Shijie Wang
Home Sweet Home at home
Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolShijie Wang is a software engineer with eight years’ experience building deep learning frameworks and core AI infrastructure, currently based in Beijing. He has worked at Alibaba’s DAMO Academy on Qwen core development and previously contributed as a core developer at OneFlow, where he implemented operators, optimized kernels, and fixed subtle bugs in broadcast ops and data type conversion. His background blends research and engineering—from research stints at NTU and HUST to production work at CONTEK and Alibaba—demonstrating fluency in both ML internals and backend systems. An active open-source contributor, he has hands-on experience improving OneFlow’s performance and functionality, including linalg.inv and COCO reader enhancements. Colleagues describe him as quietly persistent—now doing "home sweet home" work while navigating non-compete constraints—bringing pragmatic problem-solving to complex ML systems.
8 years of coding experience
5 years of employment as a software developer
Master's degree, Master's degree at Huazhong University of Science and Technology
Github Skills (13)
neural-network10
cuda10
machine-learning10
deeplearning-ai10
c-language10
deep-learning10
deep-neural-networks10
cprogramming-language10
python10
ml10
pytorch9
distributed-systems8
tensorflow8
Programming languages (4)
Github contributions (5)
Oneflow-Inc/oneflow
Jan 2020 - Aug 2022
OneFlow is a deep learning framework designed to be user-friendly, scalable and efficient.
Role in this project:
Back-end & ML Engineer
Back-end & ML EngineerContributions:154 reviews, 157 commits, 81 PRs in 2 years 7 months
Contributions summary:Shijie's contributions primarily involve modifying and testing core components of the OneFlow deep learning framework. They addressed merge conflicts, fixed bugs related to broadcast operations (like multiplication, floor modulo, and NaN handling), and implemented new functionalities such as the linalg.inv operator and enhancements for the COCO dataset reader. Furthermore, they have worked on internal components like data type conversion as well as various kernel optimizations, showcasing their understanding of the framework's internals and performance considerations.
caffe2tensorflowframework-learningchainerdeep-learning-framework
simonJJJ/transformers
Jul 2024 - Dec 2024
🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.
Contributions:1 review, 1 PR, 89 pushes in 4 months
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.