crosslang_embed/main.py at master · pathway/crosslang_embed · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
import crosslang_embed

phrase_list = ['hi', 'hello', 'bon nuit', 'bonjourno', 'boujour', 'bon soir', 'good night',
               'good evening']


mph = crosslang_embed.MultilangPhrase()

# translate and embed phrases
mph.process_phrases(phrase_list)

# Show languages and translations for all phrases ---

print(mph.dfmain)

# Find matches for any single phrase ---

print("find_matches_index:","salut")
df = mph.find_matches_index("salut", )
print( df[ df.dist<1.2].head(30) )


print("find_matches_index:","good evening")
df = mph.find_matches_index("good evening",)
print( df[ df.dist<1.2].head(10) )


# Plot phrases to see clustering ---

mph.umap_embeddings()

mph.plot_embeddings()