Kazutoshi Hirose, Ryota Uematsu, Kota Ando, Kentaro Orimo, Kodai Ueyoshi, Masayuki Ikebe, Tetsuya Asai, Shinya Takamaeda-Yamazaki, Masato Motomura. Logarithmic Compression for Memory Footprint Reduction in Neural Network Training. In Fifth International Symposium on Computing and Networking, CANDAR 2017, Aomori, Japan, November 19-22, 2017. pages 291-297, IEEE Computer Society, 2017. [doi]
Abstract is missing.