Chaofan Lin (in Chinese: 林超凡)

I'm an undergraduate at ACM Honors Class, Shanghai Jiao Tong University, advised by Prof. Yong Yu. Now I'm very fortunate to work with Yuqing Yang in the System Group of Microsoft Research Asia. Previously I'm a research intern at Apache TVM community, working closely with Tianqi Chen and focusing on the design of relax (a graph-level IR for tvm) training workflow.

My research interest mainly lies in computer systems and architecture. I enjoy crafting systems and/or exploring efficient algorithms to bridge the gap between efficiency (low-level) and usability (user-side).

I believe there is no silver bullet in designing and architecting. Hence there is no perfect solution for all ad-hoc reallife problems. However, I will try my best to narrow the gap between theory and practice, hoping that one day everyone can appreciate and benefit from those great ideas in Computer Science.


Publications

[Publications]

Projects

apache/tvm Deep Learning Compilers, Graph-level IR, Automatic Differentiation.
Working on the TVM Unity compiler. Developed an algorithm of automatic differentiation (AD) and a training workflow in Relax IR.
Click [Projects] to see a full list including research projects, course projects and my hobby projects.

Talks & Teaching

The Journey

Internship | Microsoft Research Asia
June'20, 2023 - Present

Research topics: Distributed system, Large Language Models (LLMs) serving system.

Internship | Catalyst, Carnegie Mellon University
July'22, 2022 - June'23, 2023

Research topics: TVM, Deep learning compilers, Graph-level IR for deep learning.

B.S. in Computer Science | Shanghai Jiao Tong University
Sep'1, 2020 - Present

Member of ACM Class, Zhiyuan College. Got a distinguished GPA (1/?) during my B.S.

Miscellaneous


Copyright© 2022 Chaofan Lin | Last Edited, Nov. 20, 2023 | Total views: | Source code: Lambda, feel free to use and modify. 闽ICP备2022013458号