Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Mingxuan Wang

Mingxuan Wang

LightSeq2: Accelerated Training for Transformer-based Models on GPUs


Oct 27, 2021
Xiaohui Wang, Ying Xiong, Xian Qian, Yang Wei, Lei Li, Mingxuan Wang

Figure 1 for LightSeq2: Accelerated Training for Transformer-based Models on GPUs
Figure 2 for LightSeq2: Accelerated Training for Transformer-based Models on GPUs
Figure 3 for LightSeq2: Accelerated Training for Transformer-based Models on GPUs
Figure 4 for LightSeq2: Accelerated Training for Transformer-based Models on GPUs
* 12 pages, 17 figures 
  Access Paper or Ask Questions

LightSeq: Accelerated Training for Transformer-based Models on GPUs


Oct 12, 2021
Xiaohui Wang, Ying Xiong, Xian Qian, Yang Wei, Lei Li, Mingxuan Wang

Figure 1 for LightSeq: Accelerated Training for Transformer-based Models on GPUs
Figure 2 for LightSeq: Accelerated Training for Transformer-based Models on GPUs
Figure 3 for LightSeq: Accelerated Training for Transformer-based Models on GPUs
Figure 4 for LightSeq: Accelerated Training for Transformer-based Models on GPUs
* 12 pages, 17 figures 
  Access Paper or Ask Questions

The Volctrans GLAT System: Non-autoregressive Translation Meets WMT21


Sep 24, 2021
Lihua Qian, Yi Zhou, Zaixiang Zheng, Yaoming Zhu, Zehui Lin, Jiangtao Feng, Shanbo Cheng, Lei Li, Mingxuan Wang, Hao Zhou

Figure 1 for The Volctrans GLAT System: Non-autoregressive Translation Meets WMT21
Figure 2 for The Volctrans GLAT System: Non-autoregressive Translation Meets WMT21
Figure 3 for The Volctrans GLAT System: Non-autoregressive Translation Meets WMT21
Figure 4 for The Volctrans GLAT System: Non-autoregressive Translation Meets WMT21
* 10 pages, 5 figures, WMT2021 
  Access Paper or Ask Questions

Learning Kernel-Smoothed Machine Translation with Retrieved Examples


Sep 21, 2021
Qingnan Jiang, Mingxuan Wang, Jun Cao, Shanbo Cheng, Shujian Huang, Lei Li

Figure 1 for Learning Kernel-Smoothed Machine Translation with Retrieved Examples
Figure 2 for Learning Kernel-Smoothed Machine Translation with Retrieved Examples
Figure 3 for Learning Kernel-Smoothed Machine Translation with Retrieved Examples
Figure 4 for Learning Kernel-Smoothed Machine Translation with Retrieved Examples
* EMNLP 2021 
  Access Paper or Ask Questions

UniST: Unified End-to-end Model for Streaming and Non-streaming Speech Translation


Sep 15, 2021
Qianqian Dong, Yaoming Zhu, Mingxuan Wang, Lei Li

Figure 1 for UniST: Unified End-to-end Model for Streaming and Non-streaming Speech Translation
Figure 2 for UniST: Unified End-to-end Model for Streaming and Non-streaming Speech Translation
Figure 3 for UniST: Unified End-to-end Model for Streaming and Non-streaming Speech Translation
Figure 4 for UniST: Unified End-to-end Model for Streaming and Non-streaming Speech Translation

  Access Paper or Ask Questions

Multilingual Translation via Grafting Pre-trained Language Models


Sep 11, 2021
Zewei Sun, Mingxuan Wang, Lei Li

Figure 1 for Multilingual Translation via Grafting Pre-trained Language Models
Figure 2 for Multilingual Translation via Grafting Pre-trained Language Models
Figure 3 for Multilingual Translation via Grafting Pre-trained Language Models
Figure 4 for Multilingual Translation via Grafting Pre-trained Language Models
* Accepted in EMNLP 2021 (Findings) 
  Access Paper or Ask Questions

