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 Wei Lin

Wei Lin

M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining


Oct 25, 2021
Junyang Lin, An Yang, Jinze Bai, Chang Zhou, Le Jiang, Xianyan Jia, Ang Wang, Jie Zhang, Yong Li, Wei Lin, Jingren Zhou, Hongxia Yang

Figure 1 for M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining
Figure 2 for M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining
Figure 3 for M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining
Figure 4 for M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining
* 14 pages, 4 figures 
  Access Paper or Ask Questions

Binary Code based Hash Embedding for Web-scale Applications


Aug 24, 2021
Bencheng Yan, Pengjie Wang, Jinquan Liu, Wei Lin, Kuang-Chih Lee, Jian Xu, Bo Zheng

Figure 1 for Binary Code based Hash Embedding for Web-scale Applications
Figure 2 for Binary Code based Hash Embedding for Web-scale Applications
Figure 3 for Binary Code based Hash Embedding for Web-scale Applications
Figure 4 for Binary Code based Hash Embedding for Web-scale Applications
* CIKM 2021, 5 pages; The first two authors contributed equally to this work 
  Access Paper or Ask Questions

Learning Effective and Efficient Embedding via an Adaptively-Masked Twins-based Layer


Aug 24, 2021
Bencheng Yan, Pengjie Wang, Kai Zhang, Wei Lin, Kuang-Chih Lee, Jian Xu, Bo Zheng

Figure 1 for Learning Effective and Efficient Embedding via an Adaptively-Masked Twins-based Layer
Figure 2 for Learning Effective and Efficient Embedding via an Adaptively-Masked Twins-based Layer
Figure 3 for Learning Effective and Efficient Embedding via an Adaptively-Masked Twins-based Layer
Figure 4 for Learning Effective and Efficient Embedding via an Adaptively-Masked Twins-based Layer
* CIKM 2021, 5 pages; The first two authors contributed equally to this work 
  Access Paper or Ask Questions

Boosting the Convergence of Reinforcement Learning-based Auto-pruning Using Historical Data


Jul 16, 2021
Jiandong Mu, Mengdi Wang, Feiwen Zhu, Jun Yang, Wei Lin, Wei Zhang

Figure 1 for Boosting the Convergence of Reinforcement Learning-based Auto-pruning Using Historical Data
Figure 2 for Boosting the Convergence of Reinforcement Learning-based Auto-pruning Using Historical Data
Figure 3 for Boosting the Convergence of Reinforcement Learning-based Auto-pruning Using Historical Data
Figure 4 for Boosting the Convergence of Reinforcement Learning-based Auto-pruning Using Historical Data

  Access Paper or Ask Questions

Exploring Sparse Expert Models and Beyond


Jun 14, 2021
An Yang, Junyang Lin, Rui Men, Chang Zhou, Le Jiang, Xianyan Jia, Ang Wang, Jie Zhang, Jiamang Wang, Yong Li, Di Zhang, Wei Lin, Lin Qu, Jingren Zhou, Hongxia Yang

Figure 1 for Exploring Sparse Expert Models and Beyond
Figure 2 for Exploring Sparse Expert Models and Beyond
Figure 3 for Exploring Sparse Expert Models and Beyond
Figure 4 for Exploring Sparse Expert Models and Beyond
* 16 pages, 8 figures 
  Access Paper or Ask Questions

Harmless Overparametrization in Two-layer Neural Networks


Jun 09, 2021
Huiyuan Wang, Wei Lin


  Access Paper or Ask Questions

Explicit Semantic Cross Feature Learning via Pre-trained Graph Neural Networks for CTR Prediction


May 17, 2021
Feng Li, Bencheng Yan, Qingqing Long, Pengjie Wang, Wei Lin, Jian Xu, Bo Zheng

Figure 1 for Explicit Semantic Cross Feature Learning via Pre-trained Graph Neural Networks for CTR Prediction
Figure 2 for Explicit Semantic Cross Feature Learning via Pre-trained Graph Neural Networks for CTR Prediction
Figure 3 for Explicit Semantic Cross Feature Learning via Pre-trained Graph Neural Networks for CTR Prediction
Figure 4 for Explicit Semantic Cross Feature Learning via Pre-trained Graph Neural Networks for CTR Prediction
* SIGIR 2021, 5 pages; The first two authors contributed equally to this work; Pengjie Wang gave a lot of guidance in this work 
  Access Paper or Ask Questions

Towards a Better Tradeoff between Effectiveness and Efficiency in Pre-Ranking: A Learnable Feature Selection based Approach


