@inproceedings{zhang-etal-2026-hyperadalora,
    title = "{H}yper{A}da{L}o{RA}: Accelerating {L}o{RA} Rank Allocation During Training via Hypernetworks without Sacrificing Performance",
    author = "Zhang, Hao  and
      Li, Zhenjia  and
      Gao, Yifan  and
      Xiao, Xi  and
      Zhang, Heng  and
      Zhang, Shuyang  and
      Xiaoxincc  and
      Huang, Bo  and
      Wu, Yuhang  and
      Wang, Tianyang  and
      Xu, Hao",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.findings-acl.1648/",
    pages = "32937--32949",
    ISBN = "979-8-89176-395-1",
    abstract = "Parameter-Efficient Fine-Tuning (PEFT), especially Low-Rank Adaptation (LoRA), has emerged as a promising approach to fine-tuning large language models(LLMs) while reducing computational and memory overhead. However, LoRA assumes a uniform rank \textit{r} for each incremental matrix, not accounting for the varying significance of weight matrices across different modules and layers. AdaLoRA leverages Singular Value Decomposition (SVD) to parameterize updates and employs pruning of singular values to introduce dynamic rank allocation, thereby enhancing adaptability. However, during the training process, it often encounters issues of slow convergence speed and high computational overhead. To address this issue, we propose HyperAdaLoRA, a novel framework that accelerates the convergence of AdaLoRA by leveraging a hypernetwork. Instead of directly optimizing the components of Singular Value Decomposition $(P, \Lambda, Q)$, HyperAdaLoRA employs a hypernetwork based on attention mechanisms to dynamically generate these parameters. By pruning the outputs of the hypernetwork that generates the singular values, dynamic rank allocation is achieved. Comprehensive experiments on various datasets and models demonstrate that our method achieves faster convergence without sacrificing performance. Moreover, our method generalizes well to other LoRA-based approaches, highlighting its strong generalization capability."
}