Zobrazeno 1 - 10
of 21
pro vyhledávání: '"Majumder, Rangan"'
Autor:
Chen, Qi, Geng, Xiubo, Rosset, Corby, Buractaon, Carolyn, Lu, Jingwen, Shen, Tao, Zhou, Kun, Xiong, Chenyan, Gong, Yeyun, Bennett, Paul, Craswell, Nick, Xie, Xing, Yang, Fan, Tower, Bryan, Rao, Nikhil, Dong, Anlei, Jiang, Wenqi, Liu, Zheng, Li, Mingqin, Liu, Chuanjie, Li, Zengzhong, Majumder, Rangan, Neville, Jennifer, Oakley, Andy, Risvik, Knut Magne, Simhadri, Harsha Vardhan, Varma, Manik, Wang, Yujing, Yang, Linjun, Yang, Mao, Zhang, Ce
Recent breakthroughs in large models have highlighted the critical significance of data scale, labels and modals. In this paper, we introduce MS MARCO Web Search, the first large-scale information-rich web dataset, featuring millions of real clicked
Externí odkaz:
http://arxiv.org/abs/2405.07526
This technical report presents the training methodology and evaluation results of the open-source multilingual E5 text embedding models, released in mid-2023. Three embedding models of different sizes (small / base / large) are provided, offering a b
Externí odkaz:
http://arxiv.org/abs/2402.05672
In this paper, we introduce a novel and simple method for obtaining high-quality text embeddings using only synthetic data and less than 1k training steps. Unlike existing methods that often depend on multi-stage intermediate pre-training with billio
Externí odkaz:
http://arxiv.org/abs/2401.00368
Modern search engines are built on a stack of different components, including query understanding, retrieval, multi-stage ranking, and question answering, among others. These components are often optimized and deployed independently. In this paper, w
Externí odkaz:
http://arxiv.org/abs/2310.14587
Autor:
Yang, Nan, Ge, Tao, Wang, Liang, Jiao, Binxing, Jiang, Daxin, Yang, Linjun, Majumder, Rangan, Wei, Furu
We propose LLMA, an LLM accelerator to losslessly speed up Large Language Model (LLM) inference with references. LLMA is motivated by the observation that there are abundant identical text spans between the decoding result by an LLM and the reference
Externí odkaz:
http://arxiv.org/abs/2304.04487
Autor:
Sun, Hao, Liu, Xiao, Gong, Yeyun, Dong, Anlei, Lu, Jingwen, Zhang, Yan, Yang, Linjun, Majumder, Rangan, Duan, Nan
Knowledge distillation is often used to transfer knowledge from a strong teacher model to a relatively weak student model. Traditional methods include response-based methods and feature-based methods. Response-based methods are widely used but suffer
Externí odkaz:
http://arxiv.org/abs/2212.05225
Autor:
Wang, Liang, Yang, Nan, Huang, Xiaolong, Jiao, Binxing, Yang, Linjun, Jiang, Daxin, Majumder, Rangan, Wei, Furu
This paper presents E5, a family of state-of-the-art text embeddings that transfer well to a wide range of tasks. The model is trained in a contrastive manner with weak supervision signals from our curated large-scale text pair dataset (called CCPair
Externí odkaz:
http://arxiv.org/abs/2212.03533
Autor:
Zhou, Kun, Gong, Yeyun, Liu, Xiao, Zhao, Wayne Xin, Shen, Yelong, Dong, Anlei, Lu, Jingwen, Majumder, Rangan, Wen, Ji-Rong, Duan, Nan, Chen, Weizhu
Sampling proper negatives from a large document pool is vital to effectively train a dense retrieval model. However, existing negative sampling strategies suffer from the uninformative or false negative problem. In this work, we empirically show that
Externí odkaz:
http://arxiv.org/abs/2210.11773
Autor:
Lin, Zhenghao, Gong, Yeyun, Liu, Xiao, Zhang, Hang, Lin, Chen, Dong, Anlei, Jiao, Jian, Lu, Jingwen, Jiang, Daxin, Majumder, Rangan, Duan, Nan
Knowledge distillation is an effective way to transfer knowledge from a strong teacher to an efficient student model. Ideally, we expect the better the teacher is, the better the student. However, this expectation does not always come true. It is com
Externí odkaz:
http://arxiv.org/abs/2209.13335
Autor:
Wang, Liang, Yang, Nan, Huang, Xiaolong, Jiao, Binxing, Yang, Linjun, Jiang, Daxin, Majumder, Rangan, Wei, Furu
In this paper, we propose SimLM (Similarity matching with Language Model pre-training), a simple yet effective pre-training method for dense passage retrieval. It employs a simple bottleneck architecture that learns to compress the passage informatio
Externí odkaz:
http://arxiv.org/abs/2207.02578