Closed dobraczka closed 3 years ago
I have proposed a simple fix I have been using for this problem in my pull request.
Thank you for highlighting this!
I actually wanted to support this functionality a while ago, but then forgot about it. I've just fixed it all in b70fc9f. You can now restrict the number of simultaneous jobs per job array with the max_simultaneous_jobs
config option. And we now correctly handle job arrays with %X
.
I also got this error, I use newer version of seml (see my versions in https://github.com/TUM-DAML/seml/issues/107).
It would be great if you could open a new issue for this and provide more context on your error message. Your error can't be due to the same underlying bug since it has long been fixed.
Expected Behavior
Experiments are running with SEML, while other seperate SLURM throttled job arrays are running.
seml db status
should return the statusActual Behavior
An error is thrown
Steps to Reproduce the Problem
seml [db_name] status
Error message:
Specifications
Details
- Version: 0.3.4 - Python version: Python 3.7.1 - Platform: CentOS 7 - Anaconda environment (`conda list`): -