Parameter-Efficient Fine-Tuning of State Space Models

TL;DR

This paper investigates parameter-efficient fine-tuning methods for deep State Space Models, finds limitations of existing methods, and proposes a new tailored approach called Sparse Dimension Tuning (SDT) that improves performance.

Contribution

It introduces Sparse Dimension Tuning (SDT), a novel PEFT method specifically designed for SSM modules, enhancing fine-tuning efficiency and effectiveness.

Findings

LoRA outperforms other PEFT methods on SSMs

LoRA fails on SSM modules but still surpasses alternatives

SDT combined with LoRA achieves state-of-the-art results

Abstract

Deep State Space Models (SSMs), such as Mamba (Gu & Dao, 2024), have become powerful tools for language modeling, offering high performance and linear scalability with sequence length. However, the application of parameter-efficient fine-tuning (PEFT) methods to SSM-based models remains largely underexplored. We start by investigating two fundamental questions on existing PEFT methods: (i) How do they perform on SSM-based models? (ii) Which parameters should they target for optimal results? Our analysis shows that LoRA and its variants consistently outperform all other PEFT methods. While LoRA is effective for linear projection matrices, it fails on SSM modules-yet still outperforms other methods applicable to SSMs, indicating their limitations. This underscores the need for a specialized SSM tuning approach. To address this, we propose Sparse Dimension Tuning (SDT), a PEFT method tailored for SSM modules. Combining SDT for SSMs with LoRA for linear projection matrices, we achieve state-of-the-art performance across extensive experiments.