Batch Tiling on Attention: Efficient Mixture of Experts Training on Wafer-Scale Processors.
Daria Soboleva, tienne Goffinet, Hui Zeng, Sangamesh Ragate, Elif Albuz, Natalia Vassilieva
Browse the full SC paper archive.
Daria Soboleva, tienne Goffinet, Hui Zeng, Sangamesh Ragate, Elif Albuz, Natalia Vassilieva
Browse the full SC paper archive.