Kill joblib code in python 3

Question

I have used joblib module in python 3.6.0 to parallelize my code and ran it on my Linux machine. It has now started the process on 20 threads, as expected. But I am unable to kill all of them at once. For instance, if I kill any of the same 20 processes in matlab, the whole thing gets killed. Is there any such way in python 3?

EDIT: My code looks something like this

import scipy, numpy     # All the required libraries
from joblib import Parallel, delayed
# Required parameters
a=1
b=2 ...
if __name__ == '__main__':
    Parallel(n_jobs=20)(delayed(<some_function>)((a,b,j)) for j in range(20))

The some_function uses the parameters and saves using scipy.savetxt command.