Secoco: Self-Correcting Encoding for Neural Machine Translation


Aug 27, 2021
Tao Wang, Chengqi Zhao, Mingxuan Wang, Lei Li, Hang Li, Deyi Xiong

Figure 1 for Secoco: Self-Correcting Encoding for Neural Machine Translation
Figure 2 for Secoco: Self-Correcting Encoding for Neural Machine Translation
Figure 3 for Secoco: Self-Correcting Encoding for Neural Machine Translation
Figure 4 for Secoco: Self-Correcting Encoding for Neural Machine Translation
* 6 pages, 2 figures, 3 tables 
  Access Paper or Ask Questions

Language Tags Matter for Zero-Shot Neural Machine Translation


Jun 15, 2021
Liwei Wu, Shanbo Cheng, Mingxuan Wang, Lei Li

Figure 1 for Language Tags Matter for Zero-Shot Neural Machine Translation
Figure 2 for Language Tags Matter for Zero-Shot Neural Machine Translation
Figure 3 for Language Tags Matter for Zero-Shot Neural Machine Translation
Figure 4 for Language Tags Matter for Zero-Shot Neural Machine Translation
* 7 pages, 3 figures, Accepted by the Findings of ACL2021 
  Access Paper or Ask Questions

Contrastive Learning for Many-to-many Multilingual Neural Machine Translation


Jun 09, 2021
Xiao Pan, Mingxuan Wang, Liwei Wu, Lei Li

Figure 1 for Contrastive Learning for Many-to-many Multilingual Neural Machine Translation
Figure 2 for Contrastive Learning for Many-to-many Multilingual Neural Machine Translation
Figure 3 for Contrastive Learning for Many-to-many Multilingual Neural Machine Translation
Figure 4 for Contrastive Learning for Many-to-many Multilingual Neural Machine Translation
* ACL2021 
  Access Paper or Ask Questions

Learning Shared Semantic Space for Speech-to-Text Translation


Jun 03, 2021
Chi Han, Mingxuan Wang, Heng Ji, Lei Li

Figure 1 for Learning Shared Semantic Space for Speech-to-Text Translation
Figure 2 for Learning Shared Semantic Space for Speech-to-Text Translation
Figure 3 for Learning Shared Semantic Space for Speech-to-Text Translation
Figure 4 for Learning Shared Semantic Space for Speech-to-Text Translation
* 9 pages, 5 figures, Accepted by Findings of ACL 2021 
  Access Paper or Ask Questions

Learning Language Specific Sub-network for Multilingual Machine Translation


May 19, 2021
Zehui Lin, Liwei Wu, Mingxuan Wang, Lei Li

Figure 1 for Learning Language Specific Sub-network for Multilingual Machine Translation
Figure 2 for Learning Language Specific Sub-network for Multilingual Machine Translation
Figure 3 for Learning Language Specific Sub-network for Multilingual Machine Translation
Figure 4 for Learning Language Specific Sub-network for Multilingual Machine Translation
* To appear at ACL2021 
  Access Paper or Ask Questions

The Volctrans Neural Speech Translation System for IWSLT 2021


May 16, 2021
Chengqi Zhao, Zhicheng Liu, Jian Tong, Tao Wang, Mingxuan Wang, Rong Ye, Qianqian Dong, Jun Cao, Lei Li

Figure 1 for The Volctrans Neural Speech Translation System for IWSLT 2021
Figure 2 for The Volctrans Neural Speech Translation System for IWSLT 2021
Figure 3 for The Volctrans Neural Speech Translation System for IWSLT 2021
Figure 4 for The Volctrans Neural Speech Translation System for IWSLT 2021

  Access Paper or Ask Questions

Autocorrect in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation


Apr 21, 2021
Tao Wang, Chengqi Zhao, Mingxuan Wang, Lei Li, Deyi Xiong

Figure 1 for Autocorrect in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
Figure 2 for Autocorrect in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
Figure 3 for Autocorrect in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
Figure 4 for Autocorrect in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
* 8 pages, 3 figures, 7 tables 
  Access Paper or Ask Questions

End-to-end Speech Translation via Cross-modal Progressive Training


Apr 21, 2021
Rong Ye, Mingxuan Wang, Lei Li

Figure 1 for End-to-end Speech Translation via Cross-modal Progressive Training
Figure 2 for End-to-end Speech Translation via Cross-modal Progressive Training
Figure 3 for End-to-end Speech Translation via Cross-modal Progressive Training
Figure 4 for End-to-end Speech Translation via Cross-modal Progressive Training

