ERROR: KIM Model name not found

OpenKIM
1

I ran the lammps simulation on the supercomputer platform and now want to use the following potential function by openkim
EDIP_LAMMPS_Marks_2000_C__MO_374144505645_000

kim-api and driven models were installed in the conda environment, exactly as recommended on the website

$ conda install -c conda-forge kim-api=2.3.0
kim-api-collections-management install user EDIP_LAMMPS_Marks_2000_C__MO_374144505645_000

Then I wrote a short input file for testing, indicating that the model file could not be found

  using 1 OpenMP thread(s) per MPI task
kim init EDIP_LAMMPS_Marks_2000_C__MO_374144505645_000 metal
#=== KIM is looking for 'Portable Models' in these directories ===
#  1: .
#  2: /public/home/jujingsheng/.kim-api/portable-models-dir
#  3: /public/software/apps/lammps/intelmpi/29Sep2021/external_libs/lib/kim-api/portable-models
#=== KIM is looking for 'Simulator Models' in these directories ===
#  1: .
#  2: /public/home/jujingsheng/.kim-api/simulator-models-dir
#  3: /public/software/apps/lammps/intelmpi/29Sep2021/external_libs/lib/kim-api/simulator-models
ERROR: KIM Model name not found (src/KIM/kim_init.cpp:234)
Last command: kim init EDIP_LAMMPS_Marks_2000_C__MO_374144505645_000 metal

Could you please tell me which step is wrong?
Thank you so much!
Here is the input file, the scheduling file, and the output kim.log

in.lammps (816 Bytes)
kim.log (25.7 KB)
slurmLammps.slurm (565 Bytes)

2

Hi!

My first suspicion is that you accidentally ended up with two installations of the KIM API. Please see Multiple KIM installations or cannot find libkim-api.so for info, and let me know if you need additional help!

Cheers, ilia

3

Hi, thank you for response!
I tried adding paths in ~/.bashrc and in the load script, respectively

vi  ~/.bashrc
export LD_LIBRARY_PATH=/public/home/jujingsheng/anaconda3/bin:${PATH}
source  bashrc

In addition, add directly to the scheduling file

#!/bin/bash
#SBATCH -J lammps
#SBATCH -N 1
#SBATCH --ntasks-per-node=48
#SBATCH -p short

module purge
**export LD_LIBRARY_PATH=/public/home/jujingsheng/anaconda3/bin/kim-api-collections-management**
module load compiler/intel/2021.3.0
module load mpi/intelmpi/2021.3.0
...

But the result of the error is the same as before:

LAMMPS (29 Sep 2021 - Update 3)
OMP_NUM_THREADS environment is not set. Defaulting to 1 thread. (src/comm.cpp:98)
  using 1 OpenMP thread(s) per MPI task
kim init EDIP_LAMMPS_Marks_2000_C__MO_374144505645_000 metal
#=== KIM is looking for 'Portable Models' in these directories ===
#  1: .
#  2: /public/home/jujingsheng/.kim-api/portable-models-dir
#  3: /public/software/apps/lammps/intelmpi/29Sep2021/external_libs/lib/kim-api/portable-models
#=== KIM is looking for 'Simulator Models' in these directories ===
#  1: .
#  2: /public/home/jujingsheng/.kim-api/simulator-models-dir
#  3: /public/software/apps/lammps/intelmpi/29Sep2021/external_libs/lib/kim-api/simulator-models
ERROR: KIM Model name not found (src/KIM/kim_init.cpp:234)
Last command: kim init EDIP_LAMMPS_Marks_2000_C__MO_374144505645_000 metal

So help is still needed :pray:
list.txt (2.5 KB)

4

Hi,

The LD_LIBRARY_PATH should end in /lib, not /bin or /bin/kim-api-collections-management, so can you try

/public/home/jujingsheng/anaconda3/lib
5

Dear ilia：
I successfully ran the built-in model of the supercomputer:

`Sim_LAMMPS_LJcut_AkersonElliott_Alchemy_PbAu.`

I added the following environment variables to ~/.bashrc and then ‘souce’

export LD_LIBRARY_PATH=/public/home/jujingsheng/anaconda3/lib

Then I changed the init to the previously imported model, and the same error occurred. I want to know if there is any other step that might have gone wrong :pray: :pray:

7

After I modified the environment variable, it can run successfully, thanks for the previous advice!

LD_LIBRARY_PATH=/public/home/jujingsheng/anaconda3/lib:$LD_LIBRARY_PATH

1 Like