deepmd-kit-2.1.2-cpu can't find pair style deepmd?

[Xian-qi]

I installed deepmd-kit-2.1.2-cpu-Linux-x86_64.sh, but i find a error lammps can't find pair style deepmd ? I use lmp -h, can't see it on the list

[lnnbig]

I also found this bug. It seems that USER-DEEPMD has not been included in this version of LAMMPS.

[Lewis-YL]

Have you solved this problem?

[njzjz]

Please activate the conda environment first:

conda activate /path/to/deepmd-kit

For more information, read conda documentation Activating an environment.

[Xian-qi]

I‘m sure the env is activated.

[njzjz]

What's the output of echo $LAMMPS_PLUGIN_PATH after the env is activated? Does LAMMPS log something like plugins loaded?

[Xian-qi]

echo $LAMMPS_PLUGIN_PATH : /XXXX/XX/deepmd-kit-2.1.2/deepmd-kit/lib/deepmd_lmp， and i found the Installed packages not include USER-DEEPMD when i use lmp -h

[njzjz]

You should be able to use pair_style deepmd when LAMMPS_PLUGIN_PATH is set.

The lmp -h message is maintained by LAMMPS and not what we maintain. We can't do anything on our side.

[Xian-qi]

Yes, I can use deepmd now. Thank you for your help.

[Lewis-YL]

How do you solve the problem in the end?

[chtchelkatchev]

I got the same error in deepmd 2.2.3 installed from stand alone distribution. Conda activate has been done.
echo $LAMMPS_PLUGIN_PATH gives correct path /s/ls4/users/nms/SOFT_new_centos/deepmd-kit-2.2.3-cuda11.6_gpu/lib/deepmd_lmp

---

Message when trying run lammps:
using 1 OpenMP thread(s) per MPI task
Open of file /s/ls4/users/nms/SOFT_new_centos/deepmd-kit-2.2.3-cuda11.6_gpu/lib/deepmd_lmp/dpplugin.so failed: /s/ls4/users/nms/SOFT_new_centos/deepmd-kit-2.2.3-cuda11.6_gpu/lib/deepmd_lmp/dpplugin.so: undefined symbol: MPI_Bcast
Loaded 0 plugins from /s/ls4/users/nms/SOFT_new_centos/deepmd-kit-2.2.3-cuda11.6_gpu/lib/deepmd_lmp
Reading data file ...
orthogonal box = (0 0 0) to (10.024128 10.024128 10.024128)
1 by 1 by 1 MPI processor grid
reading atoms ...
512 atoms

---

ERROR: Unrecognized pair style 'deepmd' (src/force.cpp:271)

[njzjz]

Hi @chtchelkatchev, this error was reported in #2739 and has been fixed. Please re-download the files that were released yesterday.

[chtchelkatchev]

Thank You! Everything works well now.