When running dpgen, the dp command always goes into sleep mode

[SEU-NiuWenLong]

When I run the methane example given on the official website, if I set the number of training steps in the parameter file to tens of thousands of steps, the dp command will go to sleep and not work. Why? And the CPU of the server I use is 24 cores. What I set up in the machine settings file is all 24 cores, but the CPU utilization is very low.
参数文件如下：
{
"type_map": ["H","C"],
"mass_map": [1,12],
"init_data_prefix": "../",
"init_data_sys": ["init/CH4.POSCAR.01x01x01/02.md/sys-0004-0001/deepmd"],
"sys_configs_prefix": "../",
"sys_configs": [
["init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale-1.000/00000*/POSCAR"],
["init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale-1.000/00001*/POSCAR"],
["init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale-1.000/00002*/POSCAR"],
["init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale-1.000/00001*/POSCAR"],
["init/CH4.POSCAR.01x01x01/01.scale_pert/sys-0004-0001/scale-1.000/00002*/POSCAR"]
],
"_comment": " that's all ",
"numb_models": 4,
"default_training_param": {
"model": {
"type_map": ["H","C"],
"descriptor": {
"type": "se_a",
"sel": [16,4],
"rcut_smth": 0.5,
"rcut": 5.0,
"neuron": [120,120,120],
"resnet_dt": true,
"axis_neuron": 12,
"seed": 1
},
"fitting_net": {
"neuron": [25,50,100],
"resnet_dt": false,
"seed": 1
}
},
"learning_rate": {
"type": "exp",
"start_lr": 0.001,
"decay_steps": 5000
},
"loss": {
"start_pref_e": 0.02,
"limit_pref_e": 2,
"start_pref_f": 1000,
"limit_pref_f": 1,
"start_pref_v": 0.0,
"limit_pref_v": 0.0
},
"training": {
"stop_batch": 20000,
"disp_file": "lcurve.out",
"disp_freq": 1000,
"numb_test": 4,
"save_freq": 1000,
"save_ckpt": "model.ckpt",
"disp_training": true,
"time_training": true,
"profiling": false,
"profiling_file": "timeline.json",
"_comment": "that's all"
}
},
"model_devi_dt": 0.002,
"model_devi_skip": 0,
"model_devi_f_trust_lo": 0.05,
"model_devi_f_trust_hi": 0.15,
"model_devi_e_trust_lo": 10000000000.0,
"model_devi_e_trust_hi": 10000000000.0,
"model_devi_clean_traj": true,
"model_devi_jobs": [
{"sys_idx": [0],"temps": [100],"press": [1.0],"trj_freq": 10,"nsteps": 300,"ensemble": "nvt","_idx": "00"},
{"sys_idx": [1],"temps": [100],"press": [1.0],"trj_freq": 10,"nsteps": 3000,"ensemble": "nvt","_idx": "01"},
{"sys_idx": [2],"temps": [100],"press": [1.0],"trj_freq": 10,"nsteps": 3000,"ensemble": "nvt","_idx": "02"},
{"sys_idx": [3],"temps": [100],"press": [1.0],"trj_freq": 10,"nsteps": 3000,"ensemble": "nvt","_idx": "03"},
{"sys_idx": [4],"temps": [100],"press": [1.0],"trj_freq": 10,"nsteps": 3000,"ensemble": "nvt","_idx": "04"}
],
"fp_style": "vasp",
"shuffle_poscar": false,
"fp_task_max": 30,
"fp_task_min": 5,
"fp_pp_path": "./",
"fp_pp_files": ["POTCAR_H","POTCAR_C"],
"fp_incar": "./INCAR_methane"
}
{
"api_version": "1.0",
"deepmd_version": "2.2.1",
"train": [
{
"command": "dp",
"machine": {
"batch_type": "Shell",
"context_type": "local",
"local_root": "./",
"remote_root": "/home/gcniu/work/deepmd/dpgen_example/run/temp"
},
"resources": {
"number_node": 1,
"cpu_per_node": 24,
"gpu_per_node": 0,
"group_size": 4
}
}
],
"model_devi": [
{
"command": "mpirun -np 24 lmp -i input.lammps",
"machine": {
"batch_type": "Shell",
"context_type": "local",
"local_root": "./",
"remote_root": "/home/gcniu/work/deepmd/dpgen_example/run/temp"
},
"resources": {
"number_node": 1,
"cpu_per_node": 24,
"gpu_per_node": 0,
"group_size": 4
}
}
],
"fp": [
{
"command": "mpirun -np 24 vasp_std >& log",
"machine": {
"batch_type": "Shell",
"context_type": "local",
"local_root": "./",
"remote_root": "/home/gcniu/work/deepmd/dpgen_example/run/temp"
},
"resources": {
"number_node": 1,
"cpu_per_node": 24,
"gpu_per_node": 0,
"group_size": 1
}
}
]
}

[SEU-NiuWenLong]

such as 9969 gcniu 20 0 5275656 252368 106908 R 33.6 0.1 6:47.56 lmp ,the utilization rate of CPU is 33% at this time.

[njzjz]

The cpu_per_node only controls the resources allocated from the batch system (and Shell does not support this feature). You still need to control the usage of the program manually. For DeePMD-kit, see https://deepmd.rtfd.io/parallelism/ for details.