arXiv：cs.LG（机器学习，全量分类）· Constance Douwes, Paul Magron, Romain Serizel·· 14 小时前AI 评分38
深度学习音频开发的能耗代价：开发阶段能耗是模型训练的 3 至 256 倍
Exposing the Cost of Deep Learning Audio Development
AI 导读
一项基于 Grid5000 计算平台活动日志的研究估算，深度学习音频项目开发阶段的能耗是单独训练最佳模型所需能耗的 3 至 256 倍。研究以 LORIA 实验室 Multispeech 团队的四个音频项目为案例，对比了架构原型设计与密集实验的整体能耗与已报告模型训练能耗。作者据此呼吁对深度学习音频项目全生命周期的能耗进行更系统的报告。
正文
Abstract:The environmental impact of deep learning has attracted increasing attention over the past decade. Existing studies mainly focus on the energy and carbon emissions of model training and inference, while the whole development phase is often overlooked. Yet, architecture prototyping and intensive experiments are conducted during this stage, which is highly energy-demanding. In this article, we propose a methodology to estimate these costs, based on activity logs from the Grid5000 shared computing platform used by the LORIA laboratory. As a case-study, we focus on audio projects developed in the Multispeech research team. We evaluate the overall energy cost of four projects, and we compare them to those of training the reported models. Our results show that the energy required for the development phase is 3 to 256 times greater than that required to train the best-performing model alone. These results advocate for a more systematic reporting of energy consumption across the entire life cycle of deep learning-based audio projects.
| Comments: | 5 pages, 2 figures, 1 table |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Sound (cs.SD) |
| Cite as: | arXiv:2610.01619 [cs.LG] |
| (or arXiv:2610.01619v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.01619 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Constance Douwes [view email]
[v1]
Thu, 1 Oct 2026 12:55:25 UTC (64 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org