A Graph Theoretic Framework of Recomputation Algorithms for Memory-Efficient Backpropagation

NeurIPS 2019 Mitsuru KusumotoTakuya InoueGentaro WatanabeTakuya AkibaMasanori Koyama

Recomputation algorithms collectively refer to a family of methods that aims to reduce the memory consumption of the backpropagation by selectively discarding the intermediate results of the forward propagation and recomputing the discarded results as needed. In this paper, we will propose a novel and efficient recomputation method that can be applied to a wider range of neural nets than previous methods... (read more)

PDF Abstract

Results from the Paper


  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.