Yucheng Lu

Yucheng Lu
Yucheng Lu
Assistant Professor of Computer Science, NYU Shanghai
Email
yl5541@nyu.edu
Homepage Link
https://www.yucheng-lu.me/

Yucheng Lu is an Assistant Professor of Computer Science at NYU Shanghai, where he leads the HeavyBall Research group. His research centers on hardware–software co-design for efficient machine learning systems, encompassing training and inference engines, GPU kernel optimization, and large-scale AI applications. Prior to joining NYU Shanghai, he was a Research Engineer at Together AI and held research internships at Google, Microsoft, and Amazon. He received his Ph.D. in Computer Science from Cornell University in 2023, advised by Chris De Sa.

 

Select Publications

  • Jue Wang*, Yucheng Lu*, Binhang Yuan, Beidi Chen, Percy Liang, Chris De Sa, Chris Ré, Ce Zhang. CocktailSGD: Fine-tuning Foundation Models over 500Mbps Networks. In the Fortieth International Conference on Machine Learning (ICML) 2023.
  • Yucheng Lu, Conglong Li, Minjia Zhang, Chris De Sa, Yuxiong He. Maximizing Communication Efficiency for Large-scale Training via 0/1 Adam. In the Eleventh International Conference on Learning Representations (ICLR) 2023.
  • Yucheng Lu, Wentao Guo, Chris De Sa. GraB: Finding Provably Better Data Permutations than Random Reshuffling. In the Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS) 2022.
  • Yucheng Lu, Chris De Sa. Optimal Complexity in Decentralized Training. In the Thirty-eighth International Conference on Machine Learning (ICML) 2021.
  • Yucheng Lu, Chris De Sa. Moniqua: Modulo Quantized Communication in Decentralized SGD. In the Thirty-seventh International Conference on Machine Learning (ICML) 2020.

 

Education

  • Ph.D., Computer Science
    Cornell University
  • B.Eng., Electronic Engineering
    Shanghai Jiao Tong University
Research Interests
  • Machine Learning Systems