I need to use a non-spacy language model (russian; syntagrus). There is a special pipeline component to read non-spacy languages into spacy (spacy-stanza), but the biggest problem for me is,
"where do I download the models"?
I'm forwarded from one page to the next to arrive at the starting page (or a
page that eppears to have little to do with the language model (https://universaldependencies.org/)
And if you have experience on how to load them into spacy...
spacy-stanza lets use any loaded Stanza model in spaCy. You'll have to download the data via Stanza, load it and then pass it to the spaCy wrapper. The list of models is published: Available Models & Languages - Stanza See here for the documentation on loading models: Download Models - Stanza I think you can just do it via the language code? Universal Dependencies is likely the data source the model was trained on, which is why the page linked to it.
sorry for the late response
you also offered the solution by releasing the official spaCy ru model 2 days after I asked the question. But it's good to know for other non-supported languages