Yuxiao Guo
Associate Researcher 2
Chaoyang District, Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolYuxiao Guo is an Associate Researcher with 10 years of experience in machine learning and systems engineering, currently based in Chaoyang District, Beijing and working at Microsoft. He has a strong background in deep learning infrastructure, demonstrated by contributions to the high-profile Microsoft Cognitive Toolkit (CNTK) where he added multi-GPU support and refactored ResNet implementations to improve maintainability and distributed training performance. His experience includes internships at NVIDIA and Baidu, giving him practical exposure to performance-sensitive engineering such as APEX development and app server work. Trained in computer science and software engineering at the University of Electronic Science and Technology, he blends research-oriented thinking with production-grade coding and a knack for optimizing model evaluation pipelines. An understated strength is his ability to reorganize core libraries for long-term scalability, not just short-term feature delivery.
10 years of coding experience
Computer Science, Computer Science at University of Electronic Science and Technology
Github Skills (11)
neural-network10
machine-learning10
c-language10
cntk10
deep-learning10
distributed-training10
cprogramming-language10
batch-normalization10
python6
dotnet-core5
csharp5
Programming languages (5)
Github contributions (5)
microsoft/CNTK
May 2016 - Jan 2018
Microsoft Cognitive Toolkit (CNTK), an open source deep-learning toolkit
Role in this project:
ML Engineer
ML EngineerContributions:54 commits, 13 PRs, 266 pushes in 1 year 8 months
Contributions summary:Yuxiao contributed to the Microsoft Cognitive Toolkit (CNTK) by adding support for multi-GPU processing within the PBN (probably "Post Batch Normalization") evaluation process, modifying the `SimpleEvaluator.h` and `EvalActions.cpp` files to include multi-GPU support. They also refactored the ResNet implementation, moving the model definitions into a dedicated file to keep the project organized, demonstrating an understanding of deep learning model architecture. The user's changes involve modifications to core library files, and adding batch normalization evaluation, showing a focus on deep learning model optimization and distributed training.
pytorchpythondeep-learningc-plus-plusmachine-learning
yuxiaoguo/DLEngine
Jun 2023 - Jun 2024
A deep learning framework with PyTorch
Contributions:168 PRs, 115 pushes, 158 branches in 1 year
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.