paper-with-me

홈 › Papers

Dynamic Base model Shift for Delta Compression

2025-05-16 · Chenyu Huang, Peng Ye, Shenghe Zheng, Xiaohui Wang, Lei Bai, Tao Chen, Wanli Ouyang

Transformer-based models with the pretrain-finetune paradigm bring about significant progress, along with the heavy storage and deployment costs of finetuned models on multiple tasks. Delta compression attempts to lower the costs by reducing the redundancy of delta parameters (i.e., the difference between the finetuned and pre-trained model weights) through pruning or quantization. However, existing methods by default employ the pretrained model as the base model and compress the delta parameters for every task, which may causes significant performance degradation, especially when the compression rate is extremely high. To tackle this issue, we investigate the impact of different base models on the performance of delta compression and find that the pre-trained base model can hardly be optimal. To this end, we propose Dynamic Base Model Shift (DBMS), which dynamically adapts the base model to the target task before performing delta compression. Specifically, we adjust two parameters, which respectively determine the magnitude of the base model shift and the overall scale of delta compression, to boost the compression performance on each task. Through low-cost learning of these two parameters, our DBMS can maintain most of the finetuned model's performance even under an extremely high compression ratio setting, significantly surpassing existing methods. Moreover, our DBMS is orthogonal and can be integrated with a variety of other methods, and it has been evaluated across different types of models including language, vision transformer, and multi-modal models.

📄 PDF Abstract BibTeX arXiv:2505.11344

Code (0)

등록된 구현이 없습니다.

Tasks

model

Methods 이 논문이 사용한 방법론

Pruning 설명 없음
BASE 설명 없음

Similar Papers 제목 키워드 기반

Delta Decompression for MoE-based LLMs Compression

2025-02-24 · Hao Gu, Wei Li, Lujun Li, Qiyuan Zhu 외

Mixture-of-Experts (MoE) architectures in large language models (LLMs) achieve exceptional performance, but face prohibitive storage and memory requirements. To address these challenges, we present $D^2$-MoE, a new delta…

DiversityMixture-of-Experts

ImPart: Importance-Aware Delta-Sparsification for Improved Model Compression and Merging in LLMs

2025-04-17 · Yan Yang, Yixia Li, Hongru Wang, Xuetao Wei 외

With the proliferation of task-specific large language models, delta compression has emerged as a method to mitigate the resource challenges of deploying numerous such models by effectively compressing the delta model pa…

Model CompressionQuantization

DeltaDQ: Ultra-High Delta Compression for Fine-Tuned LLMs via Group-wise Dropout and Separate Quantization

2024-10-11 · Yanfeng Jiang, Zelan Yang, Bohua Chen, Shen Li 외

Large language models achieve exceptional performance on various downstream tasks through supervised fine-tuning. However, the diversity of downstream tasks and practical requirements makes deploying multiple full-parame…

DiversityQuantization

D-QRELO: Training- and Data-Free Delta Compression for Large Language Models via Quantization and Residual Low-Rank Approximation

2026-04-18 · Junlin Li, Shuangyong Song, Guodong Du, Ngai Wong 외 arxiv

Supervised Fine-Tuning (SFT) accelerates taskspecific large language models (LLMs) development, but the resulting proliferation of finetuned models incurs substantial memory overhead. Delta compression addresses this by …

Breaking the Compression Ceiling: Data-Free Pipeline for Ultra-Efficient Delta Compression

2025-05-19 · Xiaohui Wang, Peng Ye, Chenyu Huang, Shenghe Zheng 외

With the rise of the fine-tuned--pretrained paradigm, storing numerous fine-tuned models for multi-tasking creates significant storage overhead. Delta compression alleviates this by storing only the pretrained model and …