CHEN Yuhan

Brief Biography

I am CHEN Yuhan, first year MPhil student in Data Science and Analysis Thrust of The Hong Kong University of Science and Technology (Guangzhou) supervised by Prof. CHU Xiaowen. I obtained my B.Eng. in Computer Science and Engineering from Sun Yat-sen University in 2023, supervised by Prof. CHEN Gang.

Research Interests

  • CUDA & Ascend C kernel speedup
  • LLM training & inference speedup
  • High performance ML system, multi-GPU system

Publications

  1. Towards Efficient and Reliable LLM Serving: A Real-World Workload Study, Yuxin Wang*, Yuhan Chen*, Zeyu Li, Zhenheng Tang, Rui Guo, Xin Wang, Qiang Wang, Amelie Chi Zhou, Xiaowen Chu**, arXiv preprint arXiv:2401.17644, 2024 Paper BurstGPT

Teaching

  • DSAA5013 - Advanced Machine Learning, Spring 2024, Teaching Assistant, HKUST(GZ)

Selected Competitions & Awards

  • TOP 8 (SS#6), Guangdong University Overwatch League 2024, DPS - Team HKUST(GZ)
  • Second place, Overwatch Metal League 2024 Guangzhou Division, DPS - Team Guangzhou Red Bird
  • First prize, 17th Challenge Cup National College Students' Extracurricular Academic Science and Technology Contest (Top 0.01% nationally)
  • Special prize, 16th Challenge Cup Guangdong College Students' Extracurricular Academic Science and Technology Contest

Experiences

Cost-effective All-in-One Machine for Large Language Models to Tackle Smart City Challenges
High-Performance Fine-Tuning Optimization For Large Language Models
2nd Cohort RBM Project
Supervised by Prof. CHU Xiaowen (Academic) & Dr. HUANG Tongge (Project)

WIP
Research and Analysis of Real-time Performance Metrics of RTOS
OS Kernel Research and Development Intern at ACOINFO Co., Ltd.
Supervised by Prof. CHEN Gang & Mr. WANG Dongfang (Technical Director of ACOINFO)
  • Conducted research and analysis of real-time performance metrics for real-time operating system, as part of the RTOS Key Research and Development Plans of Guangdong Province.
  • Developed a real-time testing program for SylixOS and evaluated its performance on x86, Loongson, ARM, and other platforms. Ported it to MS-RTOS, successfully deployed and evaluated its performance on a STM32F412RET6 platform.

Read More

Latest update on Feb. 27th, 2024.

Here are the change logs of this site. ( •̀ ω •́ )

简介

我是陈禹翰,香港科技大学(广州) 数据科学与分析学域硕士一年级学生,导师是褚晓文教授。我于2023年在中山大学获得计算机科学与技术工学学士学位,导师是陈刚教授

研究兴趣

  • CUDA 和 Ascend C 算子加速
  • 大语言模型训练和推理优化加速
  • 高性能机器学习系统,多卡、集群和分布式系统

Publications

  1. Towards Efficient and Reliable LLM Serving: A Real-World Workload Study, Yuxin Wang*, Yuhan Chen*, Zeyu Li, Zhenheng Tang, Rui Guo, Xin Wang, Qiang Wang, Amelie Chi Zhou, Xiaowen Chu**, arXiv preprint arXiv:2401.17644, 2024 Paper BurstGPT

教学经历

  • DSAA5013 - Advanced Machine Learning, 2024 春, 助教, 香港科技大学(广州)

部分竞赛获奖

  • 2024守望先锋广东高校联赛八强(瑞士轮#6) 输出 - 香港科技大学(广州)队
  • 2024守望先锋金属联赛广州赛区亚军 输出 - 广州红鸟队
  • 第十七届挑战杯全国大学生课外学术科技作品竞赛一等奖
  • 第十六届挑战杯广东省大学生课外学术科技作品竞赛特等奖

项目经历

Cost-effective All-in-One Machine for Large Language Models to Tackle Smart City Challenges
High-Performance Fine-Tuning Optimization For Large Language Models
2nd Cohort RBM Project
Supervised by Prof. CHU Xiaowen (Academic) & Dr. HUANG Tongge (Project)

WIP
广东省重点研发计划实时操作系统子项目实时操作系统实时性指标研究分析
翼辉信息有限公司操作系统内核研发实习
由陈刚教授和王东方技术总监指导

负责研究SylixOS和MS-RTOS内核实时性,分别针对两款操作系统编写内核测试程序测试不同平台上实时性能,优化内核实时性。

推荐阅读

最后更新于2024年2月27日。

这里是博客的维护记录( •̀ ω •́ )