Model Merging in Pre-training of Large Language Models

Model merging has emerged as a promising technique for enhancing large language models, though its application in large-scale pre-training remains relatively unexplored. In this paper, we present a comprehensive investigation of model merging techniques during the pre-training process. Through extensive experiments with both dense and Mixture-of-Experts (MoE) architectures ranging from millions to over 100 billion parameters, we demonstrate that merging checkpoints trained with constant learning rates not only achieves significant performance improvements but also enables accurate prediction of annealing behavior. These improvements lead to both more efficient model development and significantly lower training costs. Our detailed ablation studies on merging strategies and hyperparameters provide new insights into the underlying mechanisms while uncovering novel applications. Through comprehensive experimental analysis, we offer the open-source community practical pre-training guidelines for effective model merging.

Authors

• Yunshui Li Shenzhen Key Laboratory for High Performance Data Mining, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences University of Chinese Academy of Sciences
• Yiyuan Ma
• Shen Yan University of Southern California
• Chaoyi Zhang School of Computer Science, University of Sydney
• Jing Liu
• Jianqiao Lu
• Ziwen Xu
• Mengzhao Chen MAC Lab, Department of Artificial Intelligence, Xiamen University
• Minrui Wang
• Shiyi Zhan
• Jin Ma IPS Search, Tencent PCG
• Xunhao Lai
• Yao Luo
• Xingyan Bin
• Hongbin Ren
• Mingji Han
• Wenhao Hao
• Bairen Yi
• LingJun Liu
• Bole Ma
• Xiaoying Jia
• zhou Xun
• liang xiang
• Yonghui Wu

Keywords

No keywords are indexed for this paper.