paper-with-me

홈 › Papers

Enhancing Knowledge Distillation of Large Language Models through Efficient Multi-Modal Distribution Alignment

2024-09-19 · Tianyu Peng, Jiajun Zhang

Knowledge distillation (KD) is an effective model compression method that can transfer the internal capabilities of large language models (LLMs) to smaller ones. However, the multi-modal probability distribution predicted by teacher LLMs causes difficulties for student models to learn. In this paper, we first demonstrate the importance of multi-modal distribution alignment with experiments and then highlight the inefficiency of existing KD approaches in learning multi-modal distributions. To address this problem, we propose Ranking Loss based Knowledge Distillation (RLKD), which encourages the consistency of the ranking of peak predictions between the teacher and student models. By incorporating word-level ranking loss, we ensure excellent compatibility with existing distillation objectives while fully leveraging the fine-grained information between different categories in peaks of two predicted distribution. Experimental results demonstrate that our method enables the student model to better learn the multi-modal distributions of the teacher model, leading to a significant performance improvement in various downstream tasks.

📄 PDF Abstract BibTeX arXiv:2409.12545

Code (1)

pty72/rlkd 공식 구현 pytorch

Tasks

Knowledge DistillationModel Compression

Methods 이 논문이 사용한 방법론

Knowledge Distillation A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions.…

Similar Papers 제목 키워드 기반

QCRD: Quality-guided Contrastive Rationale Distillation for Large Language Models

2024-05-14 · Wei Wang, Zhaowei Li, Qi Xu, Yiqing Cai 외

The deployment of large language models (LLMs) faces considerable challenges concerning resource constraints and inference efficiency. Recent research has increasingly focused on smaller, task-specific models enhanced by…

Contrastive LearningDenoisingDiversityKnowledge Distillation+1

Powerful Teachers Matter: Text-Guided Multi-view Knowledge Distillation with Visual Prior Enhancement

2026-03-25 · Xin Zhang, Jianyang Xu, Hao Peng, Dongjing Wang 외 arxiv

Knowledge distillation transfers knowledge from large teacher models to smaller students for efficient inference. While existing methods primarily focus on distillation strategies, they often overlook the importance of e…

Knowledge Distillation

DistilCSE: Effective Knowledge Distillation For Contrastive Sentence Embeddings

2021-12-10 · Chaochen Gao, Xing Wu, Peng Wang, Jue Wang 외

Large-scale contrastive learning models can learn very informative sentence embeddings, but are hard to serve online due to the huge model size. Therefore, they often play the role of "teacher", transferring abilities to…

Contrastive LearningKnowledge DistillationSemantic Textual SimilaritySentence+4

Feature Alignment and Representation Transfer in Knowledge Distillation for Large Language Models

2025-04-18 · Junjie Yang, Junhao Song, Xudong Han, Ziqian Bi 외

Knowledge distillation (KD) is a technique for transferring knowledge from complex teacher models to simpler student models, significantly enhancing model efficiency and accuracy. It has demonstrated substantial advancem…

image-classificationImage ClassificationKnowledge DistillationLanguage Modeling+7

Multi-Stage Balanced Distillation: Addressing Long-Tail Challenges in Sequence-Level Knowledge Distillation

2024-06-19 · YuHang Zhou, Jing Zhu, Paiheng Xu, Xiaoyu Liu 외

Large language models (LLMs) have significantly advanced various natural language processing tasks, but deploying them remains computationally expensive. Knowledge distillation (KD) is a promising solution, enabling the …

Knowledge Distillation