explosion / sense2vec

🦆 Contextually-keyed word vectors
https://explosion.ai/blog/sense2vec-reloaded
MIT License
1.62k stars 240 forks source link

different results on the demo than through the spacy script #147

Open shanihadar opened 2 years ago

shanihadar commented 2 years ago

Hello, When using the demo with the word "Height", the sense "auto" and the year "2019" the results are amazing while running through spacy the results are completly different.

example: import spacy nlp_spacy = spacy.load("en_core_web_sm") s2v = nlp_spacy.add_pipe("sense2vec") s2v.from_disk("s2v_reddit_2019_lg")

doc = nlpspacy("Height") assert doc[:].text == "Height" freq = doc[:]..s2vfreq vector = doc[:]..s2v_vec mostsimilar = doc[:]..s2v_most_similar(3)

Why is that? and how can I get the same results as in the demo? Thanks.