  Access Paper or Ask Questions

Serial or Parallel? Plug-able Adapter for multilingual machine translation


Apr 16, 2021
Yaoming Zhu, Jiangtao Feng, Chengqi Zhao, Mingxuan Wang, Lei Li

Figure 1 for Serial or Parallel? Plug-able Adapter for multilingual machine translation
Figure 2 for Serial or Parallel? Plug-able Adapter for multilingual machine translation
Figure 3 for Serial or Parallel? Plug-able Adapter for multilingual machine translation
Figure 4 for Serial or Parallel? Plug-able Adapter for multilingual machine translation
* 13 pages 
  Access Paper or Ask Questions

Auto Correcting in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation


Mar 30, 2021
Tao Wang, Chengqi Zhao, Mingxuan Wang, Lei Li, Deyi Xiong

Figure 1 for Auto Correcting in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
Figure 2 for Auto Correcting in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
Figure 3 for Auto Correcting in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
Figure 4 for Auto Correcting in the Process of Translation -- Multi-task Learning Improves Dialogue Machine Translation
* 8 pages, 3 figures, 7 tables 
  Access Paper or Ask Questions

Finding Sparse Structure for Domain Specific Neural Machine Translation


Dec 19, 2020
Jianze Liang, Chengqi Zhao, Mingxuan Wang, Xipeng Qiu, Lei Li

Figure 1 for Finding Sparse Structure for Domain Specific Neural Machine Translation
Figure 2 for Finding Sparse Structure for Domain Specific Neural Machine Translation
Figure 3 for Finding Sparse Structure for Domain Specific Neural Machine Translation
Figure 4 for Finding Sparse Structure for Domain Specific Neural Machine Translation
* Accepted to AAAI 2021 
  Access Paper or Ask Questions

NeurST: Neural Speech Translation Toolkit


Dec 18, 2020
Chengqi Zhao, Mingxuan Wang, Lei Li

Figure 1 for NeurST: Neural Speech Translation Toolkit
Figure 2 for NeurST: Neural Speech Translation Toolkit
Figure 3 for NeurST: Neural Speech Translation Toolkit

  Access Paper or Ask Questions

Reciprocal Supervised Learning Improves Neural Machine Translation


Dec 05, 2020
Minkai Xu, Mingxuan Wang, Zhouhan Lin, Hao Zhou, Weinan Zhang, Lei Li

Figure 1 for Reciprocal Supervised Learning Improves Neural Machine Translation
Figure 2 for Reciprocal Supervised Learning Improves Neural Machine Translation
Figure 3 for Reciprocal Supervised Learning Improves Neural Machine Translation
Figure 4 for Reciprocal Supervised Learning Improves Neural Machine Translation

  Access Paper or Ask Questions

On the Sentence Embeddings from Pre-trained Language Models


Nov 02, 2020
Bohan Li, Hao Zhou, Junxian He, Mingxuan Wang, Yiming Yang, Lei Li

Figure 1 for On the Sentence Embeddings from Pre-trained Language Models
Figure 2 for On the Sentence Embeddings from Pre-trained Language Models
Figure 3 for On the Sentence Embeddings from Pre-trained Language Models
Figure 4 for On the Sentence Embeddings from Pre-trained Language Models
* EMNLP 2020 
  Access Paper or Ask Questions

The Volctrans Machine Translation System for WMT20


Oct 28, 2020
Liwei Wu, Xiao Pan, Zehui Lin, Yaoming Zhu, Mingxuan Wang, Lei Li

Figure 1 for The Volctrans Machine Translation System for WMT20
Figure 2 for The Volctrans Machine Translation System for WMT20
Figure 3 for The Volctrans Machine Translation System for WMT20
Figure 4 for The Volctrans Machine Translation System for WMT20

  Access Paper or Ask Questions

LightSeq: A High Performance Inference Library for Sequence Processing and Generation


