Boxiang Wang

Boxiang Wang is a deep learning algorithm engineer at NVIDIA, specializing in large-scale large language model (LLM) training. He contributes to NVIDIA NeMo Framework and Megatron-Core, focusing on long-context LLM training, state-of-the-art model architectures, and advanced model parallelism techniques. Previously, he was a founding engineer and researcher at ColossalAI. He holds a bachelor's degree in Engineering from Nanyang Technological University and a master's degree in Computational Science and Engineering from Harvard University.