Waarom schiet de effectieve contextlengte van LLM's tekort?
Why Does the Effective Context Length of LLMs Fall Short?
October 24, 2024
Auteurs: Chenxin An, Jun Zhang, Ming Zhong, Lei Li, Shansan Gong, Yao Luo, Jingjing Xu, Lingpeng Kong
cs.AI
Samenvatting
Vorderingen in gedistribueerde training en efficiënte aandachtsmechanismen hebben aanzienlijk de contextvenstergroottes van grote taalmodellen (LLM's) vergroot. Recent werk onthult echter dat de effectieve contextlengtes van open-source LLM's vaak tekortschieten, meestal niet meer dan de helft van hun trainingslengtes. In dit werk schrijven we deze beperking toe aan de links-scheve frequentieverdeling van relatieve posities gevormd in LLM's pretraining- en post-trainingfasen, wat hun vermogen belemmert om effectief verre informatie te verzamelen. Om deze uitdaging aan te gaan, introduceren we ShifTed Rotray position embeddING (STRING). STRING verplaatst goed getrainde posities om de oorspronkelijke ineffectieve posities te overschrijven tijdens inferentie, waardoor de prestaties binnen hun bestaande trainingslengtes worden verbeterd. Experimentele resultaten tonen aan dat STRING de prestaties van de nieuwste grootschalige modellen, zoals Llama3.1 70B en Qwen2 72B, aanzienlijk verbetert met meer dan 10 punten op populaire lange-context benchmarks RULER en InfiniteBench, waardoor nieuwe state-of-the-art resultaten worden behaald voor open-source LLM's. Vergeleken met commerciële modellen behaalt Llama 3.1 70B met \method zelfs betere prestaties dan GPT-4-128K en overtreft duidelijk Claude 2 en Kimi-chat.
English
Advancements in distributed training and efficient attention mechanisms have
significantly expanded the context window sizes of large language models
(LLMs). However, recent work reveals that the effective context lengths of
open-source LLMs often fall short, typically not exceeding half of their
training lengths. In this work, we attribute this limitation to the left-skewed
frequency distribution of relative positions formed in LLMs pretraining and
post-training stages, which impedes their ability to effectively gather distant
information. To address this challenge, we introduce ShifTed Rotray position
embeddING (STRING). STRING shifts well-trained positions to overwrite the
original ineffective positions during inference, enhancing performance within
their existing training lengths. Experimental results show that without
additional training, STRING dramatically improves the performance of the latest
large-scale models, such as Llama3.1 70B and Qwen2 72B, by over 10 points on
popular long-context benchmarks RULER and InfiniteBench, establishing new
state-of-the-art results for open-source LLMs. Compared to commercial models,
Llama 3.1 70B with \method even achieves better performance than GPT-4-128K and
clearly surpasses Claude 2 and Kimi-chat.Summary
AI-Generated Summary
DeepSeek-R1: Het stimuleren van redeneervermogen in LLM's via Reinforcement LearningDeepSeek-R1: Incentivizing Reasoning Capability in LLMs via
Reinforcement Learning
DeepSeek-R1: Het stimuleren van redeneervermogen in LLM's via Reinforcement Learning
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via
Reinforcement Learning
