How to reproduce the behaviour
>>> nlp = spacy.load('pt_core_news_sm')
>>> doc=nlp('''Feito.''')
>>> for token in doc:
... print(token.text, token.lemma_, token.pos_)
...
Feito Feito VERB
. . PUNCT
And it is not exclusively when starting a sentence:
>>> nlp = spacy.load('pt_core_news_sm')
>>> doc=nlp('''Eu havia Feito o trabalho.''')
>>> for token in doc:
... print(token.text, token.lemma_, token.pos_)
...
Eu Eu PRON
havia haver AUX
Feito Feito VERB
o o DET
trabalho trabalhar NOUN
. . PUNCT
It should return "Fazer" as in:
>>> nlp = spacy.load('pt_core_news_sm')
>>> doc=nlp('''feito.''')
>>> for token in doc:
... print(token.text, token.lemma_, token.pos_)
...
feito fazer VERB
. . PUNCT
Your Environment
- Operating System: Windows
- Python Version Used: Python 3.8.5
- spaCy Version Used: 2.3.5
- Environment Information:
My example includes a verb, but I've also tested using nouns and the same bug occurs.
How to reproduce the behaviour
And it is not exclusively when starting a sentence:
It should return "Fazer" as in:
Your Environment
My example includes a verb, but I've also tested using nouns and the same bug occurs.