LIFL: A Lightweight, Event-driven Serverless Platform for Federated Learning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Qi, Shixiong, Ramakrishnan, K. K., Lee, Myungjin
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908367801286656
author Qi, Shixiong
Ramakrishnan, K. K.
Lee, Myungjin
author_facet Qi, Shixiong
Ramakrishnan, K. K.
Lee, Myungjin
contents Federated Learning (FL) typically involves a large-scale, distributed system with individual user devices/servers training models locally and then aggregating their model updates on a trusted central server. Existing systems for FL often use an always-on server for model aggregation, which can be inefficient in terms of resource utilization. They may also be inelastic in their resource management. This is particularly exacerbated when aggregating model updates at scale in a highly dynamic environment with varying numbers of heterogeneous user devices/servers. We present LIFL, a lightweight and elastic serverless cloud platform with fine-grained resource management for efficient FL aggregation at scale. LIFL is enhanced by a streamlined, event-driven serverless design that eliminates the individual heavy-weight message broker and replaces inefficient container-based sidecars with lightweight eBPF-based proxies. We leverage shared memory processing to achieve high-performance communication for hierarchical aggregation, which is commonly adopted to speed up FL aggregation at scale. We further introduce locality-aware placement in LIFL to maximize the benefits of shared memory processing. LIFL precisely scales and carefully reuses the resources for hierarchical aggregation to achieve the highest degree of parallelism while minimizing the aggregation time and resource consumption. Our experimental results show that LIFL achieves significant improvement in resource efficiency and aggregation speed for supporting FL at scale, compared to existing serverful and serverless FL systems.
format Preprint
id arxiv_https___arxiv_org_abs_2405_10968
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LIFL: A Lightweight, Event-driven Serverless Platform for Federated Learning
Qi, Shixiong
Ramakrishnan, K. K.
Lee, Myungjin
Distributed, Parallel, and Cluster Computing
Machine Learning
Federated Learning (FL) typically involves a large-scale, distributed system with individual user devices/servers training models locally and then aggregating their model updates on a trusted central server. Existing systems for FL often use an always-on server for model aggregation, which can be inefficient in terms of resource utilization. They may also be inelastic in their resource management. This is particularly exacerbated when aggregating model updates at scale in a highly dynamic environment with varying numbers of heterogeneous user devices/servers. We present LIFL, a lightweight and elastic serverless cloud platform with fine-grained resource management for efficient FL aggregation at scale. LIFL is enhanced by a streamlined, event-driven serverless design that eliminates the individual heavy-weight message broker and replaces inefficient container-based sidecars with lightweight eBPF-based proxies. We leverage shared memory processing to achieve high-performance communication for hierarchical aggregation, which is commonly adopted to speed up FL aggregation at scale. We further introduce locality-aware placement in LIFL to maximize the benefits of shared memory processing. LIFL precisely scales and carefully reuses the resources for hierarchical aggregation to achieve the highest degree of parallelism while minimizing the aggregation time and resource consumption. Our experimental results show that LIFL achieves significant improvement in resource efficiency and aggregation speed for supporting FL at scale, compared to existing serverful and serverless FL systems.
title LIFL: A Lightweight, Event-driven Serverless Platform for Federated Learning
topic Distributed, Parallel, and Cluster Computing
Machine Learning
url https://arxiv.org/abs/2405.10968