DeepSeek-AI, Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, Xiaokang Zhang, Xingkai Yu, Yu Wu, Z. F. Wu, Zhibin Gou, Zhihong Shao, Zhuoshu Li, Ziyi Gao, Aixin Liu, Bing Xue, Bingxuan Wang, Bochao Wu, Bei Feng, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, Fuli Luo, Guangbo Hao, Guanting Chen, Guowei Li, H. Zhang, Han Bao, Hanwei Xu, Haocheng Wang, Honghui Ding, Huajian Xin, Huazuo Gao, Hui Qu, Hui Li, Jianzhong Guo, Jiashi Li, Jiawei Wang, Jingchang Chen, Jingyang Yuan, Junjie Qiu, Junlong Li, J. L. Cai, Jiaqi Ni, Jian Liang, Jin Chen, Kai Dong, Kai Hu, Kaige Gao, Kang Guan, Kexin Huang, Kuai Yu, Lean Wang, Lecong Zhang, Liang Zhao, Litong Wang, Liyue Zhang, Lei Xu, Leyi Xia, Mingchuan Zhang, Minghua Zhang, Minghui Tang, Meng Li, Miaojun Wang, Mingming Li, Ning Tian, Panpan Huang, Peng Zhang, Qiancheng Wang, Qinyu Chen, Qiushi Du, Ruiqi Ge, Ruisong Zhang, Ruizhe Pan, Runji Wang, R. J. Chen, R. L. Jin, Ruyi Chen, Shanghao Lu, Shangyan Zhou, Shanhuang Chen, Shengfeng Ye, Shiyu Wang, Shuiping Yu, Shunfeng Zhou, Shuting Pan, S. S. Li, Shuang Zhou, Shaoqing Wu, Shengfeng Ye, Tao Yun, Tian Pei, Tianyu Sun, T. Wang, Wangding Zeng, Wanjia Zhao, Wen Liu, Wenfeng Liang, Wenjun Gao, Wenqin Yu, Wentao Zhang, W. L. Xiao, Wei An, Xiaodong Liu, Xiaohan Wang, Xiaokang Chen, Xiaotao Nie, Xin Cheng, Xin Liu, Xin Xie, Xingchao Liu, Xinyu Yang, Xinyuan Li, Xuecheng Su, Xuheng Lin, X. Q. Li, Xiangyue Jin, Xiaojin Shen, Xiaosha Chen, Xiaowen Sun, Xiaoxiang Wang, Xinnan Song, Xinyi Zhou, Xianzu Wang, Xinxia Shan, Y. K. Li, Y. Q. Wang, Y. X. Wei, Yang Zhang, Yanhong Xu, Yao Li, Yao Zhao, Yaofeng Sun, Yaohui Wang, Yi Yu, Yichao Zhang, Yifan Shi, Yiliang Xiong, Ying He, Yishi Piao, Yisong Wang, Yixuan Tan, Yiyang Ma, Yiyuan Liu, Yongqiang Guo, Yuan Ou, Yuduan Wang, Yue Gong, Yuheng Zou, Yujia He, Yunfan Xiong, Yuxiang Luo, Yuxiang You, Yuxuan Liu, Yuyang Zhou, Y. X. Zhu, Yanhong Xu, Yanping Huang, Yaohui Li, Yi Zheng, Yuchen Zhu, Yunxian Ma, Ying Tang, Yukun Zha, Yuting Yan, Z. Z. Ren, Zehui Ren, Zhangli Sha, Zhe Fu, Zhean Xu, Zhenda Xie, Zhengyan Zhang, Zhewen Hao, Zhicheng Ma, Zhigang Yan, Zhiyu Wu, Zihui Gu, Zijia Zhu, Zijun Liu, Zilin Li, Ziwei Xie, Ziyang Song, Zizheng Pan, Zhen Huang, Zhipeng Xu, Zhongyu Zhang, Zhen Zhang•Jan 22, 2025•3685
Technisch Rapport Qwen2.5Qwen2.5 Technical Report
Technisch Rapport Qwen2.5
Qwen2.5 Technical Report
Qwen, An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, Keming Lu, Keqin Bao, Kexin Yang, Le Yu, Mei Li, Mingfeng Xue, Pei Zhang, Qin Zhu, Rui Men, Runji Lin, Tianhao Li, Tingyu Xia, Xingzhang Ren, Xuancheng Ren, Yang Fan, Yang Su, Yichang Zhang, Yu Wan, Yuqiong Liu, Zeyu Cui, Zhenru Zhang, Zihan Qiu•Dec 19, 2024•36311
MiniMax-01: Schalen van Foundation Modellen met Bliksem AandachtMiniMax-01: Scaling Foundation Models with Lightning Attention
MiniMax-01: Schalen van Foundation Modellen met Bliksem Aandacht
MiniMax-01: Scaling Foundation Models with Lightning Attention
MiniMax, Aonian Li, Bangwei Gong, Bo Yang, Boji Shan, Chang Liu, Cheng Zhu, Chunhao Zhang, Congchao Guo, Da Chen, Dong Li, Enwei Jiao, Gengxin Li, Guojun Zhang, Haohai Sun, Houze Dong, Jiadai Zhu, Jiaqi Zhuang, Jiayuan Song, Jin Zhu, Jingtao Han, Jingyang Li, Junbin Xie, Junhao Xu, Junjie Yan, Kaishun Zhang, Kecheng Xiao, Kexi Kang, Le Han, Leyang Wang, Lianfei Yu, Liheng Feng, Lin Zheng, Linbo Chai, Long Xing, Meizhi Ju, Mingyuan Chi, Mozhi Zhang, Peikai Huang, Pengcheng Niu, Pengfei Li, Pengyu Zhao, Qi Yang, Qidi Xu, Qiexiang Wang, Qin Wang, Qiuhui Li, Ruitao Leng, Shengmin Shi, Shuqi Yu, Sichen Li, Songquan Zhu, Tao Huang, Tianrun Liang, Weigao Sun, Weixuan Sun, Weiyu Cheng, Wenkai Li, Xiangjun Song, Xiao Su, Xiaodong Han, Xinjie Zhang, Xinzhu Hou, Xu Min, Xun Zou, Xuyang Shen, Yan Gong, Yingjie Zhu, Yipeng Zhou, Yiran Zhong, Yongyi Hu, Yuanxiang Fan, Yue Yu, Yufeng Yang, Yuhao Li, Yunan Huang, Yunji Li, Yunpeng Huang, Yunzhi Xu, Yuxin Mao, Zehan Li, Zekang Li, Zewei Tao, Zewen Ying, Zhaoyang Cong, Zhen Qin, Zhenhua Fan, Zhihang Yu, Zhuo Jiang, Zijia Wu•Jan 14, 2025•2826