paper-with-me

홈 › Papers

Basis Sharing: Cross-Layer Parameter Sharing for Large Language Model Compression

2024-10-02 · Jingcun Wang, Yu-Guang Chen, Ing-Chao Lin, Bing Li, Grace Li Zhang

Large Language Models (LLMs) have achieved remarkable breakthroughs. However, the huge number of parameters in LLMs require significant amount of memory storage in inference, which prevents their practical deployment in many applications. To reduce memory storage of LLMs, singular value decomposition (SVD) provides a promising solution to approximate weight matrices for compressing LLMs. In this paper, we take a step further to explore parameter sharing across different layers with SVD to achieve more effective compression for LLMs. Specifically, weight matrices in different layers are decomposed and represented as a linear combination of a set of shared basis vectors and unique coefficients. The types of weight matrices and the layer selection for basis sharing are examined when compressing LLMs to maintain the performance. Comprehensive experiments demonstrate that Basis Sharing outperforms state-of-the-art SVD-based compression approaches and parameter sharing techniques, especially under large compression ratios. Code is available at: https://github.com/TUDa-HWAI/Basis_Sharing

📄 PDF Abstract BibTeX arXiv:2410.03765

Code (1)

tuda-hwai/basis_sharing 공식 구현 pytorch

Tasks

Language ModelingLanguage ModellingLarge Language ModelModel Compression

Methods 이 논문이 사용한 방법론

SET Dynamic Sparse Training method where weight mask is updated randomly periodically

Similar Papers 제목 키워드 기반

Learning Deeply Shared Filter Bases for Efficient ConvNets

2021-01-01 · Woochul Kang, Daeyeon Kim

Recently, inspired by repetitive block structure of modern ConvNets, such as ResNets, parameter-sharing among repetitive convolution layers has been proposed to reduce the size of parameters. However, naive sharing of …

Rethinking Parameter Sharing as Graph Coloring for Structured Compression

2025-11-10 · Boyang Zhang, Daning Cheng, Yunquan Zhang arxiv

Modern deep models have massive parameter sizes, leading to high inference-time memory usage that limits practical deployment. Parameter sharing, a form of structured compression, effectively reduces redundancy, but exis…

ASLoRA: Adaptive Sharing Low-Rank Adaptation Across Layers

2024-12-13 · Junyan Hu, Xue Xiao, Mengqi Zhang, Yao Chen 외

As large language models (LLMs) grow in size, traditional full fine-tuning becomes increasingly impractical due to its high computational and storage costs. Although popular parameter-efficient fine-tuning methods, such …

parameter-efficient fine-tuning

EPAS: Efficient Training with Progressive Activation Sharing

2026-01-27 · Rezaul Karim, Maryam Dialameh, Yang Liu, Boxing Chen 외 arxiv

We present a novel method for Efficient training with Progressive Activation Sharing (EPAS). This method bridges progressive training paradigm with the phenomenon of redundant QK (or KV ) activations across deeper layers…

Continual Pretraining

A Multi-lingual Multi-task Architecture for Low-resource Sequence Labeling

2018-07-01 · ACL 2018 7 · Ying Lin, Shengqi Yang, Veselin Stoyanov, Heng Ji

We propose a multi-lingual multi-task architecture to develop supervised models with a minimal amount of labeled data for sequence labeling. In this new architecture, we combine various transfer models using two layers o…

Abstractive Text SummarizationMachine TranslationMulti-Task LearningPart-Of-Speech Tagging+2