• SPARKLE: A Unified Single-Loop Primal-Dual Framework for Decentralized Bilevel Optimization
  Great Bay University, Dongguan, China
  August, 2025
  

• A Memory Efficient Subspace Optimization Method for Training Large Language Models
  PKU Workshop on Optimization Theory and Methods
  Peking University, Beijing, China
  June, 2025
  

• Subspace Optimization for Large Language Models with Convergence Guarantees
  The NUS-PKU-SJTU Workshop on Data Science and Machine Learning
  National University of Singapore, Singapore
  November, 2024
  

• A Mathematics-Inspired Learning-to-Optimize Framework for Decentralized Optimization
  ORSC 2024
  Guiyang, China
  October, 2024
  

• Efficient Optimization for Deep Learning: Part I
  Adaptive SGD
  Fudan University, Shanghai, China
  June, 2024
  

• Efficient Optimization for Deep Learning: Part II
  Fudan University, Shanghai, China
  June, 2024
  

• Efficient Optimization for Deep Learning: Part III
  Mixed-Precision Training
  Fudan University, Shanghai, China
  June, 2024
  

• Towards Decentralized Optimization over Digraphs: Effective metrics, lower bound, and optimal algorithms
  International Workshop on Stochastic Optimization
  Xiangtan, China
  April, 2024
  

• Asynchronous Diffusion Learning with Agent Subsampling and Local Updates
  IEEE International Conference on Acoustics, Speech and Signal Processing
  Seoul, Korea
  April, 2024
  

• Distributed Machine Learning: Part I
  Machine Learning Summer School
  OIST, Okinawa, Japan
  March, 2024
  

• Distributed Machine Learning: Part II
  Machine Learning Summer School
  OIST, Okinawa, Japan
  March, 2024
  

• Achieving Linear Speedup with Network-Independent Learning Rates in Decentralized Optimization
  The 62nd IEEE Conference on Decision and Control, Singapore
  December, 2023
  

• A Unified Algorithm for Non-IID Decentralized Learning
  Shenzhen University, Shenzhen, China
  November, 2023
  

• Lower Bounds and Accelerated Algorithms in Distributed Optimization with Communication Compression
  Chinese Academy of Sciences, Beijing, China
  November, 2023
  

• Accelerating Decentralized Deep Training with Sparse and Effective Topologies
  Chinese Academy of Sciences, Beijing, China
  November, 2023