Parallel Progress

Need to keep an eye on the progress of parallelized tasks? If you’re using Python’s concurrent.futures module, one way to do it is with tqdm, a nice package for generating progress bars. Here’s how it looks.

Progress bar indicating completion of parallel tasks

And here’s the code for

from concurrent.futures import ProcessPoolExecutor, as_completed
import time

from tqdm import tqdm

def nap():

def main():
    with ProcessPoolExecutor() as executor:
        futures = [executor.submit(nap) for i in range(100)]

        kwargs = {
            'total': len(futures),
            'unit': 'nap',
            'unit_scale': True,
            'leave': True

        for f in tqdm(as_completed(futures), **kwargs):

if __name__ == '__main__':

Since as_completed() does not provide a __len__ method yielding the count of Future instances to be executed, we need to tell tqdm how much work we expect to be done using the total parameter.