Open bustrofedico opened 5 years ago
We use the Penn Treebank to train -- here is one page that explains the labels used: http://www.surdeanu.info/mihai/teaching/ista555-fall13/readings/PennTreebankConstituents.html
You can also access parser._label_vocab
and parser._tag_vocab
(for the newer models only); these are private but you can print them to get a sense of what the full set of labels is.
Thanks Nikita! Exactly what I needed.
Can I assume the "Treebank-3" https://catalog.ldc.upenn.edu/LDC99T42 was used in training?
That's correct
Hi, what is the full tagset used by the parser? Thanks!