cumulator 0.0.5

a tool to quantify and communicate the carbon footprint of machine learning methods

An API to evaluate the carbon footprint of computation and communcation of machine learning methods (or any other computing system).

Background

Current researchers produce methods and tools to evaluate and optimize the efficiency (from both time and spatial scales) of large-scale ML computations.

Aim

Raise awareness about the carbon footprint of machine learning methods and to encourage further optimization and the rationale use of AI-powered tools

Method

Create Cumulator, a simple API to evaluate the carbon footprint of communication and computation of a machine learning models which provides effortless integration within any python framework.

• Free software: MIT license

Installation

Use the following command:

pip install cumulator

Functionalities

At the moment Cumulator has the following functionalities:

• Chronometer activation and deactivation

• Time aggregation (cumulative time of activation/deactivation) per instance of the cumulator class

• Display of the carbon footprint

Hence, to compare n different network topologies, one can create n cumulator instance and display the relative carbon footprint after computation.

Use cases

Cumulator was integrated within the Alg-E platform

ChangeLog

• 07.06.2020: 0.0.2 added communication costs and cleaned src/

• 21.05.2020: 0.0.1 deployment on PypI and integration with Alg-E

Links

• GitHub: https://github.com/tristantreb/cumulator

• PyPI: https://pypi.org/project/cumulator/

Changelog

0.0.0 (2020-05-14)

• First release on PyPI.