浏览条目

浏览/检索结果: 共135条,第1-10条 帮助

限定条件                
已选(0)清除 条数/页:   排序方式:
A Systematic Study of Cross-Layer KV Sharing for Efficient LLM Inference 预印本
2024
作者:  Wu, You;  Wu, Haoyi;  Tu, Kewei
收藏  |  浏览/下载:8/0  |  提交时间:2024/11/11
Learning Robust Named Entity Recognizers From Noisy Data With Retrieval Augmentation 预印本
2024
作者:  Ai, Chaoyi;  Jiang, Yong;  Huang, Shen;  Xie, Pengjun;  Tu, Kewei
Adobe PDF(822Kb)  |  收藏  |  浏览/下载:76/0  |  提交时间:2024/08/14
Dependency Transformer Grammars: Integrating Dependency Structures into Transformer Language Models 预印本
2024
作者:  Zhao, Yida;  Lou, Chao;  Tu, Kewei
Adobe PDF(559Kb)  |  收藏  |  浏览/下载:92/2  |  提交时间:2024/08/14
用于大语言模型的层压缩键值缓存方法、系统、设备及介质 专利
申请号:CN202410423760.7,申请日期: 2024-06-25,类型:发明申请,状态:实质审查
发明人:  屠可伟;  吴昊一
Adobe PDF(731Kb)  |  收藏  |  浏览/下载:136/1  |  提交时间:2024/06/25
Sparser is Faster and Less is More: Efficient Sparse Attention for Long-Range Transformers 预印本
2024
作者:  Lou, Chao;  Jia, Zixia;  Zheng, Zilong;  Tu, Kewei
Adobe PDF(2152Kb)  |  收藏  |  浏览/下载:93/2  |  提交时间:2024/07/08
Layer-Condensed KV Cache for Efficient Inference of Large Language Models 预印本
2024
作者:  Wu, Haoyi;  Tu, Kewei
Adobe PDF(590Kb)  |  收藏  |  浏览/下载:115/2  |  提交时间:2024/06/17
句法语言模型的无监督训练方法及装置 专利
申请号:CN202410296243.8,申请日期: 2024-05-10,类型:发明申请,状态:实质审查
发明人:  胡翔;  武威;  屠可伟
Adobe PDF(829Kb)  |  收藏  |  浏览/下载:154/0  |  提交时间:2024/05/10
Potential and Limitations of LLMs in Capturing Structured Semantics: A Case Study on SRL 预印本
2024
作者:  Cheng, Ning;  Yan, Zhaohui;  Wang, Ziming;  Li, Zhijie;  Yu, Jiaming
Adobe PDF(680Kb)  |  收藏  |  浏览/下载:106/1  |  提交时间:2024/06/17
Using Interpretation Methods for Model Enhancement 预印本
2024
作者:  Chen, Zhuo;  Jiang, Chengyue;  Tu, Kewei
Adobe PDF(1566Kb)  |  收藏  |  浏览/下载:129/0  |  提交时间:2024/05/15
Frame Semantic Role Labeling Using Arbitrary-Order Conditional Random Fields 会议论文
PROCEEDINGS OF THE AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, Vancouver, BC, Canada, February 20, 2024 - February 27, 2024
作者:  Ai, Chaoyi;  Tu, Kewei
Adobe PDF(615Kb)  |  收藏  |  浏览/下载:167/4  |  提交时间:2024/04/26