Biography

I am a year 3 PhD student at The Chinese university of Hong Kong, Shenzhen (CUHKSZ), supervised by Prof. Xiao Li. Before that, I obtained my bachelor degree in computer science from CUHKSZ in 2021.

My research interest lies in understanding modern large-scale deep learning models through its empirical behaviors, including its optimization and generalization properties. Based on thses insights, my objective is to develop effective, explainable, and applicable model. Recently, I am interested in foundation models.

Contact: qijunluo at link.cuhk.edu.cn

Publications and Preprints

• BAdam: A Memory Efficient Full Parameter Training Method for Large Language Models. [arXiv]
  Qijun Luo, Hengxu Yu Xiao Li.

• Finite-Time Analysis of Fully Decentralized Single-Timescale Actor-Critic. [arXiv] [Openreview]
  Accepted by Transactions on Machine Learning Research, 2023.
  Qijun Luo, Xiao Li.

• Batch Group Normalization. [arXiv]
  Xiaoyun Zhou, Jiacheng Sun, Nanyang Ye, Xu Lan, Qijun Luo, Bolin Lai, Pedro Esperanca, Guangzhong Yang, Zhenguo Li.

• Relaxed Conditional Image Transfer for Semi-supervised Domain Adaptation. [arXiv]
  Qijun Luo*, Zhili Liu*, Lanqing Hong, Chongxuan Li, Kuo yang, Liyuan Wang, Fengwei Zhou, Guilin Li, Zhenguo Li, Jun Zhu.

* Equal contribution

Research Experiences

• The Chinese University of Hong Kong, Shen Zhen, Research Assistant, 2021.1-2021.5
• Huawei Noah’s Ark Lab, Research Intern, 2019.10-2020.9
• Shenzhen Research Institute of Big Data, Research Assistant, 2019.6-2019.8

Awards

• PhD fellowship, Shenzhen research institute of big data, 2022, 2023.
• Teaching assistant award, school of data science, CUHKSZ, 2023.

Teaching Experiences

• MDS5210: Machine Learning, 2022 spring & fall, 2023 fall. Teaching Assistant
• MAT3007: Optimization I, 2021 fall & summer. Teaching Assistant