Jiamin Li

she/her/hers

Ph.D. in Computer Science

Email: jiaminli.icy [at] gmail.com

Resume

Linkedin, Google Scholar

About

I am a researcher at Microsoft Research Vancouver, working on AI infrastructure. I received my Ph.D. in Computer Science in Feb 2024 and B.S. in Computer Science in 2019, both from City University of Hong Kong. My Ph.D. advisors are Prof. Cong Wang and Prof. Hong Xu. In 2023, I was a visiting researcher at The University of Texas at Austin, advised by Prof. Aditya Akella.

Research Topics

My recent research interest is in the area of Systems for Machine Learning, with a specific focus on the following topics:

• Systems tailored for large language models: Optimizing LLM training and inference

• Resource scheduling in GPU clusters: Designing efficient scheduling algorithms for DNN workloads

• Distributed training & inference: Accelerating large-scale distributed DNN tasks

• Adaptive & sparse computation: Exploring new computing paradigms to effectively scale DNN models

Publications

Conference Proceedings

• [C4] Jiamin Li, Qiang Su, Yitao Yang, Yimin Jiang, Cong Wang, Hong Xu "Adaptive Gating in Mixture-of-Experts based Language Models", Empirical Methods in Natural Language Processing (EMNLP), 2023. [Paper][Slides]

• [C3] Jiamin Li, Yimin Jiang, Yibo Zhu, Cong Wang, Hong Xu, "Accelerating Distributed MoE Training and Inference with Lina", USENIX Annual Technical Conference (ATC), 2023. [Paper][Slides][Poster]

• [C2] Jiamin Li, Hong Xu, Yibo Zhu, Zherui Liu, Chuanxiong Guo, Cong Wang, "Lyra: Elastic Scheduling for Deep Learning Clusters", ACM European Conference on Computer Systems (EuroSys), 2023. [Paper][Slides][Poster]

• [C1] Kaiwei Mo, Chen Chen, Jiamin Li, Hong Xu, Chun Jason Xue, "Two-Dimensional Learning Rate Decay: Towards Accurate Federated Learning with Non-IID Data", IEEE International Joint Conference on Neural Networks (IJCNN), 2021.

Journals

• [J2] Libin Liu, Hong Xu, Zhixiong Niu, Jingzong Li, Wei Zhang, Peng Wang, Jiamin Li, Jason Xue Chun, Cong Wang, "ScaleFlux: Efficient Stateful Scaling in NFV", IEEE Transactions on Parallel and Distributed Systems (TPDS), 2022.

• [J1] Libin Liu, Chengxi Gao, Peng Wang, Hongming Huang, Jiamin Li, Hong Xu, Wei Zhang, "Bottleneck-Aware Non-Clairvoyant Coflow Scheduling with Fai", IEEE Transactions on Cloud Computing (TCC), 2021.

Research & Work Experience

Visiting Researcher @ UTNS, The University of Texas at Austin

• Supervised by Prof. Aditya Akella.

• Apr 2023 - Dec 2023, Austin, TX, United States

Research Intern @ MLSys Team of AI Lab, ByteDance

• Supervised by Dr. Yibo Zhu.

• May 2019 - May 2021, Beijing & Shenzhen, China

Part-time Research Assistant @ NetX, City University of Hong Kong

• Supervised by Prof. Hong Xu.

• June 2018 - May 2019, Kowloon Tong, Hong Kong

Software Developer @ Group Human Resources, Jardine Matheson

• May 2017 - May 2018, Central & Quarry Bay, Hong Kong

Teaching Assistant

CS2311, Computer Programming: Sem A, 2022/2023, Sem A, 2019/2020

CS4296 & CS5296, Cloud Computing: Sem B, 2021/2022, Sem B, 2019/2020

CS4394 & CS5294, Information Security and Management: Sem A, 2021/2022

CS4293 & CS6290, Topics on Computer Security: Sem B, 2020/2021

CS5222, Computer Networks and Internets: Sem A, 2020/2021

Professional Services

Artifact Evaluation Committee: ACM SOSP 2023, MLSys 2023, ACM EuroSys 2023, ACM CoNEXT 2022, USENIX OSDI 2023, USENIX ATC 2023, ACM EuroSys 2024, USENIX FAST 2024

Technical Program Committee: IEEE IJCNN 2023

Reviewer: IEEE/ACM Transactions on Networking (ToN), IEEE Transactions on Services Computing

Misc

Student Travel Grant, USENIX ATC, 2023

ML and Systems Rising Stars, ML Commons, 2023

Research Activity Funds, City University of Hong Kong, 2023

Student Travel Grant, ACM EuroSys, 2023

Full Postgraduate Studentship, City University of Hong Kong, 2019-2024

Dean’s List (College of Engineering), City University of Hong Kong, 2015-2019