[E029] Trained models cause noun_chunks requires the dependency parse, which requires a statistical model to be installed and loaded

,
1

Hello Team,

I have successfully trained my first (combined) models. Thank you for your great support here on the forum. :muscle:

Although it's not perfect yet, and that's ok, I'm able to recognize my entities in text. The next step would be to describe their relationships. My first attempt has been to do something like you've done here: https://github.com/explosion/spaCy/blob/master/examples/information_extraction/entity_relations.py
: extract_currency_relations

My model is unfortunately throwing the following error:

import spacy
nlp = spacy.load(".../_models/mymodel")
doc = nlp("My sentence")
list(doc.noun_chunks)

[E029] noun_chunks requires the dependency parse, which requires a statistical model to be installed and loaded. For more info, see the documentation

I've searched through this forum and found only the following threat that deals with the accesibility of noun_chunks (Can't merge non-disjoint spans when using terms.train-vectors). I also looked at SpaCy github and SpaCy docs, but I didn't manage to solve my problem.

Can you please help?

2

Solved.

I used the en_core_web_lg to train my model instead of the en_vectors_web_lg and it works now.