About machine.json using lsf

DM0815 commented 1 year ago

When I use lsf queue system to conduct dpgen in logining node of server cluster.After submitting the command,it reminds "RuntimeError: Meet errors will handle unexpected submission state." and suggest me to see the remote_root.But there are no mistake information in work dir. And in dp task dir, the jobs is still runing, the train.log is ok. And I can the jobs in queue system. I don't know where wrong, can you give me some hints. machine.jsons and mistake informarion attached.

machine.json: { "api_version": "1.0", "_deepmd_version": "2.1.0", "train" : { "command": "dp", "machine": { "batch_type": "LSF", "context_type": "local", "local_root" : "./", "remote_root":"/public/home/dmeng/DPGEN/0316testlsf/tmp" }, "resources": { "number_node": 1, "cpu_per_node": 8, "gpu_per_node": 0, "queue_name":"normal", "group_size": 2, "_batch_type": "LSF", "_kwargs": {}, "source_list":["/public/home/dmeng/anaconda3/bin/activate deepmd"] } }, "model_devi": { "command": "lmp -i input.lammps -v restart 0", "machine": { "batch_type": "LSF", "context_type": "local", "local_root" : "./", "remote_root":"/public/home/dmeng/DPGEN/0316testlsf/tmp"

  },
  "resources": {
    "number_node": 1,
    "cpu_per_node": 8,
    "gpu_per_node": 0,
     "queue_name":"normal",
    "group_size": 100,
    "_batch_type": "LSF",
     "_kwargs": {},
    "source_list":["/public/home/dmeng/anaconda3/bin/activate deepmd"]
  }
},

"fp": { "command": "ulimit -s unlimited && mpirun -n 8 /public/home/dmeng/softwares/vasp.5.4/bin/vasp_std", "machine": { "batch_type": "LSF", "context_type": "local", "local_root" : "./", "remote_root":"/public/home/dmeng/DPGEN/0316testlsf/tmp" }, "resources": { "number_node": 1, "cpu_per_node": 8, "gpu_per_node": 0, "queue_name":"normal", "group_size": 50, "_batch_type": "LSF", "_kwargs": {}, "source_list": ["/public/softwares/intel/oneapi/setvars.sh"] } } }