Failed to use remote supercomputing resources

[Saberve]

When I run dpgen init_surf on the local server and submit the task to the remote supercomputer, the computing task is terminated。

The relevant modules have been installed on the supercomputer. But when I add the module load command to the machine.json file, the error message is repeated。

[HuangJiameng]

Maybe it is the reason of those errors?

[Saberve]

Thank you for your answer。But I checked the ID of the supercomputing user used and it is correct.
And the error message looks like looking for user and these modules on my local server.
it's confusing.

[HuangJiameng]

@KZHIWEI Could you please take a look? THX!

[AnguseZhang]

@Saberve This problem is not directly to DP-GEN. You should check the slurm scripts and try to find how to successfully run a job on your cluster manually, not via DP-GEN. And after you work out, you can modify the machine.json according to the experience.

[AnguseZhang]

One tip is that some clusters forbid the users to submit job from the computing node, and neither does DP-GEN require.