Bridging the Gap between Training and Inference for Neural Machine Translation from qun Watch Video
Preview(s):
Gallery
Play Video: (Note: The default playback of the video is HD VERSION. If your browser is buffering the video slowly, please play the REGULAR MP4 VERSION or Open The Video below for better experience. Thank you!)
Description: Authors: Wen Zhang, Yang Feng, Fandong Meng, Di You, Qun LiunnNeural Machine Translation (NMT) generates target words sequentially in the way of predicting the next word conditioned on the context words. At training time, it predicts with the ground truth words as context while at inference it has to generate the entire sequence from scratch. This discrepancy of the fed context leads to error accumulation among the way. Furthermore, word-level training requires strict matching between the ge
Play Video: (Note: The default playback of the video is HD VERSION. If your browser is buffering the video slowly, please play the REGULAR MP4 VERSION or Open The Video below for better experience. Thank you!)