Changyue Liao

I received my M.S. from Zhejiang University. I am interested in language models, machine learning, and efficient AI systems.

Blog / Github / Google Scholar / LinkedIn

Research

My interests broadly lie in machine learning and AI systems. I care about how language models reason, learn from data, and run reliably and efficiently in real-world systems.

Language Models and Reasoning

How can models understand language, represent knowledge, and solve problems that require multi-step reasoning?

Efficient AI Systems

How can we make model training and inference faster, more reliable, and easier to use?

Publications

Ratel: Optimizing Holistic Data Movement to Fine-tune 100B Model on a Consumer GPU.

Changyue Liao, Mo Sun, Zihan Yang, Jun Xie, Kaiqi Chen, Binhang Yuan, Fei Wu, Zeke Wang.

IEEE International Conference on Data Engineering (ICDE), 2025

SmartReduce: SmartSwitch-Aware SmartNIC-Centric Sparse AllReduce

Yingtao Li, Hongjing Huang, Changyue Liao, Zihan Yang, Mo Sun, Xueying Zhu, Jie Zhang, Bo Zhao, Zeke Wang.

International Conference for High Performance Computing, Networking, Storage, and Analysis (SC), 2026

DisDP: Disaggregating Compute, Network, and Storage for Model-Sharded Data-Parallel Training.

Mo Sun, Zihan Yang, Changyue Liao, Yingtao Li, Jie Zhang, Kaiqi Chen, Fei Wu, Zeke Wang.

International Symposium on Computer Architecture (ISCA), 2026

Acknowledged Work

Selected projects that acknowledge my support and contributions.

VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo — acknowledged for contributions to the VeOmni project.
SAIL-VL2 Technical Report — acknowledged for valuable support and contributions.

Blogs

More writing coming soon.

All posts

Representative Work

Tiny3D

Research and engineering explorations related to 3D learning.

[code]

A BERT-based project for the complete fill-in-the-blank task.

[code]

ASC-RACE

Models and experiments for reading comprehension.

[code]

Miscellaneous

I enjoy breaking complicated questions into small, testable ones.
I write about AI, systems engineering, and things I learn along the way.
This is a good place for hobbies, personal stories, or a line of humor.