/python-serialization-benchmark

Python single big object serialization benchmark

Primary LanguagePythonMIT LicenseMIT

Python 3.8.11 serialization benchmark

This repo is a companion to my blog post Read it for more context.

Create a new venv with python 3.8.11 and install the deps with

poetry install

Run the benchmark with

pytest --benchmark-save=results.txt --benchmark-save-data

it will generate the results in .benchmarks dir, as a JSON.

My results, generated with the above commands ( and transformed from json to a csv) are in this csv file. Mind that the columns mention following numbers of fields: single, 100, 1k ... while the blog post mentions 4, 400, 4k ... the 4x is correct, 1, 100 1k .. where the input to the generator returning the test data and pytest used it by default as the column names.