Paul O'Leary McCann

@explosion-bot please test_gpu

You can click on the red X to see why a build failed, though in this case the most recent build succeeded so you're fine. We'll take a look at...

Very late reply here, but just to follow up - the method for merging vocabs by calling `from_disk` on the same vocab multiple times may work, but it's not the...

Any ideas about what the issue was with lots of files? I've heard of git having issues with that but I think we have

Following internal discussion, it was pointed out that with some major changes to the way `Span`s work in v4 this approach alone won't be sufficient for ensuring consistency. I'm going...

@ezorita Had a quick look at the training data (OntoNotes) and it looks like "for-profit" is consistently annotated as a prepositional phrase with three tokens, while "non-profit" is a single...

@peterolson Sorry for the late reply, but thanks for reporting this. It does seem that the zh trf model really avoids recognizing short sentences. We took a look at our...

@narayanacharya6 Cannot reproduce with 3.2. Can you upgrade and try again? Also include your model versions (`spacy info`). Note that `’s` and `'s` are not the same, and the non-ASCII...

@saurav-chakravorty If you have a question it's better to open a new Discussion than to post in an unrelated thread. The training code is not public, partly because the training...

@mgrojo Thanks for pointing that out! If you'd like to open a PR we'd be happy to review it.