CoLLM: Continuous Adaptation for SLO-Aware LLM Serving on Shared GPU Clusters

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Huang, Shaoyuan, Zhao, Yunfeng, Yan, Na, Zhang, Tiancheng, Wang, Xiaokai, Wang, Xiaofei, Wang, Wenyu, Deng, Yansha
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914570954604544
author Huang, Shaoyuan
Zhao, Yunfeng
Yan, Na
Zhang, Tiancheng
Wang, Xiaokai
Wang, Xiaofei
Wang, Wenyu
Deng, Yansha
author_facet Huang, Shaoyuan
Zhao, Yunfeng
Yan, Na
Zhang, Tiancheng
Wang, Xiaokai
Wang, Xiaofei
Wang, Wenyu
Deng, Yansha
contents As Large Language Models (LLMs) are increasingly adopted in edge intelligence to power domain-specific applications and personalized services, the quality and efficiency of the LLM post-training phase-including fine-tuning and inference, have become critical due to constrained resources. Although recent advances in federated parameter-efficient fine-tuning (FL PEFT) and low-latency inference have improved individual task performance, fine-tuning and inference are still handled as isolated workloads, which overlooks their interdependence and results in redundant deployments and delayed improvement in inference quality. To address these limitations, we introduce a new co-execution framework and instantiate it with CoLLM, a system that unifies FL PEFT and inference on shared edge replicas and model parameters. CoLLM addresses key challenges at both replica and cluster levels through: (1) an intra-replica model sharing mechanism that enables real-time model parameter reuse via unmerged inference and shadow adapter strategies; and (2) a two-timescale inter-replica coordination algorithm that adaptively balances fine-tuning and inference workloads to jointly optimize long-term model quality gains and short-term inference efficiency. Extensive evaluation across diverse LLMs and real-world traces show that CoLLM consistently outperforms state-of-the-art LLM systems, achieving up to 3x higher goodput, demonstrating its effectiveness in enabling seamless LLM post-training for edge intelligence.
format Preprint
id arxiv_https___arxiv_org_abs_2604_16400
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle CoLLM: Continuous Adaptation for SLO-Aware LLM Serving on Shared GPU Clusters
Huang, Shaoyuan
Zhao, Yunfeng
Yan, Na
Zhang, Tiancheng
Wang, Xiaokai
Wang, Xiaofei
Wang, Wenyu
Deng, Yansha
Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Machine Learning
As Large Language Models (LLMs) are increasingly adopted in edge intelligence to power domain-specific applications and personalized services, the quality and efficiency of the LLM post-training phase-including fine-tuning and inference, have become critical due to constrained resources. Although recent advances in federated parameter-efficient fine-tuning (FL PEFT) and low-latency inference have improved individual task performance, fine-tuning and inference are still handled as isolated workloads, which overlooks their interdependence and results in redundant deployments and delayed improvement in inference quality. To address these limitations, we introduce a new co-execution framework and instantiate it with CoLLM, a system that unifies FL PEFT and inference on shared edge replicas and model parameters. CoLLM addresses key challenges at both replica and cluster levels through: (1) an intra-replica model sharing mechanism that enables real-time model parameter reuse via unmerged inference and shadow adapter strategies; and (2) a two-timescale inter-replica coordination algorithm that adaptively balances fine-tuning and inference workloads to jointly optimize long-term model quality gains and short-term inference efficiency. Extensive evaluation across diverse LLMs and real-world traces show that CoLLM consistently outperforms state-of-the-art LLM systems, achieving up to 3x higher goodput, demonstrating its effectiveness in enabling seamless LLM post-training for edge intelligence.
title CoLLM: Continuous Adaptation for SLO-Aware LLM Serving on Shared GPU Clusters
topic Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2604.16400