Summary of From Computation to Consumption: Exploring the Compute-energy Link For Training and Testing Neural Networks For Sed Systems, by Constance Douwes et al.
From Computation to Consumption: Exploring the Compute-Energy Link for Training and Testing Neural Networks for SED Systems
by Constance Douwes, Romain Serizel
First submitted to arxiv on: 8 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Sound (cs.SD)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper investigates the environmental impact of machine learning models, particularly neural networks, by examining their energy requirements. It focuses on several neural network architectures used in sound event detection systems, using an audio tagging task as a example. The authors measure the energy consumption for training and testing small to large architectures, establishing relationships between energy consumption, floating-point operations, number of parameters, and GPU/memory utilization. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper looks at how much energy machine learning models use, especially neural networks. It wants to understand how these models affect the environment. The researchers study different types of neural networks used in sound detection systems. They measure how much energy it takes to train and test small and big architectures, and find connections between energy usage, calculations, number of parameters, and computer memory. |
Keywords
» Artificial intelligence » Event detection » Machine learning » Neural network
