deepmodeling / deepmodeling/AIS-Square

Can't use models from AIS Square

Open
#154 1 comment 0 reactions 0 assignees View on GitHub
Dominant language
No language data
Stars
15
Forks
10
PR merge metrics
No merged PRs in 30d

Description

### Bug summary

After attempting to use the H2O-DPLR-model in the ASE and LAMMPS (w/ ASE interface), I receive the following error:

`
RuntimeError: model in graph (version 0.0) is incompatible with the model (version 1.1) supported by the current code.See https://deepmd.rtfd.io/compatability/ for details.`

### Input Files, Running Commands, Error Log, etc.

ASE input:

```
from ase import Atoms
from deepmd.calculator import DP

water = Atoms(
"H2O",
positions=[(0.7601, 1.9270, 1), (1.9575, 1, 1), (1.0, 1.0, 1.0)],
cell=[100, 100, 100],
calculator=DP(model="H2O-SCAN0-model.pb"),
)
print(water.get_potential_energy())
print(water.get_forces())
```

LAMMPS ASE input:
```

from ase.calculators.lammpsrun import LAMMPS
from ase import Atoms
water = Atoms(
"H2O",
positions=[(0.7601, 1.9270, 1), (1.9575, 1, 1), (1.0, 1.0, 1.0)],
cell=[100, 100, 100],
)

parameters = {'pair_style': 'deepmd H2O-DPLR-model.pb',
'pair_coeff': '* * O H'}
files = ['H2O-DPLR-model.pb']
lammps = LAMMPS(parameters=parameters, files=files, keep_tmp_files=True, tmp_dir='./debug')
water.calc = lammps
print(water.get_potential_energy())
```

Contributor guide

No contributing guide indexed for this repository

Research direction

Start by reproducing the ASE and LAMMPS examples with the referenced H2O-SCAN0-model.pb and H2O-DPLR-model.pb files, then read the compatibility guidance linked in the error. Compare the model graph version 0.0 with the version 1.1 supported by the current code; done means the AIS Square models load successfully or their incompatibility is clearly documented.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.