Leapfrogging for parallelism in deep neural networks
Abstract
We present a technique, which we term leapfrogging, to parallelize back- propagation in deep neural networks. We show that this technique yields a savings of 1-1/k of a dominant term in backpropagation, where k is the number of threads (or gpus).
0
Turn this paper into a lesson
ArcXiv compiles a structured reading guide from this paper's metadata: plain-English importance, contributions, prerequisite concepts, which sections to read first, flashcards, and a quiz. Grounded in the abstract, never invented.