smartmultiprocessing 0.0.1

Memory-managed parallel computing for medium-sized computational tasks.

SmartMultiprocessing

A drop-in replacement for Python's multiprocessing library with many extra features, including: memory management, smart task scheduling, a pause button, a GUI, and more.

SmartMultiprocessing is ideal for use cases beyond typical multiprocessing use, like:

• Running code where memory usage is different depending on input values
• Benchmarking experimental code where memory usage is not well understood
• Running code on a machine where it's important to be able to pause execution/change thread and memory use on the fly
• Being able to monitor the performance of code without having to write your own monitoring library from scratch

SmartMultiprocessing is in active development!

This project is a spin-out from code that @emilyhunt wrote during her PhD. It's currently in alpha / active development - check back here soon for the first production-ready versions!