ckpt.pt

5. Monitor

Training progress is usually written to lcurve.out. Check for:

• decreasing validation RMSE
• NaN or exploding losses
• train/validation divergence
• learning-rate schedule behaving as expected

6. Freeze and test

Read the selected model reference before choosing the output format. For conventional PyTorch models, a typical flow is:

dp --pt freeze -o model.pth
dp test -m model.pth -s /path/to/test_system -n 30

DPA4/SeZM and DPA4C checkpoints instead freeze to .pt2; follow the selected model reference. DPA4C may additionally be compressed for deployment. Adjust the backend and output format for other model families.

Agent checklist

• [ ] Model was selected before reading model-specific details.
• [ ] Only the selected model reference was loaded.
• [ ] Training/validation data paths exist or are clearly marked as placeholders.
• [ ] type_map matches the data and model/pretrained checkpoint.
• [ ] Virial loss is enabled only when virial labels are available and desired.
• [ ] Backend command matches the selected model and installed DeePMD-kit environment.
• [ ] The generated input.json is valid JSON.
• [ ] Training was monitored via lcurve.out or equivalent logs.
• [ ] Final model was frozen and tested when requested.

References

• Training documentation
• Training input documentation
• Model documentation