"SWARM Parallelism: Training Large Models Can Be Surprisingly ..."

Max Ryabinin et al. (2023)

> Home

Details and statistics

DOI: —

access: open

type: Conference or Workshop Paper

metadata version: 2023-08-28

- view
  - electronic edition @ mlr.press (open access)
- export record
  dblp key:
  - conf/icml/RyabininDDB23
- ask others
- share record
  persistent URL:
  - https://dblp.org/rec/conf/icml/RyabininDDB23
Max Ryabinin, Tim Dettmers, Michael Diskin, Alexander Borzunov:
SWARM Parallelism: Training Large Models Can Be Surprisingly Communication-Efficient. ICML 2023: 29416-29440

manage site settings

To protect your privacy, all features that rely on external API calls from your browser are turned off by default. You need to opt-in for them to become active. All settings here will be stored as cookies with your web browser. For more information see our F.A.Q.