Value Error while training deepMD

[magnamad]

I created the raw files ( force.raw ,energy.raw , coord.raw ,type.raw , type_map.raw and nopbc) for each system and converted the force.raw , enrgy,raw and coord.raw to .npy files using raw_to_set.sh ( https://github.com/deepmodeling/deepmd-kit/blob/master/data/raw/raw_to_set.sh ) . While training using dp train input.json , I am facing the following error. Kindly , let me know where the issue lies. I have attached my input file in the end

Thanks ,
Maddy

DEEPMD INFO    build with tf lib:
DEEPMD INFO    ---Summary of the training---------------------------------------
DEEPMD INFO    running on:           Saavan-ib
DEEPMD INFO    computing device:     cpu:0
DEEPMD INFO    Count of visible GPU: 0
DEEPMD INFO    num_intra_threads:    0
DEEPMD INFO    num_inter_threads:    0
DEEPMD INFO    -----------------------------------------------------------------
DEEPMD INFO    ---Summary of DataSystem: training     -----------------------------------------------
DEEPMD INFO    found 193 system(s):
List of those systems
DEEPMD INFO    ---Summary of DataSystem: validation   -----------------------------------------------
DEEPMD INFO    found 176 system(s):
List of those systems
DEEPMD INFO    training without frame parameter
DEEPMD INFO    data stating... (this step may take long time)
DEEPMD ERROR   cannot reshape array of size 67 into shape (47,1)
DEEPMD ERROR   This error may occur when your label mismatch it's name, i.e. you might store global tensor in `atomic_tensor.npy` or atomic tensor in `tensor.npy`.
Traceback (most recent call last):
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/utils/data.py", line 557, in _load_data
    data = np.reshape(data, [nframes, ndof])
  File "<__array_function__ internals>", line 180, in reshape
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/numpy/core/fromnumeric.py", line 298, in reshape
    return _wrapfunc(a, 'reshape', newshape, order=order)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/numpy/core/fromnumeric.py", line 57, in _wrapfunc
    return bound(*args, **kwds)
ValueError: cannot reshape array of size 67 into shape (47,1)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/chemadh/softwares/deepmd-kit/bin/dp", line 10, in <module>
    sys.exit(main())
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/entrypoints/main.py", line 631, in main
    train_dp(**dict_args)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/entrypoints/train.py", line 164, in train
    _do_work(jdata, run_opt, is_compress)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/entrypoints/train.py", line 273, in _do_work
    model.build(train_data, stop_batch, origin_type_map=origin_type_map)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/train/trainer.py", line 501, in build
    self.model.data_stat(data)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/model/ener.py", line 104, in data_stat
    all_stat = make_stat_input(data, self.data_stat_nbatch, merge_sys=False)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/model/model_stat.py", line 45, in make_stat_input
    stat_data = data.get_batch(sys_idx=ii)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/utils/data_system.py", line 396, in get_batch
    b_data = self.get_batch_standard(sys_idx)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/utils/data_system.py", line 421, in get_batch_standard
    b_data = self.data_systems[self.pick_idx].get_batch(
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/utils/data.py", line 257, in get_batch
    self._load_batch_set(self.train_dirs[self.set_count % self.get_numb_set()])
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/utils/data.py", line 404, in _load_batch_set
    self.batch_set = self._load_set(set_name)
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/utils/data.py", line 452, in _load_set
    data["find_" + kk], data[kk] = self._load_data(
  File "/home/chemadh/softwares/deepmd-kit/lib/python3.10/site-packages/deepmd/utils/data.py", line 562, in _load_data
    raise ValueError(str(err_message) + ". " + explanation)
ValueError: cannot reshape array of size 67 into shape (47,1). This error may occur when your label mismatch it's name, i.e. you might store global tensor in `atomic_tensor.npy` or atomic tensor in `tensor.npy`.

This is my input file

{
  "model": {
    "type_map": [
      "A",
      "B",
      "C",
      "D",
      "E"
    ],
    "descriptor": {
      "type": "se_e2_a",
      "sel": "auto",
      "rcut": 6,
      "rcut_smth": 1,
      "neuron": [
        25,
        50,
        100
      ],
      "axis_neuron": 12,
      "activation_function": "tanh",
      "trainable": true,
      "seed": 1
    },
    "fitting_net": {
      "type": "ener",
      "neuron": [
        240,
        240,
        240
      ],
      "resnet_dt": true,
      "seed": 100
    }
  },
  "learning_rate": {
    "type": "exp",
    "start_lr": 0.001,
    "stop_lr": 3.51e-8,
    "decay_steps": 1000
  },
  "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
  },
  "training": {
    "training_data": {
      "systems": [
        "sorted_train/file1",

      ],
      "set_prefix": "set",
      "batch_size": "auto"
    },
    "validation_data": {
      "systems": [
        "sorted_validation/file1",
     
      ],
      "set_prefix": "set",
      "batch_size": [
        "auto"
      ],
      "numb_btch": 3
    },
    "numb_steps": 1000000,
    "seed": 42,
    "disp_file": "lcurve.out",
    "disp_freq": 100,
    "save_freq": 1000,
    "disp_training": true,
    "time_training": true,
    "profiling": true,
    "enable_profiler": true
  }
}

[wanghan-iapcm]

please check if the number of frames are consistent across all the .npy files.

[magnamad]

I had some issue while preprocessing to .raw files. Thank you for your input.