?
SWARM Parallelism: Training Large Models Can Be Surprisingly Communication-Efficient
P. 29416–29440.
Ryabinin Max, Dettmers T., Diskin Michael, Borzunov Alexander
В книге
Vol. 202: International Conference on Machine Learning, 23-29 July 2023, Honolulu, Hawaii, USA. , PMLR, 2023.