Changmin Zhao
Deep Learning Software Intern at Intel Corporation
Wuhan, Hubei, China
Summary
🤩
Rockstar🎓
Top SchoolChangmin Zhao is a deep learning software engineer with nine years of experience, currently interning at Intel where he focuses on accelerating local LLM inference and model conversion for Intel XPU platforms. He specializes in back-end engineering and ML tooling, notably contributing to the popular intel/ipex-llm project by integrating GPTQ models and converting model scripts to streamline deployment within the BigDL-LLM ecosystem. With a Master’s in Computer Technology from Huazhong University of Science and Technology and a strong engineering undergraduate background, he blends rigorous academic training with practical open-source impact. Known as a "domesticated programmer," he pairs pragmatic scripting and code-quality improvements with deep model-format expertise to make advanced LLMs more usable on consumer and edge hardware.
9 years of coding experience
Master, Computer Technology, Master, Computer Technology at Huazhong University of Science and Technology
Github Skills (11)
transformers10
pytorch10
machine-learning10
model-conversion10
python10
llm10
scripting9
gpu9
shell9
script9
sh9
Programming languages (5)
Github contributions (5)
intel/ipex-llm
Nov 2020 - Jan 2023
Accelerate local LLM inference and finetuning (LLaMA, Mistral, ChatGLM, Qwen, 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, GraphRAG, DeepSpeed, Axolotl, etc
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:250 reviews, 151 commits, 269 PRs in 2 years 2 months
Contributions summary:Changmin primarily contributed to the model conversion process within the repository, specifically focusing on integrating and converting different model types, including those from GPTQ, into a format compatible with the BigDL-LLM framework. They worked on converting model scripts into shell scripts for efficient model conversion. The user's work encompassed various aspects of model conversion, including modifying existing model conversion tools and integrating GPTQ models, which improved model deployment and usability. They also addressed code quality and documentation issues.
llm-inferencellama2pythonfinetuningllama
leonardozcm/ASFM-Net-Pytorch
Dec 2021 - May 2023
This is a unofficial implements of ASFM-Net, hope it works...
Contributions:3 releases, 8 PRs, 27 pushes in 1 year 4 months
dotnetcsharp
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.