en_vectors_web_lg loading issue

, ,
1

hi @ines,
I am using prodigy manual, ner.teach and train while annotating. Now i have 60k dataset so i need to use spacy to train rather prodigy.

I can run perfectly run prodigy train command with vectors model
prodigy train ner art_dataset62 en_vectors_web_lg --output prodigy_model --n-iter 8

SPacy train throws error:
spacy train en_vectors_web_lg spacy_model .\dataset62.json .\eval1.json --pipeline ner

in get_lang_class
raise ImportError(Errors.E048.format(lang=lang, err=err))
ImportError: [E048] Can't import language en_vectors_web_lg from spacy.lang: No module named 'spacy.lang.en_vectors_web_lg'

python -m spacy validate
:heavy_check_mark: Loaded compatibility table

====================== Installed models (spaCy v2.2.3) ======================
:information_source: spaCy installation: C:\anaconda3\lib\site-packages\spacy

TYPE NAME MODEL VERSION
package en-vectors-web-lg en_vectors_web_lg 2.1.0 :heavy_check_mark:
package en-core-web-sm en_core_web_sm 2.2.5 :heavy_check_mark:
package en-core-web-lg en_core_web_lg 2.2.5 :heavy_check_mark:

why prodigy command with vectors is working fine and spacy is not?. I don't understand

2

Hi! I think the problem is that you're passing in the vectors model name as the language code when you're training with spacy train. Instead, I think you want to be passing them in as --vectors. See the documentation of spacy train here: https://spacy.io/api/cli#train