English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

To raise awareness of the huge impact Deep Learning (DL) has on theenvironment, several works have tried to estimate the energy consumption andcarbon footprint of DL-based systems across their life cycle. However, theestimations for energy consumption in the training stage usually rely onassumptions that have not been thoroughly tested. This study aims to move pastthese assumptions by leveraging the relationship between energy consumption andtwo relevant design decisions in DL training; model architecture, and trainingenvironment. To investigate these relationships, we collect multiple metricsrelated to energy efficiency and model correctness during the models' training.Then, we outline the trade-offs between the measured energy consumption and themodels' correctness regarding model architecture, and their relationship withthe training environment. Finally, we study the training's power consumptionbehavior and propose four new energy estimation methods. Our results show thatselecting the proper model architecture and training environment can reduceenergy consumption dramatically (up to 80.72%) at the cost of negligibledecreases in correctness. Also, we find evidence that GPUs should scale withthe models' computational complexity for better energy efficiency. Furthermore,we prove that current energy estimation methods are unreliable and proposealternatives 2x more precise.

How to use model architecture and training environment to estimate the energy consumption of DL training