• search
    search
  • reviewers
    reviewers
  • feeds
    feeds
  • assignments
    assignments
  • settings
  • logout

PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel.

Yanli ZhaoAndrew GuRohan VarmaLiang LuoChien-Chin HuangMin XuLess WrightHamid ShojanazeriMyle OttSam ShleiferAlban DesmaisonCan BaliogluPritam DamaniaBernard NguyenGeeta ChauhanYuchen HaoAjit MathewsShen Li
Published in: Proc. VLDB Endow. (2023)
Keyphrases