Paper Detail
Jingwei Zuo, Xinze Feng, Zien Liu, Kaijian Wang, Fanjiang Ye, Ye Cao, Zhuang Wang, Yuke Wang
Low-Rank Adaptation (LoRA) is now the dominant method for parameter-efficient fine-tuning of large language models, but achieving a high-quality adapter often requires systematic hyperparameter tuning because LoRA performance is highly sensitive to configuration choices. In practice, this leads to many concurrent LoRA jobs, often spanning heterogeneous tasks in multi-tenant environments. Existing systems largely handle these jobs independently, which both wastes computation on weak candidates and leaves GPUs underutilized. We present ALTO (Adaptive LoRA Tuning and Orchestration), a co-designed training system that accelerates LoRA hyperparameter tuning while enabling efficient cluster sharing across heterogeneous tasks. The central insight behind ALTO is that when multiple tuning jobs run concurrently over a shared frozen backbone, they expose optimization opportunities that single-job designs cannot exploit. Building on this, ALTO monitors loss trajectories to terminate unpromising configurations early, uses fused grouped GEMM together with a new rank-local adapter parallelism to co-locate surviving adapters and reclaim freed GPU capacity, and combines intra-task and inter-task scheduling to improve multi-task placement by leveraging the predictable duration of LoRA jobs. Extensive evaluation shows that ALTO achieves up to $13.8\times$ speedup over state-of-the-art without sacrificing adapter quality.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@article{zuo2026alto,
title = {ALTO: Adaptive LoRA Tuning and Orchestration for Heterogeneous LoRA Training Workloads},
author = {Jingwei Zuo and Xinze Feng and Zien Liu and Kaijian Wang and Fanjiang Ye and Ye Cao and Zhuang Wang and Yuke Wang},
year = {2026},
abstract = {Low-Rank Adaptation (LoRA) is now the dominant method for parameter-efficient fine-tuning of large language models, but achieving a high-quality adapter often requires systematic hyperparameter tuning because LoRA performance is highly sensitive to configuration choices. In practice, this leads to many concurrent LoRA jobs, often spanning heterogeneous tasks in multi-tenant environments. Existing systems largely handle these jobs independently, which both wastes computation on weak candidates an},
url = {https://arxiv.org/abs/2604.05426},
keywords = {cs.LG, cs.AI, cs.DC},
eprint = {2604.05426},
archiveprefix = {arXiv},
}
{}