Heyang Sun
AI Engineer at Intel Corporation
Hong Kong Island, Hong Kong
Summary
🤩
Rockstar🎓
Top SchoolHeyang Sun is an AI engineer with six years of experience specializing in deploying and optimizing LLMs on Intel hardware, and is a core author of the popular IPEX LLM open-source project (6k+ stars). He combines hands-on ML engineering and DevOps skills—building Kubernetes integrations, SGX device plugins, CI/CD for model training, and distributed CPU/MPU inference pipelines using OneCCL and Intel MPI. At Intel he delivered practical optimizations like NF4 QLoRA, low-bit speculative decoding, and tensor-parallel inference used by major customers including Azure, Tencent, and Alibaba Cloud. Comfortable across system-level C++ work and higher-level ML tooling, he has moved models from research code into production on Xeon, Arc, Flex/Max GPUs and AI-PCs. Notably, he focuses on compute-aware finetuning and inference on constrained/heterogeneous Intel XPU setups—a niche that often gets overlooked in cloud-first AI teams. Based in Hong Kong, he prefers engineering roles focused on systems and model deployment rather than pure data analysis.
6 years of coding experience
1 year of employment as a software developer
Bachelor of Engineering - BE Software Engineering, Bachelor of Engineering - BE Software Engineering at Southeast University
Github Skills (12)
transformers10
kubernetes10
pytorch10
gpu10
devops10
kubernetes-pods10
llm10
mlops9
deepspeed9
docker8
cicd8
dockers8
Programming languages (5)
Github contributions (5)
intel/ipex-llm
Sep 2021 - Jan 2023
Accelerate local LLM inference and finetuning (LLaMA, Mistral, ChatGLM, Qwen, DeepSeek, Mixtral, Gemma, Phi, MiniCPM, Qwen-VL, MiniCPM-V, etc.) on Intel XPU (e.g., local PC with iGPU and NPU, discrete GPU such as Arc, Flex and Max); seamlessly integrate with llama.cpp, Ollama, HuggingFace, LangChain, LlamaIndex, vLLM, DeepSpeed, Axolotl, etc.
Role in this project:
DevOps Engineer & ML Engineer
DevOps Engineer & ML EngineerContributions:283 reviews, 132 commits, 509 PRs in 1 year 4 months
Contributions summary:Heyang primarily contributed to the deployment and finetuning of Large Language Models (LLMs) on Intel XPU platforms, with a focus on Kubernetes integration, including the addition of SGX device plugins and addressing related deployment issues. They also worked on optimizing and configuring QLoRA finetuning on CPU platforms and integrated deepspeed for distributed CPU inference. The user's contributions include improvements to various examples, and the setup of CI/CD pipelines for model training and deployment, leveraging tools like OneCCL and Intel MPI.
llm-inferencellama2pythonfinetuningllama
Uxito-Ada/BigDL
Nov 2021 - Nov 2024
Building Large-Scale AI Applications for Distributed Big Data
Contributions:938 pushes, 270 branches in 3 years 1 month
machine-learningai-applicationssparkbig-datadistributed-computing
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.