Frank Seide, Hao Fu, Jasha Droppo, Gang Li, Dong Yu. On parallelizability of stochastic gradient descent for speech DNNS. In IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2014, Florence, Italy, May 4-9, 2014. pages 235-239, IEEE, 2014. [doi]
Abstract is missing.