Do you know the solution for this issue? ----> PT-PT vs PT-BR

[CROSSEO]

I just start using spacy in my models and I would like to know if it's possible to have the spacy.load("pt-pt") and not only spacy.load("pt").

The reason is that in portuguese you have "pt-pt" that stands for Portuguese of Portugal and you have "pt-br" that stands for portuguese of Brazil.

And when mixing this 2 languages is like mixing en-uk and en-us, that you have for example: British: Grey vs American: Gray.

Or

Can anyone explain me a better solution to solve this challenge?

[svlandeg]

Hi!

In fact in spaCy, there is only one "class" for English as well; en, which attempts to cover both American and English grammar. Likewise, the aim of the pt model is to cover both versions of Portuguese. Some related discussion can be found here: #2759

Brazilian and Portuguese grammars are the same so in theory, we don't need to have two separate languages. The differences would be in the meaning of some words. The changes that I made are valid for both.