Liang Zhao
Associate Professor
Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolLiang Zhao is an associate professor at Tsinghua University with 7 years of professional experience and a PhD in Inorganic Chemistry from The Chinese University of Hong Kong. His background blends deep academic research—anchored by a postdoctoral stint at the University of Utah—with hands-on machine learning engineering, evidenced by contributions to the widely used OpenDILab DI-engine reinforcement learning framework. Liang has implemented and adapted environments and replay-buffer instrumentation (including TensorBoard logging) to improve training observability and self-play workflows. Based in Beijing, he bridges chemistry-driven scientific rigor and practical ML system design, bringing a curiosity for applying computational methods to complex problems beyond traditional laboratory boundaries.
7 years of coding experience
Bachelor of Science (BS), Chemistry, Bachelor of Science (BS), Chemistry at Peking UniversityThe Chinese University of Hong Kong (CUHK)
Github Skills (4)
reinforcement-learning10
python10
pytorch10
imitation-learning8
Programming languages (4)
Github contributions (5)
opendilab/DI-engine
Sep 2021 - Aug 2022
OpenDILab Decision AI Engine. The Most Comprehensive Reinforcement Learning Framework B.P.
Role in this project:
ML Engineer
ML EngineerContributions:5 reviews, 18 commits, 15 PRs in 11 months
Contributions summary:The user, ZLX, primarily contributed to the development of reinforcement learning components and pipelines within the DI-engine framework. They added and modified replay buffer functionalities, specifically integrating TensorBoard logging for enhanced monitoring and debugging of the naive and advanced replay buffers. ZLX also implemented and integrated the Slime Volley environment, including both vs. bot training and self-play training capabilities, while adapting the environment to DI-engine's requirements.
multiagent-reinforcement-learningpythonoffline-rldecision-aipytorch-rl
LuciusMos/DeepSpeedExamples
Apr 2023 - May 2023
Example models using DeepSpeed
Contributions:110 pushes in 24 days
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.