Efficient GPU-based training of recurrent neural network language models using spliced sentence bunch

Xie Chen, Yongqiang Wang, Xunying Liu, Mark J. F. Gales, Philip C. Woodland