AMPeD: An Analytical Model for Performance in Distributed Training of Transformers

Moolchandani, DikshaDikshaMoolchandaniKundu, JoyjitJoyjitKunduRuelens, FrederikFrederikRuelensVrancx, PeterPeterVrancxEvenblij, TimonTimonEvenblijPerumkunnil, ManuManuPerumkunnil2023-08-082023-07-272023-07-282023-08-082023N/AWOS:001017781500028https://imec-publications.be/handle/20.500.12860/42204AMPeD: An Analytical Model for Performance in Distributed Training of TransformersProceedings paper10.1109/ISPASS57527.2023.00037979-8-3503-9739-0WOS:001017781500028Computer science/information technology