Yilun Sun
React Developer at RightCapital
Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolYilun Sun is a React-focused full-stack developer with eight years of experience building web applications and shipping front-end features across startups and product teams. Currently at RightCapital and previously at DataPipeline and Amélio, he combines practical React expertise with backend contributions that improve performance and interoperability. He has notable open-source experience in high-profile ML projects like PaddlePaddle—adding multi-dtype support to collective communication APIs—and in vLLM, where he fixed model-loading and integration bugs for LLM inference. His background spans UX-driven interactive installations (an Unreal Engine exhibit at Lenovo) to teaching and robotics work optimizing algorithms for LEGO and VEX platforms, reflecting a blend of systems thinking and hands-on prototyping. Comfortable across the stack, Yilun brings reliability, a knack for debugging complex integrations, and an appetite for contributing to impactful open-source AI tooling.
8 years of coding experience
5 years of employment as a software developer
Master of Science - MS, Applied Computer Science, Master of Science - MS, Applied Computer Science at Concordia UniversityBachelor of Computer Science, Computer Science and Tecnology, Bachelor of Computer Science, Computer Science and Tecnology at Beijing University of Technology
English, French, Chinese
Stackoverflow
11reputation
1kreached
0answers
1question
Github Skills (22)
debugging10
debug10
distributed-training10
python10
machine-learning10
inference10
llm10
paddlepaddle10
pytorch9
apidoc9
api9
deep-learning9
openai9
gpt8
cprogramming-language8
Programming languages (14)
Github contributions (5)
PaddlePaddle/Paddle
Aug 2022 - Jan 2023
PArallel Distributed Deep LEarning: Machine Learning Framework from Industrial Practice （『飞桨』核心框架，深度学习&机器学习高性能单机、分布式训练和跨平台部署）
Role in this project:
Back-end Developer
Back-end DeveloperContributions:26 reviews, 37 commits, 13 PRs in 4 months
Contributions summary:Yilun primarily contributed to the PaddlePaddle project by implementing and completing basic data type support for the `all_reduce` collective communication API in eager mode. They modified and added code to the `python/paddle/distributed/collective.py` and test files to enable and verify the functionality for different data types. The user's changes involved modifying various methods to support various data types like float16, float32, float64, int32, int64, int8, uint8, bool and bfloat16 for the collective communication APIs.
pytorchpythonparalleldeep-learningpaddlepaddle
vllm-project/vllm
Jun 2023 - Mar 2025
A high-throughput and memory-efficient inference and serving engine for LLMs
Role in this project:Backend Engineer
Backend EngineerContributions:12 reviews, 8 PRs, 44 comments in 1 year 8 months
Contributions summary:Yilun primarily contributed to bug fixes and improvements within the VLLM project, addressing issues related to model loading, configuration, and API behavior. Their work included correcting errors in the handling of Bloom and GPTBigCode models, fixing a typo, and aligning the `max_tokens` behavior with OpenAI's specifications. They also fixed an issue related to Ray integration and documented code blocks.
amdcudadeepseekgpthpu
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.