boudinfl/pke

LookupError: Can't find any language named 'hinglish'

alexrao02 opened this issue · 1 comments

Thanks for ur work! Here is a problem about the package.
I run the version 0729bca in colab(I have to run this version because of demands).
And I get this error:

LookupError: Can't find any language named 'hinglish'

How can I fix it?

ygorg commented

Hi, thank you for this issue. pke does not support 'hinglish' language, there is no spacy model for this language and no stemmer.
You can specify language='en' to treat these document as english.
Alternatively use the latest version and preprocess your documents as such using custom tokenizer and postagger:

excp_msg += "\n\t[[('The', 'DET'), ('brown', 'ADJ'), ('fox', 'NOUN'), ('.', 'PUNCT')]]"