README.md

---
license: cc-by-4.0
---

# Description

DPA-3.1-3M is trained using a multitask strategy on the [OpenLAM datasets](https://www.aissquare.com/datasets/detail?pageType=datasets&name=OpenLAM-TrainingSet-v1&id=308). The benchmark results can be found at the [LAMBench website.](https://www.aissquare.com/openlam?tab=Benchmark)

Please ensure the usage of the correct code version (**v3.1.0**) corresponding to this model.

# How to use
## Installation

To use the pretrained model, you need to first install the corresponding version of DeePMD-kit. You can try easy installation:

```bash
pip install torch torchvision torchaudio
pip install git+https://github.com/deepmodeling/deepmd-kit@v3.1.0
```

For other installation options, please visit the [Releases page](https://github.com/deepmodeling/deepmd-kit/releases/tag/v3.1.0) to download the off-line package for **deepmd-kit devel**, and refer to the [official documentation](https://docs.deepmodeling.com/projects/deepmd/en/stable/install/easy-install.html) for off-line installation instructions.

## Use the pretrained model

The pretrained model can be used directly for prediction tasks, such as serving as an ASE calculator.

This model is pretrained in a multi-task manner, featuring a unified backbone (referred to as the unified descriptor and the unified fitting net) and an dataset encoding vector for different tasks. For detailed information, refer to the [DPA-3 paper](https://arxiv.org/abs/2506.01686).

The first step involves selecting a specific dataset id from the model to make predictions. To list the available fitting nets (`model-branch`) in this pretrained model, use the following command:

```bash
dp --pt show DPA-3.1-3M.pt model-branch
```

This will generate an output similar to the following:

```plaintext
Available model branches are ['Domains_Alloy', 'Domains_Anode', 'Domains_Cluster',
'Domains_Drug', 'Domains_FerroEle', 'Domains_SSE_PBE', 'Domains_SemiCond', 'H2O_H2O_PD', 
'Metals_AlMgCu', 'Metals_Sn', 'Metals_Ti', 'Metals_V', 'Metals_W', 'Others_HfO2', 
'Domains_SSE_PBESol', 'Domains_Transition1x', 'Metals_AgAu_PBED3', 'Others_In2Se3', 
'MP_traj_v024_alldata_mixu', 'Alloy_tongqi', 'SSE_ABACUS', 'Hybrid_Perovskite', 
'solvated_protein_fragments', 'Electrolyte', 'ODAC23', 'Alex2D', 'Omat24', 'SPICE2', 'OC20M',
'OC22', 'Organic_Reactions', 'RANDOM'], where 'RANDOM' means using a randomly initialized 
fitting net.
```

Select a dataset id that closely matches your system. Ensure that the elements in your system are included in the corresponding task if you are conducting direct predictions or simulations. For more information on the pretrained datasets, refer to [below](https://www.aissquare.com/models/detail?pageType=models&name=DPA-2.3.1-v3.0.0rc0&id=287#data-used-for-pretraining).

Assuming you choose `H2O_H2O-PD`, you can first freeze the model branch from the multi-task pretrained model:

```bash
dp --pt freeze -c DPA-3.1-3M.pt -o frozen_model.pth --model-branch  H2O_H2O-PD
```

Then you can use the following Python code for prediction or optimization:

```python
## Compute potential energy
from ase import Atoms
from deepmd.calculator import DP as DPCalculator
dp = DPCalculator("frozen_model.pth")
water = Atoms('H2O', positions=[(0.7601, 1.9270, 1), (1.9575, 1, 1), (1., 1., 1.)], cell=[100, 100, 100])
water.calc = dp
print(water.get_potential_energy())
print(water.get_forces())

## Run BFGS structure optimization
from ase.optimize import BFGS
dyn = BFGS(water)
dyn.run(fmax=1e-6)
print(water.get_positions())
```

For more advanced usages like fine-tuning and zero-shot inference, please refer to the following [model card](https://www.aissquare.com/models/detail?pageType=models&name=DPA-2.3.1-v3.0.0rc0&id=287) .