Skip to content

Topic

LLM training efficiency

Work aimed at cutting the time, compute and memory needed to train large language models, spanning optimizer design, kernel implementation and scheduling.

Current clusters