Currently, the spaCy model defaults to en. However, en is not an option anymore. Reference: https://spacy.io/models/en
The three English pretrained models are:
en_core_web_sm
en_core_web_md
en_core_web_lg
This PR changes the default to en_core_web_md which is the smallest model with GloVe vectors, and has a reasonable download size of 48MB, vs the lg model which is 746MB.
Currently, the spaCy model defaults to
en. However,enis not an option anymore. Reference: https://spacy.io/models/enThe three English pretrained models are:
en_core_web_smen_core_web_mden_core_web_lgThis PR changes the default to
en_core_web_mdwhich is the smallest model with GloVe vectors, and has a reasonable download size of 48MB, vs thelgmodel which is 746MB.