joblib-progress 1.0.3

A contextmanager to track progress of joblib execution

joblib-progress

A contextmanager to track progress of joblib execution using rich.progress.

Why

The vanilla multiprocessing does not work when an object to multiprocess is not pickle-able. The joblib solves this, but then its progress is not tracked nicely. This library solves that tracking issue with joblib.

Install

> pip install joblib-progreess

Usage

If you know the number of items

import time

from joblib import Parallel, delayed
from joblib_progress import joblib_progress


def slow_square(i):
    time.sleep(i / 2)
    return i ** 2

with joblib_progress("Calculating square...", total=10):
    Parallel(n_jobs=4)(delayed(slow_square)(number) for number in range(10))

If you don't know the number of items

with joblib_progress("Calculating square..."):
    Parallel(n_jobs=4)(delayed(slow_square)(number) for number in range(10))

Acknowledgments

The idea of using joblib.parallel.BatchCompletionCallBack is referenced from https://stackoverflow.com/a/58936697/5133167