Yueying (Lisa) Li MIT
时间： 2024-01-10 18:15-2024-01-10 19:30
地点：腾讯会议链接： https://meeting.tencent.com/dm/VDqJWzaCNSfd #腾讯会议会议号：295-829-408
Scaling application's performance while improving system resource efficiencies has become an increasingly important agenda for both cloud providers and users. With the rise of emerging interactive applications such as LLM and microservices, users must build applications that satisfy microsecond scale tail latency service level objectives; and with emphasis on sustainability, cloud providers aim to improve user experience while reducing their resource footprints.
In this talk, I will discuss the design of a general-purpose, efficient and adaptive frameworks for efficient resource allocation and scheduling. First, I will introduce LibPreemptible, a fast, scalable and hardware-assisted user-space scheduling library that is designed for microsecond-scale workloads. Second, I will introduce FastAgent, a LLM-powered Agent System that enables memory-aware request scheduling. Finally, I will present PolicyShedder, a data driven offline RL framework for efficient, safe and rate limiter for reliable distributed system.
Lisa Li is a CS PhD student in Cornell and a visiting PhD in MIT. She is interested in distributed systems, with current focus on efficiency and reliability problems in cloud computing. She also works on reinforcement learning, and efficient LLM serving. She worked at Apple designing CPU after graduation from SJTU and UM with dual degree in ECE, CE, and minor in Math. She has received Top Scholar Award from University of Washington, Cornell Fellowship, and nominated to Microsoft Research Graduate Fellowship by Cornell. She is passionate about mentorship in CS community and serves on CASA committee and CALM (a long term mentorship program).