If you need to run multiple single core jobs, and don't want to waste your budget or computing resources by running a single core job on a whole compute node, you can create a slurm batch script something like this. In this case we are going to run a serial python spplications, with every python app in its own directory.

!/bin/bash 
#SBATCH --job-name=calculation # job name 
#SBATCH -o calc.out # output file name for all the single core jobs
#SBATCH -n 4  # 4 cores
#SBATCH --ntasks-per-node=4

module load Python 

srun -n1 --exclusive python directory1/calculation1.py &
srun -n1 --exclusive python directory2/calculation2.py &
srun -n1 --exclusive python directory3/calculation3.py &
srun -n1 --exclusive python directory4/calculation4.py &

wait