3 ms·
It's actually not at all clear if they couldn't be used. Things like Google's word2vec model are derived from copyrighted data and seem to be fine. Perhaps dis
by nl 10y ago
It's actually not at all clear if they couldn't be used. Things like Google's word2vec model are derived from copyrighted data and seem to be fine.
Perhaps distribution of the subtitles is illegal under copyright law, but I suspect the trained model would be fine.
It would be an interesting case anyway.