CLAIJun 16, 2024

ShareLoRA: Parameter Efficient and Robust Large Language Model Fine-tuning via Shared Low-Rank Adaptation

arXiv:2406.10785v210 citations
Originality Incremental advance
AI Analysis

This method addresses the need for efficient and robust fine-tuning in resource-constrained environments, offering incremental improvements over existing techniques like LoRA.

The paper tackles the problem of parameter-efficient fine-tuning for large language models by introducing ShareLoRA, which shares low-rank weight matrices across layers, achieving a 44% to 96% reduction in trainable parameters compared to standard LoRA while maintaining or improving performance, with up to 1.2% average accuracy gains in various tasks.

In this paper, we introduce \textbf{Share}d \textbf{Lo}w \textbf{R}ank \textbf{A}daptation (ShareLoRA), a Large Language Model (LLM) fine-tuning technique that balances parameter efficiency, adaptability, and robustness without compromising performance. By strategically sharing the low-rank weight matrices across different layers, ShareLoRA achieves 44\% to 96\% reduction in trainable parameters compared to standard LoRA, alongside a substantial decrease in memory overhead. This efficiency gain scales with model size, making ShareLoRA particularly advantageous for resource-constrained environments. Importantly, ShareLoRA not only maintains model performance but also exhibits robustness in both classification and generation tasks across diverse models, including RoBERTa, GPT-2, and LLaMA series (1, 2, and 3). It consistently outperforms LoRA in zero-shot, few-shot, and continual fine-tuning scenarios, achieving up to 1.2\% average accuracy improvement, and enhanced generalization across domains. In continual learning settings, ShareLoRA achieves 1.2\% higher accuracy on GSM8K, 0.6\% on HumanEval, and 0.5\% on both MMLU and MMLU-Pro. Our results demonstrate that ShareLoRA supports high-quality fine-tuning while offering strong generalization and continual adaptation across various model scales and diverse tasks.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes