Software Engineering undergraduate at Wuhan University of Technology.
Incoming M.S. student at Wuhan University, starting in 2027.
Interested in LLM training and post-training, with a secondary focus on LLM systems and prior research experience in AI for Science.
- LLM Training & Post-training: language model training, reasoning, alignment, and model capability improvement.
- LLM Systems: efficient LLM inference and serving, with a focus on memory management, caching, scheduling, and runtime efficiency.
Hands-on study of language model training, post-training, and architecture through reproducible experiments.
Studying caching, I/O, and scheduling for modern LLM serving systems.
PyTorch · Transformers · vLLM · SGLang · Slurm · OpenCL