Error running lammps

[Xi-yuanWang]

I successfully compile the lammps stable_7Aug2019 and its deepmd-kit module , but when I try to use dpgen, lammps throw the following error.

LAMMPS (3 Mar 2020)
OMP_NUM_THREADS environment is not set. Defaulting to 1 thread. (src/comm.cpp:94)
using 1 OpenMP thread(s) per MPI task
Reading data file ...
triclinic box = (0 0 0) to (9.95439 9.80314 9.83274) with tilt (0.248331 0.192711 -0.147738)
1 by 1 by 1 MPI processor grid
reading atoms ...
5 atoms
read_data CPU = 0.000724077 secs
Changing box ...
triclinic box = (0 0 0) to (9.95439 9.80314 9.83274) with tilt (0.248331 0.192711 -0.147738)
Summary of lammps deepmd module ...

Info of deepmd-kit:
installed to: /nfs-share/home/1800011848/miniconda3
source: v1.1.4-210-gbc01883
source brach: HEAD
source commit: bc01883
source commit at: 2020-05-23 23:34:57 +0800
build float prec: double
build with tf inc: /nfs-share/home/1800011848/miniconda3/include;/nfs-share/home/1800011848/miniconda3/include
build with tf lib: /nfs-share/home/1800011848/miniconda3/lib/libtensorflow_cc.so;/nfs-share/home/1800011848/miniconda3/lib/libtensorflow_framework.so
set tf intra_op_parallelism_threads: 0
set tf inter_op_parallelism_threads: 0
Info of lammps module:
use deepmd-kit at: /nfs-share/home/1800011848/miniconda3
source: v1.1.4-210-gbc01883
source branch: HEAD
source commit: bc01883
source commit at: 2020-05-23 23:34:57 +0800
build float prec: double
build with tf inc: /nfs-share/home/1800011848/miniconda3/include;/nfs-share/home/1800011848/miniconda3/include
build with tf lib: /nfs-share/home/1800011848/miniconda3/lib/libtensorflow_cc.so;/nfs-share/home/1800011848/miniconda3/lib/libtensorflow_framework.so
2020-09-03 07:59:11.057731: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 AVX512F FMA
2020-09-03 07:59:11.079920: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2600000000 Hz
2020-09-03 07:59:11.082976: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55c7aff28cd0 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-09-03 07:59:11.083037: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version
Invalid argument: Input 0 of node DescrptSeA was passed float from Reshape_2:0 incompatible with expected double.

The input file is
variable NSTEPS equal 300
variable THERMO_FREQ equal 10
variable DUMP_FREQ equal 10
variable TEMP equal 100.000000
variable PRES equal -1.000000
variable TAU_T equal 0.100000
variable TAU_P equal 0.500000

units metal
boundary p p p
atom_style atomic

neighbor 1.0 bin

box tilt large
read_data conf.lmp
change_box all triclinic
mass 1 1.000000
mass 2 12.000000
pair_style deepmd ../graph.001.pb ../graph.002.pb ../graph.003.pb ../graph.000.pb out_freq ${THERMO_FREQ} out_file model_devi.out
pair_coeff

thermo_style custom step temp pe ke etotal press vol lx ly lz xy xz yz
thermo ${THERMO_FREQ}
dump 1 all custom ${DUMP_FREQ} traj/*.lammpstrj id type x y z

velocity all create ${TEMP} 681862
fix 1 all nvt temp ${TEMP} ${TEMP} ${TAU_T}

timestep 0.002000
run ${NSTEPS}

[njzjz]

You are using the double precision LAMMPS with a float precision model. If the model is not what you want, please check the installation of deepmd-kit.

[Xi-yuanWang]

Then how to control the precision of a model or which model is in double precision?

[njzjz]

I guess you installed with conda install? If you executed conda install deepmd-kit=*=*cpu -c deepmodeling, it's double precision; if you executed conda install deepmd-kit=*=*cpu_low -c deepmodeling, that's float precision.

If you used pip to install and didn't change setup.py, it should be double precision:

deepmd-kit/setup.py

Line 58 in 567bcff

'-DFLOAT_PREC:STRING=high',

[konokcbhowmik]

Sir, I am facing the problem in windows 10.
OMP_NUM_THREADS environment is not set. Defaulting to 1 thread. (src/comm.cpp:98)
using 1 OpenMP thread(s) per MPI task

What should I do?

[Xi-yuanWang]

It works，thank you.