May 17, 2021
Xu Ma, Pengjie Wang, Hui Zhao, Shaoguo Liu, Chuhan Zhao, Wei Lin, Kuang-Chih Lee, Jian Xu, Bo Zheng

Figure 1 for Towards a Better Tradeoff between Effectiveness and Efficiency in Pre-Ranking: A Learnable Feature Selection based Approach
Figure 2 for Towards a Better Tradeoff between Effectiveness and Efficiency in Pre-Ranking: A Learnable Feature Selection based Approach
Figure 3 for Towards a Better Tradeoff between Effectiveness and Efficiency in Pre-Ranking: A Learnable Feature Selection based Approach
Figure 4 for Towards a Better Tradeoff between Effectiveness and Efficiency in Pre-Ranking: A Learnable Feature Selection based Approach

  Access Paper or Ask Questions

M6: A Chinese Multimodal Pretrainer


Mar 02, 2021
Junyang Lin, Rui Men, An Yang, Chang Zhou, Ming Ding, Yichang Zhang, Peng Wang, Ang Wang, Le Jiang, Xianyan Jia, Jie Zhang, Jianwei Zhang, Xu Zou, Zhikang Li, Xiaodong Deng, Jie Liu, Jinbao Xue, Huiling Zhou, Jianxin Ma, Jin Yu, Yong Li, Wei Lin, Jingren Zhou, Jie Tang, Hongxia Yang

Figure 1 for M6: A Chinese Multimodal Pretrainer
Figure 2 for M6: A Chinese Multimodal Pretrainer
Figure 3 for M6: A Chinese Multimodal Pretrainer
Figure 4 for M6: A Chinese Multimodal Pretrainer
* 12 pages, technical report 
  Access Paper or Ask Questions

Neural Delay Differential Equations


Feb 22, 2021
Qunxi Zhu, Yao Guo, Wei Lin

Figure 1 for Neural Delay Differential Equations
Figure 2 for Neural Delay Differential Equations
Figure 3 for Neural Delay Differential Equations
Figure 4 for Neural Delay Differential Equations
* Accepted as a poster in ICLR 2021 (submitted 28 Sep 2020, revised 22 Nov 2020, accepted 08 Jan 2021) 
  Access Paper or Ask Questions

Focusing More on Conflicts with Mis-Predictions Helps Language Pre-Training


Dec 16, 2020
Chen Xing, Wencong Xiao, Yong Li, Wei Lin

Figure 1 for Focusing More on Conflicts with Mis-Predictions Helps Language Pre-Training
Figure 2 for Focusing More on Conflicts with Mis-Predictions Helps Language Pre-Training
Figure 3 for Focusing More on Conflicts with Mis-Predictions Helps Language Pre-Training
Figure 4 for Focusing More on Conflicts with Mis-Predictions Helps Language Pre-Training

  Access Paper or Ask Questions

EasyTransfer -- A Simple and Scalable Deep Transfer Learning Platform for NLP Applications


Nov 23, 2020
Minghui Qiu, Peng Li, Hanjie Pan, Chengyu Wang, Ang Wang, Cen Chen, Yaliang Li, Dehong Gao, Jun Huang, Yong Li, Jun Yang, Deng Cai, Wei Lin

Figure 1 for EasyTransfer -- A Simple and Scalable Deep Transfer Learning Platform for NLP Applications
Figure 2 for EasyTransfer -- A Simple and Scalable Deep Transfer Learning Platform for NLP Applications
Figure 3 for EasyTransfer -- A Simple and Scalable Deep Transfer Learning Platform for NLP Applications
* 4 pages 
  Access Paper or Ask Questions

INT8 Winograd Acceleration for Conv1D Equipped ASR Models Deployed on Mobile Devices


Oct 28, 2020
Yiwu Yao, Yuchao Li, Chengyu Wang, Tianhang Yu, Houjiang Chen, Xiaotang Jiang, Jun Yang, Jun Huang, Wei Lin, Hui Shu, Chengfei Lv

Figure 1 for INT8 Winograd Acceleration for Conv1D Equipped ASR Models Deployed on Mobile Devices
Figure 2 for INT8 Winograd Acceleration for Conv1D Equipped ASR Models Deployed on Mobile Devices
Figure 3 for INT8 Winograd Acceleration for Conv1D Equipped ASR Models Deployed on Mobile Devices
Figure 4 for INT8 Winograd Acceleration for Conv1D Equipped ASR Models Deployed on Mobile Devices

  Access Paper or Ask Questions

FusionStitching: Boosting Memory Intensive Computations for Deep Learning Workloads


Sep 23, 2020
Zhen Zheng, Pengzhan Zhao, Guoping Long, Feiwen Zhu, Kai Zhu, Wenyi Zhao, Lansong Diao, Jun Yang, Wei Lin

