About me
My Name is Yang Jing，
Currently, I'm a Senior Engineer at Huawei's xxx CTO Office, working on Large-scale clusters LLM Training and System-Algorithm Co-Design.
What I focus on:
- LLM Code Agent: Building Benchmark Agent Coding in Real World;
- LLM Training: Experiences and lesssions in training LLM （not Pangu : ）, Currently obsessed experimenting with data synthesis , model design, and training strategies tricks to boost LLM Performance ;
- LLM Infra : Deep into the Ascend software stack - MindSpeed-LLM/RL, vLLM-Ascend - optimizing inference and RL training pipelines. Currently obsessed with pushing RL training efficiency on 910B/C Cluster ;
- Network: Optimized 10K+ card trainining clusters, including large scale 910B/C clusters. Tuning HCCL and UB-Mesh configs to squeeze every bit of performance from the network fabric;