Error while using sys_probs

[krishnapitike]

Hello, I see the following error when I use sys_probs to manually set the probabilities of the data. The sum of all probabilities is equal to 1. I see the following error. The input file is provided at the end. Thanks in advance for your help.

error1:
File "~/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/utils/data_system.py", line 480, in _process_sys_probs
assert assigned_sum_prob <= 1, "the sum of assigned probability should be less than 1"
AssertionError: the sum of assigned probability should be less than 1

However if I set the total sum of probabilities to slightly smaller than 1, then I see the following error:
error2:
File "~/anaconda3/envs/deepmd/lib/python3.9/site-packages/deepmd/utils/data_system.py", line 488,
in _process_sys_probs
assert np.sum(ret_prob) == 1, "sum of probs should be 1"
AssertionError: sum of probs should be 1

input file:
{
"_comment": " model parameters",
"model": {
"type_map": ["Fe"],
"descriptor" :{
"type": "se_e2_a",
"sel": [128],
"rcut_smth": 0.50,
"rcut": 6.40,
"neuron": [25, 50, 100, 200],
"resnet_dt": false,
"axis_neuron": 16,
"seed": 1,
"_comment": " end descriptor"
},
"fitting_net" : {
"neuron": [240, 240, 240, 240],
"resnet_dt": true,
"seed": 2,
"_comment": " end fitting_net"
},
"_comment": " end model"
},

"learning_rate" :{
"type":			"exp",
"decay_steps":		5000,
"start_lr":		0.001,
"stop_lr":		3.51e-8,
"_comment":		" end learning_rate"
},

"loss" :{
"type":			"ener",
"start_pref_e":		0.02,
"limit_pref_e":		1,
"start_pref_f":		1000,
"limit_pref_f":		1,
"start_pref_v":		0.2,
"limit_pref_v":		1,
"_comment":		" end loss"
},

"training" : {
"training_data": {
    "systems":		[
                             "data_npy/Fe19_tr/",
                             "data_npy/Fe37_tr/",
                             "data_npy/Fe30_tr/",
                             "data_npy/Fe34_tr/",
                             "data_npy/Fe40_tr/",
                             "data_npy/Fe129_tr/",
                             "data_npy/Fe16_tr/",
                             "data_npy/Fe32_tr/",
                             "data_npy/Fe45_tr/",
                             "data_npy/Fe54_tr/",
                             "data_npy/Fe25_tr/",
                             "data_npy/Fe15_tr/",
                             "data_npy/Fe126_tr/",
                             "data_npy/Fe20_tr/",
                             "data_npy/Fe127_tr/",
                             "data_npy/Fe13_tr/",
                             "data_npy/Fe42_tr/",
                             "data_npy/Fe41_tr/",
                             "data_npy/Fe17_tr/",
                             "data_npy/Fe36_tr/",
                             "data_npy/Fe22_tr/",
                             "data_npy/Fe21_tr/",
                             "data_npy/Fe43_tr/",
                             "data_npy/Fe33_tr/",
                             "data_npy/Fe27_tr/",
                             "data_npy/Fe26_tr/",
                             "data_npy/Fe18_tr/",
                             "data_npy/Fe128_tr/",
                             "data_npy/Fe23_tr/",
                             "data_npy/Fe31_tr/",
                             "data_npy/Fe2_tr/",
                             "data_npy/Fe29_tr/",
                             "data_npy/Fe38_tr/",
                             "data_npy/Fe39_tr/",
                             "data_npy/Fe12_tr/",
                             "data_npy/Fe24_tr/",
                             "data_npy/Fe44_tr/",
                             "data_npy/Fe28_tr/",
                             "data_npy/Fe11_tr/",
                             "data_npy/Fe14_tr/",
                             "data_npy/Fe35_tr/",
                             "data_npy/Fe130_tr/"
                            ],
        "sys_probs":        [
                             0.01 ,
                             0.02 ,
                             0.01 ,
                             0.01 ,
                             0.02 ,
                             0.1 ,
                             0.01 ,
                             0.01 ,
                             0.02 ,
                             0.2 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.1 ,
                             0.01 ,
                             0.02 ,
                             0.02 ,
                             0.01 ,
                             0.02 ,
                             0.01 ,
                             0.01 ,
                             0.02 ,
                             0.02 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.02 ,
                             0.02 ,
                             0.01 ,
                             0.01 ,
                             0.02 ,
                             0.01 ,
                             0.01 ,
                             0.01 ,
                             0.02 ,
                             0.1
                            ],
        "batch_size":       1,
    "_comment":		" end training_data"
},
"validation_data":{
    "systems":		[
                             "data_npy/Fe19_te/",
                             "data_npy/Fe37_te/",
                             "data_npy/Fe30_te/",
                             "data_npy/Fe34_te/",
                             "data_npy/Fe40_te/",
                             "data_npy/Fe129_te/",
                             "data_npy/Fe16_te/",
                             "data_npy/Fe32_te/",
                             "data_npy/Fe45_te/",
                             "data_npy/Fe54_te/",
                             "data_npy/Fe25_te/",
                             "data_npy/Fe15_te/",
                             "data_npy/Fe126_te/",
                             "data_npy/Fe20_te/",
                             "data_npy/Fe127_te/",
                             "data_npy/Fe13_te/",
                             "data_npy/Fe42_te/",
                             "data_npy/Fe41_te/",
                             "data_npy/Fe17_te/",
                             "data_npy/Fe36_te/",
                             "data_npy/Fe22_te/",
                             "data_npy/Fe21_te/",
                             "data_npy/Fe43_te/",
                             "data_npy/Fe33_te/",
                             "data_npy/Fe27_te/",
                             "data_npy/Fe26_te/",
                             "data_npy/Fe18_te/",
                             "data_npy/Fe128_te/",
                             "data_npy/Fe23_te/",
                             "data_npy/Fe31_te/",
                             "data_npy/Fe2_te/",
                             "data_npy/Fe29_te/",
                             "data_npy/Fe38_te/",
                             "data_npy/Fe39_te/",
                             "data_npy/Fe12_te/",
                             "data_npy/Fe24_te/",
                             "data_npy/Fe44_te/",
                             "data_npy/Fe28_te/",
                             "data_npy/Fe11_te/",
                             "data_npy/Fe14_te/",
                             "data_npy/Fe35_te/",
                             "data_npy/Fe130_te/"
                            ],
    "batch_size":	1,
    "numb_btch":	1,
    "_comment":		" end validation_data"
},
"numb_steps":		2000000,
"seed":			3,
"disp_file":		"lcurve.out",
"disp_freq":		100,
"save_freq":		1000,
"_comment":		" end training"
},

"_comment":			" end input.json"

}

[njzjz]

This has been fixed in #1919.