showSidebars ==
showTitleBreadcrumbs == 1
node.field_disable_title_breadcrumbs.value ==

PhD Dissertation Defense by DU Cunxiao | Towards Faster Inference of Transformers: Strategies for Accelerating Decoding Processes

Please click here if you are unable to view this page.

 
 

Towards Faster Inference of Transformers: Strategies for Accelerating Decoding Processes

DU Cunxiao

PhD Candidate 
School of Computing and Information Systems 
Singapore Management University 
 

FULL PROFILE

Research Area

Dissertation Committee

Research Advisor

Dissertation Committee Member

External Member

  • Zhaopeng TU, Principal Researcher, Tencent AI Lab
 

Date

7 June 2024 (Friday)

Time

4:00pm – 5:00pm

Venue

Meeting room 5.1, Level 5
School of Computing and Information Systems 1, Singapore Management University, 80 Stamford Road Singapore 178902

Please register by 6 June 2024.

We look forward to seeing you at this research seminar.

 

ABOUT THE TALK

This thesis delves into the acceleration and optimization of Transformer inference, a subject of increasing importance with the emergence of Large Language Models (LLMs). The study primarily addresses the challenges posed by the inherent property of Transformers during inference: autoregressive nature.

 

ABOUT THE SPEAKER

DU Cunxiao is a PhD Candidate supervised by Prof. Jing JIANG. His research interests are LLM and MLsys. His life creed is LESS IS MORE.

© Copyright 2026 by Singapore Management University. All Rights Reserved.