Zhiyu Li

Zhiyu Li is a senior deep learning engineer at NVIDIA. He specializes in efficient, high-performance large language model (LLM) training systems and infrastructure, including pretraining and post-training frameworks. He holds an M.S. in Computer Science from Carnegie Mellon University.
Zhiyu Li

Posts by Zhiyu Li

Agentic AI / Generative AI

Scale Bitwise-Deterministic Pretraining with NVIDIA Megatron Core

Bitwise determinism makes large-scale pretraining easier to debug, validate, and resume reproducibly. These benefits become especially valuable when training... 11 MIN READ