Combination with a pairwise potential

[RealStephenZhang]

Summary
Dear developers,

I am trying to use DeepMD in my campus sever(weiming2). It works well when directly training DP model, but when I try to use 'Interpolation or combination with a pairwise potential', I have the following error when doing: dp train input.json

The error I got is:
Traceback (most recent call last):
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/bin/dp", line 10, in
sys.exit(main())
^^^^^^
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/deepmd_cli/main.py", line 635, in main
deepmd_main(args)
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/deepmd/entrypoints/main.py", line 74, in main
train_dp(**dict_args)
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/deepmd/entrypoints/train.py", line 146, in train
jdata = normalize(jdata)
^^^^^^^^^^^^^^^^
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/deepmd/utils/argcheck.py", line 2008, in normalize
data = base.normalize_value(data, trim_pattern="_*")
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 554, in normalize_value
self.traverse_value(
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 342, in traverse_value
self._traverse_sub(
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 372, in _traverse_sub
subarg.traverse(value, key_hook, value_hook, sub_hook, variant_hook, path)
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 324, in traverse
self.traverse_value(
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 342, in traverse_value
self._traverse_sub(
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 372, in _traverse_sub
subarg.traverse(value, key_hook, value_hook, sub_hook, variant_hook, path)
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 324, in traverse
self.traverse_value(
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 347, in traverse_value
self._traverse_sub(
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 371, in _traverse_sub
for subarg in self.flatten_sub(value, path).values():
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 296, in flatten_sub
sub_dicts.extend(
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 297, in
vrnt.flatten_sub(value, path) for vrnt in self.sub_variants.values()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 809, in flatten_sub
choice = self.get_choice(argdict, path)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/lustre/home/2201110238/software/deepmd-kit-2.2.7-cuda118/lib/python3.11/site-packages/dargs/dargs.py", line 796, in get_choice
raise ArgumentValueError(
dargs.dargs.ArgumentValueError: [at location model/models/1] get invalid choice pairtab for flag key type.

DeePMD-kit Version
2.2.7

TensorFlow Version
2.14.0

Python Version, CUDA Version, GCC Version, LAMMPS Version, etc
Python3.11,

Details
The input file in json is:
{
"_comment1": " model parameters",
"model": {
"type": "linear_ener",
"weights": "sum",
"models": [
{
"_comment": "Here uses a delta DP model",
"type": "frozen",
"model_file": "delta.pb"
},
{
"type": "pairtab",
"tab_file": "morse.txt",
"rcut": 5.0,
"sel": 800
}
]
},

"learning_rate": {
"type": "exp",
"decay_steps": 5000,
"start_lr": 0.001,
"stop_lr": 3.51e-8,
"_comment5": "that's all"
},

"loss": {
"type": "ener",
"start_pref_e": 0.02,
"limit_pref_e": 1,
"start_pref_f": 1000,
"limit_pref_f": 1,
"start_pref_v": 0,
"limit_pref_v": 0,
"_comment6": " that's all"
},

"training": {
"training_data": {
"systems": [
"../data_delta0/training_data/"
],
"batch_size": "auto",
"_comment7": "that's all"
},
"validation_data": {
"systems": [
"../data_delta0/validation_data/"
],
"batch_size": 1,
"numb_btch": 3,
"_comment8": "that's all"
},
"numb_steps": 0,
"seed": 10,
"disp_file": "lcurve.out",
"disp_freq": 100,
"save_freq": 1000,
"_comment9": "that's all"
},

"_comment10": "that's all"
}

I have tried to revise this input file according to https://github.com/deepmodeling/deepmd-kit/tree/6dafa5076ec88557dbc1bbf68e46cbc91d1ae86c/examples/water/d3 . But this error still exists.
I would really appreciate it if anyone could help.

[wanghan-iapcm]

This input file is only supported by v3 or devel. you are using deepmd-kit v2.2.7

[njzjz]

It's new in v2.2.8. You can find the available version in a905817.