Heavyweight Statistical Alignment to Guide Neural Translation

Table 1

Some basic statistics of the datasets.

English-Vietnamese	Training	Development	Testing
Sentence pairs	42026	1482	1527
Average lengths	19.2–26.2	17.8–24.5	20.6–28.3
Words	806456–1099205	26315–36276	31513–43286
Dictionaries	36672–16441	4981–2720	6211–3462