Figure 1 for FusionStitching: Boosting Memory Intensive Computations for Deep Learning Workloads
Figure 2 for FusionStitching: Boosting Memory Intensive Computations for Deep Learning Workloads
Figure 3 for FusionStitching: Boosting Memory Intensive Computations for Deep Learning Workloads
Figure 4 for FusionStitching: Boosting Memory Intensive Computations for Deep Learning Workloads

  Access Paper or Ask Questions

One-shot Text Field Labeling using Attention and Belief Propagation for Structure Information Extraction


Sep 09, 2020
Mengli Cheng, Minghui Qiu, Xing Shi, Jun Huang, Wei Lin

Figure 1 for One-shot Text Field Labeling using Attention and Belief Propagation for Structure Information Extraction
Figure 2 for One-shot Text Field Labeling using Attention and Belief Propagation for Structure Information Extraction
Figure 3 for One-shot Text Field Labeling using Attention and Belief Propagation for Structure Information Extraction
Figure 4 for One-shot Text Field Labeling using Attention and Belief Propagation for Structure Information Extraction
* ACMMM 2020 
 * 9 pages 
  Access Paper or Ask Questions

Grasping Detection Network with Uncertainty Estimation for Confidence-Driven Semi-Supervised Domain Adaptation


Aug 20, 2020
Haiyue Zhu, Yiting Li, Fengjun Bai, Wenjie Chen, Xiaocong Li, Jun Ma, Chek Sing Teo, Pey Yuen Tao, Wei Lin

Figure 1 for Grasping Detection Network with Uncertainty Estimation for Confidence-Driven Semi-Supervised Domain Adaptation
Figure 2 for Grasping Detection Network with Uncertainty Estimation for Confidence-Driven Semi-Supervised Domain Adaptation
Figure 3 for Grasping Detection Network with Uncertainty Estimation for Confidence-Driven Semi-Supervised Domain Adaptation
Figure 4 for Grasping Detection Network with Uncertainty Estimation for Confidence-Driven Semi-Supervised Domain Adaptation
* 6 pages, 7 figures, accepted in IROS 2020 
  Access Paper or Ask Questions

Pixel-wise Crowd Understanding via Synthetic Data


Aug 03, 2020
Qi Wang, Junyu Gao, Wei Lin, Yuan Yuan

Figure 1 for Pixel-wise Crowd Understanding via Synthetic Data
Figure 2 for Pixel-wise Crowd Understanding via Synthetic Data
Figure 3 for Pixel-wise Crowd Understanding via Synthetic Data
Figure 4 for Pixel-wise Crowd Understanding via Synthetic Data
* Accepted by IJCV. arXiv admin note: text overlap with arXiv:1903.03303 
  Access Paper or Ask Questions

Interactive Feature Generation via Learning Adjacency Tensor of Feature Graph


Jul 29, 2020
Yuexiang Xie, Zhen Wang, Yaliang Li, Bolin Ding, Nezihe Merve Gürel, Ce Zhang, Minlie Huang, Wei Lin, Jingren Zhou

Figure 1 for Interactive Feature Generation via Learning Adjacency Tensor of Feature Graph
Figure 2 for Interactive Feature Generation via Learning Adjacency Tensor of Feature Graph
Figure 3 for Interactive Feature Generation via Learning Adjacency Tensor of Feature Graph
Figure 4 for Interactive Feature Generation via Learning Adjacency Tensor of Feature Graph

  Access Paper or Ask Questions

Auto-MAP: A DQN Framework for Exploring Distributed Execution Plans for DNN Workloads


Jul 08, 2020
Siyu Wang, Yi Rong, Shiqing Fan, Zhen Zheng, LanSong Diao, Guoping Long, Jun Yang, Xiaoyong Liu, Wei Lin

Figure 1 for Auto-MAP: A DQN Framework for Exploring Distributed Execution Plans for DNN Workloads
Figure 2 for Auto-MAP: A DQN Framework for Exploring Distributed Execution Plans for DNN Workloads
Figure 3 for Auto-MAP: A DQN Framework for Exploring Distributed Execution Plans for DNN Workloads
Figure 4 for Auto-MAP: A DQN Framework for Exploring Distributed Execution Plans for DNN Workloads

  Access Paper or Ask Questions

Graph Structural-topic Neural Network


Jul 04, 2020
Qingqing Long, Yilun Jin, Guojie Song, Yi Li, Wei Lin

Figure 1 for Graph Structural-topic Neural Network
Figure 2 for Graph Structural-topic Neural Network
Figure 3 for Graph Structural-topic Neural Network
Figure 4 for Graph Structural-topic Neural Network

