Is there some way to decode the hash of spaCy token entity types without accessing the ent_type_
attribute?
For example:
for token in doc:
hash_val = token.ent_type
string_repr = some_map[hash_val]
It's stored in the vocab.
nlp.vocab.strings[token.ent_type] == token.ent_type_
Not sure why you would need to do that though.