Currently a Master's student at Beihang University. My work focuses on distributed LLM training systems, especially long-sequence RLHF optimization, fine-grained scheduling, KV Cache management, and pipeline parallelism.
- On X (@0xHongliang), I share high-signal observations on AI infrastructure and systems.
- On GitHub, I build practical tools and systems that I use myself.
- Personal site: chlience.com.
- GitHub: Chlience
- X: @0xHongliang
- Website: chlience.com
Make large-scale AI systems more efficient. Write things worth reading.