AwesomeCodingBoy
Student at PEKING UNIVERSITY
Summary
🤩
RockstarAwesomeCodingBoy is a student and seasoned developer with seven years of hands-on experience focusing on backend systems and machine learning engineering. He contributes actively to open-source ML tooling, notably improving OpenPPL/ppq—the offline neural network quantization tool—by implementing Subgraph Optimization, Matrix Factorization, channel-split algorithms, and MSE calibration while fixing CUDA histogram and batch-norm merge bugs. Comfortable shipping core algorithmic changes and practical engineering fixes, he bridges research-grade quantization methods with production-ready tooling. Based in China and studying at Peking University, he combines academic grounding with real-world impact on a widely used quantization project. An uncommon strength is his focus on both algorithmic innovations and meticulous bug resolution, ensuring theoretical improvements translate into robust, deployable code.
7 years of coding experience
Github Skills (14)
neural-network10
algorithm10
quantization10
numerical-optimization10
code-optimization10
algorithms10
pytorch10
deep-learning10
onnx10
optimisation10
python10
optimization10
caffe9
cuda9
Programming languages (3)
Github contributions (5)
OpenPPL/ppq
Jan 2022 - Jan 2023
PPL Quantization Tool (PPQ) is a powerful offline neural network quantization tool.
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:4 releases, 7 reviews, 136 commits in 1 year
Contributions summary:AwesomeCodingBoy primarily contributed to bug fixes related to the merging of batch normalization layers within the quantization process. They also updated the project to version 0.5.3, adding a channel split algorithm, MSE calibration methods, and fixing bugs within the CUDA histogram and batch normalization merging processes. Further contributions include core updates related to the addition of a Subgraph Optimization Algorithm, Matrix Factorization Algorithm, and refining network analysis for neural network quantization.
cudapytorchdeep-learningonnxnetwork-quantization
ZhangZhiPku/ppq
Dec 2021 - Jun 2023
Contributions:127 pushes, 177 branches in 1 year 6 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.