Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Models

Publikation: KonferencebidragPaperForskningfagfællebedømt

Standard

Carbontracker : Tracking and Predicting the Carbon Footprint of Training Deep Learning Models. / Anthony, Lasse F. Wolff; Kanding, Benjamin; Selvan, Raghavendra.

2020. Paper præsenteret ved ICML Workshop on "Challenges in Deploying and monitoring Machine Learning Systems".

Publikation: KonferencebidragPaperForskningfagfællebedømt

Harvard

Anthony, LFW, Kanding, B & Selvan, R 2020, 'Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Models', Paper fremlagt ved ICML Workshop on "Challenges in Deploying and monitoring Machine Learning Systems", 17/07/2020. <http://arxiv.org/pdf/2007.03051v1>

APA

Anthony, L. F. W., Kanding, B., & Selvan, R. (2020). Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Models. Paper præsenteret ved ICML Workshop on "Challenges in Deploying and monitoring Machine Learning Systems". http://arxiv.org/pdf/2007.03051v1

Vancouver

Anthony LFW, Kanding B, Selvan R. Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Models. 2020. Paper præsenteret ved ICML Workshop on "Challenges in Deploying and monitoring Machine Learning Systems".

Author

Anthony, Lasse F. Wolff ; Kanding, Benjamin ; Selvan, Raghavendra. / Carbontracker : Tracking and Predicting the Carbon Footprint of Training Deep Learning Models. Paper præsenteret ved ICML Workshop on "Challenges in Deploying and monitoring Machine Learning Systems".11 s.

Bibtex

@conference{2b01c02c2c4744c7bf2e1fe12917c488,
title = "Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Models",
abstract = " Deep learning (DL) can achieve impressive results across a wide variety of tasks, but this often comes at the cost of training models for extensive periods on specialized hardware accelerators. This energy-intensive workload has seen immense growth in recent years. Machine learning (ML) may become a significant contributor to climate change if this exponential trend continues. If practitioners are aware of their energy and carbon footprint, then they may actively take steps to reduce it whenever possible. In this work, we present Carbontracker, a tool for tracking and predicting the energy and carbon footprint of training DL models. We propose that energy and carbon footprint of model development and training is reported alongside performance metrics using tools like Carbontracker. We hope this will promote responsible computing in ML and encourage research into energy-efficient deep neural networks. ",
keywords = "cs.CY, cs.LG, eess.SP, stat.ML",
author = "Anthony, {Lasse F. Wolff} and Benjamin Kanding and Raghavendra Selvan",
note = "Accepted to be presented at the ICML Workshop on {"}Challenges in Deploying and monitoring Machine Learning Systems{"}, 2020. Source code at this link https://github.com/lfwa/carbontracker/; ICML Workshop on {"}Challenges in Deploying and monitoring Machine Learning Systems{"} ; Conference date: 17-07-2020",
year = "2020",
language = "English",
url = "https://icml.cc/Conferences/2020/Schedule?showEvent=5738",

}

RIS

TY - CONF

T1 - Carbontracker

T2 - ICML Workshop on "Challenges in Deploying and monitoring Machine Learning Systems"

AU - Anthony, Lasse F. Wolff

AU - Kanding, Benjamin

AU - Selvan, Raghavendra

N1 - Accepted to be presented at the ICML Workshop on "Challenges in Deploying and monitoring Machine Learning Systems", 2020. Source code at this link https://github.com/lfwa/carbontracker/

PY - 2020

Y1 - 2020

N2 - Deep learning (DL) can achieve impressive results across a wide variety of tasks, but this often comes at the cost of training models for extensive periods on specialized hardware accelerators. This energy-intensive workload has seen immense growth in recent years. Machine learning (ML) may become a significant contributor to climate change if this exponential trend continues. If practitioners are aware of their energy and carbon footprint, then they may actively take steps to reduce it whenever possible. In this work, we present Carbontracker, a tool for tracking and predicting the energy and carbon footprint of training DL models. We propose that energy and carbon footprint of model development and training is reported alongside performance metrics using tools like Carbontracker. We hope this will promote responsible computing in ML and encourage research into energy-efficient deep neural networks.

AB - Deep learning (DL) can achieve impressive results across a wide variety of tasks, but this often comes at the cost of training models for extensive periods on specialized hardware accelerators. This energy-intensive workload has seen immense growth in recent years. Machine learning (ML) may become a significant contributor to climate change if this exponential trend continues. If practitioners are aware of their energy and carbon footprint, then they may actively take steps to reduce it whenever possible. In this work, we present Carbontracker, a tool for tracking and predicting the energy and carbon footprint of training DL models. We propose that energy and carbon footprint of model development and training is reported alongside performance metrics using tools like Carbontracker. We hope this will promote responsible computing in ML and encourage research into energy-efficient deep neural networks.

KW - cs.CY

KW - cs.LG

KW - eess.SP

KW - stat.ML

M3 - Paper

Y2 - 17 July 2020

ER -

ID: 255786102