Metraipolis › archive › dc_8331bb8d7dd8
DynSL: Resource-Adaptive Model Training across Heterogeneous Edge Devices — arXiv:2611.03209
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.03209
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T11:12:45.908Z
- SHA-256 of the archived text
b1ed9a5a62a103719a671bab0ed1bd75aeb6b81b03656829ff71e3c817d05263- Stable URL of this page
https://metraipolis.com/documento/dc_8331bb8d7dd8
How to cite
DynSL: Resource-Adaptive Model Training across Heterogeneous Edge Devices — arXiv:2611.03209 — arXiv (en), https://arxiv.org/abs/2611.03209 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8331bb8d7dd8 (Metraipolis archive dc_8331bb8d7dd8, sha256 b1ed9a5a62a103719a671bab0ed1bd75aeb6b81b03656829ff71e3c817d05263)
BibTeX
@misc{dc8331bb8d7dd8,
title = {DynSL: Resource-Adaptive Model Training across Heterogeneous Edge Devices — arXiv:2611.03209},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.03209},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8331bb8d7dd8; sha256 b1ed9a5a62a103719a671bab0ed1bd75aeb6b81b03656829ff71e3c817d05263},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03209. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Training models over data distributed across edge devices presents a trade-off: centralizing data simplifies training but raises privacy and regulatory concerns, while full local training may exceed the devices' compute and memory capacity. Split learning addresses this challenge by partitioning the model between clients and the server, allowing part of the computation to be offloaded. However, existing approaches typically rely on static split configurations or simplified assumptions that fail to capture the heterogeneity and complex dynamics of real deployments. We present DynSL, a framework that dynamically allocates model layers between clients and the server by continuously monitoring system metrics and periodically solving an optimization problem that accounts for client and server computation and communication. When beneficial, the system reconfigures the split, enabling the offloading strategy to adapt to changing device capabilities and system conditions. Extensive experiments on a testbed of 24 heterogeneous edge devices show that DynSL reduces training time by up to $\sim\!$30% over the strongest adaptive baseline and by more than 50% over a suboptimal uniform split, with no loss -- and sometimes a gain -- in accuracy. Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.DC, nuovi lavori del giorno). Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC) ; Machine Learning (cs.LG); Networking and Internet Architecture (cs.NI). Pagina del lavoro: https://arxiv.org/abs/2611.03209 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-03T11:00:07.167Z dalla pagina https://arxiv.org/list/cs.DC/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).