Create dataset loader for Open subtitles

[SamuelCahyawijaya]

NusaCatalogue: https://indonlp.github.io/nusa-catalogue/card.html?open_subtitles

Dataset	open_subtitles
Description	OpenSubtitles is a large parallel corpora from Movie and TV Subtitles. The corpora contains parallel sentences between English and Indonesian. The dataset contains 9827 documents, 56.8K sentences, 9.7M English tokens and 60.9M Indonesian tokens.
License	GNU General Public License v3.0

[haryoa]

#self-assign