Can not use deep-potencial with ASE in deepmdkit-1.1.4

[ShuSOfficial]

I am trying to use deep-potencial with ase, while the version of deepmdkit installed is 1.1.4.
My input file is as follows:

from ase import Atoms
from deepmd.calculator import DP

water = Atoms('In2Se3',
              positions=[(0.333629403078083, 0.666864220718582, 0.529711279564291),
                         (0.666963320410938, 0.333530883627878, 0.388632458639199),
                         (0.000296626117675, 0.000197550471460, 0.347102195162897),
                         (0.666962862659665, 0.333530886564587, 0.473601305147610),
                         (1.000296087588707, 0.000197553939909, 0.573622813485869)],
              cell=[[4.1011633576, 0.0, 0.0],[-2.0505818476, 3.5517116518, 0.0],[0.0, 0.0, 30.0000000000]],
              calculator=DP(model="frozen_model.pb"))
print(water.get_potential_energy())

The frozen_model.pb is a deep potencial trained by deepmdkit-1.1.4. By the way, an error ocurred:

Traceback (most recent call last):
  File "test_In2Se3.py", line 12, in <module>
    print(water.get_potential_energy())
  File "/home/liushiLab/zhouwenjie/.conda/envs/deepmd_cpu114/lib/python3.7/site-packages/ase/atoms.py", line 700, in get_potential_energy
    energy = self._calc.get_potential_energy(self)
  File "/home/liushiLab/zhouwenjie/.conda/envs/deepmd_cpu114/lib/python3.7/site-packages/ase/calculators/calculator.py", line 642, in get_potential_energy
    energy = self.get_property('energy', atoms)
  File "/home/liushiLab/zhouwenjie/.conda/envs/deepmd_cpu114/lib/python3.7/site-packages/ase/calculators/calculator.py", line 695, in get_property
    self.calculate(atoms, [name], system_changes)
  File "/home/liushiLab/zhouwenjie/.conda/envs/deepmd_cpu114/lib/python3.7/site-packages/deepmd/calculator.py", line 44, in calculate
    atype = [self.type_dict[k] for k in symbols]
  File "/home/liushiLab/zhouwenjie/.conda/envs/deepmd_cpu114/lib/python3.7/site-packages/deepmd/calculator.py", line 44, in <listcomp>
    atype = [self.type_dict[k] for k in symbols]
KeyError: 'In'

I have done a search with the keyword "ase". But, I actually didn't see a similar error.

[njzjz]

You didn't give type_map in the model. In this situation, you need to set type_dict argument when creating an ASE calculator.

[ShuSOfficial]

Where do I need to add the type_dict argument ? Can you give an example ?

[njzjz]

Please read documentation https://docs.deepmodeling.com/projects/deepmd/en/master/api_py/deepmd.html#deepmd.calculator.DP