Federated Fine-Tuning of LLMs on the Very Edge: The Good, the Bad, the Ugly

Herbert Woisetschläger, Alexander Erben, Shiqiang Wang, Ruben Mayer, Hans Arno Jacobsen

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations

Abstract

With the emergence of AI regulations, such as the EU AI Act, requirements for simple data lineage, enforcement of low data bias, and energy efficiency have become a priority for everyone offering AI services. Being pre-trained on versatile and a vast amount of data, large language models and foundation models (FMs) offer a good basis for building high-quality deep learning pipelines. Fine-tuning can further improve model performance on a specific downstream task, which requires orders of magnitude less data than pre-training. Often, access to high-quality and low-bias data for model fine-tuning is limited due to technical or regulatory requirements. Federated learning (FL), as a distributed and privacy-preserving technique, offers a well-suited approach to significantly expanding data access for model fine-tuning. Yet, this data is often located on the network edge, where energy, computational, and communication resources are significantly more limited than in data centers. In our paper, we conduct an end-to-end evaluation for fine-tuning the FLAN-T5 FM family on the network edge. We study energy efficiency potentials throughout FL systems - on clients, in communication, and on the server. Our analysis introduces energy efficiency as a real-time metric to assess the computational efficiency of an FL system. We show the stark need for further improvements in communication efficiency when working with FMs and demonstrate the importance of adaptive FL optimizers for FM training.

Original languageEnglish
Title of host publicationProceedings of the 8th Workshop on Data Management for End-to-End Machine Learning, DEEM 2024 - In conjunction with the 2024 ACM SIGMOD/PODS Conference
PublisherAssociation for Computing Machinery, Inc
Pages39-50
Number of pages12
ISBN (Electronic)9798400706110
DOIs
StatePublished - 9 Jun 2024
Externally publishedYes
Event8th Workshop on Data Management for End-to-End Machine Learning, DEEM 2024 - Santiago, Chile
Duration: 9 Jun 2024 → …

Publication series

NameProceedings of the 8th Workshop on Data Management for End-to-End Machine Learning, DEEM 2024 - In conjunction with the 2024 ACM SIGMOD/PODS Conference

Conference

Conference8th Workshop on Data Management for End-to-End Machine Learning, DEEM 2024
Country/TerritoryChile
CitySantiago
Period9/06/24 → …

Fingerprint

Dive into the research topics of 'Federated Fine-Tuning of LLMs on the Very Edge: The Good, the Bad, the Ugly'. Together they form a unique fingerprint.

Cite this