TY - RPRT TI - SuperNeurons: FFT-based Gradient Sparsification in the Distributed Training of Deep Neural Networks AU - Linnan Wang AU - Wei Wu AU - Junyu Zhang AU - Hang Liu AU - George Bosilca AU - Maurice Herlihy AU - Rodrigo Fonseca PY - 2021 UR - https://arxiv.org/abs/1811.08596 ID - 1811.08596 ER -