We use the OpenNMT-py toolkit [12] for preprocessing our data, as well as for training and testing the models. We follow the approach of See et al.
確定！ 回上一頁
