Summary of Federated Learning Of Large Asr Models in the Real World, by Yonghui Xiao et al.
Federated Learning of Large ASR Models in the Real World
by Yonghui Xiao, Yuxin Ding, Changwan Ryu, Petr Zadrazil, Francoise Beaufays
First submitted to arxiv on: 19 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computation and Language (cs.CL); Sound (cs.SD); Audio and Speech Processing (eess.AS)
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 abstract discusses the challenges of federated learning (FL) for large machine learning models, specifically Conformer-based ASR models with over 100 million parameters. FL aims to preserve privacy while training models, but large models require significant resources, which is an obstacle for FL. The paper presents a systematic solution to train full-size ASR models using FL, achieving the first real-world application of the Conformer model trained with FL and the largest model ever trained with FL so far. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper shows that FL can improve the quality of ASR models by refining data and labels from clients. The training efficiency and model quality are demonstrated in real-world experiments. |
Keywords
» Artificial intelligence » Federated learning » Machine learning
