How do I go about setting input.json when each type_map in my dataset is different from type_map.raw?

[ZKNB]

For example,I have dataset1(type_map = [H Pt], type_map = [0 1]), another dataset2(type_map = [H C O Pt], type_map = [0 1 2 3]),their indexes don't correspond. How to write "type_map" and "training":{"systems"} in the "input.json"?

Here are two dataset

type_map.raw	type.raw
H Pt	0 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
H O	0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 2 2 2 2 2 2 2

The follow is my input.json

{
    "_comment": "that's all",
    "model": {
        "type_map": [
            "H",
            "C",
            "O",
            "Pt"
        ],
        "descriptor": {
            "type": "se_e2_a",
            "sel": [
                4,
                1
            ],
            "rcut_smth": 0.5,
            "rcut": 6.0,
            "neuron": [
                10,
                20,
                40
            ],
            "resnet_dt": false,
            "axis_neuron": 4,
            "seed": 1,
            "_comment": " that's all"
        },
        "fitting_net": {
            "neuron": [
                100,
                100,
                100
            ],
            "resnet_dt": true,
            "seed": 1,
            "_comment": " that's all"
        },
        "_comment": " that's all"
    },
    "learning_rate": {
        "type": "exp",
        "decay_steps": 5000,
        "start_lr": 0.001,
        "stop_lr": 3.51e-08,
        "_comment": "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,
        "_comment": " that's all"
    },
    "training": {
        "training_data": {
            "systems": [
                "../00.data/training_data1/",
                "../00.data/training_data2/"
            ],
            "batch_size": "auto",
            "_comment": "that's all"
        },
        "validation_data": {
            "systems": [
                "../00.data/validation_data"
            ],
            "batch_size": "auto",
            "numb_btch": 1,
            "_comment": "that's all"
        },
        "numb_steps": 1000000,
        "seed": 10,
        "disp_file": "lcurve.out",
        "disp_freq": 1000,
        "save_freq": 10000,
        "_comment": "that's all"
    }
}

Is it reasonable?

when I run dp train input.json it raise error as following

DEEPMD INFO    Calculate neighbor statistics... (add --skip-neighbor-stat to skip this step)
Traceback (most recent call last):
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/bin/dp", line 10, in <module>
    sys.exit(main())
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/entrypoints/main.py", line 473, in main
    train_dp(**dict_args)
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/entrypoints/train.py", line 94, in train
    jdata = update_sel(jdata)
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/entrypoints/train.py", line 346, in update_sel
    descrpt_data = update_one_sel(jdata, descrpt_data)
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/entrypoints/train.py", line 320, in update_one_sel
    tmp_sel = get_sel(jdata, rcut)
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/entrypoints/train.py", line 283, in get_sel
    _, max_nbor_size = get_nbor_stat(jdata, rcut)
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/entrypoints/train.py", line 258, in get_nbor_stat
    train_data = get_data(jdata["training"]["training_data"], max_rcut, type_map, None)
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/entrypoints/train.py", line 202, in get_data
    data = DeepmdDataSystem(
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/utils/data_system.py", line 79, in __init__
    DeepmdData(
  File "/export/home/liluotonggpu2/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/utils/data.py", line 57, in __init__
    assert(len(self.type_map) >= max(self.atom_type)+1)
AssertionError

[njzjz]

The type.raw file should map to the type_map.raw file when the latter is provided.

[ZKNB]

Do you mean change the chart like this?

type_map.raw	type.raw
H Pt	0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
H O	0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1

[ZKNB]

Thank you for your reply. I'm sorry I didn't express my question clearly.
I have change my dataset type.raw and the type_map.raw like this

type_map.raw	type.raw
H Pt	0 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
H O	0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 2 2 2 2 2 2 2

It's counterparts are [H-0,C-1,O-2,Pt-3].
But when I 'dp train input.json',they produce the above error.
Is my input.json setting wrong or somethings else?

[njzjz]

Do you mean change the chart like this?
type_map.raw type.raw
H Pt 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
H O 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1

This is correct.

[ZKNB]

Think you for your reply I have solve this problem.