Oct 28, 2020
Xiaohui Wang, Ying Xiong, Yang Wei, Mingxuan Wang, Lei Li

Figure 1 for LightSeq: A High Performance Inference Library for Sequence Processing and Generation
Figure 2 for LightSeq: A High Performance Inference Library for Sequence Processing and Generation
Figure 3 for LightSeq: A High Performance Inference Library for Sequence Processing and Generation
Figure 4 for LightSeq: A High Performance Inference Library for Sequence Processing and Generation
* 6 pages, 8 figures 
  Access Paper or Ask Questions

Volctrans Parallel Corpus Filtering System for WMT 2020


Oct 27, 2020
Runxin Xu, Zhuo Zhi, Jun Cao, Mingxuan Wang, Lei Li

Figure 1 for Volctrans Parallel Corpus Filtering System for WMT 2020
Figure 2 for Volctrans Parallel Corpus Filtering System for WMT 2020
Figure 3 for Volctrans Parallel Corpus Filtering System for WMT 2020
Figure 4 for Volctrans Parallel Corpus Filtering System for WMT 2020
* WMT 2020 
  Access Paper or Ask Questions

Capturing Longer Context for Document-level Neural Machine Translation: A Multi-resolutional Approach


Oct 18, 2020
Zewei Sun, Mingxuan Wang, Hao Zhou, Chengqi Zhao, Shujian Huang, Jiajun Chen, Lei Li

Figure 1 for Capturing Longer Context for Document-level Neural Machine Translation: A Multi-resolutional Approach
Figure 2 for Capturing Longer Context for Document-level Neural Machine Translation: A Multi-resolutional Approach
Figure 3 for Capturing Longer Context for Document-level Neural Machine Translation: A Multi-resolutional Approach
Figure 4 for Capturing Longer Context for Document-level Neural Machine Translation: A Multi-resolutional Approach

  Access Paper or Ask Questions

Pre-training Multilingual Neural Machine Translation by Leveraging Alignment Information


Oct 07, 2020
Zehui Lin, Xiao Pan, Mingxuan Wang, Xipeng Qiu, Jiangtao Feng, Hao Zhou, Lei Li

Figure 1 for Pre-training Multilingual Neural Machine Translation by Leveraging Alignment Information
Figure 2 for Pre-training Multilingual Neural Machine Translation by Leveraging Alignment Information
Figure 3 for Pre-training Multilingual Neural Machine Translation by Leveraging Alignment Information
Figure 4 for Pre-training Multilingual Neural Machine Translation by Leveraging Alignment Information
* EMNLP 2020 
  Access Paper or Ask Questions

SDST: Successive Decoding for Speech-to-text Translation


Sep 21, 2020
Qianqian Dong, Mingxuan Wang, Hao Zhou, Shuang Xu, Bo Xu, Lei Li

Figure 1 for SDST: Successive Decoding for Speech-to-text Translation
Figure 2 for SDST: Successive Decoding for Speech-to-text Translation
Figure 3 for SDST: Successive Decoding for Speech-to-text Translation
Figure 4 for SDST: Successive Decoding for Speech-to-text Translation

  Access Paper or Ask Questions

TED: Triple Supervision Decouples End-to-end Speech-to-text Translation


Sep 21, 2020
Qianqian Dong, Mingxuan Wang, Hao Zhou, Shuang Xu, Bo Xu, Lei Li

Figure 1 for TED: Triple Supervision Decouples End-to-end Speech-to-text Translation
Figure 2 for TED: Triple Supervision Decouples End-to-end Speech-to-text Translation
Figure 3 for TED: Triple Supervision Decouples End-to-end Speech-to-text Translation
Figure 4 for TED: Triple Supervision Decouples End-to-end Speech-to-text Translation

  Access Paper or Ask Questions

Generative Imagination Elevates Machine Translation


Sep 21, 2020
Quanyu Long, Mingxuan Wang, Lei Li

Figure 1 for Generative Imagination Elevates Machine Translation
Figure 2 for Generative Imagination Elevates Machine Translation
Figure 3 for Generative Imagination Elevates Machine Translation
Figure 4 for Generative Imagination Elevates Machine Translation

  Access Paper or Ask Questions