Scholar
Hao-Jun Michael Shi
Google Scholar ID: U1efqpIAAAAJ
Research Scientist, Meta
Citations & Impact
Citations
1,778
H-index
10
i10-index
10
Publications
16
Co-authors
25
list available
Publications
Resume (English only)
Academic Achievements
- Published an efficient PyTorch implementation of the Distributed Shampoo optimizer; contributed to the development of Facebook's open-source DLRM.
Research Experience
- Research Scientist at Meta in the AI and Systems Co-Design Training team; contributed to Facebook's open-source deep learning-based recommendation model (DLRM) and developed embedding compression techniques (QR embedding) during a prior internship at Facebook.
Education
- Ph.D. in numerical optimization, with research in neural network training algorithms, stochastic optimization (progressive batching quasi-Newton), noisy optimization (noise-tolerant quasi-Newton), and derivative-free optimization (adaptive finite-difference methods under noise).
Background
- Research interests: Exploration and implementation of scalable and distributed training algorithms and systems, particularly for deep learning. Specialization: Numerical optimization, deep learning.
Miscellany
- Enjoys eating good food, reading interesting books, playing basketball, rooting for UCLA sports teams (Go Bruins!), serving at his church, spending time with his wife, and attempting to entertain his cat Taro.
