The Carbon footprint of Machine Learning Models

Abstract

Machine Learning models are growing increasingly powerful in their abilities, whether that might be in processing natural language, tackling the intricacies of computer vision or any other number of exciting application that are emerging . But the environmental impact of machine learning models is increasingly receiving attentions. Here ,the works to focus on the carbon footprint of language models, as these models grow larger and larger, do their corresponding carbon footprints, especially when it comes to creating and training complex models. Here we will take a look at some concrete example of carbon emissions from machine learning models, will present tools that can be used to estimate the carbon footprint of a machine learning models. Finally present ideas for how to reduce the carbon footprint.