This project was born out of an apparent lack of a library for job submission to slurm clusters. We sought to write a simple, low-level library with an API that mirrors the slurm sbatch
command line tool. Plurmy is the result of that.
We envision this library being used primarily through the Slurm
class. This class uses attributes, named identically to full length #SBATCH
arguments. For example:
from plurmy import Slurm
command = 'srun -n 3 python my_super_script.py"
myjob = Slurm(time="00:10:00", partition="research", ntask=3)
myjob.submit()