paper-with-me

홈 › Papers

Parameter-Efficient Fine-Tuning of State Space Models

2024-10-11 · Kevin Galim, Wonjun Kang, Yuchen Zeng, Hyung Il Koo, Kangwook Lee

Deep State Space Models (SSMs), such as Mamba (Gu & Dao, 2024), have emerged as powerful tools for language modeling, offering high performance with efficient inference and linear scaling in sequence length. However, the application of parameter-efficient fine-tuning (PEFT) methods to SSM-based models remains largely unexplored. This paper aims to systematically study two key questions: (i) How do existing PEFT methods perform on SSM-based models? (ii) Which modules are most effective for fine-tuning? We conduct an empirical benchmark of four basic PEFT methods on SSM-based models. Our findings reveal that prompt-based methods (e.g., prefix-tuning) are no longer effective, an empirical result further supported by theoretical analysis. In contrast, LoRA remains effective for SSM-based models. We further investigate the optimal application of LoRA within these models, demonstrating both theoretically and experimentally that applying LoRA to linear projection matrices without modifying SSM modules yields the best results, as LoRA is not effective at tuning SSM modules. To further improve performance, we introduce LoRA with Selective Dimension tuning (SDLoRA), which selectively updates certain channels and states on SSM modules while applying LoRA to linear projection matrices. Extensive experimental results show that this approach outperforms standard LoRA.

📄 PDF Abstract BibTeX arXiv:2410.09016

Code (1)

furiosa-ai/ssm-peft 공식 구현 pytorch

Tasks

Language ModelingLanguage ModellingMambaparameter-efficient fine-tuningState Space Models

Methods 이 논문이 사용한 방법론

Mamba Foundation models, now powering most of the exciting applications in deep learning, are almost universally based on the Transformer architecture and its core attention module.…

Similar Papers 제목 키워드 기반

Parameter-Efficient Fine-Tuning Design Spaces

2023-01-04 · Jiaao Chen, Aston Zhang, Xingjian Shi, Mu Li 외

Parameter-efficient fine-tuning aims to achieve performance comparable to fine-tuning, using fewer trainable parameters. Several strategies (e.g., Adapters, prefix tuning, BitFit, and LoRA) have been proposed. However, t…

parameter-efficient fine-tuning

SAM-PARSER: Fine-tuning SAM Efficiently by Parameter Space Reconstruction

2023-08-28 · Zelin Peng, Zhengqin Xu, Zhilin Zeng, Xiaokang Yang 외

Segment Anything Model (SAM) has received remarkable attention as it offers a powerful and versatile solution for object segmentation in images. However, fine-tuning SAM for downstream segmentation tasks under different …

parameter-efficient fine-tuningSegmentationSemantic Segmentation

GPart: End-to-End Isometric Fine-Tuning via Global Parameter Partitioning

2026-05-14 · Paolo Mandica, Michał Brzozowski, Zuzanna Dubanowska, Neo Christopher Chung arxiv

Low-rank adaptation (LoRA) has become the dominant paradigm for parameter-efficient fine-tuning (PEFT) of large language models (LLMs). However, its bilinear structure introduces a critical limitation: the mapping from t…

parameter-efficient fine-tuningNatural Language UnderstandingMathematical Reasoning

State-offset Tuning: State-based Parameter-Efficient Fine-Tuning for State Space Models

2025-03-05 · Wonjun Kang, Kevin Galim, Yuchen Zeng, Minjae Lee 외

State Space Models (SSMs) have emerged as efficient alternatives to Transformers, mitigating their quadratic computational cost. However, the application of Parameter-Efficient Fine-Tuning (PEFT) methods to SSMs remains …

parameter-efficient fine-tuningState Space Models

How fine can fine-tuning be? Learning efficient language models

2020-04-24 · Evani Radiya-Dixit, Xin Wang

State-of-the-art performance on language understanding tasks is now achieved with increasingly large networks; the current record holder has billions of parameters. Given a language model pre-trained on massive unlabeled…

Language Modelling