  Access Paper or Ask Questions

SwapText: Image Based Texts Transfer in Scenes


Mar 18, 2020
Qiangpeng Yang, Hongsheng Jin, Jun Huang, Wei Lin

Figure 1 for SwapText: Image Based Texts Transfer in Scenes
Figure 2 for SwapText: Image Based Texts Transfer in Scenes
Figure 3 for SwapText: Image Based Texts Transfer in Scenes
Figure 4 for SwapText: Image Based Texts Transfer in Scenes
* Accepted by CVPR 2020 
  Access Paper or Ask Questions

AdaBERT: Task-Adaptive BERT Compression with Differentiable Neural Architecture Search


Jan 13, 2020
Daoyuan Chen, Yaliang Li, Minghui Qiu, Zhen Wang, Bofang Li, Bolin Ding, Hongbo Deng, Jun Huang, Wei Lin, Jingren Zhou

Figure 1 for AdaBERT: Task-Adaptive BERT Compression with Differentiable Neural Architecture Search
Figure 2 for AdaBERT: Task-Adaptive BERT Compression with Differentiable Neural Architecture Search
Figure 3 for AdaBERT: Task-Adaptive BERT Compression with Differentiable Neural Architecture Search
Figure 4 for AdaBERT: Task-Adaptive BERT Compression with Differentiable Neural Architecture Search

  Access Paper or Ask Questions

NWPU-Crowd: A Large-Scale Benchmark for Crowd Counting


Jan 10, 2020
Qi Wang, Junyu Gao, Wei Lin, Xuelong Li

Figure 1 for NWPU-Crowd: A Large-Scale Benchmark for Crowd Counting
Figure 2 for NWPU-Crowd: A Large-Scale Benchmark for Crowd Counting
Figure 3 for NWPU-Crowd: A Large-Scale Benchmark for Crowd Counting
Figure 4 for NWPU-Crowd: A Large-Scale Benchmark for Crowd Counting
* 8 pages, 3 figures 
  Access Paper or Ask Questions

RPM-Oriented Query Rewriting Framework for E-commerce Keyword-Based Sponsored Search


Oct 28, 2019
Xiuying Chen, Daorui Xiao, Shen Gao, Guojun Liu, Wei Lin, Bo Zheng, Dongyan Zhao, Rui Yan

Figure 1 for RPM-Oriented Query Rewriting Framework for E-commerce Keyword-Based Sponsored Search
Figure 2 for RPM-Oriented Query Rewriting Framework for E-commerce Keyword-Based Sponsored Search
Figure 3 for RPM-Oriented Query Rewriting Framework for E-commerce Keyword-Based Sponsored Search
* AAAI 2020 
 * 2 pages, 2 figures 
  Access Paper or Ask Questions

Characterizing Deep Learning Training Workloads on Alibaba-PAI


Oct 14, 2019
Mengdi Wang, Chen Meng, Guoping Long, Chuan Wu, Jun Yang, Wei Lin, Yangqing Jia

Figure 1 for Characterizing Deep Learning Training Workloads on Alibaba-PAI
Figure 2 for Characterizing Deep Learning Training Workloads on Alibaba-PAI
Figure 3 for Characterizing Deep Learning Training Workloads on Alibaba-PAI
Figure 4 for Characterizing Deep Learning Training Workloads on Alibaba-PAI
* Accepted by IISWC2019 
  Access Paper or Ask Questions

DL2: A Deep Learning-driven Scheduler for Deep Learning Clusters


Sep 13, 2019
Yanghua Peng, Yixin Bao, Yangrui Chen, Chuan Wu, Chen Meng, Wei Lin

Figure 1 for DL2: A Deep Learning-driven Scheduler for Deep Learning Clusters
Figure 2 for DL2: A Deep Learning-driven Scheduler for Deep Learning Clusters
Figure 3 for DL2: A Deep Learning-driven Scheduler for Deep Learning Clusters
Figure 4 for DL2: A Deep Learning-driven Scheduler for Deep Learning Clusters

  Access Paper or Ask Questions

C^3 Framework: An Open-source PyTorch Code for Crowd Counting


Jul 05, 2019
Junyu Gao, Wei Lin, Bin Zhao, Dong Wang, Chenyu Gao, Jun Wen

Figure 1 for C^3 Framework: An Open-source PyTorch Code for Crowd Counting
Figure 2 for C^3 Framework: An Open-source PyTorch Code for Crowd Counting
Figure 3 for C^3 Framework: An Open-source PyTorch Code for Crowd Counting
Figure 4 for C^3 Framework: An Open-source PyTorch Code for Crowd Counting
* Technical report 
  Access Paper or Ask Questions