Haichen Huang
Machine Learning Engineer at hpcaitech
Beijing, China
Summary
🤩
RockstarHaichen Huang is a Machine Learning Engineer based in Beijing with five years of experience building distributed training systems and parallel computing solutions for large-scale models. He contributes to Colossal-AI, implementing and optimizing 2D parallel matrix multiplication and MoE parallelization to make large AI models cheaper and faster. Comfortable in low-level system work as well as model-level optimizations, he bridges algorithmic design and production-grade distributed implementations. His GitHub record shows hands-on performance tuning, bug fixes, and example-driven documentation that help teams adopt complex parallelism techniques. Colleagues can expect a pragmatic engineer who focuses on scalable, reproducible ML training and improving developer experience around distributed workflows.
5 years of coding experience
Github Skills (11)
data-parallel10
pytorch10
machine-learning10
data-parallelism10
deep-learning10
parallelization10
python10
ai10
large-scale10
distributed-computing10
tensorflow4
Programming languages (3)
Github contributions (5)
hpcaitech/ColossalAI
Dec 2021 - Jan 2023
Making large AI models cheaper, faster and more accessible
Role in this project:
ML Engineer
ML EngineerContributions:227 reviews, 123 commits, 243 PRs in 1 year 1 month
Contributions summary:Haichen implemented and optimized machine learning models within the Colossal-AI framework. The commits reveal the implementation of 2D parallel matrix multiplication operations, a critical component for large-scale model training, and integration of MoE (Mixture of Experts) parallelization techniques. The user also added examples, fixed bugs and enhanced the existing codebase, especially focusing on improvements to distributed systems.
heterogeneous-trainingcolossal-aifinetuningdeep-learninginference
1SAA/ColossalAI
Mar 2022 - Apr 2022
Colossal-AI: A Unified Deep Learning System for Large-Scale Parallel Training
Contributions:144 pushes in 1 month
pytorchcolossal-aiparalleldeep-learningmachine-learning
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.