500AI
Back to researchers

Samyam Rajbhandari

Memory-efficient distributed training (ZeRO)

Co-authored ZeRO: foundational memory optimizations for training very large models.

Highlights

SystemsTrainingScaling
Focus: Memory-efficient distributed training (ZeRO)
Why it matters: Co-authored ZeRO: foundational memory optimizations for training very large models.
Start here
ZeRO: Memory Optimizations Toward Training Trillion Parameter Modelspaper →

Research Areas

SystemsTrainingScaling
Samyam Rajbhandari - AI Researcher Profile | 500AI