AMPeD: An Analytical Model for Performance in Distributed Training of Transformers.
Diksha Moolchandani, Joyjit Kundu, Frederik Ruelens, Peter Vrancx, Timon Evenblij, Manu Perumkunnil
Browse the full ISPASS paper archive.
Diksha Moolchandani, Joyjit Kundu, Frederik Ruelens, Peter Vrancx, Timon Evenblij, Manu Perumkunnil
Browse the full ISPASS paper archive.