Zhiyuan Chen
Software Engineer at Microsoft
Mountain View, California, United States
Summary
🤩
Rockstar🎓
Top SchoolZhiyuan Chen is a software engineer based in Mountain View with nine years of experience building robust backend systems, cloud data pipelines, and game engines across companies from startups to Microsoft. He blends strong Java/Spring and distributed systems expertise with hands-on work in cloud platforms (GCP), BigQuery/Beam pipelines, and CI/CD automation, and has applied deep probability and statistics to design and verify game math. At Microsoft he continues to scale production systems while contributing to open-source ML tooling—notably enhancements to Hugging Face’s widely used accelerate library around distributed training and multi-GPU data handling. His background includes full-stack and ETL roles, a high GPA master’s in computer engineering, and a track record of turning complex algorithms into auditable, regulator-ready implementations.
9 years of coding experience
5 years of employment as a software developer
Bachelor of Science - BS, Electronic Information Science and Technology, Bachelor of Science - BS, Electronic Information Science and Technology at East China Normal UniversityMaster's degree, Computer Engineering, 3.94, Master's degree, Computer Engineering, 3.94 at Stevens Institute of Technology
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (6)
pytorch10
machine-learning10
distributed-training10
python10
model-optimization10
mlops9
Programming languages (22)
Github contributions (5)
huggingface/accelerate
Feb 2022 - Jul 2022
🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support
Role in this project:
ML Engineer
ML EngineerContributions:1 review, 4 commits, 6 PRs in 5 months
Contributions summary:Zhiyuan contributed significantly to the `accelerate` library, which focuses on simplifying PyTorch model training and deployment. Their commits involved adding support for features like gathering objects across distributed processes and enhancing compatibility with various data types. The user also improved the library by incorporating properties related to distributed training and adding decorators for running functions on main and local processes, showing an understanding of distributed training and model optimization. Furthermore, the user addressed issues related to data loading in multi-GPU environments and contributed to several example scripts to incorporate newly added features.
deepspeedpytorch
ZhiyuanChen/.setup
Mar 2022 - Apr 2026
Contributions:34 pushes, 1 branch in 4 years 1 month
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.